JP6121154B2 - Inspection apparatus and inspection method - Google Patents

Description

  The present invention relates to an inspection apparatus and an inspection method for determining pass / fail of an inspection object based on a plurality of measurement values acquired by a plurality of measurement processes.

  For example, when inspecting whether or not an insulation defect has occurred in the inspection object, the resistance value of the inspection object is measured, and if a resistance value equal to or higher than a reference value is measured, it is determined that the insulation state of the inspection object is normal. There are known inspection apparatuses and inspection methods that determine that the insulation state of an inspection object is defective when a resistance value lower than a reference value is measured. In this case, when one measured value (resistance value) obtained by one measurement process is compared with a reference value, a measured value different from the original value is obtained due to the influence of noise or the like. In addition, there is a risk of erroneously determining that an insulation failure has occurred in an inspection object having a good insulation state, or erroneously determining that an inspection object having an insulation failure is in good insulation state. Therefore, a configuration / method for determining the quality based on a plurality of measurement values obtained by executing a measurement process for an inspection object a plurality of times is widely adopted.

  As an example, the applicant obtains a measured value using a resistance measuring device / resistance measuring method (hereinafter also simply referred to as “measuring device / measuring method”) disclosed in Japanese Patent No. 4919787, and There is a possibility that misjudgment due to the influence of noise or the like can be avoided. Specifically, in the measurement apparatus / measurement method disclosed by the applicant, the control unit controls the measurement unit to execute multiple measurements when instructed to start measurement. In this case, in this measuring apparatus / measurement method, either when the measurement abnormality occurs L times (L is a natural number of 2 or more) or when the number of measurements reaches M times (M is a natural number of L or more). The measurement of the resistance value by the measurement unit is repeatedly executed until the earlier time. When the number of measurements reaches M, the control unit calculates an average value by averaging the measured M resistance values. Thereby, according to this measuring apparatus and measuring method, it is possible to obtain an accurate measurement value with a sufficiently small measurement error, and an erroneous determination is caused by determining whether the inspection object is good or bad based on this measurement value. It becomes possible to avoid the situation.

Japanese Patent No. 4919787 (page 3-7, FIG. 1-5)

  However, the inspection apparatus / inspection method for inspecting the quality of an inspection object using the measurement apparatus / measurement method disclosed by the applicant has the following problems to be improved. That is, in the measurement apparatus and measurement method disclosed by the applicant, in order to obtain an accurate measurement value with a sufficiently small measurement error, a predetermined M number of measurements are performed and the M measurement values obtained are obtained. A configuration and a method for outputting the average value of as a final measurement value is adopted. However, in the measurement apparatus and measurement method disclosed by the applicant, for example, in any of the M measurements, a value including a high level of noise that affects the calculated average value is measured. In some cases, a measurement value (average value) that is different from the original measurement value to be measured in a normal inspection object may be calculated. In such a case, there is a possibility that a normal inspection object is erroneously determined as defective.

  In addition, for example, when some abnormality occurs in the inspection object and becomes unstable, when a large number of small measured values or large measured values that are greatly different from the above original measured values are measured, Even if the original measurement value is not measured even once, the measurement value (average value) calculated as the final measurement value may be the same value as the original measurement value (reference value). is there. In such a case, there is a risk of erroneously determining that the inspection target is normal although an abnormality has occurred in the inspection target. As described above, in the inspection apparatus / inspection method for inspecting the quality of the inspection object using the measurement apparatus / measurement method disclosed by the applicant, based on one measurement value acquired by one measurement process. Although the possibility of misjudgment can be reduced compared to the configuration / method of judging the quality of an inspection object, there is a possibility that misjudgment may occur when the above measured values are obtained. Is preferred.

  The present invention has been made in view of the problems to be solved, and a main object of the present invention is to provide an inspection apparatus and an inspection method capable of more accurately determining the quality of an inspection target.

In order to achieve the above object, an inspection apparatus according to claim 1 is configured to perform a measurement process for measuring a physical quantity to be inspected and outputting a measurement value; and controlling the measurement unit N times (N is (The natural number of 2 or more) is executed , and when the distribution state of the N measurement values output from the measurement unit is a normal distribution state defined in advance, both the inspection object and the measurement value are A processing unit that determines that the N measurement values are not normal and the distribution state of the N measurement values is not the normal distribution state, and determines that an abnormality has occurred in either the inspection object or the measurement values . In the inspection apparatus, the processing unit determines that the normal distribution state is present when the number of the measured values that are out of a predetermined normal distribution range is smaller than a predetermined first number. .

Furthermore, the inspection apparatus according to claim 2 is the inspection apparatus according to claim 1 , wherein the processing unit calculates an average value and a standard deviation of the N measured values, respectively, and calculates the average value and the standard deviation. A normal distribution is specified based on the above, and a distribution range defined in advance based on the specified normal distribution is set as the normal distribution range to determine whether the distribution state of the N measurement values is the normal distribution state. Determine.

The above inspection apparatus according to claim 3, wherein, in the inspection apparatus according to claim 1, wherein the processing unit (the M, where 3 or more natural number) M number of differences from each other are included in the measurement value range Each of the measured value distribution numbers is specified, and L (L is a natural number of M / 2 or less) of the measured value ranges including the measured values smaller than the average value of the N measured values. The distribution number of the measurement value range in the K-th order (K is each natural number less than or equal to L) in the descending order from the measurement value range in which the average value is included, and the measurement value larger than the average value are included A difference between the number of distributions of the K-th measurement value range defined in advance in ascending order from the measurement value range including the average value of the L measurement value ranges is defined in advance. It is determined that the normal distribution state when the number is less than

According to a fourth aspect of the present invention, a measurement process for measuring a physical quantity to be inspected to obtain a measurement value is executed N times (N is a natural number of 2 or more), and N distribution values of the measurement value are distributed. Are determined to be normal when the normal distribution state is defined in advance, and when the distribution state of the N measurement values is not the normal distribution state, An inspection method for determining that an abnormality has occurred in any of the measurement values, wherein the number of the measurement values that are outside the normal distribution range defined in advance is less than the first number defined in advance. It is determined that the normal distribution state .
An inspection method according to claim 5 is the inspection method according to claim 4, wherein an average value and a standard deviation of the N measurement values are respectively calculated, and a normal distribution is based on the average value and the standard deviation. And a distribution range defined in advance with the specified normal distribution as a reference is used as the normal distribution range to determine whether or not the distribution state of the N measurement values is the normal distribution state.
The inspection method according to claim 6 is the inspection method according to claim 4 or 5, wherein the number of distributions of the measurement values included in M measurement value ranges different from each other (M is a natural number of 3 or more). Are included, and the average value of the L measurement value ranges (L is a natural number of M / 2 or less) including the measurement value smaller than the average value of the N measurement values is included. The number of distributions of the measurement value range of the Kth (K is each natural number less than or equal to L) in the order counted from the measurement value range, and the L number of the measurement values that are greater than the average value are included. The difference from the number of distributions of the K-th measurement value range is smaller than a predetermined second number in the order of counting from the measurement value range including the average value in the measurement value range. Sometimes it is determined that the normal distribution state.

According to the inspection apparatus according to claim 1 and the inspection method according to claim 4 , when the distribution state of the N measurement values is a normal distribution state, it is determined that both the inspection object and the measurement value are normal, By determining that there is an abnormality in either the inspection object or the measurement value when the distribution state of the N measurement values is not the normal distribution state, for example, the inspection object or the measurement value may be abnormal. Regardless, when the measurement value with a value smaller than the reference value and the measurement value with a value larger than the reference value are measured in the same number, the average value of each measurement value is about the same as the reference value. However, since the distribution state of each measurement value does not become a normal distribution state, it can be reliably determined that an abnormality has occurred in the inspection object or the measurement value.

Further, according to the inspection apparatus and the inspection method, to determine the normal distribution when the number is less than the first number of predefined predefined measurement values are outside the normal distribution range As a result, when each measurement value becomes abnormally distributed due to an abnormality in the inspection object or measurement process, this is reliably detected, and either the inspection object or the measurement value is abnormal Can be accurately determined.

Furthermore, according to the inspection apparatus according to claim 2 and the inspection method according to claim 5, a distribution range defined in advance based on a normal distribution specified based on an average value and a standard deviation of N measurement values. By making the normal distribution range, even if the user is unfamiliar with the operation of this type of device, the normal distribution range that can accurately determine whether there is an abnormality in the inspection object or the measured value is ensured. it can be easily defined.

The inspection apparatus according to claim 3, and according to the inspection method according to claim 6, wherein the average value of the L measurements range including a small measured value than the average value of the N measurements The number of distributions of the Kth measurement value range in descending order from the included measurement value range and the average value of the L measurement value ranges including the measurement values larger than the average value of the N measurement values are When the difference from the number of distributions of the Kth measurement value range in order of counting from the included measurement value range is smaller than the second number defined in advance, it is determined that the distribution is in the normal distribution state. When each measured value becomes abnormally distributed due to an abnormality in the measurement process, this is reliably detected and accurately determined that either the inspection object or the measured value is abnormal be able to.

1 is a configuration diagram showing a configuration of an inspection apparatus 1. FIG. 4 is a display screen diagram illustrating an example of a determination result display screen 11. FIG. 10 is a flowchart of an inspection process 20. It is a display screen figure which shows another example of the measured value distribution display 13 in the determination result display screen. It is a display screen figure which shows another example of the measured value distribution display. It is a display screen figure which shows another example of the measured value distribution display.

  Hereinafter, embodiments of an inspection apparatus and an inspection method will be described with reference to the accompanying drawings.

  The inspection apparatus 1 shown in FIG. 1 is an example of an “inspection apparatus”, and is configured to be able to electrically inspect the quality of an inspection target (not shown) (for example, a resistor or a capacitor) according to an “inspection method” described later. ing. Specifically, the inspection apparatus 1 includes a measurement unit 2, an operation unit 3, a display unit 4, a control unit 5, and a storage unit 6, and, as will be described later, based on an input signal S, a physical quantity (voltage) for an inspection target. Value, current value, resistance value, power value, phase, temperature, humidity, luminance (luminous intensity), illuminance, rainfall, flow rate, etc.), and pass / fail of the inspection object based on the measurement results (for example, resistance value) The structure which inspects is adopted.

  The measurement unit 2 measures the physical quantity (analog-digital conversion) to generate measurement value data Ds under the control of the control unit 5, and outputs the generated measurement value data Ds to the control unit 5. "Process" is executed. In this case, in the inspection apparatus 1 of this example, as will be described later, N times defined in advance (for example, 3 times (example of “N = 3” which is an example of “N ≧ 2”) or 3 times Multiple times (for example, the number of times set by the user within the range of 100 times (example of “N = 100” which is an example of “N ≧ 2”)) is performed by the measurement unit 2, A configuration is adopted in which pass / fail of the inspection object is determined based on N measurement values (values of measurement value data Ds) output from the measurement unit 2 in N measurement processes. The operation unit 3 includes an operation switch for setting an operation condition of the inspection apparatus 1, an operation switch for instructing inspection start / end, and the like, and outputs an operation signal corresponding to the switch operation to the control unit 5.

  The display unit 4 includes a liquid crystal display panel as an example, and an operation condition setting screen (not shown) for setting operation conditions of the inspection apparatus 1 based on display data output from the control unit 5. In addition, a determination result display screen 11 (see FIG. 2) for displaying the measurement result is displayed. In this case, as shown in FIG. 2, in the inspection apparatus 1 of this example, “average” is obtained by averaging N measurement values (values of N measurement value data Ds) acquired by the N measurement processes described above. Is determined based on a measurement result display 12 indicating a value as a numerical value, a measurement value distribution display 13 indicating a distribution of N measurement values as a graph (in this example, a bar graph), and a distribution state of the N measurement values. A configuration is adopted in which a determination result display 14 showing the determination result is displayed side by side in the determination result display screen 11. Instead of the bar graph, a configuration (not shown) in which the distribution of N measured values is displayed on the measured value distribution display 13 by various graphs (not shown) such as a line graph, a pie chart and a band graph is adopted. You can also

  The control unit 5 comprehensively controls the inspection apparatus 1. Specifically, the control unit 5 corresponds to a “processing unit” and executes an inspection process 20 shown in FIG. 3 in response to an operation signal from the operation unit 3. In addition, the control unit 5 controls the measurement unit 2 to execute N measurement processes, and causes the storage unit 6 to store measurement value data Ds output from the measurement unit 2. Further, the control unit 5 calculates the average value by averaging the values of the N pieces of measurement value data Ds (N measurement values) stored in the storage unit 6, and uses the calculated average value as the measurement result display 12. It is displayed in the determination result display screen 11. Further, the control unit 5 specifies the distribution of the values of the N pieces of measurement value data Ds, displays the measurement value distribution display 13 along with the measurement result display 12 in the determination result display screen 11, and specifies The determination result display 14 is displayed by determining the quality of the inspection object based on the distribution state. The storage unit 6 stores the calculation result of the control unit 5, the measurement value data Ds, the reference value data D0, and the like.

  When an inspection object is inspected by the inspection apparatus 1, first, the operation switch of the operation unit 3 is operated to display an operation condition setting screen on the display unit 4 (not shown), and the operation conditions (inspection conditions, etc.) Set. Specifically, first, the number of measurement processes to be performed on one inspection object is set. In this case, as an example, it is set so that N = 100 measurement processes are executed. Next, among the N = 100 measurement value data Ds output from the measurement unit 2 by the set N = 100 measurement processes (hereinafter also simply referred to as “measurement value”), the determination is made at the time of pass / fail determination described later. A range of values to be included in the target (upper limit value and lower limit value of measurement values to be included in the determination target: information that can specify measurement values to be excluded during pass / fail determination: hereinafter also referred to as “determination target range”) is set.

  In this case, in the inspection apparatus 1 of the present example, a setting method in which the user directly designates the determination target range (for example, “± A” with respect to “center value” is set as the determination target range. “Center value” and “± A” are set by the user), and a range of “± D” centered on “average value” calculated by the control unit 5 as described later (“± D” is used) A setting method for automatically setting a value specified arbitrarily by a user as a determination target range, and a range of “± E times” with respect to a range of “standard deviation” calculated by the control unit 5 as described later (“ It is possible to set the determination target range by a setting method arbitrarily selected by the user from among the setting methods for automatically setting (± E magnification) as a determination target range (magnification arbitrarily specified by the user) It is configured. In this case, as an example, the user directly specifies the determination target range. The inspection apparatus 1 employs a configuration in which the distribution state of the measured values within the determination target range is displayed in the determination result display screen 11 as the measured value distribution display 13.

  Subsequently, it is set how many the above-mentioned determination target ranges are divided and the number of distributions of measurement values is counted, that is, how many measurement value ranges are defined and the number of distributions of measurement values is counted. In this case, in the inspection apparatus 1 of this example, as an example, a setting method in which the user directly specifies the number of divisions (number of measurement value ranges), and the number of divisions according to the number of executions of the measurement process set by the above setting operation A setting method in which (number of measurement value ranges) is automatically set, and the number of divisions (number of measurement value ranges) is automatically set according to the difference between the maximum and minimum values of N measurement values. The user can arbitrarily set the number of divisions of the determination target range (the number of measurement value ranges defined in the determination target range) by a setting method arbitrarily selected from among the setting methods. In this case, as an example, the determination target range is divided into 11 (specifying M = 11 different measurement value ranges in the determination target range), and the number of distributions of each measurement value is counted. Thus, the user directly designates “11” as the number of divisions (number of measurement value ranges = M).

  In this inspection apparatus 1, the number of measurement value distribution displays 13 is divided into a number corresponding to the number of divisions of the determination target range (number of measurement value ranges) to display the distribution of measurement values (how many bar graphs). The measurement value distribution display 13 is displayed) is set. In this case, the number of divisions of the determination target range (number of measurement value ranges defined in the determination target range = M) may be configured so as not to be changed with a predetermined fixed value.

  Next, pass / fail criteria for inspection are set. In this case, in the inspection apparatus 1 of this example, in addition to the pass / fail determination based on the distribution state of each measurement value described later, the pass / fail determination based on the average value of the N measurement values and the standard deviation of the N measurement values are performed. It is comprised so that the quality determination based on can be performed. Moreover, in this inspection apparatus 1, when it is set to execute the pass / fail determination based on the average value and the pass / fail determination based on the standard deviation in addition to the pass / fail determination based on the distribution state, when all the determination results are normal When either the inspection object and the measurement value obtained by the measurement process are determined to be normal, and either of the determination results is defective, either the inspection object or the measurement value obtained by the measurement process A configuration is adopted in which it is determined that an abnormality has occurred. In this case, in this example, it is set to execute all the above three pass / fail judgments.

  Subsequently, determination criteria for pass / fail determination based on the distribution state are set. In this case, in the inspection apparatus 1 of the present example, the number of measurement values that deviate from the normal distribution range A1 (see FIG. 2: an example of “predetermined normal distribution range”) is determined in advance as the quality determination based on the distribution state. A process of determining that the distribution state of the N measurement values is “normal distribution state” and determining that both the inspection object and the measurement value are normal when the number is equal to or less than the prescribed “first number”. The configuration to execute is adopted.

  In addition, as a setting method of the normal distribution range A1, a setting method in which the user directly designates the range (for example, using the upper limit value and the lower limit value regarding the number of distributions of the measurement values for each of the M measurement value ranges described above) A range of “± F” (“± F” is based on the normal distribution specified by the control unit 5 based on the N measured values, the average value, and the standard deviation). , Which is an example of a “predetermined distribution range”, and a setting method (a value arbitrarily specified by the user) is automatically set as the normal distribution range A1 (for each measurement value range based on the normal distribution) A configuration is adopted in which the normal distribution range A1 is set by a setting method arbitrarily selected by the user from two setting methods (the setting method for automatically setting an upper limit value and a lower limit value regarding the number of distributions). Yes. In this case, as an example, the normal distribution range A1 is set to be automatically set based on the normal distribution specified by the control unit 5. Thus, the preparation work for starting the inspection process is completed.

  Next, the inspection start switch of the operation unit 3 is operated in a state where a signal input terminal (not shown) of the measurement unit 2 and the inspection target are connected to each other using a pair of signal cables (not shown). In response to this, the control unit 5 starts the inspection process 20 shown in FIG.

  In this inspection process 20, the control unit 5 first controls the measurement unit 2 to execute the first of the N = 100 measurement processes (step 21). At this time, the measuring unit 2 is based on the input signal S input via both signal cables, and an electrical parameter between two points in the inspection target to which both signal cables are connected (in this example, resistance value). Is measured to generate measurement value data Ds, and the generated measurement value data Ds is output to the control unit 5. In response to this, the control unit 5 stores the output measurement value data Ds in the storage unit 6 and determines whether or not the execution of the measurement process of the set number of times = N = 100 is completed (step 22). ). At this time, it is determined that the first measurement process has only been completed (N = 100 measurement processes have not been completed), and the control unit 5 returns to the above-described step 21 to the measurement unit 2 for the second time. The measurement process is executed.

  On the other hand, when it is determined that the measurement process of the set number of times = N = 100 has been completed by repeatedly executing the process of step 21 described above (step 22), the control unit 5 stores it in the storage unit 6. The average value and standard deviation of each measured value are calculated (step 23). At this time, the control unit 5 determines whether or not N = 100 measurement values are included in the determination target range set prior to the inspection process, and there are measurement values outside the determination target range. In some cases, as an example, the measured value is excluded from the average value, the standard deviation, and the distribution state specification target described later. In this case, as an example, all of N = 100 measurement values are values within the determination target range, and an average value and a standard deviation are calculated based on the 100 measurement values.

  In addition, when there are more than a predetermined number of measurement values outside the determination target range (for example, 10 corresponding to 10% of the above “N”), the above N measurement processes are normally performed. It is also possible to adopt a configuration / method for displaying an error message to that effect and ending the inspection process 20 by assuming that each measurement value is abnormal due to not being performed. On the other hand, as shown in FIG. 2, the controller 5 displays the calculated average value on the measurement result display 12 as a numerical value. As a result, as shown in the figure, the average value of the measured values (in this example, the resistance value) measured by N = 100 measurement processes is “1.23456 kΩ” in the determination result display screen 11. Is displayed.

  Next, the control unit 5 calculates the normal distribution range A1 based on the average value and the standard deviation calculated in step 23 and each measurement value (N = 100 measurement values in this example) included in the determination target range. Set (step 24). Specifically, the control unit 5 first specifies a normal distribution based on the average value and the standard deviation calculated in step 23 described above. Next, the control unit 5 sets a range of “± F” based on the specified normal distribution as the normal distribution range A1. Further, as shown in FIG. 2, the control unit 5 performs the guide line L2 (normal distribution display) indicating the normal distribution specified when setting the normal distribution range A1 in the above step 24, and the upper limit value of the set normal distribution range A1. And a guide line L2b indicating the lower limit value of the normal distribution range A1 are displayed in the display area of the measured value distribution display 13, respectively.

  Subsequently, the control unit 5 sets the measurement values included in the determination target range among the N = 100 measurement values stored in the storage unit 6 to the 11 measurement value ranges set by the setting operation. Each of the measured value ranges is specified (the number of distributions of measured values included in each measured value range is counted: step 25). Next, as shown in FIG. 2, the control unit 5 determines the number of measurement values belonging to each measurement value range based on the specified result (number counted) as shown in FIG. The bar graph consisting of “bar B” when not distinguished from each other is displayed as a measurement value distribution display 13 on the measurement result display 12 and displayed in the determination result display screen 11. Further, the control unit 5 causes the determination result display screen 11 to display the line L1 (average value display) indicating the average value calculated in step 23 above the measurement value distribution display 13.

  Next, the control unit 5 determines pass / fail of the inspection object based on the average value of each measurement value calculated in step 23 and the reference value data D0 stored in the storage unit 6 (step 26). At this time, the “inspection object” recorded in the reference value data D0 due to an abnormality in the inspection object or an abnormality in the measurement process (such as a signal cable connection failure). When a large number of measurement values that are significantly different from the “reference value to be measured” are measured (when the measurement value is abnormal), the average value calculated in step 23 becomes the reference value data D0. This is a value that differs greatly from the recorded reference value. Accordingly, when the average value calculated in step 23 differs from the reference value recorded in the reference value data D0 beyond the allowable range defined in advance, the control unit 5 determines the inspection object and the measured value. It is determined that an abnormality has occurred in any of the above. On the other hand, when the calculated average value is within the allowable range with respect to the reference value, it is determined that both the inspection object and the measured value are normal (“provisional determination that there is no abnormality”).

  Subsequently, the control unit 5 determines the quality of the inspection target based on the standard deviation of each measurement value calculated in step 23 and the reference value data D0 stored in the storage unit 6 (step 27). At this time, when a large variation occurs in each measurement value measured during the above-described N = 100 measurement processes due to an inspection object or an abnormality in the measurement process, the standard deviation calculated in step 23 is Large value. Therefore, the control unit 5 determines that an abnormality has occurred in either the inspection target or the measured value when the standard deviation calculated in step 23 is a large value that exceeds a predetermined allowable range. On the other hand, when the calculated standard deviation is a value within the allowable range, it is determined that both the inspection object and the measured value are normal (“provisional determination that there is no abnormality”).

  Next, the control unit 5 determines whether or not the number of distributions of measurement values for each measurement value range specified in Step 25 is a number within the normal distribution range A1 set in Step 24. The quality is judged (step 28). At this time, when irregular variations occur in each measurement value measured during the above-mentioned N = 100 measurement processes due to an inspection object or an abnormality in the measurement process, the distribution state of each measurement value is The distribution is different from the normal distribution. Therefore, as shown in the measurement value distribution display 13 shown in FIG. 4, when there is a measurement value range having a larger number of distributions than the normal distribution range A1 (for example, the number of distributions of the measurement value range indicated by the bar Ba2), normal When there is a measurement value range with a smaller number of distributions than the distribution range A1 (for example, the number of distributions in the measurement value range indicated by the bar Bb1), there is even one measurement value range with a distribution number outside the normal distribution range A1. When: “The number of measurement values out of the normal distribution range defined in advance is the first number (the number of measurement values out of the normal distribution range defined in advance” is an example of “when the distribution state of the N measurement values is not the normal distribution state”) In this example, the control unit 5 determines that an abnormality has occurred in either the inspection object or the measurement value.

  On the other hand, when the number of measurement value distributions for each measurement value range indicated by each bar B is a number within the normal distribution range A1 as in the measurement value distribution display 13 shown in FIG. 2 (normal distribution range A1) When there is no measurement value range of the number of distributions that deviates from: “When the distribution state of N measurement values is a normal distribution state” is an example of “a measurement value that is out of the normal distribution range defined in advance” In the case where the number is smaller than a predetermined first number (in this example, one)), it is determined that both the inspection object and the measurement value are normal (“temporary not abnormal” Judgment "). It should be noted that when there is at least one measurement value range that is out of the normal distribution range A1, that is, when the number of measurement values out of the normal distribution range A1 is 1 or more (the “first number” is In the example of “1”), instead of the determination processing in step 28 described above in which it is determined that an abnormality has occurred in either the inspection target or the measurement value, the number of distributions that deviate from the normal distribution range A1 in the plurality of measurement value ranges. When the total number of measured values out of the normal distribution range A1 in the counted state is, for example, (N / 10) or more (in this example, 10: another example of “first number”), A determination process for determining that an abnormality has occurred in either the inspection object or the measurement value can also be executed.

  Subsequently, the control unit 5 determines whether or not all of the determination processing results (determination results) in steps 26 to 28 are “normal” (step 29). At this time, when any or all of the results of the determination processes in steps 26 to 28 are “abnormal”, the control unit 5 determines the inspection target and the measurement value obtained by the measurement process. It is determined that an abnormality has occurred in any of the above (abnormality determination: step 30), and the determination result (a character string “FAIL” as an example) is displayed on the determination result display screen 11 as the determination result display 14 ( This inspection process 20 is terminated. On the other hand, when all the results of the determination processes in steps 26 to 28 are “normal”, the control unit 5 determines that both the inspection target and the measurement value obtained by the measurement process are normal. Then, as shown in FIG. 2, the determination result (a character string “PASS” as an example) is displayed on the determination result display screen 11 as the determination result display 14 as shown in FIG. Exit.

  In this case, as described above, the reference value recorded in the reference value data D0 (when the inspection object is normal and the measurement process is normally executed due to an abnormality of the inspection object or the measurement process) When a large number of measurement values that are greatly different from the measurement values to be measured) are measured, the average value calculated in step 23 is a value that differs greatly from the reference value. It can be determined that an abnormality has occurred in any of the measured values. In addition, when a large variation occurs in each measurement value due to an inspection object or measurement processing abnormality, the standard deviation calculated in step 23 becomes a large value. It can be determined that an abnormality has occurred in any of them. However, even if there is an abnormality in the inspection target or measurement value, for example, the measurement value slightly smaller than the reference value and the measurement value slightly larger than the reference value are approximately the same number and average value. In the case where a larger number of measurement values than the neighboring measurement values are measured, in step 23, an average value comparable to the reference value and a standard deviation of a small value are calculated. Therefore, in this case, in steps 26 and 27, the inspection object or measurement value in which an abnormality has occurred is erroneously determined to be normal.

  On the other hand, in the inspection apparatus 1 of this example, in step 28, the quality determination process based on the number of distributions of measurement values for each measurement value range is performed. In this case, as in the above example, due to an abnormality in the inspection object or measurement value, the measurement value with a value smaller than the reference value and the measurement value with a value larger than the reference value are measured in the same number Includes the number of distributions of the measurement value region including the above-mentioned “measurement value of a value smaller than the reference value” with respect to the number of distributions of the measurement value for each measurement value region specified in step 25, and the above-mentioned “reference value” The number of distributions of the measurement value region including the “measured value of a larger value” is larger than the normal distribution range A1. Therefore, even if the inspection object and the measurement value are erroneously determined to be normal in steps 26 and 27, it can be determined in step 28 that an abnormality has occurred in either the inspection object or the measurement value.

  Thus, according to this inspection apparatus 1 and the inspection method by the inspection apparatus 1, when the distribution state of the N measurement values is the normal distribution state, it is determined that both the inspection object and the measurement value are normal. By determining that there is an abnormality in either the inspection object or the measurement value when the distribution state of the N measurement values is not a normal distribution state, for example, an abnormality occurs in the inspection object or the measurement value. Nevertheless, when the measured value of the value smaller than the reference value and the measured value of the value larger than the reference value are measured in the same number, the average value of each measured value is about the same as the reference value. In addition, since the distribution state of each measurement value does not become a normal distribution state, it can be reliably determined that an abnormality has occurred in the inspection object or the measurement value.

  Further, according to the inspection apparatus 1 and the inspection method by the inspection apparatus 1, it is normal when the number of measurement values that are out of the normal distribution range A1 defined in advance is equal to or less than the predetermined “first number”. By determining the distribution state, when each measurement value becomes an abnormal distribution state due to an abnormality in the inspection object or measurement process, this is reliably detected, and either the inspection object or the measurement value is detected. It can be accurately determined that an abnormality has occurred.

  Further, according to the inspection apparatus 1 and the inspection method by the inspection apparatus 1, a distribution range defined in advance based on a normal distribution specified based on an average value and standard deviation of N measurement values is a normal distribution range. By using A1, even if a user is unfamiliar with the operation of this type of apparatus, the normal distribution range A1 that can accurately determine whether or not there is an abnormality in the inspection object or measurement value is ensured and easy. The inspection process can be executed by specifying.

  Note that the configuration of the “inspection apparatus” and the content of the “inspection method” are not limited to the configuration of the inspection apparatus 1 and the content of the inspection method by the inspection apparatus 1. For example, a normal distribution range A1 that defines both an upper limit value and a lower limit value of the number of distributions of measurement values for each measurement value region is defined, and the number of distributions of measurement values for each measurement value region is within the normal distribution range A1. In some cases, the configuration / method for determining that both the inspection target and the measurement value obtained by the measurement process are normal has been described as an example, but the normal distribution range A2 in the measurement value distribution display 13 illustrated in FIG. When the “normal distribution range” that defines only the upper limit value of the number of distributions of measurement values for each measurement value region is defined, and the number of distributions of measurement values for each measurement value region is within the normal distribution range A2 It is also possible to adopt a configuration / method for determining that the inspection object and the measurement value are normal. By adopting such a configuration / method, when there is no abnormality in the inspection object or the measurement value, for example, the average value is large as the number of distributions of the measurement value area indicated by the bar Bb4 in FIG. Even if the number of distributions in the measurement value range that includes different measurement values is smaller than the number of normal distributions, it is erroneously determined that either the inspection target or the measurement value is abnormal Can be avoided.

  In addition, whether or not the distribution state of each measurement value is a normal distribution state defined in advance is determined based on the first number in which the number of measurement values out of the normal distribution range defined in advance is predetermined. It is not limited to the determination based on whether there are few. For example, in the same manner as in the preparatory work of the inspection process 20 described above, the set determination target range is divided into several parts to count the number of measurement value distributions, that is, the measurement value range is defined and the measurement value Whether to count the number of distributions (set M measurement value ranges), and in the inspection process, L measurement value ranges including measurement values smaller than the average value of N measurement values The number of distributions of the Kth measurement value range in descending order counting from the measurement value range including the average value, and the L measurement value ranges including the measurement value larger than the average value of the N measurement values It is possible to employ a configuration / method for determining the quality of an inspection object and measurement processing based on the difference from the number of distributions of the Kth measurement value range in ascending order from the measurement value range in which the average value is included. it can.

  Specifically, as an example, M = 11 measurement value ranges (an example of “L = 5”) are set in the same manner as in the example of the inspection process 20 described above, and the process of step 25 in the inspection process 20 is performed. In the same manner as above, the measurement value included in the determination target range among the N = 100 measurement values is specified to which measurement value range among the M = 11 measurement value ranges. To do. Thereby, as indicated by each bar B in the measurement value distribution display 13 shown in FIG. 6, the number of distributions of measurement values included in each measurement value range is counted. Next, the number of distributions indicated by the bar Ba1 (the measurement value range including the average value of the “L measurement value ranges” (in this example, the measurement value range of the distribution number indicated by the bar B0)) is large. K = number of distributions of the first measurement value range in order) and the distribution number indicated by the bar Bb1 (measurement value range including an average value of “L measurement value ranges” (in this example, the bar B0) The difference d1 with respect to the distribution number of the first measurement value range) is calculated in ascending order starting from the measurement value range of the distribution number indicated by (1).

  Similarly, the number of distributions indicated by the bar Ba2 (number of distributions in the second measurement value range in the descending order counting from the measurement value range including the average value) and the number of distributions indicated by the bar Bb2 (average value) The difference d2 from the measurement value range in which K is included and the distribution number indicated by the bar Ba3 (in descending order from the measurement value range in which the average value is included) And K = number of distributions in the third measurement value range) and the number of distributions indicated by the bar Bb3 (K = number of distributions in the third measurement value range in ascending order from the measurement value range including the average value) and Difference d3, the number of distributions indicated by the bar Ba4 (K = the number of distributions in the fourth measurement value range in descending order from the measurement value range including the average value), and the number of distributions indicated by the bar Bb4 (average value) K = 4th measurement in ascending order from the measured value range. Difference d4 from the number of distributions of the value range) and the number of distributions indicated by the bar Ba5 (K = the number of distributions of the fifth measurement value range in descending order from the measurement value range including the average value) and the bar Bb5 And the difference d5 from the number of distributions (K = number of distributions in the fifth measurement value range in ascending order from the measurement value range including the average value).

  At this time, when irregular variations occur in each measurement value measured during the above-mentioned N = 100 measurement processes due to an inspection object or an abnormality in the measurement process, the distribution state of each measurement value is The distribution is different from the normal distribution. Therefore, as in the measurement value distribution display 13 shown in FIG. 6, the number of distributions is smaller than the normal distribution, such as the number of distributions of the measurement value range indicated by the bar Bb1, or the measurement value range indicated by the bar Ba2 As a result of the number of distributions being larger than the normal distribution, such as the number of distributions, in the example shown in the figure, the differences d1 and d2 are large values. Therefore, the control unit 5 has an abnormality in either the inspection object or the measurement value when the difference d equal to or greater than the “second number” defined in advance is present in the calculated differences d1 to d5. Is determined. Further, when all of the calculated differences d1 to d5 are smaller than the “second number” (when the differences d1 to d5 are smaller than the “second number”, respectively), both the inspection target and the measured value are normal. It is determined that In this case, in the inspection apparatus 1 of this example, when the total value of the differences d1 to d5 is equal to or greater than the “second number”, it can be determined that an abnormality has occurred in either the inspection target or the measurement value. .

In this way, of the Kth measurement value range in descending order from the measurement value range including the average value among the L measurement value ranges including the measurement values smaller than the average value of the N measurement values. The number of distributions and the Kth measurement value range in ascending order from the measurement value range including the average value among the L measurement value ranges including the measurement values larger than the average value of the N measurement values. According to the inspection apparatus 1 configured to determine the normal distribution state when the difference from the distribution number is smaller than the predetermined second number, and the inspection method, the normal distribution ranges A1, A2, etc. are defined. In the same manner as in the above configuration and method for determining whether or not the distribution state of each measurement value is a normal distribution state, each measurement value is regarded as having an abnormal distribution state due to an abnormality in the inspection target or measurement process. When this happens, this is reliably detected, and no matter what is being tested or measured It is possible to accurately determine the abnormality has occurred in either.

  Furthermore, not only the pass / fail judgment based on the number of distributions of the measurement values for each measurement value range (for example, step 28), but also the pass / fail judgment based on the average value (step 26) and the pass / fail judgment based on the standard deviation (step 27). The configuration and method for determining that both the inspection object and the measured value were normal when executed and determined to be “normal” in all of these were described as an example. ”Is not an essential process for determining the quality based on the average value or the standard deviation, but only performing the quality determination based on the number of distributions of the measured values for each measurement value range. It is also possible to adopt a configuration / method for determining. Further, the configuration / method of displaying the measurement result display 12 and the measurement value distribution display 13 side by side with the determination result display 14 in the determination result display screen 11 has been described as an example. However, in the “inspection apparatus” and “inspection method” Display of the measurement result display 12 and the measurement value distribution display 13 is not an essential process. For example, only the determination result display 14 is displayed in the determination result display screen 11 or the determination result is notified by voice instead of the determination result display 14. You can do it.

DESCRIPTION OF SYMBOLS 1 Inspection apparatus 2 Measuring part 4 Display part 5 Control part 6 Memory | storage part 11 Judgment result display screen 14 Judgment result display 20 Inspection process A1, A2 Normal distribution range B0, Ba1-Ba5, Bb1-Bb5 Bar D0 Reference value data d1-d5 Difference Ds Measurement data

Claims (6)

A measurement unit that performs a measurement process of measuring a physical quantity to be inspected and outputting a measurement value;
A normal distribution state in which the measurement unit is controlled to execute the measurement process N times (N is a natural number of 2 or more) and the distribution state of the N measurement values output from the measurement unit is defined in advance. When it is determined that both the inspection object and the measurement value are normal and the distribution state of the N measurement values is not the normal distribution state, either the inspection object or the measurement value is abnormal An inspection apparatus including a processing unit that determines that the
The inspection apparatus determines that the normal distribution state is present when the number of the measurement values that are out of a predetermined normal distribution range is smaller than a predetermined first number . The processing unit calculates an average value and a standard deviation of the N measurement values, specifies a normal distribution based on the average value and the standard deviation, and predefines the reference based on the specified normal distribution. It has been examined according to claim 1, wherein the distribution of the N number of measurements the distribution range as the normal distribution range it is determined whether or not the normal distribution. The processing unit specifies the number of distributions of the measurement values included in M measurement value ranges (M is a natural number of 3 or more) that are different from each other, and is greater than the average value of the N measurement values. Of the L measurement value ranges including the small measurement values (L is a natural number of M / 2 or less), the Kth (K is The number of distributions of the measurement value range of each natural number less than or equal to L and the measurement value range including the average value of the L measurement value ranges including the measurement value larger than the average value the difference between the number of distributions of ascending order in the K-th of the measurement range, counting from is determined to be in the normal distribution when less than a second number of predefined claim 1 or 2, wherein Inspection equipment. When the measurement process for measuring the physical quantity to be inspected and obtaining the measurement value is executed N times (N is a natural number of 2 or more) and the distribution state of the N measurement values is a normal distribution state defined in advance It is determined that both the inspection object and the measurement value are normal, and an abnormality occurs in either the inspection object or the measurement value when the distribution state of the N measurement values is not the normal distribution state. An inspection method for determining that
An inspection method for discriminating that the normal distribution state is present when the number of the measured values out of a predetermined normal distribution range is smaller than a first predetermined number . The average value and standard deviation of the N measured values are calculated, respectively, a normal distribution is specified based on the average value and the standard deviation, and a pre-defined distribution range based on the specified normal distribution is determined. The inspection method according to claim 4, wherein it is determined whether or not the distribution state of the N measurement values is the normal distribution state as the normal distribution range. The number of distributions of the measurement values included in the M measurement value ranges different from each other (M is a natural number of 3 or more) is specified, and the measurement value smaller than the average value of the N measurement values is Among the L measurement values included (L is a natural number equal to or less than M / 2), the Kth (K is each natural number equal to or less than L) in descending order from the measurement value range including the average value. ) In the order of counting from the measurement value range including the average value of the L measurement value ranges including the measurement value larger than the average value. 6. The inspection method according to claim 4, wherein the normal distribution state is determined when a difference from the distribution number of the K-th measurement value range is smaller than a predetermined second number.

Images