CN103733041A - Management device, management method, program, and recording media - Google Patents

Abstract

A management device (40) acquires a measurement data group comprising a plurality of property values obtained by a plurality of measuring instruments (11) that measure the properties of an introduced workpiece, and calculates parameters for indicating measurement error between the measuring instruments on the basis of the measurement data group.

Description

Management devices, management method, program and recording medium

Technical field

The present invention relates to a kind of management devices that measuring error between measuring appliance used on production line is managed.

Background technology

In the past, on production of articles line, possessed and had the inspection step that the various characteristics of the workpiece as end article or intermediate is checked.This type of checks in step and can use in order to measure the measuring appliance of characteristic.If the measured value that measuring appliance records drops in specialized range, judge that goods are non-defective unit, if do not drop in specialized range, judge that goods are defective products.

Conventionally, measuring appliance is when measuring, longer to the processing time of 1 workpiece, therefore can carry out parallel processing with a plurality of measuring appliances.While adopting like this plurality of measuring appliances to carry out parallel processing, the measuring error between measuring appliance just can become problem.For example, when measuring 10 identical samples with 3 measuring appliances, the measured value of each measuring appliance is inequality sometimes.This measuring error between measuring appliance brings larger discrete impact will to the characteristic of goods.In addition, measuring error, because meeting is because the internal environment of measuring appliance and the subtle change of external environment condition increase, therefore has the tendency increasing in time.Therefore, the best measuring error between stage control survey device measuring appliance is proofreaied and correct etc. to processing in early days.

About the supervision method of the measuring error between measuring appliance, have and a kind ofly by histogram, come sense of vision to express the method for the measured value of each measuring appliance.For example,, if the histogram of 1 measuring appliance obvious histogram that departs from other measuring appliances as Figure 13 (a) only can infer that this measuring appliance has occurred that certain is abnormal.Yet in actual production line, do not limit (a) such only 1 the visibly different situation of histogram that only occurs Figure 13, sometimes also can be as Figure 13 (b), the histogram that occurs each measuring appliance is slightly different situation each other.For the latter's situation,, with the histogram of observing Figure 13 (b), be to be only difficult to judge that the measuring error between measuring appliance has or not.

In addition, when the histogram that has 1 measuring appliance is as Figure 13 (a) during obvious departing from, though can confirm that the measuring error between measuring appliance has or not, this has just occurred to depart from more greatly Shi Caineng at histogram and has confirmed.Therefore, the measuring error between measuring appliance has caused occurring larger economic loss.Here the reason of said economic loss is, the non-defective unit/defective products misjudgment due to the measuring error between measuring appliance increases.

In addition,, about the supervision method of the measuring error between measuring appliance, also has a kind of method that the mean value of the measured value of each measuring appliance is confirmed.For example, if as Figure 14 (a) only the histogram of 1 measuring appliance (in figure shown in CH6) depart from significantly the histogram of other measuring appliances, just the mean value of the mean value of the measured value of measuring appliance CH6 and the measured value of all measuring appliances differs " a " so.Therefore can infer that this measuring appliance has occurred that certain is abnormal.Yet in actual production line, do not limit (a) such only 1 the visibly different situation of histogram that only occurs Figure 14, sometimes also can be as Figure 14 (b), the histogram that occurs each measuring appliance is slightly different situation each other.For the latter's situation, only poor with between the mean value of measured value of each measuring appliance and the mean value of the measured value of all measuring appliances, is to be difficult to judge that the measuring error between measuring appliance has or not.

To this, ISO/TS16949 standard is to measuring systematic analysis (MSA; Measurement Systems Analysis) in addition regulation.MSA is a kind of method of managing measuring accuracy.As the typical example of MSA gimmick, surely there is following evaluation assessment R & R(GRR).

[mathematical expression 1]

$\mathrm{GRR}=\frac{\sqrt{\mathrm{<figure-callout\; id="2"\; label="E\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">E</figure-callout>}{V}^{}{}^2+A{V}^2}}{\sqrt{\mathrm{<figure-callout\; id="2"\; label="T\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">T</figure-callout>}{V}^{}{}^2}}=\frac{\sqrt{\mathrm{<figure-callout\; id="2"\; label="E\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">E</figure-callout>}{V}^{}{}^2+A{V}^2}}{\sqrt{\mathrm{<figure-callout\; id="2"\; label="P\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">P</figure-callout>}{V}^{}{}^2+\mathrm{<figure-callout\; id="2"\; label="E\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">E</figure-callout>}{V}^{}{}^2+\mathrm{<figure-callout\; id="2"\; label="A\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">A</figure-callout>}{V}^{}{}^2}}$

Wherein, TV ²the dispersion of expressing all discrete case of measurement data, PV ²the dispersion of expressing discrete case when a plurality of goods (workpiece) have been carried out measuring under identical conditions, EV ²be the dispersion of expressing repeatedly discrete case (same measuring appliance repeatedly measure same goods and the dispersion of measured value), AV ²be the dispersion of expressing the measured value discrete case between measuring appliance (every single measuring appliance same goods are taken multiple measurements and the dispersion of mean value).

And manage GRR according to following benchmark.

GRR is below 10%---qualified

GRR is 10～30%---qualified under certain condition

GRR surpasses 30%---and defective.

In addition, in patent documentation 1, disclosed in a kind of processing of the mensuration at substrate size and by asking for the random value (sum total random value) of all measured values in mensuration system, carried out the scheme of evaluating and measuring system.In patent documentation 1, to measuring the measurement result of system and the measurement result of standard test system, do respectively linear regression, then according to difference between the two, remove the random value of standard test system, thus evaluating and measuring system more accurately.

[prior art document]

Patent documentation

Patent documentation 1: Japan's license bulletin " speciallyying permit No. 4272624 instructions ", on June 3rd, 2009 is announced.

Summary of the invention

[problem to be solved by this invention]

In having used the management of the fixed GRR of above-mentioned MSA, and in the technology disclosing at patent documentation 1, need get ready for the sample of analyzing the common production stopping on production line and repeatedly carry out sample measurement.In addition, in the technology that patent documentation 1 discloses, also need to measure by standard test system, therefore bothersome.

The present invention researches and develops in order to solve the above problems, and object is to provide a kind of just can easily recognize without the common production stopping on production line management devices, management method, program and the recording medium that the measuring error between measuring appliance changes.

[in order to the technical scheme of dealing with problems]

For addressing the above problem, a plurality of measuring appliances that management devices management of the present invention is measured the characteristic of workpiece on production line, it is characterized in that: described a plurality of measuring appliances are measured the characteristic of the workpiece sending separately, and this management devices possesses: characteristic value obtaining section, obtains a plurality of characteristic values that the characteristic of the described workpiece sending are measured by described a plurality of measuring appliances; Operational part, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate the parameter of expressing the measuring error between described a plurality of measuring appliances.

In addition, a plurality of measuring appliances that management method of the present invention is measured the characteristic of workpiece on production line in order to management, it is characterized in that: described a plurality of measuring appliances are measured the characteristic of the workpiece sending separately, and this management method comprises: characteristic value is obtained step, obtain a plurality of characteristic values that the characteristic of the described workpiece sending are measured by described a plurality of measuring appliances; Calculation step, according to a plurality of characteristic values of obtaining in described characteristic value is obtained step, calculates the parameter of expressing the measuring error between described a plurality of measuring appliances.

In addition, a plurality of measuring appliances that management devices management of the present invention is measured the characteristic of workpiece on production line, is characterized in that: possess: characteristic value obtaining section, obtains a plurality of characteristic values that obtain by described a plurality of measuring appliances; Operational part, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate the parameter of expressing the measuring error between described a plurality of measuring appliances; The 1st value that described operational part is calculated the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be a plurality of characteristic values of obtaining of described characteristic value obtaining section, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation of a plurality of characteristic values of obtaining of described characteristic value obtaining section.

In addition, a plurality of measuring appliances that management method of the present invention is measured the characteristic of workpiece on production line in order to management, is characterized in that: comprise: characteristic value is obtained step, obtain a plurality of characteristic values that obtain by described a plurality of measuring appliances; Calculation step, according to a plurality of characteristic values of obtaining in described characteristic value is obtained step, calculate the parameter of expressing the measuring error between described a plurality of measuring appliances, in described calculation step, the 1st value of calculating the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be described characteristic value obtain a plurality of characteristic values of obtaining in step, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation of a plurality of characteristic values of obtaining of described characteristic value obtaining section.

[invention effect]

As mentioned above, the effect of management devices of the present invention is, without the common production stopping on production line, just can easily recognize the variation of the measuring error between measuring appliance.

Accompanying drawing explanation

Fig. 1 is the schematic configuration schematic diagram of the measuring appliance error management system of an embodiment of the present invention.

Fig. 2 is the number of units schematic diagram of the equipment that has of the step of each on production line shown in Fig. 1.

Fig. 3 is the illustration of the institute of database shown in Fig. 1 canned data.

Fig. 4 is the illustration of the operation information that shown in Fig. 1, operation recording device of transit is stored.

Fig. 5 is the processing flow chart of the measurement data obtaining section that shown in Fig. 1, management devices possesses.

Fig. 6 is the schematic diagram of measured complete batch in timing statistics section.

Fig. 7 is the illustration of the specification value storage part institute canned data that shown in Fig. 1, management devices possesses.

Fig. 8 is the processing flow chart of the operational part that shown in Fig. 1, management devices possesses.

Fig. 9 is the distribution plan of standard deviation TV and standard deviation PV'.

Figure 10 is a schematic diagram that storage is routine of the storage part that shown in Fig. 1, management devices possesses.

Figure 11 is the schematic diagram of the display case of the chart Graphics Processing portion that shown in Fig. 1, management devices possesses.

Figure 12 is the illustration that is shown with the chart of content of operation information.

Figure 13 is the diagram of histogrammic 2 kinds of examples of the measured value of each measuring appliance.

Figure 14 is the histogram of measured value of each measuring appliance and the diagram of 2 kinds of examples of mean value.

Embodiment

Below, describe by reference to the accompanying drawings an embodiment of the invention in detail.Fig. 1 is the schematic configuration schematic diagram of the measuring appliance error management system of an embodiment of the present invention.

The measuring appliance error management system 1 of present embodiment possesses production line 10, measurement data gathering-device 20, database 30, management devices 40, operation recording device of transit 50.

In present embodiment, on production line 10, comprise: the 1st～3rd production stage, in order to article of manufacture; Check step, the various characteristics of the workpiece as intermediate or end article is checked.Fig. 2 is the number of units schematic diagram of the equipment that has of each step.At this, in order to make the processing speed of each step roughly keep constant, the 1st production stage～3rd production stage has respectively 8,3,1 production equipment, checks that step has 3 measuring appliances 11.In each step, by a plurality of equipment or measuring appliance, implement parallel processing.That is, a plurality of workpiece that send from the 2nd production stage are transmitted to 3 measuring appliances 11 through distribution, and 11 pairs of characteristics that send workpiece of each measuring appliance are measured.It is 3 that the number of units of the measuring appliance 11 that inspection step has is not limited to, as long as there are many.

20 pairs of measurement data gathering-devices check a plurality of measuring appliances 11 that steps have separately measured measurement data collect, and deposit the measurement data of collection in database 30.Fig. 3 is the illustration of database 30 data of storing.For each workpiece, measurement data gathering-device 20 as shown in Figure 3, is arranged measurement data according to measuring date sequential constantly, and is deposited these measurement data in database 30.In measurement data, comprise the following items corresponding to each other: work ID, in order to identify this workpiece; Measuring appliance identifying information, has implemented to this workpiece the measuring appliance 11 checking in order to identification; The machine numbering of the end article of this workpiece; Mission Number under this workpiece; Characteristic value, the namely measurement result corresponding with this n specific character of characteristic T1～Tn; Date while measuring constantly.

In addition, measurement data gathering-device 20, according to the information that never illustrated input part input is come, deposits machine numbering and Mission Number in database 30.That is, staff, when switching batch, inputs machine numbering and Mission Number by input part.Measurement data gathering-device 20, when having received machine numbering and Mission Number, is all given unique work ID to each measurement data collected from now, and is generated measurement data, and deposit measurement data in database 30.Work ID in measurement data and the machine of receiving numbering and Mission Number are corresponding mutually.

Operation recording device of transit 50 is devices of store operation information, comprises the following two classes information of correspondence mutually in operation information: content of operation information, the content of the operation that its expression staff carries out; Timing information, it expresses date that this operation is performed regularly (constantly; Timing).Operation recording device of transit 50 carrys out store operation information according to the instruction of for example staff's input.Fig. 4 is the illustration of the operation information stored of operation recording device of transit 50.

Management devices 40 is the devices to checking that the measuring error of 11 of a plurality of measuring appliances that step has manages.As shown in Figure 1, management devices 40 possess input part 41, display part 42, obtain configuration part 43, measurement data obtaining section (characteristic value obtaining section) 44, specification value storage part 45, operational part 46, deposit data portion 47, storage part 48, chart Graphics Processing portion 49.

Input part 41 is accepted the various inputs from production line staff, and it is formed by inputting with some position device, other entering apparatus such as button, keyboard, mouses.

Display part 42 is LCD(liquid crystal display), PDP(plasma display), organic EL(electro luminescence: the electroluminescence) display unit such as display, it demonstrates the various information such as word, image according to the demonstration data of receiving.

Obtain configuration part 43, setting with reference to condition when measurement data obtaining section 44 is obtained to measurement data group.Obtain configuration part 43 according to the information that is input to input part 41, the statistics moment, timing statistics section are set.In addition, obtain configuration part 43 when the update instruction of having received from measurement data obtaining section 44, by from statistics constantly through the moment after a timing statistics section, be set as new statistics constantly.

Measurement data obtaining section 44 obtains measurement data group from database 30.Concrete processing about measurement data obtaining section 44, describes below in conjunction with Fig. 5.Fig. 5 is the processing flow chart of measurement data obtaining section 44.

First, measurement data obtaining section 44 judges whether current time has arrived and obtains statistics that configuration part 43 sets (S1) constantly.

If current time has arrived statistics constantly (being "Yes" at S1), measurement data obtaining section 44 goes out from specific database 30: with the constantly corresponding measurement data of the measurement date of the front a statistical time constantly in section from statistics.Then, measurement data obtaining section 44 is according to measuring date sequential constantly, arrange by the specific measurement data going out, and judge whether that measurement data and a measurement data thereafter exist different on Mission Number.When measuring that the date, sequential was constantly arranged, last workpiece that exists different measurement data on Mission Number to be exactly batch from a rear measurement data.Therefore, there are different on Mission Number by judging whether measurement data and a measurement data thereafter in measurement data obtaining section 44, just can judge from statistics, whether the front a statistical time constantly has measured complete batch in section.Thus, measurement data obtaining section 44 is specific goes out N the measurement data Mission Number (S2) mutually different from the Mission Number of a rear measurement data.

For example, as shown in Figure 6, in front a statistical time constantly in section from statistics, measured complete if Mission Number is " M2 ", " M3 " 2 batches, by N specific be 2, specificly go out " M2 ", " M3 " these 2 Mission Numbers.

Then, measurement data obtaining section is respectively for each person of specific N the Mission Number going out in step S2, from database, read the measurement data group who is formed by all measurement data with this Mission Number, and the measurement data group who reads is exported to operational part 46(S3～S6).

It is "No" at S4 that the measurement data group corresponding with the specific Mission Number going out in step S2 is all exported to operational part 46() after, measurement data obtaining section 44 is just to obtaining configuration part 43 output update instruction.Thus, from statistics constantly through the moment after a timing statistics section be set to new statistics (S7) constantly.

As described above, measurement data obtaining section 44, in front a statistical time constantly in section measured complete batch from statistics, is exported to operational part 46 by the measurement data group of each batch.

Specification value storage part 45 is in order to store specification value, and specification value is distinguished non-defective unit and defective products in order to the various characteristics of measuring for measuring appliance 11.The specification value storage part 45 storage higher limit of specification value and at least one party of lower limit.The specification value storage part 45 storage specification value set to each characteristic.Fig. 7 is the illustration of 45 canned datas of specification value storage part.At this, specification value storage part is to store specification value according to the information that inputs to input part.Therefore staff can, by input part 41 input specification values, set the specification value of various characteristics.

The measurement data group that operational part 46 is used each batch obtaining from measurement data obtaining section 44, calculates respectively the parameter that can be easy to recognize 11 measuring error of measuring appliance for each characteristic.; by the same goods of repeated measurement and the standard deviation of measured value be made as EV; and by the standard deviation of the measured value of a plurality of goods measured under identical conditions (; while the standard deviation of expression goods discrete case) being made as PV; operational part 64 is calculated the PV ' that meets following formula according to measurement data group, and uses the PV calculating ^'calculate the parameter of expressing discrete case between measuring appliance.

PV′ ²＝PV ²+EV ²

Concrete processing about operational part 46, describes below in conjunction with Fig. 8.Fig. 8 is the processing flow chart of operational part 46.

First, operational part 46, according to measurement data group, is calculated statistic (S11) with regard to each characteristic.Particularly, operational part 46 is with regard to each characteristic, all calculates mean value and the standard deviation of the characteristic value that all measuring appliances 11 record.At this, the characteristic value of measured j the workpiece of i measuring appliance is designated as to xij.In addition, the number that checks the measuring appliance 11 that step has is designated as to a, and the number of the measured workpiece of i measuring appliance 11 is designated as to ni.So, the mean value ave(x of the characteristic value that all measuring appliances 11 are measured) and standard deviation TV can express by following mathematical expression.

[mathematical expression 2]

$\mathrm{ave}\left(x\right)\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}{x}_{\mathrm{ij}}}{N_{\mathrm{total}}}$

[mathematical expression 3]

$\mathrm{TV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x\right)-x_{\mathrm{ij}}{)}^2}{N_{\mathrm{total}}}}$

[mathematical expression 4]

$N_{\mathrm{total}}=\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}{n}_i$

Then, operational part 46, with regard to i the various characteristics that measuring appliance 11 is measured, is calculated the mean value ave(xi of characteristic value according to following mathematical expression) and standard deviation sigma (xi) is (S12).

[mathematical expression 5]

$\mathrm{ave}\left(x_i\right)=\frac{\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}\left(x_{\mathrm{ij}}\right)}{n_i}$

[mathematical expression 6]

$\mathrm{\&sigma;}\left(x_i\right)=\sqrt{\frac{\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x_i\right)-x_{\mathrm{ij}}{)}^2}{n_i}}$

Then, operational part 46 is calculated discrete Magnification.If the standard deviation of the characteristic value between measuring appliance is AV, according to discrete additivity, formula TV ²=PV ²+ EV ²+ AV ²set up.Due to PV ' ²=PV ²+ EV ²so, have TV ²=PV ' ²+ EV ².

Operational part 46 is calculated PV ', AV according to following mathematical expression.

[mathematical expression 7]

$\mathrm{TV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x\right)-x_{\mathrm{ij}}{)}^2}{N_{\mathrm{total}}}}$

[mathematical expression 8]

$\mathrm{AV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}{n}_i(\mathrm{ave}\left(x\right)-\mathrm{ave}\left(x_i\right){)}^2}{N_{\mathrm{total}}}}$

PV ' also can pass through formula (TV ²-AV ²) ^1/2calculate.In addition, also can calculate AV, PV ', TV by other operational methods such as the method for average/area method, discrete analyses.

Then, operational part 46 is calculated discrete Magnification (S13) according to the following formula.

Discrete Magnification=TV/PV '

Discrete Magnification is the value that TV obtains divided by PV '.If the discrete value 0(of the characteristic value between measuring appliance is AV=0), discrete Magnification is 1.On the other hand, if the discrete value of the characteristic value between measuring appliance increases to some extent, discrete Magnification becomes the value that is greater than 1.That is, the expressed concept of discrete Magnification is:: the standard deviation while there is measuring error between measuring appliance is several times while there is not the perfect condition of measuring error between measuring appliance.

Then, operational part 46 is read the specification value (S14) of each characteristic of storage in specification value storage part 45.Then, the specification value of operational part 46 based on reading, calculates and improves fraction defective (S15).Improving fraction defective is the value of expressing the difference between current fraction defective and the fraction defective of perfect condition.So-called perfect condition refers to that the measuring error between measuring appliance is 0.Current fraction defective can be calculated according to the distribution situation of standard deviation TV.On the other hand, if the measuring error between measuring appliance is 0, AV=0, so the measuring error between measuring appliance is that the fraction defective of 0 o'clock can be calculated according to the distribution situation of standard deviation PV '.Fig. 9 is the distribution plan of the standard deviation PV ' while there is not measuring error between standard deviation TV and measuring appliance.Known as shown in Figure 9, in the distribution of the less PV ' of standard deviation, the fraction defective that departs from specification value be reduce gradually tendency.This means when improving fraction defective when larger, by carrying out the processing such as correction of measuring appliance, reduce the measuring error between measuring appliance, just very likely reduce fraction defective.

Operational part 46 for example adopts Microsoft Excel(registered trademark) NORMDIST function calculate and improve fraction defective.NROMDIST(x, μ, σ, ture) expressed function is with regard to the normal distribution of average value mu, standard deviation sigma, and calculate stochastic variable and become the probability below x.

If establishing upper limit specification value is d1, setting limit gauge scale value is d2, and the mean value of establishing the measured characteristic value of all measuring appliances 11 is ave(x), current fraction defective f(d1, d2, ave (x), TV) can calculate by following formula.

f（d1,d2,ave(x),TV）＝NORMDIST（d2,ave(x),TV,true）＋〔1－NORMDIST（d1,ave(x),TV,true）〕

In addition, this perfect condition that is 0 due to the measuring error between measuring appliance refers to AV=0, so fraction defective f(d1 during perfect condition, d2, ave (x), PV ') can calculate by following formula.

f（d1,d2,ave(x),PV′）＝NORMDIST（d2,ave(x),PV′,true）＋〔1－NORMDIST（d1,ave(x),PV′,true）〕

So operational part 46 is calculated according to the following formula and is improved fraction defective.

Improve fraction defective=f(d1, d2, ave (x), TV)-f(d1, d2, ave (x), PV ')

In addition,, if the distribution of characteristic value submits to normal distribution distribution in addition, operational part 46 can, according to the cumulative distribution function based on this distribution, be calculated and improve fraction defective so.

Like this, operational part 46 is for each characteristic, all calculate the characteristic value that the mean value of the characteristic value that all measuring appliances 11 record and standard deviation, every single measuring appliance 11 record mean value and standard deviation, discrete Magnification, improve fraction defective.

Mean value and standard deviation, the discrete Magnification of the characteristic value that the mean value of the characteristic value that all measuring appliances 11 that operational part 46 is calculated for each characteristic record and standard deviation, every single measuring appliance 11 record, improve fraction defective, by deposit data portion 47, deposited in storage part.Deposit data portion 47 is when depositing in, and the value that these are calculated and following data are set up corresponding relation and deposited storage part 48 in, and these data are: the data group ID that becomes the measurement data group of calculating object in order to identification; The corresponding machine numbering of measurement data group; Mission Number; Measure the date constantly.At this, deposit data portion 47 can be by the moment on measurement date of any one measurement data in measurement data group, as this measurement data group's the moment on measurement date.For example, deposit data portion can the measurement data that (the 1st) records by first the measurement date constantly, measurement date of last (n) measurement data of recording constantly and n/2 (if n odd number, get (n+1)/2) measurement date of the measurement data that records either party constantly, be set as this measurement data group's the measurement date constantly.

Figure 10 is the illustration figure of 48 canned datas of storage part.As shown in figure 10, in each batch, mean value and standard deviation, the discrete Magnification of the characteristic value that the mean value of the characteristic value that all measuring appliances 11 record and standard deviation (being denoted as " all " in Figure 10), every single measuring appliance 11 record, improve fraction defective and measure the date and constantly all correspond to each other.In addition, in the drawings, " Ave " represents mean value, and " Sd " represents standard deviation.

Chart Graphics Processing portion 49 processes, thereby in display part 42, demonstrates the chart of the Temporal changes of expressing various parameters.When chart idsplay order is transfused to into input part 41, chart Graphics Processing portion 49 is from reading each each self-corresponding discrete Magnification of characteristic in the middle of storage part 48 canned datas, improve fraction defective, measuring the date constantly, and makes the chart of expressing discrete Magnification and improving the Temporal changes of fraction defective.Particularly, make that transverse axis representative is measured the date constantly and the longitudinal axis represents the chart of discrete Magnification and transverse axis representative is measured the date constantly and longitudinal axis representative improves the chart of fraction defective, and these charts are presented in display part 42.

In addition, chart Graphics Processing portion 49 also can be in the middle of storage part 48 canned datas, read each characteristic the characteristic value that records of the mean value of each self-corresponding characteristic value being recorded by all measuring appliances 11 and standard deviation, every single measuring appliance 11 mean value and standard deviation and measure the date constantly, and make the chart of the Temporal changes of expressing each numerical value, and these charts are presented in display part 42.

Figure 11 is the schematic diagram of a display case of chart Graphics Processing portion 49.In Figure 11, (a) represent the Temporal changes of the discrete Magnification of each characteristic, (b) represent the Temporal changes that improves fraction defective of each characteristic.In addition, (c) and (f) with regard to characteristic T1, represent the mean value of each measuring appliance and the Temporal changes of standard deviation respectively, (d) and (g) with regard to characteristic T2, represent the mean value of each measuring appliance and the Temporal changes of standard deviation respectively, (e) and (h) with regard to characteristic T3, represent the mean value of each measuring appliance and the Temporal changes of standard deviation respectively.

Chart Graphics Processing portion 49 is also from operation recording device of transit 50 read operation information.So chart Graphics Processing portion 49 can, in each chart shown in Figure 11, at the display position place constantly of date shown in operation timing information, demonstrate the content of operation information corresponding with this operation timing information.Figure 12 is the illustration that is shown with the chart of content of operation information.As shown in figure 12, content of operation information shows along time shaft, therefore by confirming discrete Magnification and improve fraction defective to be content of operation information shown on the position of larger variation, just can easily recognize discrete Magnification and improve the reason of changes of fraction defective.

As mentioned above, a plurality of measuring appliances 11 in order to measuring workpieces characteristic on 40 pairs of production lines of management devices of present embodiment manage.On production line, a plurality of workpiece are passed to a plurality of measuring appliances 11 through distribution, and a plurality of measuring appliances 11 are measured the characteristic of the workpiece sending separately.At this, the measurement data obtaining section 44 of management devices 40 obtains measurement data group, and this measurement data group comprises by the characteristic of 11 pairs of workpiece that send of a plurality of measuring appliances and measures and a plurality of characteristic values of obtaining.And the measurement data group of operational part 46 based on obtaining, calculates the parameter of the measuring error of expressing 11 of a plurality of measuring appliances.

Like this, in the operation of production line, operational part 46, according to the characteristic of the workpiece sending being measured by a plurality of measuring appliances a plurality of characteristic values that obtain, is calculated the parameter of the measuring error of expressing 11 of a plurality of measuring appliances.Its result, just can easily recognize the measuring error of 11 of measuring appliances without the common production stopping on production line.

The 1st value of expressing mutual relationship between the 1st standard deviation (PV ') and the 2nd standard deviation (TV) is calculated by budget portion 46, and using the 1st value as described parameter.The 1st standard deviation is measurement data group, the standard deviation when the measuring error of 11 of a plurality of measuring appliances is taken as to 0.The 2nd standard deviation is this measurement data group's standard deviation.

Conventionally, at the scene of manufacturing, by standard deviation sigma, carry out step management.For example, step Capability index Cpk can be by formula Cpk=(upper limit specification value-undergage scale value)/6 σ express.It is 3.4/1000000ths that the well-known part of 6 σ (six-sigma) is to drop on this extraneous probability of (mean value-6 σ)～(mean value+6 σ), its slogan is " even if implement the operation of 1,000,000 times, the incidence of defective products is also suppressed in 3,4 times ".So, for staff, standard deviation is a kind of parameter being very familiar to.Therefore staff has very high consciousness to the decline degree of standard deviation.

And the GRR shown in [mathematical expression 1], shown in it: in order to express the standard deviation of AV and the corresponding dispersion of EV, with respect in order to express the ratio of the standard deviation of all discrete case.Yet this,, for being familiar with the staff of standard deviation, is difficult to recognize intuitively the measuring error between measuring appliance.

For example, if GRR is 30%, staff learns that the standard deviation that the measuring error between measuring appliance affects is 30%.Now, staff may be misinterpreted as long as the measuring error between measuring appliance is adjusted to 0, just can to the full extent current standard deviation be reduced to its 70%(100-30=70%).This is due to staff, to take for the cause of TV=PV+AV+EV.That is, takeing for PV '/TV=1-AV/TV calculates.In fact, the additivity of standard deviation and be false, but is only set up discrete phase additivity here, namely TV only ²=PV ²+ AV ²+ EV ²set up.Therefore when GRR is 30%, in fact according to following mathematical expression 9, if the measuring error between measuring appliance is adjusted to 0, current standard deviation maximum can be reduced to its 0.954 times, yet staff is but difficult to recognize this situation.

[mathematical expression 9]

$\frac{P{V}^{\&prime;}}{\mathrm{TV}}=\sqrt{\frac{T{V}^2-A{V}^2}{\mathrm{<figure-callout\; id="2"\; label="T\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">T</figure-callout>}{V}^{}{}^2}}=\sqrt{1-\frac{A{V}^2}{\mathrm{<figure-callout\; id="2"\; label="T\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">T</figure-callout>}{V}^{}{}^2}}$

As mentioned above, for staff, GRR be a kind of be difficult to recognize measuring error between measuring appliance have or not and current standard deviation between the parameter of relation.Namely there is following problem: staff is difficult to recognize that to be worth to make standard deviation to reduce according to which kind of of GRR how many.

And in present embodiment, staff can recognize the 1st value of expressing mutual relationship between the 1st standard deviation (PV ') and the 2nd standard deviation (TV).The 1st standard deviation is here measurement data group, the standard deviation when the measuring error of 11 of a plurality of measuring appliances is got to 0.The 2nd standard deviation is measurement data group's standard deviation.At this, the 2nd standard deviation (TV) is that its state is surrounded by the measuring error of 11 of measuring appliances in being from the standard deviation of the characteristic value of a plurality of measuring appliances 11 acquisitions.Therefore by confirming the 1st value, just can easily learn than the measuring error between a plurality of measuring appliances and be taken as the state of 0 o'clock, there are how many variations in standard deviation.

Particularly, by the same goods of repeated measurement and the standard deviation of measured value be made as EV, and when the standard deviation of the measured value of a plurality of goods is made as to PV, operational part 64 is calculated and is met formula PV ' according to measurement data group ²=PV ²+ EV ²pV ', and with the PV ' calculating, calculate the parameter of expressing discrete case between measuring appliance.Then, chart Graphics Processing portion 49 makes display part 42 display parameter.

Like this, in the operation of production line, operational part 46 is according to the measurement data group who has comprised a plurality of characteristic values that the characteristic of the workpiece sending are measured by a plurality of measuring appliances, with meeting PV ' ²=PV ²+ EV ²pV ' carry out calculating parameter.That is, make measuring appliance with regard to same sample carry out standard deviation EV that repeated measurement obtains itself, without operational part 46, calculate.Therefore, can use the characteristic value obtaining in production line operation, demonstrate the parameter of expressing discrete case between a plurality of measuring appliances 11.

Its result, just can easily recognize the measuring error of 11 of measuring appliances without the common production stopping on production line.

Operational part 46 is discrete Magnification by TV/PV ', as parameter, calculates.As above, the expressed concept of this discrete Magnification is:: the standard deviation while there is measuring error between measuring appliance is several times while there is not the perfect condition of measuring error between measuring appliance.Therefore, staff, by observing discrete Magnification, just can easily recognize the measuring error existing between measuring appliance causes standard deviation to increase how many degree.

Although in the above description, that operational part 46 is calculated is discrete Magnification TV/PV ', also can not calculate discrete Magnification TV/PV ', but calculate its inverse, is PV '/TV.The expressed concept of PV '/TV is:: the standard deviation while there is not the perfect condition of measuring error between measuring appliance is several times while there is the current state of measuring error between measuring appliance.If adopt PV '/TV, staff is by confirming PV '/TV, just can easily recognize by the measuring error between measuring appliance being adjusted to 0 just to make standard deviation reduce how many degree.

In addition, operational part 46 is calculated the 2nd value according to the good no characteristic value specification limit of confession judgement workpiece, and using it as parameter.Mutual relationship between the 2nd two kinds of value representations fraction defective, these two kinds of fraction defectives are: the measuring error that a plurality of measuring appliances are 11 is got fraction defective, that draw according to measurement data group at 0 o'clock; And, the fraction defective drawing according to measurement data group.

Particularly, operational part 46 is fallen standard deviation the probability outside the normal distribution specification limit of TV, and the fraction defective while being taken as TV as standard deviation is calculated; Operational part 46 is also fallen standard deviation the probability outside the normal distribution specification limit of PV ', and the fraction defective that is taken as PV ' time as standard deviation is calculated.Then, calculate and improve fraction defective and be used as the 2nd value.Improving fraction defective is: fraction defective when standard deviation is taken as TV and standard deviation are taken as the difference between the fraction defective in PV ' time.That is, improving fraction defective is: the difference between fraction defective during this perfect condition that between current fraction defective and measuring appliance, measuring error is 0.

Thus, staff, by confirming the 2nd value (improving fraction defective), just can easily recognize by the measuring error between measuring appliance being adjusted to 0 and just can make fraction defective how much decline.

In addition, operational part 46 is calculated mean value and the standard deviation of the measured characteristic value of all measuring appliances, also calculates mean value and the standard deviation of the measured characteristic value of every single measuring appliance.And chart Graphics Processing portion 49 demonstrates the chart (referring to Figure 11) of the Temporal changes of expressing each value.

Like this, can both show discrete Magnification and improve fraction defective, showing again mean value and the standard deviation of the characteristic value that every single measuring appliance is measured, thereby can carry out various analyses.

For example, by confirming Figure 11 (a), the discrete Magnification of known characteristic T2 carries out transition with high value state.In addition, chart (d), (g) have represented that every single measuring appliance is with regard to the mean value of the measured measured value of characteristic T2 and the migration situation of standard deviation, by confirming chart (d), (g), known measuring appliance a～c separately corresponding standard deviation is roughly the same, but the corresponding mean value of measuring appliance a compared with other, both are large.According to this phenomenon, can infer that deviation has probably appearred in the corrector bias of measuring appliance a when measurement characteristics T2.

In addition, by confirming Figure 11 (a), the discrete Magnification of known characteristic T3 carries out transition with high value state.Chart (e), (h) have represented that every single measuring appliance is with regard to the mean value of the measured measured value of characteristic T3 and the migration situation of standard deviation, by confirming chart (e), (h), the corresponding mean value of known measuring appliance c is maximum, and the corresponding standard deviation of measuring appliance c is minimum.Generally speaking, if mean value is larger, standard deviation also can be larger.In addition, also the corresponding standard deviation of known only measuring appliance c is different from the timeliness transition of the standard deviation of other 2 measuring appliances.According to this phenomenon, can infer measuring appliance c and occur that certain is abnormal.Especially for the characteristic of measuring by contact workpiece as load groove (load cell), if occurred and the improper abnormal conditions that contact of workpiece, the Temporal changes of standard deviation will be different from the tendency of the other.Therefore can infer and this type of abnormal generation.

In addition,, by confirming Figure 11 (b), the fraction defective (2011/3/28 to 2011/3/29,2011/4/4 to 20114/10) during only specific that improves of known characteristic T3 increases to some extent.Chart (e) has represented that every single measuring appliance is with regard to the mean value of the measured measured value of characteristic T3 and the migration situation of standard deviation, by confirming chart (e), the measuring error of known these 3 measuring appliances has relatively big difference each other, and mean value all during this is specific, decline has occurred.According to this phenomenon, can recognize that the distribution of characteristic T3 is just close to lower limit.Especially can recognize, the measured characteristic T3 of measuring appliance b that mean value is lower causes fraction defective to increase to some extent and measuring appliance b need to proofread and correct.

Chart Graphics Processing portion 49 is from operation information pen recorder, specific operation timing information during going out to represent in chart, and in chart, corresponding display position in the timing shown in the above-mentioned specific operation timing information going out, demonstrates the content of operation information corresponding with this operation timing information.

Thus, just can easily infer discrete Magnification according to content of operation information and improve the reason of changes of fraction defective.

In the above description, measurement data obtaining section 41 from statistics the front a statistical time constantly measured each complete batch in section measurement data group, export to operational part 46.But measurement data obtaining section 44 obtains measurement data group's method and is not limited thereto.For example, measurement data obtaining section 41 also can be take each batch as unit, to dividing into groups in the measurement data that the front a statistical time constantly records in section from statistics, and with each Zu Wei unit, the measurement data group who belongs to this group is exported to operational part 46.Or measurement data obtaining section 44 also can, using all measurement data the front a statistical time constantly records in section from statistics as 1 measurement data group, be exported to operational part 46.

Although not only calculate discrete Magnification but also calculate the scheme of improving fraction defective above-mentioned middle explanation, also can adopt the scheme of only calculating certain side wherein and showing.

It is the respective embodiments described above that the present invention is not limited to, and can in the scope shown in claim, carry out various changes, suitably combines the technical scheme of describing in different embodiments and the embodiment obtaining is also included in the technical scope of the present invention.

In addition, ROM (read-only memory)) and RAM(random access memory central processing unit) etc. each portion of management devices 40 in the respective embodiments described above can realize by following scheme: by CPU(central processing unit: arithmetic element is carried out ROM(read only memory:: random access memory) etc. the program of storing in storage unit, and the communication units such as the output units such as the output units such as keyboard, display or interface circuit are controlled.Therefore, possess and have the computing machine of these unit only to need to read and carry out this program to recording the recording medium of said procedure, just can realize various functions and the various processing of the line management device of present embodiment.In addition, by said procedure being recorded in packaged type recording medium, just can realize above-mentioned various functions and various processing by computing machine arbitrarily.

About this recording medium, can be the not shown storer of processing for microcomputer, such as the program medium that has ROM etc.Also can be the not shown program reading device arranging as external memory by inserting, the program medium that just can be read.

Which kind of scheme no matter, stored program is preferably visited and is carried out by microprocessor.In addition, preferably in the following ways: the program of reading is performed after being downloaded to the program storage area of microcomputer again.Here, for the program of carrying out this download, pre-deposit in host apparatus.

In addition, said procedure media be can be separated with main frame recording medium, such as having: tape and cartridge tape etc. band class; The dish class that comprises the CDs such as the magnetic plates such as pliability dish and hard disk and CD-ROM, MO, MD, DVD, CD-R; The card such as IC-card (comprising storage card) class; EPROM (Erasable Programmable Read Only Memory)), EEPROM(Electrically Erasable Programmable Read Only Memory or comprise mask model ROM, EPROM(Erasable Programmable Read Only Memory:: EEPROM (Electrically Erasable Programmable Read Only Memo)), the nature static of the semiconductor memory such as flash rom hold recording medium of program etc.

In addition, if system can accessing Internet etc. communication network, preferably adopt just as hold to dynamic the recording medium of program from downloaded program.

In addition, if desired as described above from downloaded program, for the program of carrying out this download, preferably pre-deposited host apparatus, or the program of installing from other recording mediums.

As mentioned above, a plurality of measuring appliances that management devices management of the present invention is measured the characteristic of workpiece on production line, it is characterized in that: described a plurality of measuring appliances are measured the characteristic of the workpiece sending separately, and this management devices possesses: characteristic value obtaining section, obtains a plurality of characteristic values that the characteristic of the described workpiece sending are measured by described a plurality of measuring appliances; Operational part, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate the parameter of expressing the measuring error between described a plurality of measuring appliances.

In such scheme, in production line operation, according to a plurality of characteristic values that the characteristic of the workpiece sending are measured by a plurality of measuring appliances, calculate the parameter of expressing the measuring error between a plurality of measuring appliances.Its result, just can easily recognize the measuring error between measuring appliance without the common production stopping on production line.

In addition, in management devices of the present invention, also can adopt following scheme: the 1st value that described operational part is calculated the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be a plurality of characteristic values of obtaining of described characteristic value obtaining section, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation of a plurality of characteristic values of obtaining of described characteristic value obtaining section.

In addition, a plurality of measuring appliances that management devices management of the present invention is measured the characteristic of workpiece on production line, is characterized in that: possess: characteristic value obtaining section, obtains a plurality of characteristic values that obtain by described a plurality of measuring appliances; Operational part, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate the parameter of expressing the measuring error between described a plurality of measuring appliances; The 1st value that described operational part is calculated the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be a plurality of characteristic values of obtaining of described characteristic value obtaining section, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation of a plurality of characteristic values of obtaining of described characteristic value obtaining section.

Conventionally, at the scene of manufacturing, by standard deviation sigma, carry out step management, therefore, for staff, standard deviation is a kind of parameter being very familiar to.So staff has very high consciousness to the decline degree of standard deviation.

On the other hand, above-mentioned GRR is shown: in order to express the standard deviation of AV and the corresponding dispersion of EV, with respect in order to express the ratio of the standard deviation of all dispersions.Yet this,, for being familiar with the staff of standard deviation, is difficult to recognize intuitively the measuring error between measuring appliance.That is to say, for staff, GRR be a kind of be difficult to recognize measuring error between measuring appliance have or not and current standard deviation between the parameter of relation, so staff is difficult to recognize that to be worth to make standard deviation to reduce according to which kind of of GRR how many.

But by above-mentioned scheme, staff can recognize the 1st value of expressing mutual relationship between the 1st standard deviation and the 2nd standard deviation.The 1st standard deviation is standard deviation drawn when the measuring error between a plurality of measuring appliances is got to 0, and the 2nd standard deviation is the standard deviation of a plurality of characteristic values of obtaining of described characteristic value obtaining section.At this, the 2nd standard deviation is the standard deviation of the characteristic value that obtains from a plurality of measuring appliances, and its state is surrounded by the measuring error between measuring appliance in being.Therefore by confirming the 1st value, just can easily learn than the measuring error between a plurality of measuring appliances and be taken as the state of 0 o'clock, there are how many variations in standard deviation.

In addition, in management devices of the present invention, also can adopt following scheme: be set with to judge whether workpiece is the specification limit of the characteristic value of non-defective unit, and if characteristic value falls in described specification limit and just judges that workpiece is in the situation of non-defective unit, described operational part is calculated the 2nd value based on described specification limit and is used as described parameter, wherein, described the 2nd value is expressed the mutual relationship between two kinds of fraction defectives, described two kinds of fraction defectives are: the measuring error between described a plurality of measuring appliances is being got 0 o'clock, the fraction defective that a plurality of characteristic values of obtaining according to described characteristic value obtaining section draw, and, the fraction defective that a plurality of characteristic values of obtaining according to described characteristic value obtaining section draw.

By such scheme, staff, as long as confirm the 2nd value, just can easily recognize by the measuring error between measuring appliance being adjusted to 0 and just can make fraction defective how much decline.

In addition, in management devices of the present invention, also can adopt following scheme: by the same workpiece of repeated measurement and the standard deviation of measured value be made as EV, and when the standard deviation of the measured value of a plurality of workpiece measured under identical conditions is made as to PV, a plurality of characteristic values that described operational part is obtained according to described characteristic value obtaining section, will meet formula PV ' ²=PV ²+ EV ²pV ' as described the 1st standard deviation, calculate, and calculate described the 2nd standard deviation TV, then TV/PV ' or PV '/TV are calculated as described the 1st value.

Can also adopt following scheme: the characteristic value of j the workpiece being recorded by i measuring appliance in the middle of a plurality of characteristic values that described characteristic value obtaining section is obtained is made as xij, and the number of described a plurality of measuring appliances is made as to a, when the number of the workpiece of also i measuring appliance being measured is made as ni, described operational part is according to following mathematical expression

[mathematical expression 10]

${\mathrm{PV}}^{\&prime;}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x\right)-x_{\mathrm{ij}}{)}^2}{N_{\mathrm{total}}}}$

[mathematical expression 11]

$\mathrm{TV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x\right)-x_{\mathrm{ij}}{)}^2}{N_{\mathrm{total}}}}$

[mathematical expression 12]

$\mathrm{ave}\left(x\right)=\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}{x}_{\mathrm{ij}}}{N_{\mathrm{total}}}$

[mathematical expression 13]

$\mathrm{ave}\left(x_i\right)=\frac{\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}\left(x_{\mathrm{ij}}\right)}{n_i}$

[mathematical expression 14]

$N_{\mathrm{total}}=\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}{n}_i$

Calculate PV ' and TV.

Or also can adopt following scheme: the characteristic value of j the workpiece being recorded by i measuring appliance in the middle of a plurality of characteristic values that described characteristic value obtaining section is obtained is made as xij, and the number of described a plurality of measuring appliances is made as to a, and the number of the workpiece that i measuring appliance measured is while being made as ni, described operational part is according to following mathematical expression

[mathematical expression 15]

$\mathrm{TV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x\right)-x_{\mathrm{ij}}{)}^2}{N_{\mathrm{total}}}}$

[mathematical expression 16]

$\mathrm{AV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}{n}_1(\mathrm{ave}\left(x\right)-\left(x_i\right){)}^2}{N_{\mathrm{total}}}}$

[mathematical expression 17]

$\mathrm{ave}\left(x\right)=\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}{x}_{\mathrm{ij}}}{N_{\mathrm{total}}}$

[mathematical expression 18]

$\mathrm{ave}\left(x_i\right)=\frac{\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}\left(x_{\mathrm{ij}}\right)}{n_i}$

[mathematical expression 19]

$N_{\mathrm{total}}=\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}{n}_i$

[mathematical expression 20]

$P{V}^{\&prime;}=\sqrt{\mathrm{<figure-callout\; id="2"\; label="T\; V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">T</figure-callout>}{V}^{}{}^2+A{\mathrm{<figure-callout\; id="2"\; label="V"\; filenames="HDA0000466020150000081.png"\; state="\{\{state\}\}">V</figure-callout>}}^2}$

Calculate PV ' and TV.

In such scheme, in production line operation, according to a plurality of characteristic values that the characteristic of the workpiece sending are measured by a plurality of measuring appliances, with meeting PV ' ²=PV ²+ EV ²pV ' carry out calculating parameter.That is, make measuring appliance with regard to same sample carry out standard deviation EV that repeated measurement obtains itself, without calculating.Therefore, can demonstrate the parameter of expressing dispersion between a plurality of measuring appliances with the characteristic value obtaining in production line operation.

In addition, as described above, for staff, GRR be a kind of be difficult to recognize measuring error between measuring appliance have or not and current standard deviation between the parameter of relation, so staff is difficult to recognize that to be worth to make standard deviation to reduce according to which kind of of GRR how many.

But in such scheme, TV/PV ' or PV '/TV are calculated as described parameter.The expressed concept of TV/PV ' is:: the current standard deviation while there is measuring error between measuring appliance is several times while there is not this perfect condition of measuring error between measuring appliance.In addition, the expressed concept of PV '/TV is:: the standard deviation while there is not this perfect condition of measuring error between measuring appliance is several times of current state that have measuring error between measuring appliance.Therefore, staff is by observing TV/PV ' or PV '/TV, just can easily recognize the measuring error existing between measuring appliance causes standard deviation to increase how many degree, or can easily recognize by the measuring error between measuring appliance being adjusted to 0 just to make standard deviation reduce how many degree.

In addition, in management devices of the present invention, described operational part preferably carries out following processing (1) and (2): (1) by the same workpiece of repeated measurement and the standard deviation of measured value be made as EV, and when the standard deviation of the measured value of a plurality of workpiece measured under identical conditions is made as to PV, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate and meet formula PV ' ²=PV ²+ EV ²pV ', and calculate the standard deviation TV of a plurality of characteristic values that described characteristic value obtaining section obtains; (2) based on described specification limit, using improving fraction defective, as described the 2nd value, calculate, wherein, the described fraction defective that improves is, fraction defective when standard deviation is taken as TV and standard deviation are taken as poor between the fraction defective in PV ' time.

In addition, operational part also can be with regard to the normal distribution of TV and probability outside standard deviation the is fallen specification limit fraction defective when standard deviation is taken as TV calculate, and the fraction defective that the probability with regard to the normal distribution of PV ' and outside standard deviation is fallen specification limit is taken as PV ' time as standard deviation is calculated.

By such scheme, staff improves fraction defective by confirmation, just can easily recognize by the measuring error between measuring appliance being adjusted to 0 and just can make fraction defective how much decline.

In addition, in management devices of the present invention, also can adopt following scheme: described operational part is calculated mean value and the standard deviation of a plurality of characteristic values that described characteristic value obtaining section obtains, and a plurality of characteristic values that described characteristic value obtaining section is obtained are divided into the measured characteristic value of every single measuring appliance, and calculate mean value and the standard deviation of the measured characteristic value of every single measuring appliance.Thus, by confirming these mean values and standard deviation, just can carry out various analyses.

In addition, in management devices of the present invention, also can adopt following scheme: described characteristic value obtaining section obtains characteristic value during with regard to each regulation; Described operational part is calculated described parameter during with regard to each regulation.Thus, can easily confirm the Temporal changes of parameter.

In addition, above-mentioned management devices also can be realized by computing machine.Now, make computing machine bring into play the control program of function as each portion of described control device and the computer-readable recording medium that records this program is also contained in category of the present invention.

[utilizability in industry]

The management devices that the present invention can manage for the measuring error between a plurality of measuring appliances that arrange on production line.

[description of reference numerals]

1 measuring appliance error management system

10 production lines

11 measuring appliances

20 measurement data gathering-devices

30 databases

40 management devices

41 input parts

42 display parts

43 obtain configuration part

44 measurement data obtaining sections (characteristic value obtaining section)

45 specification value storage parts

46 operational parts

47 deposit data portions

48 storage parts

49 chart Graphics Processing portions

50 operation recording device of transits (memory storage)

Claims (16)

1. a management devices, a plurality of measuring appliances that its management is measured the characteristic of workpiece on production line, this management devices is characterised in that:

Described a plurality of measuring appliance is measured the characteristic of the workpiece sending separately,

And this management devices possesses:

Characteristic value obtaining section, obtains a plurality of characteristic values that the characteristic of the described workpiece sending are measured by described a plurality of measuring appliances;

Operational part, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate the parameter of expressing the measuring error between described a plurality of measuring appliances.

2. management devices according to claim 1, is characterized in that:

The 1st value that described operational part is calculated the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be a plurality of characteristic values of obtaining of described characteristic value obtaining section, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation of a plurality of characteristic values of obtaining of described characteristic value obtaining section.

3. a management devices, a plurality of measuring appliances that its management is measured the characteristic of workpiece on production line, this management devices is characterised in that:

Possess:

Characteristic value obtaining section, obtains a plurality of characteristic values that obtain by described a plurality of measuring appliances;

Operational part, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate the parameter of expressing the measuring error between described a plurality of measuring appliances,

The 1st value that described operational part is calculated the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be a plurality of characteristic values of obtaining of described characteristic value obtaining section, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation of a plurality of characteristic values of obtaining of described characteristic value obtaining section.

4. according to the management devices described in any one in claim 1～3, it is characterized in that:

Be set with to judge whether workpiece is the specification limit of the characteristic value of non-defective unit, and if characteristic value fall in described specification limit and just judge that workpiece is in the situation of non-defective unit,

Described operational part is calculated the 2nd value based on described specification limit and is used as described parameter, wherein, described the 2nd value is expressed the mutual relationship between two kinds of fraction defectives, described two kinds of fraction defectives are: the measuring error between described a plurality of measuring appliances is being got to the fraction defective that a plurality of characteristic values of obtaining according to described characteristic value obtaining section draw at 0 o'clock; The fraction defective that a plurality of characteristic values of obtaining according to described characteristic value obtaining section draw.

5. according to the management devices described in claim 2 or 3, it is characterized in that:

By the same workpiece of repeated measurement and the standard deviation of measured value be made as EV, and when the standard deviation of the measured value of a plurality of workpiece measured under identical conditions is made as to PV,

A plurality of characteristic values that described operational part is obtained according to described characteristic value obtaining section, will meet formula PV ' ²=PV ²+ EV ²pV ' as described the 1st standard deviation, calculate, and calculate described the 2nd standard deviation TV, then TV/PV ' or PV '/TV are calculated as described the 1st value.

6. management devices according to claim 5, is characterized in that:

The characteristic value of j the workpiece being recorded by i measuring appliance in the middle of a plurality of characteristic values that described characteristic value obtaining section is obtained is made as xij, and the number of described a plurality of measuring appliances is made as to a, when the number of the workpiece of also i measuring appliance being measured is made as ni,

Described operational part is according to various below

$\mathrm{TV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x\right)-x_{\mathrm{ij}}{)}^2}{N_{\mathrm{total}}}}$

$\mathrm{ave}\left(x\right)=\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}{x}_{\mathrm{ij}}}{N_{\mathrm{total}}}$

$\mathrm{ave}\left(x_i\right)=\frac{\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}\left(x_{\mathrm{ij}}\right)}{n_i}$

Calculate PV ' and TV.

7. management devices according to claim 5, is characterized in that:

The characteristic value of j the workpiece being recorded by i measuring appliance in the middle of a plurality of characteristic values that described characteristic value obtaining section is obtained is made as xij, and the number of described a plurality of measuring appliances is made as to a, and the number of the workpiece that i measuring appliance measured is while being made as ni,

Described operational part is according to various below

$\mathrm{TV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}(\mathrm{ave}\left(x\right)-x_{\mathrm{ij}}{)}^2}{N_{\mathrm{total}}}}$

$\mathrm{AV}=\sqrt{\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}{n}_i(\mathrm{ave}\left(x\right)-\mathrm{ave}\left(x_i\right){)}^2}{N_{\mathrm{total}}}}$

$\mathrm{ave}\left(x\right)=\frac{\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}{x}_{\mathrm{ij}}}{N_{\mathrm{total}}}$

$\mathrm{ave}\left(x_i\right)=\frac{\underset{j=1}{\overset{n_i}{\mathrm{\&Sigma;}}}\left(x_{\mathrm{ij}}\right)}{n_i}$

$N_{\mathrm{total}}=\underset{i=1}{\overset{a}{\mathrm{\&Sigma;}}}{n}_i$

$P{V}^{\&prime;}=\sqrt{T{V}^2-A{V}^2}$

Calculate PV ' and TV.

8. management devices according to claim 4, is characterized in that:

Described operational part carries out following processing (1) and (2):

(1) by the same workpiece of repeated measurement and the standard deviation of measured value be made as EV, and when the standard deviation of the measured value of a plurality of workpiece measured under identical conditions is made as to PV, a plurality of characteristic values of obtaining according to described characteristic value obtaining section, calculate and meet formula PV ^{' 2}=PV ²+ EV ²pV ', and calculate the standard deviation TV of a plurality of characteristic values that described characteristic value obtaining section obtains;

(2) based on described specification limit, using improving fraction defective, as described the 2nd value, calculate, wherein, the described fraction defective that improves is, fraction defective when standard deviation is taken as TV and standard deviation are taken as poor between the fraction defective in PV ' time.

9. management devices according to claim 8, is characterized in that:

The fraction defective of the probability of described operational part with regard to the normal distribution of TV and outside standard deviation is fallen specification limit when standard deviation is taken as TV calculated, and the fraction defective of the probability with regard to the normal distribution of PV ' and outside standard deviation is fallen specification limit when standard deviation is taken as PV' calculated.

10. according to the management devices described in any one in claim 1～9, it is characterized in that:

Described operational part is not only calculated described parameter, also calculate mean value and the standard deviation of a plurality of characteristic values that described characteristic value obtaining section obtains, and a plurality of characteristic values that described characteristic value obtaining section is obtained are divided into the measured characteristic value of every single measuring appliance, and calculate mean value and the standard deviation of the measured characteristic value of every single measuring appliance.

11. according to the management devices described in any one in claim 1～10, it is characterized in that:

Described characteristic value obtaining section obtains characteristic value during with regard to each regulation;

Described operational part is calculated described parameter during with regard to each regulation.

12. 1 kinds of management methods, a plurality of measuring appliances that it is measured the characteristic of workpiece on production line in order to management, this management method is characterised in that:

Described a plurality of measuring appliance is measured the characteristic of the workpiece sending separately,

And this management method comprises:

Characteristic value is obtained step, obtains a plurality of characteristic values that the characteristic of the described workpiece sending are measured by described a plurality of measuring appliances;

Calculation step, according to a plurality of characteristic values of obtaining in described characteristic value is obtained step, calculates the parameter of expressing the measuring error between described a plurality of measuring appliances.

13. 1 kinds of management methods, a plurality of measuring appliances that it is measured the characteristic of workpiece on production line in order to management, this management method is characterised in that:

Comprise:

Characteristic value is obtained step, obtains a plurality of characteristic values that obtain by described a plurality of measuring appliances;

Calculation step, according to a plurality of characteristic values of obtaining in described characteristic value is obtained step, calculates the parameter of expressing the measuring error between described a plurality of measuring appliances,

In described calculation step, the 1st value of calculating the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be described characteristic value obtain a plurality of characteristic values of obtaining in step, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation that described characteristic value is obtained a plurality of characteristic values of obtaining in step.

14. 1 kinds of programs, it is performed by computing machine, a plurality of measuring appliances that described computer management is measured the characteristic of workpiece on production line, wherein,

Described a plurality of measuring appliance is measured the characteristic of the workpiece sending separately,

This program makes described computing machine carry out following steps:

Characteristic value is obtained step, obtains a plurality of characteristic values that the characteristic of the described workpiece sending are measured by described a plurality of measuring appliances;

Calculation step, according to a plurality of characteristic values of obtaining in described characteristic value is obtained step, calculates the parameter of expressing the measuring error between described a plurality of measuring appliances.

15. 1 kinds of programs, it is performed by computing machine, a plurality of measuring appliances that described computer management is measured the characteristic of workpiece on production line, wherein,

This program makes described computing machine carry out following steps:

Characteristic value is obtained step, obtains a plurality of characteristic values that obtain by described a plurality of measuring appliances;

Calculation step, according to a plurality of characteristic values of obtaining in described characteristic value is obtained step, calculates the parameter of expressing the measuring error between described a plurality of measuring appliances,

In described calculation step, the 1st value of calculating the mutual relationship between expression the 1st standard deviation and the 2nd standard deviation is used as described parameter, wherein, described the 1st standard deviation be described characteristic value obtain a plurality of characteristic values of obtaining in step, the standard deviation when the measuring error between described a plurality of measuring appliances is got to 0, described the 2nd standard deviation is the standard deviation that described characteristic value is obtained a plurality of characteristic values of obtaining in step.

16. 1 kinds of recording mediums, it is the computer-readable recording medium that records the program described in claims 14 or 15.

Images