JP2019012300A - Inspection tool of operation performance data - Google Patents

Abstract

To provide a novel tool that can reduce the burden of inspection of operation performance data in the renewal work of an existing production line.SOLUTION: A correlation coefficient (>0) between data of an actuator control signal and data of a sensor signal is calculated (step S10). It is determined whether or not there is a sensor signal whose correlation coefficient is larger than a threshold value (step S12). If the determination result in step S12 is affirmative, the sensor signals are rearranged in descending order of the correlation coefficient (step S14). A causal relationship between the actuator control signal and the sensor signal is investigated in order of the rearranged sensor signals (step S16). It is judged whether or not there is the sensor signal that establishes the causal relationship (step S18). When the determination result of step S18 is affirmative, the sensor signal is selected as the signal of a survey target sensor (step S20).SELECTED DRAWING: Figure 3

Description

  The present invention relates to an operation performance data survey tool.

  A plant production line is provided with actuators such as motors and valves for driving equipment. Such an actuator is automatically operated by a PLC (Programmable Logic Controller). A number of sensors are connected to the PLC together with the actuator. The PLC uses these sensors to perform automatic operation of the actuator.

  In the automatic operation by the PLC, the actuator is driven by the output of the PLC. The production line equipment operates by the driving force of the actuator. The sensor detects the driving state of the actuator or the operating state of the equipment that is operated by the driving force of the actuator. The sensor converts the detected driving state or operating state into a signal and outputs the signal to the PLC. The PLC generates and outputs a signal for driving the actuator based on a signal input from the sensor to the PLC. Such feedback-type closed loop control is widely performed in automatic operation by PLC.

  In a plant that is operating, it is common to collect operation result data of a production line. For example, Japanese Patent Application Laid-Open No. 10-069498 discloses a data collection and recording device that collects operation line data of a production line at a constant period and stores it as time-series data.

JP-A-10-069498

  In order to implement automatic operation by PLC in the production line, it is necessary to associate a signal output from the PLC to the actuator and a signal input from the sensor corresponding to the actuator to the PLC. In order to associate both signals, it is necessary to create a correspondence table between the operation direction of the actuator and the detection state of the sensor.

  This correspondence table is usually determined at the machine design stage. In the construction of a new production line, the equipment will be constructed according to the machine design. Therefore, if machine design information can be obtained, a correspondence table can be easily created in new construction.

  However, in the renewal work of the existing production line, the machine design information itself that is the basis of the correspondence table often does not exist. Even when machine design information is present, some changes may have been made to the facility at the beginning of the new installation. In such a case, the machine design information does not match the current equipment information. For this reason, it is necessary to check the connection status between the current equipment and actuator, or check the PLC source code, by tracing the cable connected to the PLC input / output module in order to investigate the operation performance data. It takes a lot of labor and time.

  The present invention has been made in view of the above-described problems, and it is an object of the present invention to provide a novel tool capable of reducing the burden of investigating operation performance data in the renewal work of an existing production line.

1st invention is the operation performance data investigation tool for achieving said objective, and has the following characteristics.
The operation result data investigation tool is configured to be accessible to a storage device in which plant operation result data is stored.
The plant includes an actuator that is driven by a control signal, and a sensor that detects a driving state of the actuator or an operating state of equipment that is operated by a driving force of the actuator.
The operation performance data includes time-series data of the control signal and time-series data serving as candidate sensor signals output from the sensor.
The operation performance data survey tool is
Using the time-series data in a predetermined period of the operation performance data, calculate a correlation coefficient between the sensor signal candidate and the control signal,
It is determined whether or not a causal relationship is established between the sensor signal candidate and the control signal in the descending order of the correlation coefficient for candidates having the correlation coefficient larger than the threshold among the sensor signal candidates. ,
When it is determined that the causal relationship is established, a correspondence table between the control signal and the sensor signal is created using the sensor signal candidates for which the causal relationship is established.

The second invention is an operation performance data survey tool for achieving the above object, and has the following features.
The operation result data investigation tool is configured to be accessible to a storage device in which plant operation result data is stored.
The plant includes an actuator that is driven by a control signal, and a sensor that detects a driving state of the actuator or an operating state of equipment that is operated by a driving force of the actuator.
The operation performance data includes time-series data of the control signal and time-series data serving as candidate sensor signals output from the sensor.
The operation performance data survey tool is
Using the time-series data in a predetermined period of the operation performance data, calculate a correlation coefficient between the sensor signal candidate and the control signal,
Determining whether or not a causal relationship is established between the sensor signal candidate and the control signal for all candidates of the sensor signal candidate for which the correlation coefficient is greater than a threshold;
When it is determined that the causal relationship is established, among the candidate sensor signals for which the causal relationship is established, a candidate having the shortest time difference between the change of the control signal and the change of the sensor signal candidate is used. Thus, the correspondence table of the control signal and the sensor signal is created.

The third invention is an operation performance data survey tool for achieving the above object, and has the following features.
The operation result data investigation tool is configured to be accessible to a storage device in which plant operation result data is stored.
The plant includes an actuator that is driven by a control signal, and a sensor that detects a driving state of the actuator or an operating state of equipment that is operated by a driving force of the actuator.
The operation performance data includes time-series data of the control signal and time-series data serving as candidate sensor signals output from the sensor.
The operation performance data survey tool is
Using the time-series data in a predetermined period of the operation performance data, calculate a correlation coefficient between the sensor signal candidate and the control signal,
It is determined whether or not a causal relationship is established between the sensor signal candidate and the control signal in the descending order of the correlation coefficient for candidates having the correlation coefficient larger than the threshold among the sensor signal candidates. ,
When it is determined that the causal relationship is established, among the candidates for the sensor signal, other candidates for which the success or failure of the causal relationship has not been determined, and the sensor signal candidate that has been determined that the causal relationship is established Determining whether there are other candidates equal in the correlation coefficient;
If there are other candidates, determine whether a causal relationship is established between the other candidates and the control signal,
When it is determined that a causal relationship is established between the other candidate and the control signal, the sensor signal is changed after the control signal is changed among all candidates of the sensor signal determined to be established. A correspondence table between the control signal and the sensor signal is created using a candidate having the shortest time difference until the signal candidate changes.

A fourth invention has the following characteristics in any one of the first to third inventions.
The predetermined period is at least three periods set in advance.
The operation performance data survey tool creates the correspondence table using the candidate having the largest number of selections when the sensor signal candidates that satisfy the causal relationship are selected in the at least three periods. It is configured.

The fifth invention has the following features in the fourth invention.
The operation performance data survey tool is configured to perform the weighting set in advance for each of the at least three periods and to evaluate the number of selections.

The sixth invention has the following characteristics in the fourth invention.
The operation performance data survey tool is
In each of the at least three periods, the sensor signal in which a causal relationship is established between the sensor signal candidate and the control signal for all candidates of the sensor signal that have a correlation coefficient greater than a threshold value. Calculate the number of
Weighting according to the number of the sensor signals for which the causal relationship is established is performed on each of the at least three periods, and the selection number is evaluated.

  According to the operation result data survey tool according to the first aspect of the present invention, whether or not a causal relationship is established between the control signal and the control signal in the descending order of the correlation coefficient for sensor signal candidates whose correlation coefficient is greater than the threshold value. It is possible to determine and select an appropriate sensor signal. Accordingly, it is possible to reduce the burden of investigating the operation result data in the renewal work of the existing production line.

  According to the operation performance data survey tool according to the second invention, a candidate having the shortest time difference among all candidates having a correlation coefficient larger than the threshold value can be selected as an appropriate sensor signal. Accordingly, it is possible to reduce the burden of investigating the operation result data in the renewal work of the existing production line.

  According to the operation result data survey tool according to the third aspect of the invention, whether or not a causal relationship is established between the control signal and the control signal in the descending order of the correlation coefficient for candidate sensor signals having a correlation coefficient larger than a threshold value. If there is a sensor signal candidate for which a causal relationship is established between the control signal and the control signal, is the causal relationship established between another candidate for which the causal relationship has not been determined and the control signal? You can determine whether or not. If there is another candidate of the sensor signal that has a causal relationship with the control signal, the candidate with the shortest time difference is selected from all the candidates of the sensor signal that are determined to have the causal relationship. Can be elected as a signal. Accordingly, it is possible to reduce the burden of investigating the operation result data in the renewal work of the existing production line.

  According to the operation performance data tool according to the fourth invention, the reliability of the selection result of an appropriate sensor signal can be increased.

  According to the operation performance data tool according to the fifth aspect of the invention, an appropriate sensor signal can be selected even when there is a specific circumstance during a predetermined period.

  According to the operation performance data tool according to the sixth aspect of the invention, an appropriate sensor signal can be selected even when the singular data is included in the time-series data within the predetermined period.

It is a figure explaining the structural example of the investigation tool which concerns on Embodiment 1 of this invention, and the structural example of the manufacturing line of the plant to which the said investigation tool is applied. It is the figure which showed an example of the time series data of an actuator control signal and a sensor signal. In Embodiment 1 of this invention, it is a figure explaining an example of the signal selection and the correspondence table preparation process which an arithmetic unit performs. FIG. 4 is a diagram for specifically explaining the processing after step S14 in FIG. 3. In Embodiment 2 of this invention, it is a figure explaining an example of the signal selection and the correspondence table preparation process which an arithmetic unit performs. FIG. 6 is a diagram for specifically explaining the processing after step S <b> 22 in FIG. 5. In Embodiment 3 of this invention, it is a figure explaining an example of the signal selection and the correspondence table preparation process which an arithmetic unit performs. It is a figure explaining the process after step S30 of FIG. In Embodiment 4 of this invention, it is a figure explaining an example of the signal selection and correspondence table preparation processing which an arithmetic unit performs. It is a figure explaining the process of step S44. It is a figure explaining the process of step S44. In Embodiment 5 of this invention, it is a figure explaining an example of the signal selection and correspondence table preparation processing which an arithmetic unit performs. In Embodiment 7 of this invention, it is a figure explaining an example of the signal selection and correspondence table preparation processing which an arithmetic unit performs. In Embodiment 8 of this invention, it is a figure explaining an example of the signal selection and a correspondence table preparation process which an arithmetic unit performs.

  Embodiments of the present invention will be described below with reference to the drawings. However, in the embodiment shown below, when referring to the number of each element, quantity, quantity, range, etc., unless otherwise specified or clearly specified in principle, the reference However, the present invention is not limited to this number. Further, the structures, steps, and the like described in the embodiments below are not necessarily essential to the present invention unless otherwise specified or clearly specified in principle.

Embodiment 1 FIG.
First, the investigation tool according to Embodiment 1 of the present invention will be described with reference to FIGS. FIG. 1 is a diagram illustrating a configuration example of a survey tool according to Embodiment 1 of the present invention and a configuration example of a plant production line to which the survey tool is applied.

  A survey tool 10 shown in FIG. 1 is a support tool for investigating operation performance data of an existing production line. Before describing the configuration of the survey tool 10, the configuration of the production line 20 that is the subject of the update work will be described. The production line 20 is an existing production line. The production line 20 includes an actuator 22, a plurality of sensors 24, a PLC 26, other equipment 28, and a data collection device 30.

  The actuator 22 is connected to the output module of the PLC 26. However, the actuator 22 may be connected to the PLC 26 via a network. The actuator 22 is driven by a control signal from the PLC 26. Various facilities (not shown) of the production line 20 are operated by the driving force of the actuator 22.

  The sensors 24 are connected to the input modules of the PLC 26, respectively. The sensors 24 may be connected to the PLC 26 via a network. The sensor 24 includes a sensor that detects the driving state of the actuator 22 or the operating state of equipment that operates by the driving force of the actuator 22 (hereinafter also referred to as “survey target sensor”). The sensor 24 converts the driving state of the actuator 22 and other devices 28 or the operation state of the equipment into a signal and outputs the signal to the PLC 26.

  The PLC 26 grasps the driving state of the actuator 22 and other devices 28 or the operation state of the equipment based on the signal from the sensor 24 and performs various controls. In particular, the PLC 26 generates and outputs a signal for driving the actuator 22 based on an input signal from the investigation target sensor. That is, the PLC 26 controls the driving of the actuator 22 by feedback-type closed loop control based on the input signal from the investigation target sensor.

  The data collection device 30 is connected to the PLC 26 via a network. The data collection device 30 is a PC, for example, and includes a storage device 32. In FIG. 1, the storage device 32 is built in the data collection device 30. However, the storage device 32 may be provided outside the data collection device 30. The data collection device 30 collects operation result data of the production line 20 in time series and stores it in the storage device 32.

  The operation result data includes time-series data of signals (hereinafter also referred to as “sensor signals”) input from the sensor 24 to the PLC 26, and signals (hereinafter also referred to as “actuator control signals”) output from the PLC 26 to the actuator 22. )) And time series data of signals exchanged between the PLC 26 and the other device 28 (hereinafter also referred to as “other signals”).

  Next, the configuration of the survey tool 10 will be described. The investigation tool 10 is configured to be able to access the operation result data stored in the storage device 32. The investigation tool 10 is, for example, a PC, and includes an arithmetic device 12, a cache memory 14, and a main memory 16.

  The arithmetic device 12 reads a part of the operation result data in the storage device 32 and transfers it to the cache memory 14 to perform a predetermined process. However, the arithmetic unit 12 may transfer all or part of the operation result data in the storage device 32 to the main memory 16 and further transfer a part thereof to the cache memory 14 to perform predetermined processing.

  The main memory 16 stores at least a correspondence table created as a result of predetermined processing by the arithmetic device 12. The correspondence table shows the correspondence between actuator control signals and sensor signals. In the renewal work of the production line 20, the repair and maintenance of the various facilities of the actuator 22 and the production line 20 are performed based on this correspondence table.

  In the first embodiment, the predetermined processing by the arithmetic device 12 is performed as follows. First, the arithmetic unit 12 classifies the operation result data stored in time series into sensor signal data, actuator control signal data, and other signal data.

FIG. 2 is a diagram illustrating an example of time-series data of the actuator control signal and the sensor signal. As shown in FIG. 2, the actuator control signal data V is maintained at V ₁ from time t _h to time t _i, and changes from V ₁ to V ₂ at time t _i . Data V is kept at V ₂ from time t _i to time t _j and changes from V ₂ to V ₃ at time t _j . Similarly, the data V is kept at V ₃ from time t _j to time t _k and changes from V ₃ to V ₁ at time t _k . The data P of the sensor signal (i), the data Q of the sensor signal (ii), the data R of the sensor signal (iii), and the data S of the sensor signal (iv) change following the change of the data V, or It changes without following the change of data V.

  As shown in FIG. 2, it is not known which of the sensor signals (i) to (iv) corresponds to the signal of the investigation target sensor only by separating the time series data from the operation result data. Therefore, the arithmetic unit 12 selects a signal of the investigation target sensor from these candidates and performs a process for creating the correspondence table described above. A specific example of the signal selection / correspondence table creation processing will be described with reference to FIGS.

  FIG. 3 is a diagram illustrating an example of signal selection / correspondence table creation processing performed by the arithmetic device 12 in the first embodiment of the present invention. In the example shown in FIG. 3, the arithmetic unit 12 first calculates a correlation coefficient between the actuator control signal data and the sensor signal data (step S10). In calculating the correlation coefficient, the arithmetic unit 12 uses the actuator control signal data as a reference, and sets the period of one month, for example, or a period until the change of the actuator control signal data reaches a total of 100 times as one break. Cut out. Subsequently, the arithmetic device 12 calculates a correlation coefficient for each sensor signal using data within the cut-out period.

  Subsequent to step S10, the arithmetic unit 12 determines whether there is a sensor signal for which the correlation coefficient calculated in step S10 is greater than a threshold (however, threshold> 0) (step S12). The threshold is set to 0.8, for example. It is determined that the sensor signal whose correlation coefficient is equal to or less than the threshold has a low relationship with the actuator control signal. When the determination result of step S12 is negative, that is, when the correlation coefficients of all the sensor signals are determined to be equal to or less than the threshold value, the arithmetic device 12 determines that no investigation target sensor exists and ends the process. .

When the determination result of step S12 is affirmative, the arithmetic unit 12 rearranges the sensor signals in descending order of the correlation coefficient (step S14). The sensor signals to be rearranged are limited to those having a correlation coefficient larger than a threshold value. FIG. 4 is a diagram for specifically explaining the processing after step S14 in FIG. In the example shown in FIG. 4, the correlation coefficient r ₁ of the sensor signal (i) and the correlation coefficient r ₃ of the sensor signal (iii) are larger than the threshold, and the correlation coefficient r ₂ of the sensor signal (ii) and the sensor signal correlation coefficient r ₄ of (iv) is less than the threshold value. Therefore, when the process of step S14 is performed, the data R of the sensor signal (iii) is arranged next to the data P of the sensor signal (i). The data Q of the sensor signal (ii) and the data S of the sensor signal (iv) are not subject to rearrangement.

  Subsequent to step S14, the arithmetic unit 12 investigates the causal relationship between the actuator control signal and the sensor signal in the order of the sensor signals rearranged in step S14 (step S16). That is, the arithmetic unit 12 investigates the causal relationship in the order of the data P of the sensor signal (i) and the data R of the sensor signal (iii) shown in FIG. For example, the arithmetic device 12 investigates the causal relationship by determining whether or not the actuator control signal data V has changed a predetermined time before the time when the sensor signal data has changed. The signal of the investigation target sensor is considered to change with a delay with respect to the change of the actuator control signal. For this reason, if the actuator control signal data V changes a predetermined time before the time when the sensor signal data changes, it can be estimated that there is a causal relationship between the two signals.

  Subsequent to step S16, the arithmetic unit 12 determines whether or not there is a sensor signal that satisfies the causal relationship investigated in step S16 (step S18). The process of step S16 is performed in order from the sensor signal having the largest correlation coefficient. Therefore, when a sensor signal that has a causal relationship is not found in the process of step S16, the arithmetic unit 12 determines that there is no such sensor signal, and ends the process.

  When the determination result of step S18 is affirmative, the arithmetic unit 12 selects the sensor signal as a signal of the investigation target sensor and creates a correspondence table (step S20). In other words, when a sensor signal with a causal relationship is found in the investigation in step S16, a correspondence table is created based on the sensor signal data. The created correspondence table is stored in the main memory 16. In the example shown in FIG. 4, the sensor signal (i) is selected as the signal of the investigation target sensor. Accordingly, the data P of the sensor signal (i) is written in the correspondence table.

  As described above, according to the signal selection / correspondence table creation processing shown in FIG. 3, the causal relationship is investigated in descending order of the correlation coefficient with the actuator control signal, and the signal of the investigation target sensor can be selected. Accordingly, it is possible to reduce the burden of investigating the operation result data in the renewal work of the existing production line.

Embodiment 2. FIG.
Next, an investigation tool according to Embodiment 2 of the present invention will be described with reference to FIGS. Note that descriptions overlapping with the contents of the first embodiment are omitted as appropriate.

  The survey tool according to the second embodiment is different from the survey tool according to the first embodiment in the signal selection process performed by the arithmetic device 12. FIG. 5 is a diagram illustrating an example of signal selection / correspondence table creation processing performed by the arithmetic device 12 in the second embodiment of the present invention. In the example shown in FIG. 5, the arithmetic unit 12 first performs the processes of steps S10 and S12. The processing of these steps is as described in FIG.

When the determination result of step S12 is affirmative, the arithmetic unit 12 investigates the causal relationship between the actuator control signal and the sensor signal (step S22). In the first embodiment, the causal relationship is investigated in order from the sensor signal having the largest correlation coefficient in step S16 in FIG. In contrast, in the second embodiment, the causal relationship is investigated for all sensor signals having a correlation coefficient larger than the threshold value in step S22. FIG. 6 is a diagram for specifically explaining the processing after step S22 in FIG. In the example shown in FIG. 6, the correlation coefficients (that is, correlation coefficients r ₁ and r ₃ ) between the sensor signal (i) and the sensor signal (iii) are larger than the threshold value. Therefore, the arithmetic unit 12 investigates the causal relationship between the data P of the sensor signal (i) and the data R of the sensor signal (iii) shown in FIG. The causal relationship investigation method is as described in step S16 in FIG. The data Q of the sensor signal (ii) and the data S of the sensor signal (iv) are not subject to investigation.

  Subsequent to step S22, the arithmetic unit 12 determines whether there is a sensor signal that satisfies the causal relationship investigated in step S22 (step S24). When a sensor signal that satisfies the causal relationship is not found in the process of step S22, the arithmetic unit 12 determines that there is no such sensor signal, and ends the process.

  When the determination result in step S24 is affirmative, the arithmetic unit 12 determines the length of the sensor signal for which the causal relationship has been established from the time when the data of the actuator control signal changes until the time when the data of the sensor signal changes (hereinafter referred to as the sensor signal data). Simply referred to as “time difference”) (step S26). In the example shown in FIG. 6, it is determined that the data P of the sensor signal (i) and the data R of the sensor signal (iii) are both causally related to the data V of the actuator control signal. Therefore, in step S26, a time difference is calculated for the data of these sensor signals. This time difference is calculated as an average value using the number of time differences confirmed within the cutout period.

  Subsequent to step S26, the arithmetic unit 12 selects a sensor signal having the shortest time difference calculated in step S26 as a signal of the investigation target sensor, and creates a correspondence table (step S28). The created correspondence table is stored in the main memory 16. In the example shown in FIG. 6, the sensor signal (iii) is selected as the signal of the investigation target sensor. Therefore, the data R of the sensor signal (iii) is written in the correspondence table.

  In the signal selection process (see FIG. 3) of the first embodiment, the causal relationship is investigated in order from the sensor signal having the largest correlation coefficient. For this reason, once a sensor signal with a causal relationship is found, the remaining sensor signals are not investigated. Therefore, when there are a plurality of sensor signals having the same correlation coefficient, there is a possibility that a sensor signal that is not suitable as the signal of the investigation target sensor is selected. In this regard, according to the signal selection process shown in FIG. 5, the signal of the investigation target sensor is selected based on the time difference calculated in step S24. Therefore, according to the signal selection process of the second embodiment, even when there are a plurality of sensor signals having the same correlation coefficient, it is possible to select an appropriate signal for the investigation target sensor. .

Embodiment 3 FIG.
Next, an investigation tool according to Embodiment 3 of the present invention will be described with reference to FIGS. Note that description overlapping with the contents of the first and second embodiments is omitted as appropriate.

  The investigation tool according to the third embodiment is characterized in that the arithmetic device 12 performs a signal selection process that combines the signal selection processes of the first and second embodiments. FIG. 7 is a diagram for explaining an example of signal selection / correspondence table creation processing performed by the arithmetic device 12 in Embodiment 3 of the present invention. In the example illustrated in FIG. 7, the arithmetic device 12 first performs steps S <b> 10 to S <b> 18. The processing of these steps is as described in FIG. That is, the processing so far is the same as the signal selection processing in the first embodiment.

  If the determination result of step S18 is affirmative, the arithmetic unit 12 determines whether there is another sensor signal having a correlation coefficient equal to the sensor signal with which the causal relationship is established in the process of step S16 (step S16). S30). When the determination result of step S30 is negative, the arithmetic unit 12 selects the sensor signal found in step S18 as the signal of the investigation target sensor and creates a correspondence table (step S32).

  If the determination result in step S30 is affirmative, the arithmetic unit 12 determines whether or not a causal relationship is established between the other sensor signal discovered in step S30 and the actuator control signal (step S34). ). For example, the arithmetic device 12 investigates the causal relationship by determining whether or not the data of the actuator control signal has changed a predetermined time before the time when the data of the other sensor signal has changed. When the determination result in step S34 is negative, it can be estimated that the other sensor signal found in step S30 is not the signal of the investigation target sensor. Therefore, in this case, the arithmetic unit 12 performs the process of step S32.

FIG. 8 is a diagram for explaining processing after step S30 in FIG. In the example shown in FIG. 8, the correlation coefficient of the sensor signal (v) is equal to the correlation coefficient of the sensor signal (i) (that is, the correlation coefficient r ₁ ). Therefore, the arithmetic unit 12 investigates the causal relationship with respect to the data T of the sensor signal (v) as the other sensor signal described above. The causal relationship investigation method is as described in step S16 in FIG. In the example shown in FIG. 8, it is determined that the data T of the sensor signal (v) has a causal relationship with the data V of the actuator control signal.

  When the determination result of step S34 is affirmative, the arithmetic unit 12 calculates a time difference for the sensor signal in which the causal relationship is established (step S36). As described above, in the example shown in FIG. 8, it is determined that the data P of the sensor signal (i) and the data T of the sensor signal (v) are both causally related to the data V of the actuator control signal. Therefore, in step S36, a time difference is calculated for the data of these sensor signals. The time difference calculation method is as described in step S26 in FIG.

  Subsequent to step S36, the arithmetic unit 12 selects a sensor signal having the shortest time difference calculated in step S36 as a signal of the investigation target sensor, and creates a correspondence table (step S38). The created correspondence table is stored in the main memory 16. That is, the processing in steps S36 and S38 is the same as the signal selection processing in the second embodiment. In the example shown in FIG. 8, the sensor signal (v) is selected as the signal of the investigation target sensor. Accordingly, the data T of the sensor signal (v) is written in the correspondence table.

  As described in the second embodiment, in the signal selection process of the first embodiment (see FIG. 3), when a plurality of sensor signals having the same correlation coefficient exist, An unsuitable sensor signal may be selected. On the other hand, in the signal selection process of the second embodiment (see FIG. 5), there is a possibility that a sensor signal having a small correlation coefficient is selected as a signal of the investigation target sensor. In this regard, according to the signal selection process shown in FIG. 7, the time difference can be calculated only when there are a plurality of sensor signals having the same correlation coefficient and having a causal relationship. Therefore, according to the signal selection process of the third embodiment, it is possible to select a more appropriate signal as the investigation target sensor.

Embodiment 4 FIG.
Next, an investigation tool according to Embodiment 4 of the present invention will be described with reference to FIGS. Note that description overlapping with the contents of the first to third embodiments is omitted as appropriate.

  In the survey tool according to the fourth embodiment, in the signal selection process performed by the arithmetic device 12, the signal of the survey target sensor is determined by providing three segmentation periods and by majority decision based on the provisional selection results in each segmentation period. It is characterized by selecting. FIG. 9 is a diagram for explaining an example of signal selection / correspondence table creation processing performed by the arithmetic device 12 in Embodiment 4 of the present invention. In the example illustrated in FIG. 9, the arithmetic device 12 first performs the processes of steps S <b> 10 to S <b> 18. The processing of these steps is as described in FIG. That is, the processing so far is the same as the signal selection processing in the first embodiment.

  When the determination result of step S18 is affirmative, the arithmetic unit 12 determines whether or not the survey is completed in all the setting periods (step S40). For example, when one section is one month, all of the set periods are three sections (three months in total). Further, when the period until the number of changes in the actuator control signal data reaches a total of 100 times is one section, all of the set periods are three sections (total 300 times). When the determination result of step S40 is negative, the arithmetic unit 12 changes the investigation period for signal selection to the next cutout period (step S42), and performs the processes after step S10 again.

  When the determination result of step S40 is affirmative, the arithmetic unit 12 creates a correspondence table using the sensor signal data selected by majority vote (step S44). The arithmetic unit 12 compares the sensor signals in which the causal relationship is established in each investigation period. Then, the arithmetic device 12 selects the sensor signal with the largest number of sensor signal selections as the signal of the investigation target sensor, and creates a correspondence table. The created correspondence table is stored in the main memory 16.

  10 to 11 are diagrams for explaining the processing in step S44. 10 to 11 are diagrams showing examples of time-series data in sections different from the sections shown in FIG. 2 or FIG. Note that the sensor signals shown in FIGS. 10 to 11 are rearranged in the descending order of the correlation coefficient, like the sensor signals shown in FIG.

In the example shown in FIG. 10, the correlation coefficient r ₆ of the correlation coefficient r ₅ and the sensor signal of the sensor signal (iii) (i) is greater than the threshold value. Further, it is determined that the data R of the sensor signal (iii) having the correlation coefficient r ₅ (> r ₆ ) has a causal relationship with the data V of the actuator control signal. Therefore, in the example shown in FIG. 10, the sensor signal (iii) is provisionally selected as a candidate signal for the investigation target sensor. On the other hand, in the example shown in FIG. 11, larger than a correlation coefficient r ₁₀ threshold of the correlation coefficient r ₉ and the sensor signal (iii) of the sensor signal (i). Further, it is determined that the data R of the sensor signal (i) having the correlation coefficient r ₉ (> r ₁₀ ) has a causal relationship with the data V of the actuator control signal. Therefore, in the example shown in FIG. 11, the sensor signal (i) is provisionally selected as a candidate signal for the investigation target sensor.

  Assuming that the total three sections shown in FIG. 4, FIG. 10, and FIG. 11 are the above-described set periods, the sensor signal (i) selected in the sections shown in FIG. 4 and FIG. It will be elected as a signal. Therefore, according to the process of step S44, the data P of the sensor signal (i) is written in the correspondence table.

  As described above, according to the signal selection / correspondence table creation process shown in FIG. 9, it is possible to select the signal of the investigation target sensor by majority with three or more cutout periods. Therefore, it becomes possible to improve the reliability of the selection result of the signal of the investigation target sensor.

  By the way, in the said Embodiment 3, the setting period was set to 3 areas. However, since the reliability of the selection result increases as the cutout period increases, the set period may be longer than three sections. Needless to say, such a modification that increases the set period is similarly applied to the set periods of the fourth to eighth embodiments described below.

Embodiment 5. FIG.
Next, an investigation tool according to Embodiment 5 of the present invention will be described with reference to FIG. Note that description overlapping with the contents of Embodiments 1 to 4 is omitted as appropriate.

  The survey tool according to the fourth embodiment selects a signal of the survey target sensor by a majority vote based on the signal selection process of the first embodiment. On the other hand, the survey tool according to the fifth embodiment is characterized in that the signal of the survey target sensor is selected by a majority decision based on the signal selection process of the second embodiment. FIG. 12 is a diagram for explaining an example of signal selection / correspondence table creation processing performed by the arithmetic unit 12 in the fifth embodiment of the present invention. In the example shown in FIG. 12, the arithmetic unit 12 first performs the processes of steps S10, S12, and S22 to S26. The processing of these steps is as described in FIG. That is, the processing so far is the same as the signal selection processing of the second embodiment.

  Subsequent to step S26, the arithmetic unit 12 determines whether or not the survey has been completed in the entire setting period (step S46). The process of step S46 is the same as the process of step S40 described in FIG. When the determination result of step S46 is negative, the arithmetic unit 12 changes the investigation period for signal selection to the next cutout period (step S48). The process of step S48 is the same as the process of step S42 described in FIG.

  If the determination result of step S46 is affirmative, the arithmetic unit 12 creates a correspondence table using the sensor signal data selected by majority vote (step S50). The arithmetic unit 12 compares the number of sensor signals for which a causal relationship is established in each of the investigation periods. Then, the arithmetic unit 12 selects the sensor signal data having the largest number of sensor signals as the signal of the investigation target sensor, and creates a correspondence table. The created correspondence table is stored in the main memory 16.

  As described above, according to the signal selection / correspondence table creation process shown in FIG. 12, the same effects as those obtained by the signal selection process of the fourth embodiment can be obtained.

Embodiment 6 FIG.
Next, an investigation tool according to Embodiment 6 of the present invention will be described with reference to FIGS. 7 and 12. Note that description overlapping with the contents of the first to fifth embodiments is omitted as appropriate.

  The survey tool according to the fourth embodiment selects a signal of the survey target sensor by a majority vote based on the signal selection process of the first embodiment. The survey tool according to the fifth embodiment selects the signal of the survey target sensor by majority vote based on the signal selection process of the second embodiment. On the other hand, the investigation tool according to the sixth embodiment is characterized in that the signal of the investigation target sensor is selected by a majority decision based on the signal selection processing of the third embodiment.

  The process of step S38 shown in FIG. 7 is replaced with the process of steps S46 to S50 described in FIG. Then, the signal of the investigation target sensor can be selected by a majority decision based on the signal selection process of the third embodiment.

Embodiment 7 FIG.
Next, with reference to FIG. 13, the investigation tool according to Embodiment 7 of the present invention will be described. Note that description overlapping with the contents of the first to sixth embodiments is omitted as appropriate.

  The investigation tools according to Embodiments 4 to 6 handle the sensor signal data in the cut-out period in the same row in the signal selection process. On the other hand, the survey tool according to the seventh embodiment is characterized in that the selection result of the signal of the survey target sensor is weighted by giving superiority or inferiority to the cutout period. FIG. 13 is a diagram illustrating an example of signal selection / correspondence table creation processing performed by the arithmetic device 12 in the seventh embodiment of the present invention. In the example illustrated in FIG. 13, the arithmetic device 12 first performs steps S <b> 10 to S <b> 42. The processing of these steps is as described in FIG. That is, the processing so far is the same as the signal selection processing in the fourth embodiment.

  When the determination result of step S40 is affirmative, the arithmetic unit 12 weights the selection result of the cutout period in accordance with the superiority or inferiority of the cutout period (step S52). For example, when time-series data is cut out with one month as one section, a larger weight is given to the selection result of a new cutting-out period in terms of time. That is, weighting is performed on the assumption that data in a new cutout period is more reliable in terms of time. In addition, when time-series data is cut out by the cumulative number of changes in the actuator control signal data, a large weight is given to the selection result of the cut-out period with a larger number of included data. In other words, the extraction period with a larger number of data is weighted as having higher data reliability.

  Subsequent to step S52, the arithmetic unit 12 creates a correspondence table using the sensor signal data selected by majority vote (step S54). The process of step S54 is the same as the process of step S44 described in FIG.

  As described above, according to the signal selection / correspondence table creation process shown in FIG. Therefore, even when there is a specific situation in the cutout period, it is possible to select a more appropriate signal as the investigation target sensor in consideration thereof.

  By the way, in FIG. 13, it demonstrated as an example the weighting based on the signal selection process of the said Embodiment 4. FIG. However, it is needless to say that the weighting of the seventh embodiment can be similarly applied to the selection result of the clipping period based on the signal selection processing of the fifth and sixth embodiments.

Embodiment 8 FIG.
Next, an investigation tool according to Embodiment 8 of the present invention will be described with reference to FIG. Note that description overlapping with the contents of Embodiments 1 to 7 is omitted as appropriate.

  In the said Embodiment 7, the universality with respect to all the cutting-out periods was ensured by attaching | subjecting superiority or inferiority to the cutting-out period with the method set beforehand. The eighth embodiment is characterized in that this superiority or inferiority is given according to the number of sensor signals selected in each cutout period. FIG. 14 is a diagram illustrating an example of the weighting setting process performed by the arithmetic device 12 in the eighth embodiment of the present invention. 14 is assumed to be executed prior to the signal selection / correspondence table creation process shown in FIG. 13 or simultaneously with the signal selection / correspondence table creation process. In the example illustrated in FIG. 14, the arithmetic device 12 first performs the processes of steps S56 to S66. The processing in steps S56 to S62 is the same as the processing in steps S10, S12, S22, and S24 described in FIG. The processes in steps S64 and S66 are the same as the processes in steps S40 and S42 described with reference to FIG.

  When the determination result of step S64 is affirmative, the arithmetic unit 12 determines whether or not the number of sensor signals for which a causal relationship is established in the determination of step S62 is different during the cutout period (step S68). If the determination result in step S68 is negative, it can be determined that there is no superiority or inferiority in the cutout period. Therefore, the arithmetic unit 12 does not set weighting for the cutout period (step S70). On the other hand, if the determination result in step S68 is affirmative, a weighting is set for the cutout period so that a larger weighting is performed as the number of sensor signals having a causal relationship is larger (step S72).

  As described above, according to the weighting setting process shown in FIG. 14, the weighting can be set in consideration of the actual data within the clipping period. Therefore, even when the specific data is included in the data within the cut-out period, it is possible to select a more appropriate signal as the investigation target sensor in consideration thereof.

DESCRIPTION OF SYMBOLS 10 ... Investigation tool 12 ... Arithmetic unit 14 ... Cache memory 16 ... Main memory 20 ... Production line 22 ... Actuator 24 ... Sensor 26 ... PLC
32 ... Storage device

Claims (6)

An actuator driven by a control signal and a sensor for detecting an operation state of the actuator or an operation state of equipment operating by the driving force of the actuator are accessible to a storage device storing plant operation result data An operational performance data survey tool,
The operation performance data includes time-series data of the control signal, and time-series data serving as candidate sensor signals output from the sensor,
The operation performance data survey tool is
Using the time-series data in a predetermined period of the operation performance data, calculate a correlation coefficient between the sensor signal candidate and the control signal,
It is determined whether or not a causal relationship is established between the sensor signal candidate and the control signal in the descending order of the correlation coefficient for candidates having the correlation coefficient larger than the threshold among the sensor signal candidates. ,
When it is determined that the causal relationship is established, the sensor signal candidate that satisfies the causal relationship is used to create a correspondence table between the control signal and the sensor signal. Operation performance data survey tool. An actuator driven by a control signal and a sensor for detecting an operation state of the actuator or an operation state of equipment operating by the driving force of the actuator are accessible to a storage device storing plant operation result data An operational performance data survey tool,
The operation performance data includes time-series data of the control signal, and time-series data serving as candidate sensor signals output from the sensor,
The operation performance data survey tool is
Using the time-series data in a predetermined period of the operation performance data, calculate a correlation coefficient between the sensor signal candidate and the control signal,
Determining whether or not a causal relationship is established between the sensor signal candidate and the control signal for all candidates of the sensor signal candidate for which the correlation coefficient is greater than a threshold;
When it is determined that the causal relationship is established, among the candidate sensor signals for which the causal relationship is established, a candidate having the shortest time difference between the change of the control signal and the change of the sensor signal candidate is used. An operation performance data survey tool configured to create a correspondence table between the control signal and the sensor signal. An actuator driven by a control signal and a sensor for detecting an operation state of the actuator or an operation state of equipment operating by the driving force of the actuator are accessible to a storage device storing plant operation result data An operational performance data survey tool,
The operation performance data includes time-series data of the control signal, and time-series data serving as candidate sensor signals output from the sensor,
The operation performance data survey tool is
Using the time-series data in a predetermined period of the operation performance data, calculate a correlation coefficient between the sensor signal candidate and the control signal,
It is determined whether or not a causal relationship is established between the sensor signal candidate and the control signal in the descending order of the correlation coefficient for candidates having the correlation coefficient larger than the threshold among the sensor signal candidates. ,
When it is determined that the causal relationship is established, among the candidates for the sensor signal, other candidates for which the success or failure of the causal relationship has not been determined, and the sensor signal candidate that has been determined that the causal relationship is established Determining whether there are other candidates equal in the correlation coefficient;
If there are other candidates, determine whether a causal relationship is established between the other candidates and the control signal,
When it is determined that a causal relationship is established between the other candidate and the control signal, the sensor signal is changed after the control signal is changed among all candidates of the sensor signal determined to be established. An operation result data survey tool configured to create a correspondence table between the control signal and the sensor signal using a candidate having the shortest time difference until the signal candidate changes. The predetermined period is at least three periods set in advance,
The operation performance data survey tool creates the correspondence table using the candidate having the largest number of selections when the sensor signal candidates that satisfy the causal relationship are selected in the at least three periods. It is comprised, The operation performance data investigation tool of any one of Claims 1 thru | or 3 characterized by the above-mentioned.   The said operation performance data investigation tool is comprised so that the weighting set beforehand with respect to each of the said at least 3 period may be performed, and the said selection number may be evaluated. Operation performance data survey tool. The operation performance data survey tool is
In each of the at least three periods, the sensor signal in which a causal relationship is established between the sensor signal candidate and the control signal for all candidates of the sensor signal that have a correlation coefficient greater than a threshold value. Calculate the number of
5. The configuration according to claim 4, wherein the selection number is evaluated by performing weighting according to the number of the sensor signals for which the causal relationship is established for each of the at least three periods. Operation performance data survey tool described in 1.

Images