Substation Pressplate Online Monitoring System and Online Monitoring Method

Abstract

The present invention discloses an on-line monitoring system and method for substation pressure plates, belonging to the technical field of pressure plate monitoring. The system includes: a pressure plate status monitoring unit for collecting the status information of the in-panel pressure plates; a pressure plate status aggregation unit for receiving the status information of the in-panel pressure plates; a management machine for obtaining the in-panel pressure plate displacement identification information; and a master station for generating alarm information. This application solves the technical problems of misoperation and omission in traditional substation pressure plate management, as well as the large workload and poor accuracy of manual verification of pressure plate status, and achieves the technical effect of real-time monitoring of the pressure plate status through intelligent monitoring and automatic alarm, reducing human errors, and ensuring the safe operation of the power system.

Description

Technical Field

[0001] The present invention relates to the technical field of platen monitoring, and particularly to an on-line monitoring system and an on-line monitoring method for substation platens. Background Art

[0002] With the continuous development of the operation mode of the power system and the gradual expansion of the power grid scale, the substation, as a key node of the power system, undertakes the tasks of power transmission and distribution. The protection platen of the substation plays a crucial role in the operation of power equipment and can ensure the correct operation of the relay protection device. However, the traditional operation of substation platen switching mainly relies on manual operation, lacking effective monitoring and management means, which is prone to operation errors or omissions, thus affecting the normal realization of the relay protection function and even possibly causing serious consequences such as large-scale power outages in the power grid. In the traditional power system, the switching process of the platen is entirely carried out by the operation and maintenance personnel according to the work regulations. However, due to the differences in personnel responsibility and operation carefulness, the switching state of the platen often cannot be effectively monitored and verified. Errors in platen switching (such as missing switching and incorrect switching) not only increase the complexity of power system operation and maintenance but also directly affect the stability and security of the power grid. This manual operation method has great potential risks, and due to problems such as long manual inspection cycles and large workloads, its accuracy and efficiency are often not guaranteed. More importantly, the existing system lacks real-time monitoring and anti-error technical means and cannot detect and correct these problems in time. Summary of the Invention

[0003] This application provides an on-line monitoring system and an on-line monitoring method for substation platens, aiming to solve the technical problems of incorrect switching, missing switching in traditional substation platen management, as well as the large workload and poor accuracy of manual verification of platen status, and achieve the technical effect of real-time monitoring of platen status through intelligent monitoring and automatic alarm, reducing human errors, and ensuring the safe operation of the power system.

[0004] In view of the above problems, this application provides an on-line monitoring system and an on-line monitoring method for substation platens.

[0005] In the first aspect disclosed in this application, an on-line monitoring system for substation platens is provided. The system includes: a platen status monitoring unit for collecting the status information of the in-panel platens; a platen status aggregation unit for receiving the status information of the in-panel platens from the platen status monitoring unit; a management machine for receiving the corresponding information from the platen status aggregation unit and performing a status evaluation on the status information of the in-panel platens to obtain the in-panel platen displacement identification information; and a master station for receiving the in-panel platen displacement identification information from the management machine, performing an abnormal analysis on the in-panel platen displacement identification information based on the circuit operation path to obtain the abnormally displaced platens, and generating an alarm message according to the abnormally displaced platens.

[0006] Another aspect disclosed in this application provides a method for online monitoring of substation pressure plates. The method includes: collecting the status information of the pressure plates in the panel; receiving the status information of the pressure plates in the panel; performing a status assessment on the status information of the pressure plates in the panel to obtain the displacement identification information of the pressure plates in the panel; receiving the displacement identification information of the pressure plates in the panel, and based on the circuit operation path, performing an abnormal analysis on the displacement identification information of the pressure plates in the panel to obtain the abnormally displaced pressure plates, and generating an alarm message according to the abnormally displaced pressure plates.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] By using the pressure plate status monitoring unit to collect the status information of the pressure plates in the panel; using the pressure plate status aggregation unit to receive the status information of the pressure plates in the panel from the pressure plate status monitoring unit; using the management machine to receive the corresponding information from the pressure plate status aggregation unit and perform a status assessment on the status information of the pressure plates in the panel to obtain the displacement identification information of the pressure plates in the panel; using the master station to receive the displacement identification information of the pressure plates in the panel from the management machine, and based on the circuit operation path, performing an abnormal analysis on the displacement identification information of the pressure plates in the panel to obtain the abnormally displaced pressure plates, and generating an alarm message according to the abnormally displaced pressure plates. This solves the technical problems of misoperation and omission in the traditional management of substation pressure plates, as well as the large workload and poor accuracy of manual verification of the pressure plate status, and achieves the technical effect of real-time monitoring of the pressure plate status through intelligent monitoring and automatic alarm, reducing human errors, and ensuring the safe operation of the power system. The above description is only an overview of the technical solutions of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the specific embodiments of this application are given below. Brief Description of the Drawings

[0009] Figure 1 This is a schematic structural diagram of a substation pressure plate online monitoring system provided by an embodiment of this application.

[0010] Figure 2 This is a schematic flowchart of a method for online monitoring of substation pressure plates provided by an embodiment of this application.

[0011] Description of the reference numerals: Pressure plate status monitoring unit 1, Pressure plate status aggregation unit 2, Management machine 3, Master station 4. Detailed Embodiments

[0012] By providing a substation pressure plate online monitoring system and an online monitoring method, this application solves the technical problems of misoperation, missing operation in traditional substation pressure plate management, as well as the large workload and poor accuracy of manual verification of the pressure plate status, achieving the technical effect of real-time monitoring of the pressure plate status through intelligent monitoring and automatic alarm, reducing human errors, and ensuring the safe operation of the power system.

[0013] Next, the technical solutions in this application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments of this application. It should be understood that this application is not limited by the example embodiments described here. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of this application. Additionally, it should be noted that for the sake of description, only the parts related to this application are shown in the accompanying drawings rather than all of them.

[0014] Embodiment 1, as Figure 1 shown, the embodiment of this application provides a substation pressure plate online monitoring system, and this system includes:

[0015] A pressure plate status monitoring unit, which is used to collect the status information of the pressure plates inside the panel.

[0016] Specifically, the pressure plate status monitoring unit is a key component for implementing the online monitoring of substation pressure plates, responsible for real-time collecting the status information of the pressure plates in each panel cabinet in the substation. This unit detects the switch status of the pressure plates (such as "inserted" or "withdrawn") through sensors (such as photoelectric sensors or magnetic sensors) connected to the pressure plates, and transmits the collected status data to the pressure plate status aggregation unit through the data bus for further processing and analysis. The above method can achieve precise monitoring of the substation pressure plate status, greatly improve the automation degree of pressure plate operation, and reduce the risk of human operation errors.

[0017] Furthermore, the pressure plate status monitoring unit is deployed in the pressure plate panel cabinet. Any one of the pressure plate status monitoring units corresponds to a row of pressure plates inside the panel. Any one of the pressure plate status monitoring units includes a photoelectric sensor, and the photoelectric sensor corresponds to the pressure plates inside the panel one by one. The steps of collecting the status information of the pressure plates inside the panel include:

[0018] Communicating with the photoelectric sensor and receiving the photoelectric induction signal; performing a binary evaluation of the status of the pressure plates inside the panel according to the photoelectric induction signal to obtain the status information of the pressure plates inside the panel, where any one of the status information of the pressure plates inside the panel is in the inserted state or the withdrawn state.

[0019] In a preferred embodiment, the pressure plate state monitoring unit is installed in the pressure plate cabinet of the substation. Each pressure plate state monitoring unit corresponds to a row of in-cabinet pressure plates in the cabinet. Each monitoring unit includes a photoelectric inductor, and the photoelectric inductor corresponds one-to-one with the in-cabinet pressure plates, and is used to collect the state information of the in-cabinet pressure plates in real time. The working principle of the pressure plate state monitoring unit is to use photoelectric induction technology. Without direct contact with the in-cabinet pressure plates, it can accurately sense the action changes of the in-cabinet pressure plates and monitor the on-off state of the in-cabinet pressure plates in real time. Specifically, when monitoring the in-cabinet pressure plates, the pressure plate state monitoring unit will first establish a wireless communication connection with the photoelectric inductor. The photoelectric inductor transmits and receives photoelectric induction signals, and transmits the received photoelectric induction signals to the pressure plate state monitoring unit through the established wireless communication. The pressure plate state monitoring unit detects whether the pressure plate is in the input state (i.e., the pressure plate is closed) or the output state (i.e., the pressure plate is open) through these photoelectric induction signals. That is, the received photoelectric induction signals are input into a predetermined evaluation model for signal analysis to judge the state of each in-cabinet pressure plate and obtain whether the in-cabinet pressure plate is in the input state or the output state. This method does not require physical contact and uses photoelectric signals to complete the monitoring of the pressure plate state, making the monitoring process safer and more efficient, reducing the error rate of manual operations, and being able to collect accurate pressure plate state data in the first time, providing a reliable basis for subsequent monitoring, analysis and alarm.

[0020] Further, the step of obtaining the state information of the in-cabinet pressure plates by performing a binary state evaluation on the in-cabinet pressure plates according to the photoelectric induction signals includes:

[0021] Obtaining a binary state evaluation model; inputting the photoelectric induction signals into the first binary state evaluation base model of the binary state evaluation model to obtain a first binary state result; until the photoelectric induction signals are input into the Nth binary state evaluation base model of the binary state evaluation model to obtain the Nth binary state result, where N is an odd number; performing a majority full connection on the first binary state result and the Nth binary state result to obtain the state information of the in-cabinet pressure plates.

[0022] In a feasible implementation, the system terminal acquires and activates a pre-constructed state binary evaluation model. This model determines the actual state of the in-panel pressure plate by performing a binary evaluation (i.e., engaged or disengaged) on the state of the in-panel pressure plate. The base model of this evaluation model is trained using optoelectronic induction signal records and in-panel pressure plate state identifiers, providing a basis for subsequent state evaluation. During actual operation, the system terminal inputs the optoelectronic induction signal into the state binary evaluation model. The first state binary evaluation base model of the state binary evaluation model determines the state of the pressure plate based on the input optoelectronic induction signal and generates a first state binary result. At the same time, the optoelectronic induction signal is gradually evaluated by the Nth state binary evaluation base model of the state binary evaluation model, where N is an odd number. Each state binary evaluation base model performs signal analysis and state evaluation at different depths to obtain the corresponding Nth state binary result. Through multiple evaluations, the model can more accurately determine the state of the in-panel pressure plate. Finally, by performing a mode full connection on the first state binary result and the Nth state binary result, the final in-panel pressure plate state information is obtained. The role of the mode full connection is to combine multiple evaluation results and select the result with the highest frequency of occurrence as the final judgment result to improve the accuracy of the evaluation.

[0023] Further, the construction steps of the first state binary evaluation base model include:

[0024] Collect optoelectronic induction signal record data and in-panel pressure plate state identifier information. Among them, the in-panel pressure plate state identifier information is 1 or 0. 1 represents that the state of the in-panel pressure plate is the engaged state, and 0 represents the disengaged state. Using the in-panel pressure plate state identifier information as supervision and the optoelectronic induction signal record data as input, train the first state binary evaluation base model.

[0025] In a feasible implementation, in order to perform accurate state binary evaluation, the system terminal collects the optoelectronic induction signal record data and the in-panel pressure plate state identification information at historical times. The optoelectronic induction signal record data is extracted from the monitoring logs of the pressure plate state monitoring unit. These signals contain the behavior information of the in-panel pressure plate at different time points. For example, whether the in-panel pressure plate is in a closed state (engaged) or an open state (disengaged). The changes in these signal data reflect the real-time state of the in-panel pressure plate's action. The in-panel pressure plate state identification information is generated by manual detection or a standardized program. The pressure plate state at each collection point is represented by 1 for the engaged state and 0 for the disengaged state. This identification data serves as the supervision data for the optoelectronic induction signals and is the basis for subsequent model training. Subsequently, the optoelectronic induction signal record data is used as the input data, and the in-panel pressure plate state identification information is used as the supervision data. Combining with a machine learning model, a first state binary evaluation base model is constructed. Taking the multi-layer perceptron (MLP) in the neural network as an example, the system terminal uses MLP to construct a pressure plate state binary evaluation model, including an input layer, a hidden layer, an output layer, etc. Then, the number of neurons in the input layer is initialized to be the same as the dimension of the optoelectronic induction signal features, and the output layer is set to have 1 neuron, which is used to predict the state of the in-panel pressure plate (engaged or disengaged). After that, by randomly initializing the weights, a network structure with one or more hidden layers is constructed. Then, the training set divided based on the optoelectronic induction signal record data and the in-panel pressure plate state identification information is input into the pressure plate state binary evaluation model for forward propagation. The data enters from the input layer, passes through the calculations and activation functions (such as ReLU) of each layer for layer-by-layer transmission, reaches the output layer, and calculates the prediction probability of whether the in-panel pressure plate is in the engaged state or the disengaged state. The prediction result is mapped between 0 and 1 through the Sigmoid activation function, which is used to represent the binary classification result of the pressure plate state. Then, the binary cross-entropy loss function is used to calculate the loss value between the prediction result and the in-panel pressure plate state identification. The loss function measures the accuracy of the model prediction. The smaller the loss value, the more accurate the model. Then, the backpropagation algorithm is used to calculate the gradients of the loss with respect to the weights of each layer layer by layer. Based on these gradients, the Adam optimizer is used to adjust the weights of the model. The optimization goal is to minimize the value of the loss function. This process is repeated for multiple epochs until the maximum number of iterations is reached or the change in the loss function meets the stopping condition. During the training process, the hyperparameters of the model (such as the learning rate, the number of neurons in the hidden layer, the training batch size, etc.) can be adjusted through cross-validation and other methods to improve the accuracy of the model. After the training is completed, the performance of the model is evaluated using the divided test set, and the accuracy of the model in predicting the pressure plate state (engaged or disengaged) is tested. If the accuracy meets the expected requirements, the current first state binary evaluation base model is output as the final evaluation model. If the accuracy does not meet the requirements, the hyperparameters are adjusted according to the evaluation results, and training and optimization are performed again;For the N - state binary evaluation base model, it is constructed through other machine learning models, such as support vector machines (SVMs), random forests, and gradient - boosting trees. Through the above process, the system terminal can use the optoelectronic induction signal data to train a reliable evaluation model, so as to automatically and accurately conduct a binary evaluation of the state of the in - panel pressure plate. The training of the model depends on a large amount of labeled data, enabling the system terminal to adaptively identify the state of the pressure plate during actual operation.

[0026] The pressure - plate state collection unit is used to receive the in - panel pressure - plate state information from the pressure - plate state monitoring unit.

[0027] Specifically, the pressure - plate state collection unit is an important component in the on - line monitoring of substation pressure plates. It is responsible for centrally receiving the pressure - plate state information collected by each pressure - plate state monitoring unit. Each pressure - plate state monitoring unit is responsible for monitoring the state of the pressure plates in the switchgear cabinet and sending the collected data to the pressure - plate state collection unit. Specifically, after receiving the data from multiple pressure - plate state monitoring units, the pressure - plate state collection unit will summarize and integrate these data, and centrally manage the pressure - plate state information transmitted by different monitoring units through bus communication. The pressure - plate state collection unit is not only an information receiver, but it can also be responsible for certain data caching work to ensure the accurate transmission of information, and transmit the summarized data to the management machine through bus communication or wireless WLAN for subsequent state analysis, anomaly detection, and alarm processing.

[0028] The management machine is used to receive the corresponding information from the pressure - plate state collection unit and conduct a state evaluation on the in - panel pressure - plate state information to obtain the in - panel pressure - plate displacement identification information.

[0029] Specifically, the management machine is the core control unit in the on - line monitoring of substation pressure plates. It is responsible for receiving all the pressure - plate state information from the pressure - plate state collection unit and conducting centralized processing and evaluation. Specifically, the management machine receives the pressure - plate state data monitored in each switchgear cabinet from the pressure - plate state collection unit. These data include the input or withdrawal state of each in - panel pressure plate. After receiving this information, the management machine will conduct a comprehensive evaluation of the in - panel pressure - plate state, analyze the current state of different in - panel pressure plates, and compare it with the stored in - panel pressure - plate state to determine whether there has been a displacement of the in - panel pressure plates (such as an abnormal change in the state of a certain pressure plate). Through this process, the management machine can generate the in - panel pressure - plate displacement identification information, that is, mark which pressure plates have changed their states and identify the specific positions of their displacements. This function is crucial for monitoring the normal operation of the pressure plates. Especially in unmanned substations, the automatic evaluation of the management machine can timely detect abnormalities in pressure - plate operations, provide accurate displacement identification information, and provide data support for subsequent anomaly analysis and alarm.

[0030] Furthermore, the management machine is configured to receive corresponding information from the platen state aggregation unit, perform a state assessment on the in-panel platen state information, and obtain in-panel platen displacement identification information. The execution steps include:

[0031] Obtain the stored in-panel platen state information, where the stored in-panel platen state information is pre-configured in-panel platen state information loaded from the master station; extract the set of in-panel platens where the in-panel platen state information is different from the stored in-panel platen state information; identify the in-panel platen displacement identification information for the set of in-panel platens.

[0032] In a preferred implementation, when performing the state assessment, the system terminal will obtain the stored in-panel platen state information. The stored in-panel platen state information refers to the platen state data that has been stored through preset configuration or historical data before the monitoring system runs. Usually, this data comes from the pre-configured file loaded from the master station or the standard state information saved during system initialization. When the monitoring system starts, the master station will load and provide a standard state information of the in-panel platens. This information contains the initial or expected state of all platens (for example, the platen should be in the inserted state or the withdrawn state). Usually, it is stored in the database or configuration file of the master station. The stored state information usually exists in the form of a data table or a state log, and the content includes the number, current state, historical change records, etc. of the in-panel platens. Once the system terminal obtains the current state information of the in-panel platens, it will compare these in-panel platen state information with the pre-stored stored in-panel platen state information to find the inconsistencies. That is to say, the system terminal needs to check whether the actual state of each in-panel platen is consistent with the pre-configured standard state. Through this comparison, the different parts between the two can be extracted. These different parts are the in-panel platens whose states have changed (such as the state of a certain in-panel platen changes from inserted to withdrawn). The system terminal adds these changed platens to the set of in-panel platens; according to the extracted set of in-panel platens, the system terminal will generate corresponding in-panel platen displacement identification information. This identification information not only indicates which in-panel platens have changed their states (such as changing from inserted to withdrawn), but also marks the specific location (in-panel platen number), type (misinsertion, missed insertion, etc.) and the time of the change (timestamp). Once the displacement identification information is generated, the system terminal will transmit the in-panel platen displacement identification information to the master station for analysis and alarm, helping the operation and maintenance personnel quickly locate the problem and repair or adjust the platen state in time.

[0033] The master station is configured to receive the in-panel platen displacement identification information from the management machine, perform an abnormal analysis on the in-panel platen displacement identification information based on the circuit operation path, obtain the abnormally displaced platens, and generate alarm information according to the abnormally displaced platens.

[0034] Specifically, the master station is an important component in the on-line monitoring of substation pressure plates. It is mainly responsible for receiving the in-panel pressure plate position change identification information from the management machine and further analyzing and processing these information for abnormalities. The master station first receives the in-panel pressure plate position change identification information from the management machine. These identification information indicate that the status of some pressure plates is inconsistent with the preset standard and may have changed. At this time, the master station will conduct further analysis based on the circuit operation path of the substation. The circuit operation path reflects the connection and functional relationship between each in-panel pressure plate and other components in the power system. The master station calibrates the status of the in-panel pressure plates by understanding the circuit layout and conducts abnormal analysis on the calibration results and the in-panel pressure plate position change identification information by comparison to determine the abnormally changed pressure plates and mark them; when the master station discovers abnormally changed pressure plates, the system terminal will generate alarm information according to the preset standards and rules. The alarm information will include important information such as the number of the pressure plates where the abnormality occurs, the type of change (such as misoperation, omission), and the possible risks to the safe operation of the power grid. The generated alarm information will be promptly transmitted to the operation and maintenance personnel so that they can quickly take measures for repair or adjustment. This helps to avoid power system failures caused by incorrect pressure plate operations and ensure the safe and stable operation of the substation and the power grid.

[0035] Furthermore, the master station is used to receive the in-panel pressure plate position change identification information from the management machine and conduct abnormal analysis on the in-panel pressure plate position change identification information based on the circuit operation path to obtain abnormally changed pressure plates. The execution steps include:

[0036] Communicate with the circuit operation control end to collect the circuit operation path; calibrate the status of the in-panel pressure plates according to the circuit operation path to obtain the calibration result of the in-panel pressure plate status; add the in-panel pressure plates where the calibration result of the in-panel pressure plate status is inconsistent with the in-panel pressure plate position change identification information to the abnormally changed pressure plates.

[0037] In a preferred embodiment, the circuit operation control terminal is a module responsible for managing and monitoring the operation status of the substation circuit. It provides information such as the connection relationships, status, and operation paths of each circuit in the power system. The master station establishes communication with the circuit operation control terminal and exchanges information with the control terminal through protocols (such as IEC 61850, MODBUS, etc.) to obtain the real-time operation path and topological structure of the circuit. The circuit operation path refers to the connection relationships of various devices (such as pressure plates, circuit breakers, relay protection devices, etc.) in the power system, including the physical connection of the circuit, the position of the pressure plates, the status of the circuit breakers, and the direction of the power flow, etc. The collected circuit operation path will provide a basis for the subsequent calibration of the pressure plate status within the screen. Subsequently, the system terminal inputs the circuit operation path into the pre-constructed in-screen pressure plate status calibration model for in-screen pressure plate status calibration, verifies whether the actual status of each in-screen pressure plate meets the circuit operation requirements, and generates the status calibration result of each in-screen pressure plate. This result will indicate the status that the in-screen pressure plate should be in under the current environment. Once the status calibration result of the in-screen pressure plate is obtained, it will be compared with the in-screen pressure plate displacement identification information. If the calibration result is inconsistent with the displacement identification information, it indicates that the status of the pressure plate has undergone an abnormal displacement. For example, according to the analysis of the circuit operation path, it is found that a certain in-screen pressure plate should be in the input state, but the displacement identification information obtained through pressure plate monitoring shows that the pressure plate is in the withdrawn state, which constitutes a status inconsistency. When the calibration result does not match the displacement identification information, the system terminal will identify these inconsistent pressure plates as abnormally displaced pressure plates, and these abnormally displaced pressure plates will be recorded and identified for subsequent processing and alarm use.

[0038] Further, performing in-screen pressure plate status calibration according to the circuit operation path to obtain the in-screen pressure plate status calibration result includes:

[0039] Performing neural network topology simulation based on the circuit topology diagram to obtain the neural network topology structure, where the neural network topology structure is the same as the circuit topology diagram, the input data is the circuit connectivity status, and the output data is the status information of each pressure plate node; configuring the circuit operation path record data; performing positive sample identification on the circuit operation path record data to obtain the in-screen pressure plate target status identification data; training the neural network topology structure with the in-screen pressure plate target status identification data as the supervision and the circuit operation path record data as the input to obtain the in-screen pressure plate status calibration model; inputting the circuit operation path into the in-screen pressure plate status calibration model and outputting the in-screen pressure plate status calibration result.

[0040] In a feasible implementation manner, the circuit topology diagram extracted from the circuit operation control terminal describes the connection relationships of various devices (such as pressure plates, circuit breakers, relay protection devices, etc.) in the power system. Each node in the topology diagram represents a device or a state, and the edges represent the electrical connections or current flow paths between the devices. By converting this circuit topology diagram into a neural network topology structure, the goal is to match the connection relationships of the circuit with the structure of the neural network. In this process, each circuit node will correspond to a neuron in the neural network, and the connection paths in the circuit will correspond to the connection weights between the neurons. The input data of this neural network topology structure is the circuit connection state (i.e., the obtained circuit operation path), and the output data is the state information of each pressure plate node (i.e., the target state identifier of the in-panel pressure plate); Subsequently, the system terminal extracts the circuit operation path configurations at multiple historical times from the circuit operation control terminal communication to configure the circuit operation path record data. These circuit operation paths refer to the actual current flow paths in the power system. Each circuit operation path contains the state information of devices such as connected pressure plates and circuit breakers. Then, positive sample identification is performed on the circuit operation path record data. The positive sample identification process is to calibrate the target state of each in-panel pressure plate by analyzing the circuit operation path record data in combination with the requirements of circuit operation, that is, to determine the state that each in-panel pressure plate should be in under a specific circuit operation state. For example, when the current needs to flow through a certain line, some in-panel pressure plates should be in the input state, and when the current is cut off, the in-panel pressure plates should be in the withdrawn state. Through this positive sample identification, the target state identifier data of each in-panel pressure plate is obtained, that is, the ideal state of each in-panel pressure plate under specific circuit operation conditions; After that, the target state identifier data of the in-panel pressure plate is used as the supervision signal, and the circuit operation path record data is used as the input data and input into the neural network topology structure for training. The cross-entropy loss function is used to measure the gap between the prediction result of the neural network and the actual target state, and then the backpropagation algorithm and an optimizer (such as the Adam optimizer) are used to adjust the weights of the neural network to minimize the loss function, thereby optimizing the performance of the neural network. Through training, the neural network can automatically learn the relationship between the circuit operation path and the pressure plate state, and can predict the correct state of each in-panel pressure plate according to the circuit operation situation; The trained neural network will be used as an in-panel pressure plate state calibration model for actual pressure plate state calibration tasks. The system terminal inputs the current circuit operation path into the in-panel pressure plate state calibration model, and the model will output the calibrated state of each in-panel pressure plate according to the circuit operation state, forming the in-panel pressure plate state calibration result. This method effectively combines the topology structure of the power system with the learning ability of the neural network, realizing automatic and accurate in-panel pressure plate state calibration.

[0041] In summary, the substation pressure plate online monitoring system provided by the embodiments of this application has the following technical effects:

[0042] The pressing plate status monitoring unit is used to collect the status information of the pressing plates inside the panel; the pressing plate status aggregation unit is used to receive the status information of the pressing plates inside the panel from the pressing plate status monitoring unit; the management machine is used to receive the corresponding information from the pressing plate status aggregation unit and perform a status evaluation on the status information of the pressing plates inside the panel to obtain the displacement identification information of the pressing plates inside the panel; the master station is used to receive the displacement identification information of the pressing plates inside the panel from the management machine, perform an abnormal analysis on the displacement identification information of the pressing plates inside the panel based on the circuit operation path, obtain the abnormally displaced pressing plates, and generate an alarm message according to the abnormally displaced pressing plates. Through the above steps, the technical problems of misoperation and missing operation existing in the traditional management of pressing plates in substations, as well as the large workload and poor accuracy of manual verification of the status of pressing plates, are solved, and the technical effect of real-time monitoring of the status of pressing plates through intelligent monitoring and automatic alarm, reducing human errors, and ensuring the safe operation of the power system is achieved.

[0043] Embodiment 2, based on the same inventive concept as the substation pressing plate online monitoring system in the foregoing embodiment, as Figure 2 shown, the embodiment of the present application provides a substation pressing plate online monitoring method, which includes:

[0044] Collect the status information of the pressing plates inside the panel; receive the status information of the pressing plates inside the panel; perform a status evaluation on the status information of the pressing plates inside the panel to obtain the displacement identification information of the pressing plates inside the panel; receive the displacement identification information of the pressing plates inside the panel, perform an abnormal analysis on the displacement identification information of the pressing plates inside the panel based on the circuit operation path, obtain the abnormally displaced pressing plates, and generate an alarm message according to the abnormally displaced pressing plates.

[0045] Further, the step of collecting the status information of the pressing plates inside the panel includes:

[0046] Communicate with the photoelectric inductor and receive the photoelectric induction signal; perform a binary status evaluation on the pressing plates inside the panel according to the photoelectric induction signal to obtain the status information of the pressing plates inside the panel, wherein any one of the status information of the pressing plates inside the panel is in the input state or the output state.

[0047] Further, the step of performing a binary status evaluation on the pressing plates inside the panel according to the photoelectric induction signal to obtain the status information of the pressing plates inside the panel includes:

[0048] Obtain a binary status evaluation model; input the photoelectric induction signal into the first binary status evaluation base model of the binary status evaluation model to obtain a first binary status result; until the photoelectric induction signal is input into the Nth binary status evaluation base model of the binary status evaluation model to obtain the Nth binary status result, where N is an odd number; perform a mode full connection on the first binary status result and the Nth binary status result to obtain the status information of the pressing plates inside the panel.

[0049] Further, the steps for constructing the first state binary evaluation base model include:

[0050] Collect the optoelectronic induction signal recording data and the in-panel busbar state identification information, where the in-panel busbar state identification information is 1 or 0. A value of 1 indicates that the in-panel busbar state is the input state, and a value of 0 indicates the withdrawal state. Using the in-panel busbar state identification information as the supervision and the optoelectronic induction signal recording data as the input, train the first state binary evaluation base model.

[0051] Further, obtaining the in-panel busbar displacement identification information includes:

[0052] Obtain the stored in-panel busbar state information, where the stored in-panel busbar state information is the pre-configured in-panel busbar state information loaded from the master station; extract the set of in-panel busbars that are different between the in-panel busbar state information and the stored in-panel busbar state information; identify the in-panel busbar displacement identification information for the set of in-panel busbars.

[0053] Further, the steps for obtaining the abnormally displaced busbars include:

[0054] Communicate with the circuit operation control end to collect the circuit operation path; perform in-panel busbar state calibration based on the circuit operation path to obtain the in-panel busbar state calibration result; add the in-panel busbars whose in-panel busbar state calibration results are inconsistent with the in-panel busbar displacement identification information to the abnormally displaced busbars.

[0055] Further, performing in-panel busbar state calibration based on the circuit operation path to obtain the in-panel busbar state calibration result includes:

[0056] Perform neural network topology simulation based on the circuit topology diagram to obtain the neural network topology structure, where the neural network topology structure is the same as the circuit topology diagram, the input data is the circuit connection state, and the output data is the state information of each busbar node; configure the circuit operation path recording data; perform positive sample identification on the circuit operation path recording data to obtain the in-panel busbar target state identification data; use the in-panel busbar target state identification data as the supervision and the circuit operation path recording data as the input to train the neural network topology structure to obtain the in-panel busbar state calibration model; input the circuit operation path into the in-panel busbar state calibration model and output the in-panel busbar state calibration result.

[0057] Any step of the above-described method can be stored as computer instructions or programs in an unrestricted computer memory and can be called and recognized by an unrestricted computer processor to implement any one of the methods in the embodiments of the present application, without further limitation here.

[0058] Furthermore, the first or second as described above may not only represent an order relationship, but may also represent a certain specific concept, and / or refer to the selection of multiple elements either individually or in whole. Obviously, those skilled in the art can make various changes and modifications to this application without departing from the scope of this application. Thus, if these modifications and variations of this application fall within the scope of this application and its equivalent technologies, this application intends to include these changes and variations.

Claims

1. Substation pressure plate online monitoring system, characterized by: include: The pressure plate status monitoring unit is used to collect the pressure plate status information inside the screen; A pressure plate status collection unit, configured to receive the pressure plate status information within the panel from the pressure plate status monitoring unit; The management machine is used to receive the corresponding information from the pressure plate status collection unit, and perform status evaluation on the pressure plate status information within the panel to obtain the displacement identification information of the pressure plate within the panel; The master station is configured to receive the displacement identification information of the pressure plate in the panel from the management machine, perform abnormal analysis on the displacement identification information of the pressure plate in the panel based on the circuit operation path, obtain the abnormal displacement pressure plate, and generate alarm information according to the abnormal displacement pressure plate; The method for receiving the displacement identification information of the pressure plate in the panel from the management machine, performing abnormal analysis on the displacement identification information of the pressure plate in the panel based on the circuit operation path, and obtaining the abnormal displacement pressure plate includes the following steps: Communicating with the circuit operation control terminal to collect the circuit operation path; Calibrate the state of the pressure plate inside the screen according to the circuit operation path to obtain the calibration result of the state of the pressure plate inside the screen; Add the panel pressure plate whose state calibration result is inconsistent with the panel pressure plate displacement identification information to the abnormal displacement pressure plate; The step of calibrating the state of the pressure plate in the panel according to the circuit operation path and obtaining the calibration result of the state of the pressure plate in the panel includes: Performing a neural network topology simulation based on the circuit topology diagram to obtain a neural network topology structure, wherein the neural network topology structure is the same as the circuit topology diagram, the input data is the circuit connectivity state, and the output data is the state information of each pressure plate node; Configure circuit operation path recording data; Performing positive sample identification on the circuit operation path record data to obtain target state identification data of the pressure plate within the panel; Using the target state identification data of the pressure plate in the panel as supervision and the circuit operation path record data as input, a neural network topology structure is trained to obtain a calibration model of the pressure plate state in the panel; The circuit operation path is input into the in-panel pressure plate state calibration model, and the in-panel pressure plate state calibration result is output.

2. The substation pressure plate online monitoring system according to claim 1, characterized in that: The pressure plate status monitoring unit is deployed in the pressure plate panel cabinet. Any one of the pressure plate status monitoring units corresponds to a pressure plate in a row of panels. Any one of the pressure plate status monitoring units includes a photoelectric sensor. The photoelectric sensor corresponds to a pressure plate in the panel one by one. The steps of collecting the pressure plate status information in the panel include: communicating with the photoelectric sensor and receiving a photoelectric sensing signal; A binary state evaluation is performed on the intra-screen pressure plate according to the photoelectric sensing signal to obtain the intra-screen pressure plate state information, wherein any intra-screen pressure plate state in the intra-screen pressure plate state information is an engaged state or an exited state.

3. The substation pressure plate online monitoring system according to claim 2, characterized in that: The steps of performing binary evaluation on the status of the pressure plate in the screen according to the photoelectric sensing signal and obtaining the status information of the pressure plate in the screen include: Obtain a state dichotomous assessment model; Inputting the photoelectric sensing signal into a first state binary evaluation base model of the state binary evaluation model to obtain a first state binary result; until the photoelectric sensing signal is input into the Nth state binary evaluation base model of the state binary evaluation model to obtain the Nth state binary result, wherein N is an odd number; Perform a majority full connection on the first state binary division result and the Nth state binary division result to obtain the screen pressure plate state information.

4. The substation pressure plate online monitoring system according to claim 3, characterized in that: The steps of constructing the first state binary evaluation base model include: Collecting photoelectric sensing signal recording data and screen pressure plate status identification information, wherein the screen pressure plate status identification information is 1 or 0, 1 indicates that the screen pressure plate is in the engaged state, and 0 indicates that it is in the withdrawn state; The first state binary evaluation base model is trained using the identification information of the pressure plate state in the screen as supervision and the photoelectric sensing signal recording data as input.

5. The substation pressure plate online monitoring system according to claim 1, characterized in that: The management machine is used to receive corresponding information from the pressure plate status collection unit, and perform status evaluation on the pressure plate status information within the panel to obtain the displacement identification information of the pressure plate within the panel. The execution steps include: Obtaining stored in-panel pressure plate status information, wherein the stored in-panel pressure plate status information is pre-configured in-panel pressure plate status information loaded from the master station; Extracting a different set of in-screen pressure plate states from the in-screen pressure plate state information and the stored in-screen pressure plate state information; The displacement identification information of the intra-panel pressure plate is identified for the intra-panel pressure plate set.

6. The method for online monitoring of transformer substation pressure plate is characterized by: The method is performed by the substation pressure plate online monitoring system according to any one of claims 1 to 5, comprising: Collect the status information of the pressure plate inside the screen; Receiving the status information of the pressure plate in the screen; Performing a status evaluation on the internal pressure plate status information to obtain internal pressure plate displacement identification information; Receive the displacement identification information of the pressure plate in the screen, perform abnormal analysis on the displacement identification information of the pressure plate in the screen based on the circuit operation path, obtain the abnormal displacement pressure plate, and generate alarm information according to the abnormal displacement pressure plate.

Images