Automatic verification system, automatic verification method, and automatic verification program

Abstract

To automatically perform verification work for a control logic installed in a control unit.SOLUTION: The present invention relates to an automatic verification system for performing automatic verification on a control logic installed in a control unit of a control target. This system is an automatic verification system comprising: The system creates an automatic verification algorithm generation program to generate automatic verification algorithms for performing automatic verification by machine learning using training data including input operations and output results from past verification performance data and a verification performance control logic. With the automatic verification algorithm generation program, the automatic verification algorithm corresponding to a verification objective control logic is generated. The automatic verification is executed using the automatic verification algorithm created in this manner.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present disclosure relates to an automatic verification system, an automatic verification method, and an automatic verification program. [Background technology]

[0002] For example, in plant facilities such as power plants, a large number of devices are operated in cooperation under the management of a control device. The control device executes pre-installed software to realize various controls. Since the amount of program required to write such software in a general programming language would be enormous, it is written as control logic in the form of a block diagram that is easy to understand visually. For example, Patent Document 1 discloses an example of control logic in a control device for controlling a power plant. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2019-2416 Summary of the Invention [Problem to be solved by the invention]

[0004] Verification of the control logic installed in the control device is performed by performing input operations on the control logic and obtaining the output results of the control logic in response to the input operations. Conventionally, such verification work has been performed manually by an operator, but because the control devices for plant equipment handle control targets consisting of an extremely large number of components, the control logic is complex and the verification work is time-consuming.

[0005] Therefore, automation of verification work is desired to reduce the burden of verification work. For example, simple, routine verification work such as unit testing of instruments installed in plant facilities (e.g., verification testing to confirm normal operation, such as range over and alarm confirmation) can be relatively easily achieved using Class 1 automation (Robotic Process Automation (RPA)). On the other hand, operational verification testing of auxiliary equipment, valves, and other devices installed in plant facilities requires feedback from the controlled system to the control logic installed in the control device, and processes such as condition setting are also required as necessary, making it atypical verification work. Such atypical verification work is difficult to handle using the aforementioned Class 1 automation. Furthermore, verification testing of turbine start / stop in plant facilities requires the overall operation of the control logic installed in the control device, making automation difficult and unrealized.

[0006] At least one embodiment of the present disclosure has been developed in consideration of the above-mentioned circumstances, and aims to provide an automatic verification system, an automatic verification method, and an automatic verification program that can automatically perform verification work on control logic installed in a control device. [Means for solving the problem]

[0007] In order to solve the above problem, an automatic verification system according to at least one embodiment of the present disclosure includes: An automatic verification system for performing automatic verification on control logic installed in a control device to be controlled, a verification target control logic acquisition unit for acquiring a verification target control logic corresponding to the automatic verification; a verification result control logic reading unit for reading the verification result control logic; an automatic verification algorithm generation program creation unit for creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using teacher data including input operations and output results in past verification performance data and the verification performance control logic; an automatic verification algorithm generation unit for generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; an automatic verification execution unit for executing automatic verification of the control logic using the automatic verification algorithm; Equipped with.

[0008] In order to solve the above problem, an automatic verification method according to at least one embodiment of the present disclosure includes: An automatic verification method for performing automatic verification on control logic installed in a control device to be controlled, comprising: acquiring a control logic to be verified that corresponds to the automatic verification; reading verification performance control logic; a step of creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using training data including input operations and output results in past verification performance data and the verification performance control logic; generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; performing automatic verification of the control logic using the automatic verification algorithm; Equipped with.

[0009] In order to solve the above problem, an automatic verification program according to at least one embodiment of the present disclosure includes: An automatic verification program for performing automatic verification on a control logic installed in a control device to be controlled, To the computer device, acquiring a control logic to be verified that corresponds to the automatic verification; reading verification performance control logic; a step of creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using training data including input operations and output results in past verification performance data and the verification performance control logic; generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; performing automatic verification of the control logic using the automatic verification algorithm; Execute the following. [Effects of the Invention]

[0010] According to at least one embodiment of the present disclosure, it is possible to provide an automatic verification system, an automatic verification method, and an automatic verification program that can automatically perform verification of control logic installed in a control device. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a schematic configuration diagram of a control system according to an embodiment. [Figure 2] FIG. 2 is a schematic diagram showing a verification operation of the control logic of FIG. 1 by manual operation by an operator. [Figure 3] 2 is a schematic diagram showing how the control logic of FIG. 1 is automatically verified by an automatic verification system according to an embodiment. [Figure 4] FIG. 4 is a block diagram showing the configuration of the automatic verification system of FIG. 3. [Figure 5] 5 is a diagram showing an example of verification item designation information acquired by a verification item designation information acquisition unit in FIG. 4. FIG. [Figure 6] 5 is a flowchart illustrating an automatic verification method implemented by the automatic verification system of FIG. 4. [Figure 7] 5 is an example of verification result data acquired by the verification result data acquisition unit of FIG. 4. [Figure 8] 5 is a diagram illustrating an example of a definition process performed on verification result data by the preprocessing unit in FIG. 4. FIG. [Figure 9] FIG. 4 is a schematic diagram simply illustrating various signals transmitted and received during verification in the control logic and plant model of FIG. 3. [Figure 10] FIG. 5 is a conceptual diagram of machine learning performed by the machine learning unit in FIG. 4. [Figure 11] 5 is a flowchart showing a method for creating an automatic verification algorithm generation program by the automatic verification algorithm generation program creation unit of FIG. 4. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, several embodiments of the present invention will be described with reference to the accompanying drawings. However, the configurations described as the embodiments or shown in the drawings are merely illustrative examples and are not intended to limit the scope of the present invention.

[0013] First, a control system 1 to which an automatic verification system 100 according to this embodiment is applied will be described with reference to Fig. 1. Fig. 1 is a schematic configuration diagram of the control system 1 according to one embodiment. The control system 1 includes a control device emulator 2, an operator station (OPS) 4, and a simulation device 6.

[0014] The control device emulator 2 is an emulator device for virtually simulating plant equipment that is controlled based on control logic CL, which is software to be verified. Specifically, the control device emulator 2 is configured to include a plant model PM that simulates the plant equipment to be controlled, and control logic CL that controls the plant model PM.

[0015] Instead of the control device emulator 2, the control system 1 may include a control device in which the control logic CL is actually installed, and an actual plant facility that is controlled by the control device.

[0016] The operator station 4 is an operator station that has various functions related to the operation and monitoring of the control system 1. In Fig. 1, a logic management unit 8, an operation unit 10, and a status confirmation unit 12 are illustrated as components for realizing the representative functions of the operator station 4.

[0017] The logic management unit 8 is configured to manage the control logic CL or the plant model PM installed in the control device emulator 2. Specifically, the logic management unit 8 can check or change the contents of the control logic CL or the plant model PM by calling up the related control logic CL or the plant model PM from the graphic display on the display screen of the operator station 4.

[0018] The operation unit 10 is configured to enable an operator to operate the control system 1 by displaying the operation terminals for operating the control system 1 as objects as a control loop plate on the display screen of the operator station 4. For example, if the controlled object is a valve, it is possible to perform opening / closing operations and opening degree operations, and if the controlled object is an auxiliary device, it is possible to perform on / off operations of the auxiliary device.

[0019] The status confirmation unit 12 is configured to confirm the control status of the control system 1. The status confirmation unit 12 allows the operator to confirm the control status by graphically displaying the control status on the display screen of the operator station 4. The control status displayed by the status confirmation unit 12 includes an entire or partial system diagram of the control system 1, a real-time display of the process status and process values, an alarm display, a trend display, etc.

[0020] The simulation device 6 is a device for managing the operation of the control device emulator 2. The simulation device 6 can manage the operation of the control device emulator 2 by setting, for example, environmental parameters such as temperature, humidity, and pressure of the surrounding environment of the plant equipment simulated by the control device emulator 2, saving and loading the initial state, starting and stopping control calculations, changing the calculation speed, and the like.

[0021] In the control system 1 having the above configuration, verification of the control logic CL, which is software installed in the control device emulator 2, has conventionally been performed by manual operation by an operator. Here, Fig. 2 is a schematic diagram showing the verification work performed by manual operation of the operator of the control logic CL of Fig. 1. Fig. 2 shows how the verification work is performed by checking the operating status (output result) of the control logic CL in response to the operation content by the status checking unit 12 when the operator manually operates the logic management unit 8 and operation unit 10 of the operator station 4 and the simulation device 6.

[0022] However, because a control system 1 such as a plant facility uses a complex control logic CL, manually verifying the control logic CL imposes a heavy workload on the operator, and automation of the verification process is desirable. Simple routine tasks, such as unit testing of instruments in the plant facility (e.g., verification tests to confirm normal operation, such as range over and alarm confirmation), can be relatively easily achieved using Class 1 automation (Robotic Process Automation (RPA)). On the other hand, operational verification of the control logic CL, such as auxiliary equipment and valves in the plant facility, involves feedback from the controlled object (plant model PM) to the control logic CL and also involves processes such as condition setting, making it a non-routine task. Verification work that includes such non-routine tasks is difficult to achieve using Class 1 automation. Furthermore, verification work related to the start and shutdown of turbines and other components in the plant facility involves the overall operation of the control logic CL, making automation highly challenging and unsuitable for Class 1 automation (Robotic Process Automation (RPA)). These challenges can be effectively addressed by the following embodiments.

[0023] FIG. 3 is a schematic diagram showing how the automatic verification system 100 according to an embodiment performs automatic verification of the control logic CL of FIG. 1, and FIG. 4 is a block diagram showing the configuration of the automatic verification system 100 of FIG.

[0024] The automatic verification system 100 is a system for performing automatic verification following the verification work that was manually performed by an operator in Fig. 2. As shown in Fig. 3, the automatic verification system 100 performs automatic verification by providing input operations to the control logic CL or the plant model PM in place of the logic management unit 8 and operation unit 10 of the operator station 4, providing input operations to the simulation device 6, and checking the output results (operating state) from the control logic CL in place of the state checking unit 12.

[0025] As shown in FIG. 4, the automatic verification system 100 includes a verification target control logic acquisition unit 102, an automatic verification algorithm generation unit 104, an automatic verification algorithm storage unit 106, an automatic verification execution unit 108, and a verification result output unit 110.

[0026] The verification target control logic acquisition unit 102 is configured to acquire the verification target control logic PJ. The verification target control logic PJ is data related to the control logic CL, including input operations and output results corresponding to the automatic verification performed by the automatic verification system 100. The data includes at least the input elements and output elements of the control logic CL, and may further include other elements between the input elements and output elements, or may be the entire control logic CL. The method for acquiring the data related to the verification target control logic is the same as that used by the verification record control logic reading unit 120 (described in detail below), but all data related to the calculation elements in the control logic CL is acquired.

[0027] The verification item designation information acquisition unit 103 is configured to acquire verification item designation information D. The verification item designation information D is information for designating verification items in the automatic verification performed by the automatic verification system 100.

[0028] Fig. 5 is a diagram showing an example of verification item designation information D acquired by the verification item designation information acquisition unit 103 in Fig. 4. In this example, an instrument alone, a control valve alone, an auxiliary device alone, and turbine startup are prepared as candidates for verification items for automatic verification, and an operator can select a verification item by selecting the corresponding check box (Fig. 5 shows an example in which auxiliary device alone and turbine startup are designated as verification items).

[0029] The automatic verification algorithm generation unit 104 is configured to generate an automatic verification algorithm AG corresponding to the control logic PJ to be verified, using an automatic verification algorithm generation program P. The automatic verification algorithm generation program P is created by an automatic verification algorithm generation program creation unit 112, which will be described later, and is stored in advance in an automatic verification algorithm generation program storage unit 124. The automatic verification algorithm generation unit 104 can acquire the automatic verification algorithm generation program P by accessing the automatic verification algorithm generation program storage unit 124.

[0030] The automatic verification algorithm generation program P generates an automatic verification algorithm AG based on the control logic PJ to be verified and the verification item specification information D. A specific method for generating the automatic verification algorithm AG by the automatic verification algorithm generation program P will be described later, but an automatic verification algorithm AG is generated for executing automatic verification including the verification items specified by the verification item specification information D. The automatic verification algorithm AG generated by the automatic verification algorithm generation unit 104 is readably stored in the automatic verification algorithm storage unit 106.

[0031] The automatic verification execution unit 108 is configured to execute automatic verification based on the automatic verification algorithm AG stored in the automatic verification algorithm storage unit 106. By executing the automatic verification algorithm AG, the automatic verification execution unit 108 can automate the verification work that has conventionally been performed manually by an operator, as described above with reference to FIG.

[0032] The verification result output unit 110 is configured to output the results of the automatic verification executed by the automatic verification execution unit 108. There are no limitations on the output method of the verification results, but for example, the results are output in a report format that can be recognized by an operator.

[0033] Next, a description will be given of an automatic verification method implemented by the automatic verification system 100 having the above configuration. Fig. 6 is a flowchart showing the automatic verification method implemented by the automatic verification system of Fig. 4.

[0034] First, the verification target control logic PJ is acquired by the verification target control logic data acquisition unit 102 (step S100). Next, the verification item designation information acquisition unit 103 acquires the verification item designation information D (step S101). As described above with reference to Fig. 5, the verification item designation information D is information for specifying the verification items to be verified by selecting the verification items for automatic verification using a checkbox method, and is input by the operator.

[0035] Next, the automatic verification algorithm generation unit 104 accesses the automatic verification algorithm generation program storage unit 124 to read the automatic verification algorithm generation program P (step S102) and generate an automatic verification algorithm AG (step S103). The automatic verification algorithm generation program P is read in step S102 by searching the automatic verification algorithm generation program storage unit 124 using, as a condition, a model created by inputting multiple computational elements, input operations, or output results of the control logic CL corresponding to the control logic PJ to be verified. In step S103, the automatic verification algorithm AG is generated by executing the automatic verification algorithm generation program P read in step S102 using the control logic PJ to be verified obtained in step S100 and the verification target specification information D obtained in step S101.

[0036] Next, the automatic verification execution unit 108 performs automatic verification by executing the automatic verification algorithm AG generated in step S103 (step S104). The result of the automatic verification executed in step S104 is output as a verification result in a predetermined format by the verification result output unit 110 (step S105).

[0037] 4, the automatic verification system 100 also includes an automatic verification algorithm generation program creation unit 112 for creating an automatic verification algorithm generation program P used to generate the automatic verification algorithm AG as described above. The automatic verification algorithm generation program creation unit 112 includes a verification performance data acquisition unit 114, a preprocessing unit 116, a training data storage unit 118, a verification performance control logic reading unit 120, a machine learning unit 122, and an automatic verification algorithm generation program storage unit 124.

[0038] The verification result data acquisition unit 114 is configured to acquire verification result data as learning data. The verification result data is data related to past verification results, and is collected from a data log stored and managed by, for example, an event data logger (recorder) not shown.

[0039] Here, Fig. 7 shows an example of the verification result data acquired by the verification result data acquisition unit 114 in Fig. 4. In this example, the actual verification target control logic is a data log relating to the operating state of the control target acquired from an event data logger (recorder), and each event is displayed in chronological order.

[0040] The preprocessing unit 116 is configured to perform preprocessing on the verification result data acquired by the verification result data acquisition unit 114 so that the data is in a format appropriate for use as training data for machine learning. The preprocessing includes a definition process for identifying parameters corresponding to input operations and output results for the control logic CL when creating the automatic verification algorithm generation program P by machine learning, as will be described later (that is, the definition process defines which of the various data included in the verification result data will be used as the input operations and output results in automatic verification).

[0041] Fig. 8 is a diagram showing an example of definition processing performed on verification performance data by the pre-processing unit 116 in Fig. 4. In this example, "loop plate operation," "manual set," "pump run operation," "manual set of plant abnormality signal," "turbine startup operation," and "power generation start operation" are defined as input operations, and "event information," "alarm information," "pump run event," "plant abnormality alarm," "turbine startup event," and "power generation start event" are defined as output results corresponding to each of them.

[0042] 7 as the actual control logic to be verified, processing may be performed for each event included in the data log to convert it into a format that can be recognized by the machine learning algorithm by inputting information about the type of test (test type, for example, information related to the test including the test target such as a meter alone, a control valve alone, an auxiliary device alone, or turbine startup, as shown in FIG. 5) and the status of the test. The verification performance data thus preprocessed by the preprocessing unit 116 is accumulated in the training data storage unit 118 as training data.

[0043] 4, the verification result control logic reading unit 120 is configured to read the control logic CL. The control logic CL read by the verification result control logic reading unit 120 is used for machine learning in the machine learning unit 122.

[0044] The machine learning unit 122 is configured to create an automatic verification algorithm generation program P by performing machine learning based on the teacher data stored in the teacher data storage unit 118 and the control logic CL read by the verification performance control logic reading unit 120. Here, the machine learning performed by the machine learning unit 122 will be described in detail.

[0045] 9 is a schematic diagram simply illustrating various signals transmitted and received during verification between the control logic CL and the plant model PM of FIG. 3. In FIG. 9, when an operation input is made to the control logic CL from the operation unit 10, the control logic CL outputs a start command. When the start command from the control logic CL is input to the plant model PM as a model input, the plant model PM attempts to start the start target in response to the start command and outputs the result as a model output. The model output is input to the control logic CL as a start signal. The control logic CL outputs an output result (start result) in accordance with the result of receiving the start signal in this way.

[0046] In this way, in the control logic CL, feedback from the plant model PM occurs, and such a control flow is characterized by (i) an input operation from the operation unit 10, (ii) a start command output from the control logic CL to the plant model PM, and (iii) an output result output from the control logic CL. The machine learning unit 122 recognizes the control logic CL in a simplified form that is a higher-level concept by using these three characteristic points of the control logic CL as keys (hereinafter, the control logic CL recognized in such a simplified form will be referred to as a "higher-level concept control logic" as appropriate).

[0047] 10 is a conceptual diagram of machine learning by the machine learning unit 122 of Fig. 4. As described above with reference to Fig. 9, the machine learning unit 122 recognizes superordinate conceptualized control logics A to E as learning data from a large amount of accumulated training data. Then, by performing machine learning based on this training data, a verification algorithm generation program P is created for generating a verification algorithm corresponding to new superordinate conceptualized control logic F (superordinate conceptualized control logic that is not training data and corresponds to unknown control logic CL).

[0048] The automatic verification algorithm generation program P thus obtained as a learning result by the machine learning unit 122 is stored in the automatic verification algorithm generation program storage unit 124. As described above, the automatic verification algorithm generation program P stored in the automatic verification algorithm generation program storage unit 124 is read out by the automatic verification algorithm generation unit 104 and used to generate the automatic verification algorithm AG.

[0049] Next, a method for creating the automatic verification algorithm generation program P in the automatic verification algorithm generation program creation unit 112 having the above configuration will be described. Fig. 11 is a flowchart showing a method for creating the automatic verification algorithm generation program P by the automatic verification algorithm generation program creation unit 112 of Fig. 4.

[0050] First, the verification result data acquisition unit 114 acquires verification result data as learning data (step S200). Next, the preprocessing unit 116 preprocesses the verification result data acquired by the verification result data acquisition unit 114 so that the data is in a format appropriate for use as training data for machine learning (step S201). In step S201, as described above, the input data and output data are defined, and the verification result data is converted into training data. The training data obtained in the preprocessing of step S201 is stored in the training data storage unit 118 (step S202).

[0051] Next, the verification performance control logic reading unit 120 reads the verification performance control logic CL (step S203). Next, the machine learning unit 122 performs machine learning based on the training data stored in the training data storage unit 118 and the verification performance control logic CL read in step S203 (step S204), and creates an automatic verification algorithm generation program (step S205).

[0052] As described above, according to each of the above embodiments, an automatic verification algorithm generation program is created by machine learning using input operations and output results from past verification performance data as training data. The automatic verification algorithm generation program generates an automatic verification algorithm corresponding to the control logic to be verified that corresponds to the automatic verification, and the automatic verification of the control logic can be realized using the automatic verification algorithm. By generating an automatic verification algorithm using the automatic verification algorithm generation program created by machine learning in this way, it is possible to automate non-routine verification work and verification work that requires advanced judgment to operate the control logic installed in the control device as a whole.

[0053] In addition, within the scope of the present disclosure, the components in the above-described embodiments may be replaced with well-known components as appropriate, and the above-described embodiments may be combined as appropriate.

[0054] The contents described in each of the above embodiments can be understood, for example, as follows.

[0055] (1) An automated verification system according to one aspect includes: An automatic verification system for performing automatic verification on control logic installed in a control device to be controlled, a verification target control logic acquisition unit for acquiring a verification target control logic corresponding to the automatic verification; a verification result control logic reading unit for reading the verification result control logic; an automatic verification algorithm generation program creation unit for creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using teacher data including input operations and output results in past verification performance data and the verification performance control logic; an automatic verification algorithm generation unit for generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; an automatic verification execution unit for executing automatic verification of the control logic using the automatic verification algorithm; Equipped with.

[0056] According to the above aspect (1), an automatic verification algorithm generation program is created by machine learning using input operations and output results from past verification performance data as training data. The automatic verification algorithm generation program generates an automatic verification algorithm corresponding to the control logic to be verified that corresponds to the automatic verification, and the automatic verification of the control logic can be realized using the automatic verification algorithm. By generating an automatic verification algorithm using the automatic verification algorithm generation program created by machine learning in this way, it is possible to automate non-routine verification work and verification work that requires advanced judgment to operate the control logic installed in the control device as a whole.

[0057] (2) In another embodiment, in the above embodiment (1), a verification item specification information acquisition unit for acquiring verification item specification information for specifying verification items in the automatic verification; The automatic verification algorithm generation unit generates the automatic verification algorithm for performing the automatic verification using the verification item designation information as the verification items.

[0058] According to the above aspect (2), the automatic verification algorithm is created to perform automatic verification including the verification items specified by the verification item designation information (for example, a meter alone, a control valve alone, an auxiliary device alone, turbine startup, etc.). By executing automatic verification according to the automatic verification algorithm created in this way, it is possible to automatically perform verification work including the specific verification items.

[0059] (3) In another aspect, in the above aspect (1) or (2), The machine learning targets the computational elements in the control flow related to the input operation and the output result in the control logic.

[0060] According to the above aspect (3), by selecting, as the learning target for machine learning, an operational element on the control flow related to the input and output operations of automatic verification from the control logic, it is possible to reduce the computational load while improving the learning accuracy.

[0061] (4) In another embodiment, in any one of the above (1) to (3), a verification performance data acquisition unit for acquiring the verification performance data; a preprocessing unit for processing the verification result data into the training data by performing preprocessing on the verification result data; Equipped with.

[0062] According to the above aspect (4), the collected verification result data is pre-processed and processed so as to be suitable as training data to be used in machine learning.

[0063] (5) In another embodiment, in any one of the above (1) to (4), The system further comprises a verification result output unit for outputting the verification result obtained by the automatic verification.

[0064] According to the above aspect (5), the verification results obtained by executing the automatic verification can be output in a predetermined format.

[0065] (6) An automatic verification method according to one aspect includes: An automatic verification method for performing automatic verification on control logic installed in a control device to be controlled, comprising: acquiring a control logic to be verified that corresponds to the automatic verification; reading verification performance control logic; a step of creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using training data including input operations and output results in past verification performance data and the verification performance control logic; generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; performing automatic verification of the control logic using the automatic verification algorithm; Equipped with.

[0066] According to the above aspect (6), an automatic verification algorithm generation program is created by machine learning using input operations and output results from past verification performance data as training data. The automatic verification algorithm generation program generates an automatic verification algorithm corresponding to the control logic to be verified that corresponds to the automatic verification, and the automatic verification of the control logic can be realized using the automatic verification algorithm. By generating an automatic verification algorithm using the automatic verification algorithm generation program created by machine learning in this way, it is possible to automate non-routine verification work and verification work that requires advanced judgment to operate the control logic installed in the control device as a whole.

[0067] (7) An automatic verification program according to one aspect of the present invention includes: An automatic verification program for performing automatic verification on a control logic installed in a control device to be controlled, To the computer device, acquiring a control logic to be verified that corresponds to the automatic verification; reading verification performance control logic; a step of creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using training data including input operations and output results in past verification performance data and the verification performance control logic; generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; performing automatic verification of the control logic using the automatic verification algorithm; Execute the following.

[0068] According to the above aspect (7), an automatic verification algorithm generation program is created by machine learning using input operations and output results from past verification performance data as training data. The automatic verification algorithm generation program generates an automatic verification algorithm corresponding to the control logic to be verified that corresponds to the automatic verification, and the automatic verification of the control logic can be realized using the automatic verification algorithm. By generating an automatic verification algorithm using the automatic verification algorithm generation program created by machine learning in this way, it is possible to automate non-routine verification work and verification work that requires advanced judgment to operate the control logic installed in the control device as a whole. [Explanation of symbols]

[0069] 1. Control System 2. Control Unit Emulator 4 Operator Station 6 Emulator Management Device 8 Logic Management Unit 10 Control section 12 Status check section 100 Automatic Verification System 102 Verification target control logic acquisition unit 104 Automatic Verification Algorithm Generation Unit 106 Automatic verification algorithm memory 108 Automatic Verification Execution Unit 110 Verification result output section 112 Automatic Verification Algorithm Generation Program Creation Department 114 Verification performance data acquisition section 116 Pretreatment section 118 Teacher data storage unit 120 Control logic reading unit 122 Machine Learning Department 124 Automatic verification algorithm generation program memory CL control logic PJ Verification target control logic D. Verification item specification information P Automatic Verification Algorithm Generator AG Automatic Verification Algorithm

Claims

1. An automatic verification system for performing automatic verification on control logic installed in a control device to be controlled, a verification target control logic acquisition unit for acquiring a verification target control logic corresponding to the automatic verification; a verification result control logic reading unit for reading the verification result control logic; an automatic verification algorithm generation program creation unit for creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using teacher data including input operations and output results in past verification performance data and the verification performance control logic; an automatic verification algorithm generation unit for generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; an automatic verification execution unit for executing automatic verification of the control logic using the automatic verification algorithm; An automated verification system comprising:

2. a verification item specification information acquisition unit for acquiring verification item specification information for specifying verification items in the automatic verification; 2. The automatic verification system according to claim 1, wherein the automatic verification algorithm generation unit generates the automatic verification algorithm for performing the automatic verification using the verification item specification information as the verification item.

3. The automatic verification system according to claim 1 , wherein the machine learning targets computation elements in a control flow related to the input operation and the output result in the control logic.

4. a verification performance data acquisition unit for acquiring the verification performance data; a preprocessing unit for processing the verification result data into the training data by performing preprocessing on the verification result data; 3. The automatic verification system according to claim 1, comprising:

5. 3. The automatic verification system according to claim 1, further comprising a verification result output unit for outputting a verification result obtained by the automatic verification.

6. An automatic verification method for performing automatic verification on control logic installed in a control device to be controlled, comprising: acquiring a control logic to be verified that corresponds to the automatic verification; reading verification performance control logic; a step of creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using training data including input operations and output results in past verification performance data and the verification performance control logic; generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; performing automatic verification of the control logic using the automatic verification algorithm; An automatic verification method comprising:

7. An automatic verification program for performing automatic verification on a control logic installed in a control device to be controlled, To the computer device, reading verification performance control logic; a step of creating an automatic verification algorithm generation program for generating an automatic verification algorithm for performing the automatic verification by machine learning using training data including input operations and output results in past verification performance data and the verification performance control logic; generating the automatic verification algorithm corresponding to the control logic to be verified using the automatic verification algorithm generation program; performing automatic verification of the control logic using the automatic verification algorithm; An automated verification program that runs

Images