A method and system for preventing misoperation of a substation remote control

Abstract

The present invention discloses a method and system for preventing misoperation of remote control of a substation, which obtains the current value, temperature value and frequency value of a monitored substation at multiple predetermined time points within a predetermined time period after the first remote control operation is performed, which are collected by a sensor group; arranges them into a current time series input vector, a temperature time series input vector and a frequency time series input vector according to the time dimension; performs time series responsiveness association coding on the current time series input vector, the temperature time series input vector and the frequency time series input vector to obtain a multi-parameter transfer association feature; and, based on the multi-parameter transfer association feature, determines whether the first remote control operation is a misoperation. In this way, abnormal situations can be identified and it can be determined whether the first remote control operation is a misoperation, so as to issue an early warning, thereby realizing real-time detection of the rationality and safety of remote control operations, and improving the accuracy and reliability of remote control operations.

Description

Technical Field

[0001] The present invention relates to the field of intelligent operation technology, and in particular to a remote control misoperation prevention method and system for a substation. Background Art

[0002] Substations are an important part of the power system, used to convert high-voltage electricity into low-voltage electricity for supply to users. During operation, substations need to be remotely controlled to monitor, control and protect the power system. Substation remote control refers to the operation of switches, circuit breakers and other equipment in substations through remote control equipment to achieve substation operation control and fault handling.

[0003] Substation remote control has the advantages of high efficiency, flexibility and safety, but there is also the risk of misoperation, which may cause damage to substation equipment, instability of the power grid and even accidents, posing a threat to the safety and stability of the power system. Therefore, how to prevent misoperation of substation remote control is an important issue. However, traditional methods to prevent misoperation mainly rely on the experience and attention of operators, but this method is subjective and limited, and cannot fully guarantee the correctness of the operation. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and system for preventing misoperation of a substation by remote control to solve the above problems.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions.

[0006] A remote control misoperation prevention method for a substation, comprising:

[0007] Acquire current values, temperature values, and frequency values ​​of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after the first remote control operation is performed;

[0008] Arranging the current values, temperature values ​​and frequency values ​​at the plurality of predetermined time points into a current timing input vector, a temperature timing input vector and a frequency timing input vector according to the time dimension;

[0009] Performing timing responsiveness association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature;

[0010] And based on the multi-parameter transfer association characteristics, determine whether the first remote control operation is an erroneous operation. In this way, it is possible to identify abnormal situations and judge whether the first remote control operation is an erroneous operation, so as to issue an early warning, thereby achieving real-time detection of the rationality and safety of the remote control operation and improving the accuracy and reliability of the remote control operation.

[0011] Preferably, performing timing responsiveness association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature includes:

[0012] By using a time series feature extractor based on a deep neural network model, feature extraction is performed on the current time series input vector, the temperature time series input vector and the frequency time series input vector to obtain a current time series multi-scale feature vector, a temperature time series multi-scale feature vector and a frequency time series multi-scale feature vector;

[0013] Calculating a responsiveness estimate of the temperature time series multi-scale feature vector relative to the current time series multi-scale feature vector to obtain a current-temperature responsiveness fusion feature matrix; and

[0014] A responsiveness transfer analysis is performed on the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer correlation feature.

[0015] Preferably, the deep neural network model is a multi-scale neighborhood feature extraction module.

[0016] Preferably, a responsiveness transfer analysis is performed on the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer association feature, including: calculating the responsiveness transfer of the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain a multi-parameter transfer association feature vector as the multi-parameter transfer association feature.

[0017] Preferably, determining whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature includes:

[0018] Passing the multi-parameter transfer associated feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether there is an abnormality in the monitored substation;

[0019] Based on the classification result, it is determined whether the first remote control operation is an erroneous operation.

[0020] Preferably, the method further includes a training step for training the temporal feature extractor based on the multi-scale neighborhood feature extraction module and the classifier.

[0021] Preferably, the training step comprises:

[0022] Acquire training data, the training data including training current values, training temperature values ​​and training frequency values ​​of the monitored substation at multiple predetermined time points within a predetermined time period after the first remote control operation is performed, and a true value of whether the monitored substation has an abnormality;

[0023] Arrange the training current values, training temperature values ​​and training frequency values ​​at the plurality of predetermined time points into a training current timing input vector, a training temperature timing input vector and a training frequency timing input vector according to the time dimension;

[0024] Passing the training current time series input vector, the training temperature time series input vector and the training frequency time series input vector through the time series feature extractor based on the multi-scale neighborhood feature extraction module to obtain a training current time series multi-scale feature vector, a training temperature time series multi-scale feature vector and a training frequency time series multi-scale feature vector;

[0025] Calculating a responsiveness estimate of the training temperature time series multi-scale feature vector relative to the training current time series multi-scale feature vector to obtain a training current-temperature responsiveness fusion feature matrix;

[0026] Calculating the responsiveness transfer of the training frequency-time series multi-scale feature vector relative to the training current-temperature responsiveness fusion feature matrix to obtain a training multi-parameter transfer associated feature vector;

[0027] Passing the training multi-parameter transfer association feature vector through the classifier to obtain a classification loss function value;

[0028] The temporal feature extractor based on the multi-scale neighborhood feature extraction module and the classifier are trained based on the classification loss function value and through the direction propagation of gradient descent, wherein, at each iteration of the training, the training multi-parameter transfer association feature vector is subjected to weight space iterative recursive directional proposal optimization.

[0029] A remote control and error-proof operation system for a substation, comprising:

[0030] A data acquisition module, used to acquire the current value, temperature value and frequency value of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after the first remote control operation is performed;

[0031] A vector arrangement module, used for arranging the current values, temperature values ​​and frequency values ​​at the plurality of predetermined time points into a current timing input vector, a temperature timing input vector and a frequency timing input vector according to the time dimension;

[0032] an association coding module, configured to perform timing responsive association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature; and

[0033] The first remote control operation determination module is used to determine whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature.

[0034] Preferably, the association coding module includes:

[0035] A feature extraction unit, configured to perform feature extraction on the current time series input vector, the temperature time series input vector, and the frequency time series input vector respectively through a time series feature extractor based on a deep neural network model to obtain a current time series multi-scale feature vector, a temperature time series multi-scale feature vector, and a frequency time series multi-scale feature vector;

[0036] a responsiveness estimation calculation unit, configured to calculate a responsiveness estimation of the temperature time series multi-scale characteristic vector relative to the current time series multi-scale characteristic vector to obtain a current-temperature responsiveness fusion characteristic matrix; and

[0037] The responsiveness transfer analysis unit is used to perform responsiveness transfer analysis on the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer correlation feature.

[0038] Preferably, the deep neural network model is a multi-scale neighborhood feature extraction module.

[0039] Compared with the prior art, the present invention discloses a method and system for preventing misoperation of a substation remote control.

[0040] ①Through remote control, the switches, circuit breakers and other equipment of the substation can be remotely controlled to achieve the distribution, switching and regulation of electric energy. For example, the closing and opening of the switch can be remotely controlled to achieve the on-off operation of the circuit to meet the operation requirements of the power system. When a substation fails, the faulty equipment can be remotely powered off or switched through remote control to isolate the fault and protect the safety of the power system. For example, when a certain equipment fails, the circuit breaker can be remotely operated to cut off the power to prevent the fault from expanding or affecting other equipment. Through remote control, the status and parameters of the substation equipment, such as current, voltage, temperature, etc., can be monitored in real time. At the same time, the equipment can be debugged and repaired to ensure the normal operation and performance optimization of the equipment;

[0041] ② The advantages of remote control operation of substations include high efficiency, flexibility and safety. It can reduce the workload and risks of manual operation and improve the accuracy and reliability of operation. At the same time, remote control operation can also realize the control of remote or inaccessible equipment, improving the convenience and flexibility of operation;

[0042] ③ In the step 110, when obtaining the current value, temperature value and frequency value of the monitored substation, it is necessary to ensure that the data collected by the sensor group is accurate and reliable. The selection and arrangement of the sensors should take into account the specific conditions and monitoring requirements of the substation. In addition, the selection of the predetermined time period should fully cover the time range in which misoperation may occur. By obtaining the current value, temperature value and frequency value at multiple predetermined time points, multi-dimensional data on the operating status of the substation can be obtained. These data can be used for subsequent time series correlation analysis to help determine whether the first remote control operation is a misoperation;

[0043] ④ In step 120, when the current value, temperature value and frequency value are arranged as a time series input vector according to the time dimension, it is necessary to ensure that the corresponding relationship of the data is accurate, the time interval of the data is consistent, and the data is arranged in chronological order. In addition, it is possible to consider standardizing or normalizing the data for subsequent analysis and processing. By arranging the data as a time series input vector according to the time dimension, the data at multiple time points can be integrated into a sequence, the time relationship of the data is retained, which is helpful for subsequent time series correlation analysis and better capture of the dynamic characteristics between the data;

[0044] ⑤ In the step 130, when performing timing responsiveness association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector, it is necessary to select an appropriate coding method and model. Commonly used methods include recurrent neural network (RNN) and convolutional neural network (CNN). Appropriate model structure and parameter settings can be selected according to specific circumstances. By performing timing responsiveness association coding on the timing input vector, transfer association features between multiple parameters can be extracted, which helps to capture the timing relationship and mutual influence between data, and provides a more comprehensive and accurate basis for determining whether the first remote control operation is an erroneous operation.

[0045] ⑥ In step 140, when determining whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature, it is necessary to establish a suitable classification model or rule determination method. The training and verification of the model requires the use of accurate label data to ensure the reliability of the judgment result. At the same time, it is necessary to set appropriate thresholds or rules according to the actual situation to balance the risks of false positives and false negatives. By determining whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature, the detection accuracy and efficiency of the erroneous operation can be improved, which helps to timely discover and correct the erroneous operation and reduce the potential risks to the substation equipment and the power system.

[0046] ⑦ By real-time monitoring of the parameter data of the substation after executing the first remote control operation, such as current value, temperature value and frequency value, and introducing data processing and analysis algorithms at the back end to perform time series analysis of the substation parameter data, the substation status and operation status are monitored based on the time series change trend of the current, temperature and frequency of the substation after executing the first remote control operation, so as to identify abnormal situations and determine whether the first remote control operation is a wrong operation, so as to issue an early warning, so as to achieve real-time detection of the rationality and safety of the remote control operation, and improve the accuracy and reliability of the remote control operation.

[0047] ⑧ The information of current and temperature can be integrated to further extract the correlation features between them, so that the relationship between current and temperature can be analyzed more comprehensively, so as to more accurately determine whether the remote control operation is a wrong operation

[0048] ⑨ Further calculate the responsiveness transfer of the frequency time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain a multi-parameter transfer correlation feature vector. By obtaining the multi-parameter transfer correlation feature vector, the correlation information between current, temperature and frequency can be comprehensively considered. In this way, the comprehensive influence between them can be more comprehensively analyzed, further improving the accuracy and reliability of remote control operation.

[0049] ⑩ Compared with the traditional rule-based method, this method based on feature vectors and classifiers can capture abnormal patterns more accurately and improve the accuracy of misoperation detection. By continuously acquiring and analyzing multi-parameter transfer-related feature vectors and performing classification judgments in real time, real-time monitoring and early warning of the substation operation status can be achieved. Once an abnormality occurs in the substation, operators can take timely measures to avoid further losses and risks. The classification results can provide operators with information about the substation status to help them make decisions and perform operations. If the classification results show that the substation is normal, the operator can continue to perform the planned operation; if the classification results show an abnormality, the operator The operator can suspend the operation and conduct further inspection and confirmation to avoid potential risks caused by misoperation. It monitors the parameter data of the substation after executing the first remote control operation in real time, such as current value, temperature value and frequency value, and introduces data processing and analysis algorithms at the back end to perform time series analysis of the substation parameter data. It monitors the status and operation of the substation based on the time series change trend of the current, temperature and frequency of the substation after executing the first remote control operation, identifies abnormal situations and determines whether the first remote control operation is a misoperation, and then issues an early warning, so as to achieve real-time detection of the rationality and safety of the remote control operation and improve the accuracy and reliability of the remote control operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1A flowchart of a method for preventing misoperation of a substation remote control provided in an embodiment of the present invention;

[0051] Figure 2 A schematic diagram of a system architecture of a method for preventing misoperation of a substation remotely controlled provided in an embodiment of the present invention;

[0052] Figure 3 A block diagram of a substation remote control anti-misoperation system provided in an embodiment of the present invention;

[0053] Figure 4 This is an application scenario diagram of a substation remote control and anti-misoperation method provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0055] A substation remote control misoperation prevention method and system, comprising:

[0056] Acquire current values, temperature values, and frequency values ​​of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after the first remote control operation is performed;

[0057] Arranging the current values, temperature values ​​and frequency values ​​at the plurality of predetermined time points into a current timing input vector, a temperature timing input vector and a frequency timing input vector according to the time dimension;

[0058] Performing timing responsiveness association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature;

[0059] And based on the multi-parameter transfer association feature, determine whether the first remote control operation is an erroneous operation, so that the abnormal situation can be identified and the first remote control operation can be judged to be an erroneous operation, so as to issue an early warning, thereby achieving real-time detection of the rationality and safety of the remote control operation and improving the accuracy and reliability of the remote control operation. The current time series input vector, the temperature time series input vector and the frequency time series input vector are time-series responsively associated and encoded to obtain the multi-parameter transfer association feature, including:

[0060] By using a time series feature extractor based on a deep neural network model, feature extraction is performed on the current time series input vector, the temperature time series input vector and the frequency time series input vector to obtain a current time series multi-scale feature vector, a temperature time series multi-scale feature vector and a frequency time series multi-scale feature vector;

[0061] Calculating a responsiveness estimate of the temperature time series multi-scale feature vector relative to the current time series multi-scale feature vector to obtain a current-temperature responsiveness fusion feature matrix; and

[0062] The frequency-time multi-scale feature vector is subjected to responsiveness transfer analysis relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer association feature. The deep neural network model is a multi-scale neighborhood feature extraction module. The frequency-time multi-scale feature vector is subjected to responsiveness transfer analysis relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer association feature, including: calculating the responsiveness transfer of the frequency-time multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain a multi-parameter transfer association feature vector as the multi-parameter transfer association feature. Based on the multi-parameter transfer association feature, determining whether the first remote control operation is an erroneous operation includes:

[0063] Passing the multi-parameter transfer associated feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether there is an abnormality in the monitored substation;

[0064] Based on the classification result, it is determined whether the first remote control operation is an erroneous operation. It also includes a training step: for training the temporal feature extractor based on the multi-scale neighborhood feature extraction module and the classifier. The training step includes:

[0065] Acquire training data, the training data including training current values, training temperature values ​​and training frequency values ​​of the monitored substation at multiple predetermined time points within a predetermined time period after the first remote control operation is performed, and a true value of whether the monitored substation has an abnormality;

[0066] Arrange the training current values, training temperature values ​​and training frequency values ​​at the plurality of predetermined time points into a training current timing input vector, a training temperature timing input vector and a training frequency timing input vector according to the time dimension;

[0067] Passing the training current time series input vector, the training temperature time series input vector and the training frequency time series input vector through the time series feature extractor based on the multi-scale neighborhood feature extraction module to obtain a training current time series multi-scale feature vector, a training temperature time series multi-scale feature vector and a training frequency time series multi-scale feature vector;

[0068] Calculating a responsiveness estimate of the training temperature time series multi-scale feature vector relative to the training current time series multi-scale feature vector to obtain a training current-temperature responsiveness fusion feature matrix;

[0069] Calculating the responsiveness transfer of the training frequency-time series multi-scale feature vector relative to the training current-temperature responsiveness fusion feature matrix to obtain a training multi-parameter transfer associated feature vector;

[0070] Passing the training multi-parameter transfer association feature vector through the classifier to obtain a classification loss function value;

[0071] The temporal feature extractor based on the multi-scale neighborhood feature extraction module and the classifier are trained based on the classification loss function value and through the direction propagation of gradient descent, wherein, at each iteration of the training, the training multi-parameter transfer association feature vector is subjected to weight space iterative recursive directional proposal optimization.

[0072] A remote control and error-proof operation system for a substation, comprising:

[0073] A data acquisition module, used to acquire the current value, temperature value and frequency value of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after the first remote control operation is performed;

[0074] A vector arrangement module, used for arranging the current values, temperature values ​​and frequency values ​​at the plurality of predetermined time points into a current timing input vector, a temperature timing input vector and a frequency timing input vector according to the time dimension;

[0075] an association coding module, configured to perform timing responsive association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature; and

[0076] The first remote control operation determination module is used to determine whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature. The association coding module includes:

[0077] A feature extraction unit, configured to perform feature extraction on the current time series input vector, the temperature time series input vector, and the frequency time series input vector respectively through a time series feature extractor based on a deep neural network model to obtain a current time series multi-scale feature vector, a temperature time series multi-scale feature vector, and a frequency time series multi-scale feature vector;

[0078] A responsiveness estimation calculation unit, used to calculate a responsiveness estimation of the temperature time series multi-scale characteristic vector relative to the current time series multi-scale characteristic vector to obtain a current-temperature responsiveness fusion characteristic matrix;

[0079] and a responsiveness transfer analysis unit for performing responsiveness transfer analysis on the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer correlation feature. The deep neural network model is a multi-scale neighborhood feature extraction module.

[0080] Substations are key components in the power system. They are used to convert high-voltage electricity into low-voltage electricity and distribute, control and protect electricity. They play an important role in the power system, ensuring the safe and reliable transmission and supply of electricity.

[0081] A substation usually consists of the following main components:

[0082] Transformer: Transformer is the core equipment of substation, which is used to convert high voltage electric energy transmitted to substation into low voltage electric energy suitable for users. Through transformer, the voltage level of electric energy can be changed to realize energy conversion between transmission and distribution.

[0083] Switchgear: Substations are equipped with various switchgear, such as circuit breakers, disconnectors and load switches, to control the on / off and distribution of electrical energy. Switchgear can switch, isolate and protect circuits to ensure the safe operation of the power system.

[0084] Protection equipment: The substation is also equipped with various protection equipment, including differential protection, overcurrent protection, overvoltage protection, etc., which are used to monitor the operating status of the power system and take protective measures when a fault or abnormal situation occurs. The protection equipment can detect and isolate faults in time to protect the safety of equipment and personnel.

[0085] Control system: The control system of the substation is used to monitor and control the operating status of the substation. It can obtain various parameters and signals in real time, and remotely control and adjust the equipment to meet the operating requirements of the power system.

[0086] Auxiliary equipment: The substation also includes some auxiliary equipment, such as power supply system, battery pack, communication equipment and monitoring instrument, etc. These equipment provide functions such as power support, data transmission and equipment status monitoring to ensure the normal operation and management of the substation.

[0087] Substation remote control operation refers to the operation of switches, circuit breakers and other equipment in the substation through remote control equipment to achieve operation control and fault handling of the substation. The operation instructions are transmitted to the substation site through wireless communication, network communication or other remote communication methods to achieve remote control operation of the equipment.

[0088] Through remote control, the switches, circuit breakers and other equipment of the substation can be remotely controlled to achieve the distribution, switching and regulation of electric energy. For example, the closing and opening of the switch can be remotely controlled to achieve the on-off operation of the circuit to meet the operation requirements of the power system. When a substation fails, the faulty equipment can be remotely powered off or switched through remote control to isolate the fault and protect the safety of the power system. For example, when a certain device fails, the circuit breaker can be remotely operated to power off to prevent the fault from expanding or affecting other equipment. Through remote control, the status and parameters of substation equipment, such as current, voltage, temperature, etc., can be monitored in real time. At the same time, equipment debugging and maintenance operations can also be carried out to ensure the normal operation and performance optimization of the equipment.

[0089] The advantages of remote control operation of substations include high efficiency, flexibility and safety, which can reduce the workload and risks of manual operation and improve the accuracy and reliability of operation. At the same time, remote control operation can also realize the control of remote or inaccessible equipment, and improve the convenience and flexibility of operation. However, factors such as operator negligence, fatigue, lack of experience or insufficient training may lead to misoperation. In order to reduce the impact of human factors, operators need to be fully trained and educated to improve their operating skills and awareness. At the same time, strict operating procedures and standards should be established to ensure the consistency and correctness of operations. Substation remote control operation usually relies on the communication system for command transmission. If there are problems such as failure, delay or interference in the communication system, misoperation may occur. In order to prevent the risk of misoperation caused by communication problems, it is necessary to ensure the reliability and stability of the communication system, carry out regular maintenance and overhaul, and establish a backup communication system to deal with possible failures. Failure of substation equipment may lead to misoperation. For example, the failure of switchgear may cause misclosing or misopening. In order to prevent misoperation caused by equipment failure, regular equipment overhaul and maintenance are required to ensure the normal operation and reliability of the equipment. At the same time, an equipment status monitoring and fault diagnosis system should be established to detect and handle equipment failures in a timely manner. In order to prevent misoperation, some security measures and verification mechanisms can be adopted. For example, a double confirmation mechanism can be set up to require operators to perform double verification before performing key operations; use technologies such as passwords, identity verification or fingerprint recognition to restrict access and operations by unauthorized personnel; and perform pre-operation confirmation and inspection before key operations to ensure the correctness of the operation.

[0090] In order to reduce the risk of misoperation, substation remote control operation needs to pay attention to the training and awareness of operators, ensure the reliability of the communication system, perform equipment maintenance and repair, and take corresponding safety measures and verification mechanisms. In addition, it is also very important to continuously monitor and improve the security of the substation remote control operation system.

[0091] In one embodiment of the present invention, Figure 1The present invention provides a flowchart of a method for preventing misoperation of a substation remote control according to an embodiment of the present invention. Figure 2 Schematic diagram of the system architecture of a method for preventing misoperation of a substation remote control provided in an embodiment of the present invention. Figure 1 and Figure 2 As shown, the method for preventing misoperation of a substation remote control according to an embodiment of the present invention includes: 110, obtaining the current value, temperature value and frequency value of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after executing a first remote control operation; 120, arranging the current value, temperature value and frequency value at the multiple predetermined time points into a current timing input vector, a temperature timing input vector and a frequency timing input vector according to the time dimension; 130, performing timing responsive association encoding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature; and, 140, determining whether the first remote control operation is a misoperation based on the multi-parameter transfer association feature.

[0092] In step 110, when obtaining the current value, temperature value and frequency value of the monitored substation, it is necessary to ensure that the data collected by the sensor group is accurate and reliable. The selection and arrangement of sensors should take into account the specific conditions and monitoring requirements of the substation. In addition, the selection of the predetermined time period should fully cover the time range in which misoperation may occur. By obtaining the current value, temperature value and frequency value at multiple predetermined time points, multi-dimensional data on the operating status of the substation can be obtained, which can be used for subsequent time series correlation analysis to help determine whether the first remote control operation is a misoperation.

[0093] In step 120, when the current value, temperature value and frequency value are arranged as a time series input vector according to the time dimension, it is necessary to ensure that the corresponding relationship of the data is accurate. Ensure that the time intervals of the data are consistent and arranged in chronological order. In addition, it is possible to consider standardizing or normalizing the data for subsequent analysis and processing. By arranging the data as a time series input vector according to the time dimension, the data at multiple time points can be integrated into a sequence, retaining the time relationship of the data, which is helpful for subsequent time series correlation analysis and better capturing the dynamic characteristics between the data.

[0094] In step 130, when performing timing responsiveness association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector, it is necessary to select an appropriate coding method and model. Commonly used methods include recurrent neural network (RNN) and convolutional neural network (CNN), and appropriate model structure and parameter settings can be selected according to specific circumstances. By performing timing responsiveness association coding on the timing input vector, the transfer association features between multiple parameters can be extracted, which helps to capture the timing relationship and mutual influence between data, and provides a more comprehensive and accurate basis for determining whether the first remote control operation is a misoperation.

[0095] In step 140, when determining whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature, it is necessary to establish a suitable classification model or rule judgment method. The training and verification of the model requires the use of accurate label data to ensure the reliability of the judgment result. At the same time, it is necessary to set appropriate thresholds or rules according to the actual situation to balance the risks of false positives and false negatives. By determining whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature, the accuracy and efficiency of detecting erroneous operations can be improved, which helps to timely discover and correct erroneous operations and reduce potential risks to substation equipment and power systems.

[0096] In response to the above technical problems, the technical concept of the present invention is to monitor the parameter data of the substation after executing the first remote control operation, such as current value, temperature value and frequency value, in real time, and introduce data processing and analysis algorithms at the back end to perform time series analysis of the substation parameter data, so as to monitor the status and operation of the substation based on the time series change trend of the current, temperature and frequency of the substation after executing the first remote control operation, identify abnormal situations and determine whether the first remote control operation is an erroneous operation, and then issue an early warning, so as to achieve real-time detection of the rationality and safety of the remote control operation and improve the accuracy and reliability of the remote control operation.

[0097] Specifically, in the technical solution of the present application, first, the current value, temperature value and frequency value of the monitored substation at multiple predetermined time points within a predetermined time period after the first remote control operation is performed are acquired by the sensor group. Next, considering that the current value, temperature value and frequency value of the monitored substation after the first remote control operation not only have their own time series dynamic change rules in the time dimension, but also have a time series collaborative correlation relationship between the current value, the temperature value and the frequency value, they jointly provide sufficient basis for the state detection of the monitored substation after the first remote control operation and the judgment of the misoperation of the first remote control operation. Therefore, in the technical solution of the present application, it is necessary to further arrange the current value, temperature value and frequency value of the multiple predetermined time points according to the time dimension into a current time series input vector, a temperature time series input vector and a frequency time series input vector, so as to respectively integrate the distribution information of the current value, the temperature value and the frequency value in the time series.

[0098] In one embodiment of the present application, the current timing input vector, the temperature timing input vector and the frequency timing input vector are subjected to timing responsiveness association encoding to obtain multi-parameter transfer association features, including: performing feature extraction on the current timing input vector, the temperature timing input vector and the frequency timing input vector respectively through a timing feature extractor based on a deep neural network model to obtain a current timing multi-scale feature vector, a temperature timing multi-scale feature vector and a frequency timing multi-scale feature vector; calculating a responsiveness estimate of the temperature timing multi-scale feature vector relative to the current timing multi-scale feature vector to obtain a current-temperature responsiveness fusion feature matrix; and performing a responsiveness transfer analysis on the frequency timing multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer association features.

[0099] Among them, the deep neural network model is a multi-scale neighborhood feature extraction module.

[0100] Then, a convolutional neural network model with excellent performance in implicit correlation feature extraction is used to perform feature mining of the current time series input vector, the temperature time series input vector and the frequency time series input vector, respectively. In particular, considering that the current value, the temperature value and the frequency value are all volatile in the time dimension, and the time series changes and fluctuations may be relatively weak, it is difficult to effectively capture and characterize the features through traditional feature extraction methods. Therefore, in the technical solution of the present invention, the current time series input vector, the temperature time series input vector and the frequency time series input vector are further subjected to feature mining in a time series feature extractor based on a multi-scale neighborhood feature extraction module, so as to extract the time series multi-scale change feature information of the current value, the temperature value and the frequency value under different time spans, respectively, so as to obtain the current time series multi-scale feature vector, the temperature time series multi-scale feature vector and the frequency time series multi-scale feature vector.

[0101] During the operation of the substation, the current and temperature are usually closely related, because the magnitude of the current is related to the load of the substation, and the increase or decrease of the load will cause the temperature of the substation to change. Therefore, in the technical solution of the present application, the responsiveness estimate of the temperature time series multi-scale feature vector relative to the current time series multi-scale feature vector is further calculated to obtain the current-temperature responsiveness fusion feature matrix. By calculating the responsiveness estimate of the temperature time series multi-scale feature vector relative to the current time series multi-scale feature vector, the sensitivity of temperature to current changes can be measured. If the temperature is very sensitive to the change of current, that is, the temperature changes rapidly with the change of current, then in the remote control operation, if it is found that the current changes greatly and the temperature changes less, it may mean that there is an abnormal situation, which may be caused by misoperation. In addition, by obtaining the current-temperature responsiveness fusion feature matrix, the information of current and temperature can be fused, and the correlation features between them can be further extracted, so that the relationship between current and temperature can be analyzed more comprehensively, so as to more accurately determine whether the remote control operation is a misoperation.

[0102] In one embodiment of the present application, a responsive transfer analysis is performed on the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer association feature, including: calculating the responsive transfer of the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain a multi-parameter transfer association feature vector as the multi-parameter transfer association feature.

[0103] During the operation of the substation, frequency is an important parameter that reflects the stability and operating status of the power system. When the frequency deviates from the standard value, the various devices in the power system may not be able to operate synchronously, thereby causing imbalance and instability in the power system, which may cause the power load to be unable to be properly supplied, and even cause a failure in the power system. In addition, it is also considered that the frequency change of the substation may be related to the change of current and temperature. For example, when the load increases, the current increases, which will cause the frequency to drop; and the change of temperature will also affect the frequency. For example, the increase in temperature will cause the loss of the equipment to increase, thereby affecting the frequency stability. Therefore, in the technical solution of the present invention, the responsiveness transfer of the frequency time series multi-scale eigenvector relative to the current-temperature responsiveness fusion feature matrix is ​​further calculated to obtain a multi-parameter transfer associated eigenvector. By obtaining the multi-parameter transfer associated eigenvector, the associated information between current, temperature and frequency can be comprehensively considered. In this way, the comprehensive influence between them can be analyzed more comprehensively, and the accuracy and reliability of remote control operation can be further improved.

[0104] In one embodiment of the present application, based on the multi-parameter transfer association feature, determining whether the first remote control operation is an erroneous operation includes: passing the multi-parameter transfer association feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether there is an abnormality in the monitored substation; and, based on the classification result, determining whether the first remote control operation is an erroneous operation.

[0105] Then, the multi-parameter transfer correlation feature vector is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the monitored substation has an abnormality. In other words, the comprehensive responsiveness time series correlation feature information between the current value, temperature value and frequency value of the monitored substation after the first remote control operation is performed is used for classification processing, so as to detect and judge whether the monitored substation has an abnormality, and based on the classification result, determine whether the first remote control operation is an erroneous operation, so as to achieve real-time detection of the rationality and safety of the remote control operation, and improve the accuracy and reliability of the remote control operation.

[0106] By classifying the multi-parameter transfer correlation feature vector through a classifier, we can get the classification result of whether the monitored substation has abnormalities. This classification result can be used to indicate whether the operation status of the substation is normal. If the classification result shows an abnormality, it may mean that the substation has a fault, equipment abnormality or other abnormal conditions, which require further inspection and processing.

[0107] Based on the classification results, it can be determined whether the first remote control operation is an erroneous operation. If the classification results show that there is an abnormality in the substation, and the execution of the first remote control operation is related to this abnormal state, it can be preliminarily determined that the first remote control operation may be an erroneous operation. This judgment based on the classification results can provide a quick preliminary screening method to help operators identify potential erroneous operation situations.

[0108] By inputting multi-parameter transfer-associated feature vectors into a classifier for classification, machine learning or statistical methods can be used to automatically identify abnormal states of substations. Compared with traditional rule-based methods, this feature vector and classifier-based method can capture abnormal patterns more accurately and improve the accuracy of misoperation detection. By continuously acquiring and analyzing multi-parameter transfer-associated feature vectors and performing classification judgments in real time, real-time monitoring and early warning of substation operation status can be achieved. Once an abnormality occurs in a substation, operators can take timely measures to avoid further losses and risks. The classification results can provide operators with information about the status of the substation to help them make decisions and perform operations. If the classification results show that the substation is normal, the operator can continue to perform the planned operation; if the classification results show an abnormality, the operator can suspend the operation and conduct further inspection and confirmation to avoid potential risks caused by misoperation.

[0109] In one embodiment of the present application, the substation remote control anti-misoperation method also includes a training step: for training the timing feature extractor based on the multi-scale neighborhood feature extraction module and the classifier. The training step includes: obtaining training data, the training data including training current values, training temperature values ​​and training frequency values ​​at multiple predetermined time points within a predetermined time period after the monitored substation performs the first remote control operation, and the real value of whether the monitored substation has an abnormality; arranging the training current values, training temperature values ​​and training frequency values ​​at the multiple predetermined time points according to the time dimension into a training current timing input vector, a training temperature timing input vector and a training frequency timing input vector; passing the training current timing input vector, the training temperature timing input vector and the training frequency timing input vector through the timing feature extractor based on the multi-scale neighborhood feature extraction module to obtain a training current timing multi-scale feature vector, a training temperature timing multi-scale feature vector and a training frequency timing multi-scale feature vector. scale feature vector; calculate the responsiveness estimate of the training temperature time series multi-scale feature vector relative to the training current time series multi-scale feature vector to obtain the training current-temperature responsiveness fusion feature matrix; calculate the responsiveness transfer of the training frequency time series multi-scale feature vector relative to the training current-temperature responsiveness fusion feature matrix to obtain the training multi-parameter transfer associated feature vector; pass the training multi-parameter transfer associated feature vector through the classifier to obtain the classification loss function value; train the time series feature extractor based on the multi-scale neighborhood feature extraction module and the classifier based on the classification loss function value and through the direction propagation of gradient descent, wherein, at each iteration of the training, the training multi-parameter transfer associated feature vector is subjected to weight space iterative recursive directional proposal optimization.

[0110] In particular, in the technical solution of the present application, the training current time series multi-scale feature vector, the training temperature time series multi-scale feature vector and the training frequency time series multi-scale feature vector respectively express the local time series correlation characteristics of the current value, the temperature value and the frequency value. Thus, the training current-temperature responsiveness fusion feature matrix obtained by calculating the responsiveness estimation of the training temperature time series multi-scale feature vector relative to the training current time series multi-scale feature vector is used to express the current-temperature feature domain response of the time series feature distribution, and the training multi-parameter transfer correlation feature vector obtained by calculating the responsiveness transfer of the training frequency time series multi-scale feature vector relative to the training current-temperature responsiveness fusion feature matrix further expresses the frequency-current temperature feature domain responsiveness transfer of the time series feature distribution. However, although the training multi-parameter transfer association feature vector basically follows the local time series association distribution, the misalignment of the current value, temperature value and frequency value in the time series direction also causes the time series feature space expression on the time series distribution of the training multi-parameter transfer association feature vector to deviate from the time series feature space expression of the current value, temperature value and frequency value respectively. This also causes the weight matrix of the classifier to be difficult to converge to the class label of the time series feature space dimension belonging to the predetermined sample when the training multi-parameter transfer association feature vector is classified by the classifier, thereby affecting the training effect of the classifier.

[0111] Therefore, when the applicant of the present application classifies the training multi-parameter transfer association feature vector through a classifier, at each iteration, the training multi-parameter transfer association feature vector is subjected to a weight space iterative recursive directional proposal optimization, which is specifically expressed as follows: the training multi-parameter transfer association feature vector is subjected to a weight space iterative recursive directional proposal optimization to obtain an optimized training multi-parameter transfer association feature vector according to the following optimization formula; wherein the optimization formula is:

[0112]

[0113]

[0114]

[0115] Among them, M1 and M2 are the weight matrices of the previous and current iterations, respectively, and V c is the training multi-parameter transfer association feature vector, represents matrix multiplication, ⊙ represents point multiplication by position, exp(·) represents vector exponential operation, and the vector exponential operation represents the calculation of the natural exponential function value raised to the power of the eigenvalues ​​at each position in the vector, V' c Represents the optimized training multi-parameter transfer association feature vector.

[0116] Here, the weight space iterative recursive directional proposal optimization can be achieved by transferring the initial training multi-parameter associated feature vector V to be classified c As an anchor point, the multi-parameter transfer correlation feature vector V corresponding to the training is iterated in the weight space based on the weight matrix c The anchor footprints under different sample time-series spatial feature distribution dimensions are obtained according to the different sample time-series spatial feature distribution directions, as oriented proposals for iterative recursion in the weight space, so as to improve the class confidence and local accuracy of the weight matrix convergence based on the prediction proposal, so as to improve the training effect of the training multi-parameter transfer association feature vector through the classifier. In this way, the state and operation of the substation can be monitored based on the time-series change trend of the current, temperature and frequency of the substation after the first remote control operation is performed, so as to identify abnormal situations and determine whether the first remote control operation is a wrong operation, so as to issue an early warning, so as to realize real-time detection of the rationality and safety of the remote control operation, and improve the accuracy and reliability of the remote control operation.

[0117] In summary, the substation remote control anti-misoperation method based on the embodiment of the present invention is explained, which monitors the parameter data of the substation after executing the first remote control operation, such as current value, temperature value and frequency value, in real time, and introduces data processing and analysis algorithms at the back end to perform time series analysis of the substation parameter data, so as to monitor the status and operation of the substation based on the time series change trend of the current, temperature and frequency of the substation after executing the first remote control operation, identify abnormal situations and determine whether the first remote control operation is an erroneous operation, and then issue an early warning, so as to achieve real-time detection of the rationality and safety of the remote control operation, and improve the accuracy and reliability of the remote control operation.

[0118] Figure 3 FIG. 1 is a block diagram of a substation remote control anti-error operation system provided in an embodiment of the present invention. Figure 3 As shown, the substation remote control anti-error operation system 200 includes: a data acquisition module 210, which is used to obtain the current value, temperature value and frequency value of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after the first remote control operation is performed; a vector arrangement module 220, which is used to arrange the current value, temperature value and frequency value of the multiple predetermined time points according to the time dimension into a current timing input vector, a temperature timing input vector and a frequency timing input vector; an association coding module 230, which is used to perform time-series responsive association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature; and a first remote control operation determination module 240, which is used to determine whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature.

[0119] In the substation remote control anti-error operation system, the association coding module includes: a feature extraction unit, which is used to perform feature extraction on the current time series input vector, the temperature time series input vector and the frequency time series input vector respectively through a time series feature extractor based on a deep neural network model to obtain a current time series multi-scale feature vector, a temperature time series multi-scale feature vector and a frequency time series multi-scale feature vector; a responsiveness estimation calculation unit, which is used to calculate the responsiveness estimation of the temperature time series multi-scale feature vector relative to the current time series multi-scale feature vector to obtain a current-temperature responsiveness fusion feature matrix; and a responsiveness transfer analysis unit, which is used to perform responsiveness transfer analysis on the frequency time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer association feature.

[0120] In the substation remote control and error-proof operation system, the deep neural network model is a multi-scale neighborhood feature extraction module.

[0121] Those skilled in the art will appreciate that the specific operations of each step in the above-mentioned substation remote control anti-misoperation system have been described in detail above. Figure 1 to Figure 2 The substation remote control anti-misoperation method has been described in detail, and therefore, its repeated description will be omitted.

[0122] As described above, the substation remote control anti-misoperation system 200 according to the embodiment of the present invention can be implemented in various terminal devices, such as a server for substation remote control anti-misoperation, etc. In one example, the substation remote control anti-misoperation system 200 according to the embodiment of the present invention can be integrated into the terminal device as a software module and / or a hardware module. For example, the substation remote control anti-misoperation system 200 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the substation remote control anti-misoperation system 200 can also be one of the many hardware modules of the terminal device.

[0123] Alternatively, in another example, the substation remote control anti-error operation system 200 and the terminal device may also be separate devices, and the substation remote control anti-error operation system 200 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0124] Figure 4 FIG. 1 is an application scenario diagram of a method for preventing misoperation of a substation remote control provided in an embodiment of the present invention. Figure 4 As shown, in this application scenario, first, the current values ​​of the monitored substation at multiple predetermined time points within a predetermined time period after the first remote control operation is performed are acquired by the sensor group (for example, Figure 4C1 as shown in the figure), temperature value (e.g. Figure 4 C2) and frequency values ​​(e.g., Figure 4 Then, the obtained current value, temperature value and frequency value are input to a server (for example, Figure 4 In S) shown in , the server is capable of processing the current value, the temperature value and the frequency value based on the substation remote control anti-misoperation algorithm to determine whether the first remote control operation is a misoperation.

[0125] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0126] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0127] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for preventing misoperation of a substation remote control, characterized in that: include: Acquire current values, temperature values, and frequency values ​​of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after the first remote control operation is performed; Arranging the current values, temperature values ​​and frequency values ​​at the plurality of predetermined time points into a current timing input vector, a temperature timing input vector and a frequency timing input vector according to the time dimension; Performing timing responsiveness association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature; and determining whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature; The current time series input vector, the temperature time series input vector and the frequency time series input vector are subjected to time series responsiveness association coding to obtain multi-parameter transfer association characteristics, including: By using a time series feature extractor based on a deep neural network model, feature extraction is performed on the current time series input vector, the temperature time series input vector and the frequency time series input vector to obtain a current time series multi-scale feature vector, a temperature time series multi-scale feature vector and a frequency time series multi-scale feature vector; Calculating a responsiveness estimate of the temperature time series multi-scale feature vector relative to the current time series multi-scale feature vector to obtain a current-temperature responsiveness fusion feature matrix; and Performing responsiveness transfer analysis on the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer correlation feature; The method further includes a training step: for training a temporal feature extractor and a classifier based on a multi-scale neighborhood feature extraction module; Wherein, the training step includes: Acquire training data, the training data including training current values, training temperature values ​​and training frequency values ​​of the monitored substation at multiple predetermined time points within a predetermined time period after the first remote control operation is performed, and a true value of whether the monitored substation has an abnormality; Arrange the training current values, training temperature values ​​and training frequency values ​​at the plurality of predetermined time points into a training current timing input vector, a training temperature timing input vector and a training frequency timing input vector according to the time dimension; Passing the training current time series input vector, the training temperature time series input vector and the training frequency time series input vector through the time series feature extractor based on the multi-scale neighborhood feature extraction module to obtain a training current time series multi-scale feature vector, a training temperature time series multi-scale feature vector and a training frequency time series multi-scale feature vector; Calculating a responsiveness estimate of the training temperature time series multi-scale feature vector relative to the training current time series multi-scale feature vector to obtain a training current-temperature responsiveness fusion feature matrix; Calculating the responsiveness transfer of the training frequency-time series multi-scale feature vector relative to the training current-temperature responsiveness fusion feature matrix to obtain a training multi-parameter transfer associated feature vector; Passing the training multi-parameter transfer association feature vector through the classifier to obtain a classification loss function value; The temporal feature extractor based on the multi-scale neighborhood feature extraction module and the classifier are trained based on the classification loss function value and by directional propagation of gradient descent, wherein, at each iteration of the training, the training multi-parameter transfer association feature vector is subjected to weight space iterative recursive directional proposal optimization; In each iteration, the training multi-parameter transfer association feature vector is subjected to weight space iterative recursive directional proposal optimization, which is specifically expressed as: the training multi-parameter transfer association feature vector is subjected to weight space iterative recursive directional proposal optimization according to the following optimization formula to obtain an optimized training multi-parameter transfer association feature vector; wherein the optimization formula is: Among them, M1 and M2 are the weight matrices of the previous and current iterations, respectively, and V c is the training multi-parameter transfer association feature vector, represents matrix multiplication, ⊙ represents point multiplication by position, exp(·) represents vector exponential operation, and the vector exponential operation represents the calculation of the natural exponential function value raised to the power of the eigenvalues ​​at each position in the vector, V' c represents the optimization training multi-parameter transfer associated feature vector; The weight space iterative recursive directional proposal optimization can be achieved by transferring the initial training multi-parameter associated feature vector V to be classified. c As an anchor point, the multi-parameter transfer correlation feature vector V corresponding to the training is iterated in the weight space based on the weight matrix c The anchor point footprints under different sample temporal spatial feature distribution dimensions are obtained according to the different sample temporal spatial feature distribution directions as directional proposals for iterative recursion in the weight space.

2. The method for preventing misoperation of a substation remote control according to claim 1, characterized in that: The deep neural network model is a multi-scale neighborhood feature extraction module.

3. The method for preventing misoperation of a substation remote control according to claim 2, characterized in that: The frequency-time multi-scale feature vector is subjected to responsiveness transfer analysis relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer association feature, including: calculating the responsiveness transfer of the frequency-time multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain a multi-parameter transfer association feature vector as the multi-parameter transfer association feature.

4. The method for preventing misoperation of a substation remote control according to claim 3, characterized in that: Determining whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature includes: Passing the multi-parameter transfer associated feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether there is an abnormality in the monitored substation; Based on the classification result, it is determined whether the first remote control operation is an erroneous operation.

5. A substation remote control anti-misoperation system, using the substation remote control anti-misoperation method according to claim 1, characterized in that: include: A data acquisition module, used to acquire the current value, temperature value and frequency value of the monitored substation collected by the sensor group at multiple predetermined time points within a predetermined time period after the first remote control operation is performed; A vector arrangement module, used for arranging the current values, temperature values ​​and frequency values ​​at the plurality of predetermined time points into a current timing input vector, a temperature timing input vector and a frequency timing input vector according to the time dimension; An association coding module, used for performing timing responsive association coding on the current timing input vector, the temperature timing input vector and the frequency timing input vector to obtain a multi-parameter transfer association feature; as well as The first remote control operation determination module is used to determine whether the first remote control operation is an erroneous operation based on the multi-parameter transfer association feature.

6. The substation remote control anti-misoperation system according to claim 5, characterized in that: The associated coding module comprises: A feature extraction unit, configured to perform feature extraction on the current time series input vector, the temperature time series input vector, and the frequency time series input vector respectively through a time series feature extractor based on a deep neural network model to obtain a current time series multi-scale feature vector, a temperature time series multi-scale feature vector, and a frequency time series multi-scale feature vector; a responsiveness estimation calculation unit, configured to calculate a responsiveness estimation of the temperature time series multi-scale characteristic vector relative to the current time series multi-scale characteristic vector to obtain a current-temperature responsiveness fusion characteristic matrix; and The responsiveness transfer analysis unit is used to perform responsiveness transfer analysis on the frequency-time series multi-scale feature vector relative to the current-temperature responsiveness fusion feature matrix to obtain the multi-parameter transfer correlation feature.

7. The substation remote control anti-misoperation system according to claim 6, characterized in that: The deep neural network model is a multi-scale neighborhood feature extraction module.

Images