A Remote Control Obstacle Analysis Method Based on a Hierarchical Control Model

Abstract

The present invention provides a remote control obstacle analysis method based on a hierarchical control model, including: realizing model description of systems, devices, terminals, and signals in various remote control processes by establishing a hierarchical control model, digitally modeling the abstract electrical information model, training the model using a neural network algorithm based on the cause-and-effect relationship of control obstacles in a known actual control system, obtaining the topological correlation relationship between the control obstacle input and output information models of the hierarchical control model, applying the model to the actual control process, obtaining the actual control message through a message capture unit and parsing the key information of the message, and based on the key information of the message and the topological correlation relationship between the control obstacle input and output information models of the hierarchical control model, real-time outputting the cause of the control obstacle when the control process is interrupted, quickly realizing the analysis and positioning of the cause of the control obstacle, which is of great significance for improving the economic and safety benefits of the work such as the operation, commissioning, acceptance, switching operation, and one-key sequence control operation of the substation.

Description

Technical Field

[0001] The present invention relates to the technical field of substation protection control, and specifically to a remote control obstacle analysis method based on a hierarchical control model. Background Art

[0002] At present, the degree of intelligence of intelligent substations is increasing day by day, and more complex and advanced functions can be realized, such as four remote functions and remote control (remote switching of setting areas, modifying settings, putting into and withdrawing soft pressure plates, etc.) functions. The integrated operation of regulation and control requires unattended operation of substations. The control operation parts of primary and secondary equipment in the station are executed at the regulation and control master station. The telecontrol equipment needs to send a large amount of real-time information of primary and secondary equipment to the regulation and control master station in a timely, correct and reliable manner, and at the same time receive the remote control commands sent from the regulation and control master station. The amount of telecontrol data exchanged between the master station and the substation has increased sharply, and the forms of interaction are also more diverse. The deployment of the remote "one-key sequence control" function has doubled the amount of remote control services, and also brought higher requirements for control obstacle analysis. During the remote operation process, when a control obstacle occurs, there is a lack of effective technical means to quickly locate it, and it takes a long time to discover the problem. Summary of the Invention

[0003] The purpose of the present invention is to provide a remote control obstacle analysis method based on a hierarchical control model. By analyzing various typical control messages of intelligent substations and combining the configuration information of configuration files such as SCD files, the core information in the control messages is obtained, and the systems, equipment, terminals and information involved in the remote control process of intelligent substations are modeled to form a hierarchical control model. Based on this model, the general laws and characteristics of obstacles in the remote control process are analyzed, and combined with the characteristics of intelligent substation control services, through a neural network algorithm, automatic analysis and positioning of control obstacles in the remote control process of intelligent substations are realized.

[0004] To achieve the above purpose, the present invention adopts the following technical solutions:

[0005] A remote control obstacle analysis method based on a hierarchical control model includes the following steps:

[0006] Step 1, establish a hierarchical control model, and decompose the remote control process into three types of control models: a control master station, a control unit, and a control terminal. Among them, the control master station is the control process initiation unit, the control unit is the control instruction receiving, processing, and forwarding unit in the control process, and the control terminal is the final execution unit in the control process;

[0007] Step 2, establish information models for the three types of control models of the control master station, the control unit, and the control terminal. The information models of the control models include control type, control method, control permission signal, control locking signal, control feedback signal, control time, and control feedback signal time;

[0008] Step 3: Use a hierarchical control model to describe the devices, systems, terminals, and signals in the actual control system;

[0009] Step 4: Based on the causal relationships between some known control obstacle causes and results in the actual control system, form model training samples, and train them through a neural network algorithm to form an association relationship between specific control obstacle cause and result information models;

[0010] Step 5: For the actual control process, monitor the message information during the control process, parse it into the corresponding information model in the control model, and calculate the control obstacle causes in the actual control process based on the association relationship between the control obstacle cause and result information models obtained through neural network algorithm training;

[0011] Step 6: Analyze whether the actual control obstacle causes obtained in Step 5 meet the expectations. When they meet the expectations, no new neural network algorithm training is performed. When they do not meet the expectations, the data in the actual control process is added to the model training samples, and new neural network algorithm training is performed on the association relationship between the control obstacle cause and result information models until the obtained control obstacle causes meet the expectations;

[0012] Step 7: Repeat Steps 4 - 6 to perform iterative training on the known control obstacles and the control obstacle processes occurring in the actual control process, obtain the topological association relationship between the control obstacle input and output information models of the hierarchical control model, and store this topological association relationship.

[0013] Step 8: Apply the stored topological association relationship between the control obstacle input and output information models to the actual control system, and capture the messages in the control system in real time through the deployed message capture unit, including control selection messages, control execution messages, control feedback messages, control permission signal status, and action messages exchanged between the control master station, control unit, and control terminal, and output the control obstacle causes in real time when the control process is interrupted.

[0014] Further, the specific content of the hierarchical control model further includes:

[0015] The complete control process includes three levels: control master station + control unit + control terminal. The actual control process includes several different control master stations, control units, and control terminals;

[0016] The control master station issues control instructions with specific control type attributes to the control unit. The control unit judges whether the control permission signal is satisfied, executes the next-level control instruction, and simultaneously feeds back control feedback signals and action signals to the upper-level control master station or control unit;

[0017] The control terminal receives the control instructions issued by the control unit, determines whether the control permission signal is satisfied, executes the control outlet action, and at the same time feeds back the action signal to the upper-level control unit;

[0018] There may be multiple control instructions between the control master station and the control unit, and between different control units;

[0019] The factors affecting the control process include the control permission signal, the action signal, and the feedback signal. If any of the control permission signal, the action signal, or the feedback signal does not meet the conditions during the whole process, control obstacles will occur;

[0020] In case of control obstacles during a single control process, it is only necessary to locate the control permission signal, the action signal, and the feedback signal to determine the fault point.

[0021] Furthermore, the control master station includes:

[0022] The substation monitoring system is used to monitor and control the primary and secondary equipment information of the substation;

[0023] The one-key sequence control host is used to implement the one-key sequence control function of the substation;

[0024] The remote dispatching monitoring system is used to monitor and control the primary and secondary equipment information of regional substations;

[0025] The centralized control master station is used to monitor and control the primary and secondary equipment information of regional substations.

[0026] Furthermore, the control unit includes:

[0027] The communication network shutdown device is used to forward the data of teleinformation, telemetry, and telecontrol information;

[0028] The protection device is used to upload the information of pressure plates, control words, and setting values, respond to telecontrol, and make the functions effective;

[0029] The measuring and control device is used to upload the information of pressure plates, control words, setting values, and telecontrol information, respond to telecontrol, and make the functions effective.

[0030] Furthermore, the control terminal includes:

[0031] The intelligent terminal is used to execute the outlet of the control instruction and upload the action result signal.

[0032] Furthermore, step 3 specifically includes the following steps:

[0033] Step 3.1, classify the equipment, systems, and terminals in the actual control system and correspond them to the control master station, control unit, and control terminal in the hierarchical control model respectively;

[0034] Step 3.2: Classify the control-related signals contained in the devices, systems, and terminals in the actual control system, and respectively correspond them to the information models of the three types of control models: the control master station, the control unit, and the control terminal.

[0035] Step 3.3: According to the process of Step 3.1 - Step 3.2, respectively realize the modeling description of the control process models of the substation monitoring system, the substation one-key sequence control host, the remote dispatching monitoring system, and the centralized control master station.

[0036] Step 3.4: Perform digital processing on the control process that has completed the modeling description of the control process, so that the information in the model can be used for mathematical calculations.

[0037] Furthermore, Step 4 specifically includes the following steps:

[0038] Step 4.1: Analyze the corresponding control process result situations that will occur when actual problems such as incorrect configuration of the control master station, non-satisfaction of the synchronization conditions of the control master station, non-satisfaction of the interlock conditions of the control master station, incorrect configuration of the control unit, non-satisfaction of the synchronization conditions of the control unit, non-satisfaction of the interlock conditions of the control unit, non-satisfaction of the remote control of the control unit, incorrect configuration of the control terminal, non-satisfaction of the interlock conditions of the control terminal, interruption of the control terminal link, non-satisfaction of the remote control of the control terminal, and withdrawal of the control terminal output pressure plate occur in the actual control system.

[0039] Step 4.2: According to the analyzed control obstacle causes and results, form model training samples, and the model training samples include specific causes and results, excluding their correlation relationships.

[0040] Step 4.3: Use the neural network algorithm, take the result of the control execution as the input layer of the neural network, take the control enable signal, action signal, and feedback signal during the control execution process as the output result of the neural network, take the topological correlation relationship between the result of the control execution and the control enable signal, action signal, and feedback signal during the control execution process as the hidden layer of the neural network, set the number of iterative training times and the target result threshold, and train the specific causes and results reflected in the samples.

[0041] Step 4.4: When the number of training times or the target result threshold meets the expectation, terminate the training process, and form the correlation relationship between the control obstacle causes and the result information model of the specific model.

[0042] Furthermore, the topological correlation relationship between the control obstacle input and output information models of the hierarchical control model describes the correlation relationship between the control enable signal, action signal, and feedback signal conditions and the control execution result at any link in the entire control process, specifically manifested as the set of control enable signals, action signals, and feedback signals corresponding when the entire control process of a single control process is executed normally, partially successfully, or completely fails.

[0043] Further, the specific content of the neural network algorithm includes:

[0044] Taking the result of control execution as the input layer of the neural network, taking the control permission signal, action signal, and feedback signal during the control execution process as the output result of the neural network, and taking the topological association relationship between the result of control execution and the control permission signal, action signal, and feedback signal during the control execution process as the hidden layer of the neural network.

[0045] Further, the specific functions of the message capture unit include:

[0046] Reading communication messages in the network link through network mirroring, including IEC104 messages, MMS messages, and GOOSE messages;

[0047] The message capture unit also has the function of message parsing, parsing the type identifier, transmission reason, ASDU address, information body address, message type, and message time information of IEC104 messages, parsing the message type, control object, control signal reference, and message time information of MMS messages, and parsing the control block reference, maintenance flag, control point, and message time information of GOOSE messages.

[0048] The present invention has the following beneficial effects compared with the prior art:

[0049] 1. The present invention provides a method for control model modeling and remote control obstacle analysis application, which can efficiently and quickly realize the location analysis of the reasons for remote control obstacles in substations;

[0050] 2. The present invention improves the work efficiency of substation operation and maintenance, repair, acceptance, switching operation, one-key sequence control operation, etc., and is of great significance for improving the economic and safety benefits of substation commissioning and acceptance work. Brief Description of the Drawings

[0051] Figure 1 It is a schematic diagram of the hierarchical control model of the present invention;

[0052] Figure 2 It is a schematic diagram of the deployment of the message capture unit of the present invention;

[0053] Figure 3 It is a flowchart of one embodiment of a method for remote control obstacle analysis based on a hierarchical control model of the present invention. Detailed Embodiments

[0054] To facilitate the understanding and implementation of the present invention by those of ordinary skill in the art, the present invention will be further described in detail below in conjunction with implementation examples. It should be understood that the implementation examples described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0055] As Figure 1 shown, the present invention provides a hierarchical control model, and its system composition mainly includes:

[0056] The complete control process includes three levels: the control master station + the control unit + the control terminal. In the actual control process, there are several different control master stations, control units, and control terminals;

[0057] Among them, the control master station includes: a substation monitoring system for realizing the monitoring and control of the primary and secondary equipment information of the substation; a one-key sequence control host for realizing the one-key sequence control function of the substation; a remote dispatching monitoring system for realizing the monitoring and control of the primary and secondary equipment information of regional substations; a centralized control master station for realizing the monitoring and control of the primary and secondary equipment information of regional substations; and other master station systems that can realize the monitoring and control of the primary and secondary equipment information of specific equipment, intervals, substations, and regional substations.

[0058] The control unit includes: a communication gateway for realizing the data forwarding of telemetry, telecontrol, and remote signaling information; a protection device for realizing the uploading of pressure plate, control word, and setting value information, remote control response, and function activation; a measurement and control device for realizing the uploading of pressure plate, control word, setting value, and remote control information, remote control response, and function activation; and other devices that can realize the uploading of monitoring information, control instruction response, control instruction forwarding, and control logic judgment.

[0059] The control terminal includes: an intelligent terminal for realizing the outlet execution of control instructions, realizing the uploading of action result signals, and other devices or terminals that can realize the outlet execution of control instructions and the uploading of action result signals.

[0060] The control master station issues control instructions with specific control type attributes to the control unit. The control unit judges whether the control permission signal is satisfied, executes the next-level control instructions, and at the same time feeds back the control feedback signal and action signal to the upper-level control master station or control unit;

[0061] The control terminal receives the control instructions issued by the control unit, judges whether the control permission signal is satisfied, executes the control outlet action, and at the same time feeds back the action signal to the upper-level control unit;

[0062] There may be multiple control instructions between the control master station and the control unit and between different control units;

[0063] The factors affecting the control process include the control permission signal, action signal, and feedback signal. If any one of the control permission signal, action signal, and feedback signal in the whole process does not meet the conditions, control obstacles will occur;

[0064] In the event of a control obstacle in a single control process, it is only necessary to locate the control permission signal, action signal, and feedback signal to determine the fault point.

[0065] According to the above model, combined with Figure 2 the example shown, such as Figure 3 the process shown, an embodiment of the present invention provides a remote control obstacle analysis method based on a hierarchical control model, including the following steps:

[0066] Step 1, establish a hierarchical control model, and decompose the remote control process into three types of control models: a control master station, a control unit, and a control terminal. Among them, the control master station is the control process initiation unit, the control unit is the control instruction receiving, processing, and forwarding unit in the control process, and the control terminal is the final execution unit of the control process;

[0067] Step 2, establish information models for the three types of control models of the control master station, control unit, and control terminal. The information model of the control model includes control type, control method, control permission signal, control locking signal, control feedback signal, control time, and control feedback signal time;

[0068] Step 3, use the hierarchical control model to describe the equipment, systems, terminals, and signals in the actual control system.

[0069] The specific content of Step 3 includes:

[0070] Step 3-1, classify the equipment, systems, and terminals in the actual control system, and respectively correspond them to the control master station, control unit, and control terminal in the hierarchical control model.

[0071] Step 3-2, classify the control-related signals included in the equipment, systems, and terminals in the actual control system, and respectively correspond them to the information models of the three types of control models of the control master station, control unit, and control terminal.

[0072] Step 3-3, according to the process of Step 3-1 - Step 3-2, respectively realize the modeling description of the control process models of monitoring systems such as substation monitoring systems, substation one-key sequence control hosts, remote dispatching monitoring systems, and centralized control master stations.

[0073] Step 3-4, perform digital processing on the control process that has completed the modeling description of the control process, so that the information in the model can be used for mathematical calculations. The various information models in the control process model can be represented by digital IDs as shown in Table 1, so that different information models in the control process model can be directly used in the mathematical formula operations of neural network algorithms.

[0074] Table 1 Model replaced with digital ID

[0075]

[0076] Step 4: Based on the known control obstacle causes and results in the actual control system, form model training samples, and train them through algorithms such as neural networks to establish the association relationship between the control obstacle cause and result information models of a specific model. Specifically, it includes:

[0077] Step 4-1: Analyze the corresponding control process result situations that occur when actual problems such as incorrect configuration of the control master station, non-satisfaction of the synchronization conditions of the control master station, non-satisfaction of the interlock conditions of the control master station, incorrect configuration of the control unit, non-satisfaction of the synchronization conditions of the control unit, non-satisfaction of the interlock conditions of the control unit, non-satisfaction of the remote operation of the control unit, incorrect configuration of the control terminal, non-satisfaction of the interlock conditions of the control terminal, interruption of the control terminal link, non-satisfaction of the remote operation of the control terminal, and withdrawal of the control terminal outlet pressure plate occur in the actual control system.

[0078] Step 4-2: Based on the analyzed control obstacle causes and results, form model training samples, which include specific causes and results but do not include their association relationships.

[0079] Step 4-3: Use neural network algorithms such as LSTM (Long Short-Term Memory) and BP (error backpropagation). Take the result of control execution as the input layer of the neural network, take the control permission signal, action signal, and feedback signal during the control execution process as the output results of the neural network, and take the topological association relationship between the result of control execution and the control permission signal, action signal, and feedback signal during the control execution process as the hidden layer of the neural network. Set the number of iterative training times and the target result threshold, and train the specific causes and results reflected in the samples.

[0080] Step 4-4: When the number of training times or the target result threshold meets the expectation, terminate the training process and establish the association relationship between the control obstacle cause and result information models of a specific model.

[0081] Step 5: For the actual control process, such as Figure 2 Deploy a message capture unit as shown, monitor the message information during the control process, parse it into the corresponding information model in the control model, and calculate the control obstacle causes in the actual control process based on the association relationship between the control obstacle cause and result information models obtained through training with the BP neural network algorithm.

[0082] Step 6. Optionally, analyze whether the actual control obstacle cause obtained from the association relationship formed in Step 5 meets the expectation. When it meets the expectation, no new BP neural network algorithm training is performed. When it does not meet the expectation, the data in the actual control process is added to the model training samples, and a new BP neural network algorithm training is performed on the association relationship between the control obstacle input and output information models until the obtained control obstacle cause meets the expectation.

[0083] Step 7. Repeat Step 4 - Step 6 to perform iterative training on the known control obstacles and the control obstacle processes occurring in the actual control process, obtain the topological association relationship between the control obstacle input and output information models of the hierarchical control model, and store the topological association relationship.

[0084] Step 8. Apply the stored association relationship of the control obstacle input and output information models to the actual control system, such as Figure 2 deploy the message capture unit as shown, and read the communication messages in the network link through network mirroring. For example, it includes remote control, remote signaling, and remote measurement service messages related to the substation control process, such as IEC104 messages, MMS messages, and GOOSE messages.

[0085] The message capture unit also has the function of message parsing. For example, it parses information such as the type identifier, transmission reason, ASDU address, information body address, message type, and message time of the IEC104 message, parses information such as the message type, control object, control signal reference, and message time of the MMS message, and parses information such as the control block reference, maintenance flag, control point, and message time of the GOOSE message.

[0086] Based on the message information captured in the control system in real time by the message capture unit and the topological association relationship between the control obstacle input and output information models of the stored hierarchical control model, the control obstacle cause at the time of interruption of the control process is output in real time, realizing the remote control obstacle analysis based on the hierarchical control model.

[0087] The present invention realizes the model description of systems, devices, terminals and signals in various remote control processes by establishing a hierarchical control model, digitally models the abstract electrical information model, trains the model using a neural network algorithm based on the cause-and-effect relationship of control obstacles in a known actual control system, obtains the topological correlation relationship between the control obstacle input and output information models of the hierarchical control model, applies the model to the actual control process, obtains the actual control message through the message capture unit and parses the key information of the message, and based on the key information of the message and the topological correlation relationship between the control obstacle input and output information models of the hierarchical control model, real-time outputs the cause of the control obstacle when the control process is interrupted, quickly realizes the analysis and positioning of the cause of the control obstacle, and is of great significance for improving the economic and safety benefits of the work such as the operation, commissioning, acceptance, switching operation, and one-key sequence control operation of the substation.

[0088] As described above, the above are only the specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A remote control obstacle analysis method based on a hierarchical control model, characterized in that It includes the following steps: Step 1: Establish a hierarchical control model, which decomposes the remote control process into three types of control models: the control master station, the control unit, and the control terminal. Among them, the control master station is the control process initiation unit, the control unit is the unit that receives, processes, and forwards control instructions during the control process, and the control terminal is the final execution unit of the control process; Step 2: Establish information models for the three types of control models of the control master station, the control unit, and the control terminal. The information model of the control model includes control type, control method, control permission signal, control interlock signal, control feedback signal, control time, and control feedback signal time; Step 3: Use the hierarchical control model to describe the equipment, systems, terminals, and signals in the actual control system; Step 4: Based on the known cause-and-effect relationship between some control obstacles in the actual control system, form model training samples, and train them through the neural network algorithm to form the association relationship between the specific control obstacle cause and the result information model; Step 5: For the actual control process, monitor the message information in the control process, parse it into the corresponding information model in the control model, and calculate the control obstacle cause in the actual control process based on the association relationship between the control obstacle cause and the result information model obtained by training with the neural network algorithm; Step 6: Analyze whether the actual control obstacle cause obtained in Step 5 meets the expectations. When it meets the expectations, no new neural network algorithm training is performed. When it does not meet the expectations, the data in the actual control process is added to the model training samples, and a new neural network algorithm training is performed on the association relationship between the control obstacle cause and the result information model until the obtained control obstacle cause meets the expectations; Step 7: Repeat Steps 4 - 6 to perform iterative training on the known control obstacles and the control obstacle processes that occur in the actual control process, obtain the topological association relationship between the control obstacle input and output information models of the hierarchical control model, and store this topological association relationship; Step 8: Apply the stored topological association relationship between the control obstacle input and output information models to the actual control system. By deploying a message capture unit, capture the messages in the control system in real time, including control selection messages, control execution messages, control feedback messages, control permission signal status, and action messages exchanged between the control master station, the control unit, and the control terminal, and output the control obstacle cause in real time when the control process is interrupted; The specific content of the described hierarchical control model further includes: The complete control process includes three levels: the control master station + the control unit + the control terminal. The actual control process includes several different control master stations, control units, and control terminals; The control master station issues control instructions with specific control type attributes to the control unit. The control unit judges whether the control permission signal is satisfied, executes the next-level control instruction, and at the same time feeds back the control feedback signal and the action signal to the upper-level control master station or control unit; The control terminal receives the control instruction issued by the control unit, judges whether the control permission signal is satisfied, and executes the control outlet action, and at the same time feeds back the action signal to the upper-level control unit; There are multiple control instructions between the control master station and the control unit, and between different control units; The factors affecting the control process include control enable signals, action signals, and feedback signals. If any of the control enable signals, action signals, or feedback signals in the entire process do not meet the conditions, control obstacles will occur. In the event of a control obstacle in a single control process, it is only necessary to locate the control enable signal, action signal, and feedback signal to determine the fault point.

2. The remote control obstacle analysis method based on the hierarchical control model according to claim 1, characterized in that The control master station includes: A substation monitoring system for monitoring and controlling the information of primary and secondary equipment in the substation. A one-key sequence control host for implementing the one-key sequence control function of the substation. A remote dispatching monitoring system for monitoring and controlling the information of primary and secondary equipment in regional substations. A centralized control master station for monitoring and controlling the information of primary and secondary equipment in regional substations.

3. The remote control obstacle analysis method based on the hierarchical control model according to claim 1, characterized in that The control unit includes: A communication network shutdown device for data forwarding of teleinformation, telemetry, and telecontrol information. A protection device for uploading pressure plate, control word, and setting value information, remote control response, and function activation. A measurement and control device for uploading pressure plate, control word, setting value, and remote control information, remote control response, and function activation.

4. The remote control obstacle analysis method based on a hierarchical control model according to claim 1, wherein The control terminal includes: An intelligent terminal for implementing the export execution of control instructions and uploading the action result signal.

5. The remote control obstacle analysis method based on a hierarchical control model according to claim 1, characterized in that, Step 3 specifically includes the following steps: Step 3.1: Classify the equipment, systems, and terminals in the actual control system and correspond them to the control master station, control unit, and control terminal in the hierarchical control model respectively. Step 3.2: Classify the control-related signals included in the equipment, systems, and terminals in the actual control system and correspond them to the information models of the three control models of the control master station, control unit, and control terminal respectively. Step 3.3: According to the process of Step 3.1 - Step 3.2, respectively implement the modeling description of the control process models of the substation monitoring system, substation one-key sequence control host, remote dispatching monitoring system, and centralized control master station. Step 3.4: Perform digital processing on the control process that has completed the modeling description of the control process so that the information in the model can be used for mathematical calculations.

6. The remote control obstacle analysis method based on a hierarchical control model according to claim 1, characterized in that, Step 4 specifically includes the following steps: Step 4.1: Analyze the corresponding control process result situations that will occur when there are actual problems such as incorrect configuration of the control master station, non-satisfaction of the synchronization conditions of the control master station, non-satisfaction of the interlock conditions of the control master station, incorrect configuration of the control unit, non-satisfaction of the synchronization conditions of the control unit, non-satisfaction of the interlock conditions of the control unit, non-satisfaction of the remote control of the control unit, incorrect configuration of the control terminal, non-satisfaction of the interlock conditions of the control terminal, interruption of the control terminal link, non-satisfaction of the remote control of the control terminal, and withdrawal of the control terminal outlet pressure plate in the actual control system. Step 4.2: According to the analyzed control obstacle causes and results, form model training samples, where the model training samples include specific causes and results and do not include their correlation relationships. Step 4.3: Use the neural network algorithm, set the number of iterative training times and the target result threshold, and train the specific causes and results reflected in the samples. Step 4.4: When the number of training times or the target result threshold meets the expectation, terminate the training process and form the correlation relationship between the control obstacle causes and the result information model of the specific model.

7. The remote control obstacle analysis method based on the hierarchical control model according to claim 1, wherein The topological association relationship between the control obstacle input and the output information model of the hierarchical control model describes the association relationship between the control enable signal, action signal, feedback signal conditions and the control execution result in any link of the entire control process. Specifically, it is manifested as the set of control enable signals, action signals, and feedback signals corresponding to the normal execution of all control processes, partial successful execution, and total execution failure in a single control process.

8. The remote control obstacle analysis method based on the hierarchical control model according to claim 1, wherein The specific content of the neural network algorithm includes: Taking the result of control execution as the input layer of the neural network, taking the control enable signal, action signal, and feedback signal during the control execution process as the output result of the neural network, and taking the topological association relationship between the result of control execution and the control enable signal, action signal, and feedback signal during the control execution process as the hidden layer of the neural network.

9. The remote control obstacle analysis method based on the hierarchical control model according to claim 1, characterized in that The specific functions of the message capture unit include: Reading the communication messages in the network link through network mirroring, including IEC104 messages, MMS messages, and GOOSE messages; The message capture unit also has the function of message parsing, parsing the type identifier, transmission reason, ASDU address, information body address, message type, and message time information of the IEC104 message, parsing the message type, control object, control signal reference, and message time information of the MMS message, and parsing the control block reference, maintenance flag, control point, and message time information of the GOOSE message.

Images