A substation switching misoperation key factor adaptive identification method and system

Abstract

The application discloses a substation switching misoperation key factor adaptive identification method and system, relates to the adaptive identification technical field, and comprises the following steps: adopting a hierarchical structure system to process the system architecture, constructing a hierarchical semantic feature extraction model for the text data of the substation operation ticket, proposing an improved LSTM-based time sequence feature extraction network for the substation equipment state data, designing a dynamic fusion method based on an attention mechanism, proposing a substation operation knowledge distillation framework, and identifying the substation switching based on gradient accumulation quantitative evaluation. The application realizes unified representation and analysis of various data, applies the knowledge distillation technology to the substation switching operation field, designs a teacher-student network framework, compresses a complex expert experience model into a lightweight practical model, proposes a soft label generation method based on temperature scaling, effectively retains the fine-grained information of expert decision-making, and realizes real-time adaptive identification of risk factors.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of adaptive recognition, in particular to a substation switching misoperation key factor adaptive recognition method and system. BACKGROUND

[0002] With the rapid development of the power system, the concept of smart grid is increasingly accepted. Smart grid not only needs efficient data collection and processing capability, but also needs powerful fault diagnosis and prediction function to ensure the safe and stable operation of the power system. In this context, the intelligent management level of the substation as a key node in the power system directly affects the efficiency and reliability of the entire power system. In the operation process of the substation, especially in the switching operation, a large amount of data processing and complex operation procedures are involved. The correctness and timeliness of these operations are crucial to the safety of the power system. However, with the continuous expansion of the scale of the power system and the increase of technical complexity, traditional manual operation and simple automation methods have been unable to meet the needs of modern power systems, and it is urgent to develop more intelligent and efficient substation operation management technology.

[0003] Existing substation switching operation recognition methods mostly rely on single modal data analysis, such as only based on historical operation records or device state monitoring data. This method has obvious limitations. First, single modal data is difficult to fully reflect the real situation of substation operation, which may lead to misjudgment. Second, traditional methods lack effective knowledge transfer mechanism and cannot fully utilize existing operation experience and knowledge to improve the processing effect of new tasks. Finally, existing methods usually do not consider the mutual influence between different operation factors, which cannot provide accurate operation risk assessment. Therefore, the present application proposes a substation switching misoperation key factor adaptive recognition method, which builds a hierarchical system architecture, uses an improved LSTM model and attention mechanism for deep fusion of multi-modal data, and realizes effective transfer of operation knowledge through a knowledge distillation framework, thereby improving the accuracy and reliability of substation switching operation recognition and providing strong technical support for the safe and stable operation of the power system. SUMMARY

[0004] In view of the problems existing in the prior art substation switching misoperation key factor adaptive recognition method and system, the present application is proposed.

[0005] Therefore, the purpose of the present application is to provide a substation switching misoperation key factor adaptive recognition method and system. In view of the fact that misoperation is often caused by the interaction of multiple factors, the present application proposes a substation switching misoperation key factor adaptive recognition method and system to solve the problem.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the embodiments of the present application provide a substation switching misoperation key factor adaptive identification method, which comprises: constructing a system architecture, processing the system architecture by adopting a hierarchical structure system, constructing a hierarchical semantic feature extraction model for text data of a substation operation ticket; proposing an improved LSTM-based time sequence feature extraction network for substation equipment state data, designing a dynamic fusion method based on an attention mechanism; proposing a substation operation knowledge distillation framework, and identifying the substation switching based on gradient accumulation quantitative evaluation.

[0008] As a preferred scheme of the substation switching misoperation key factor adaptive identification method, the construction of the system architecture comprises adopting a hierarchical structure system.

[0009] The structure system comprises a data acquisition layer, a feature extraction layer, a knowledge transfer layer and a decision output layer, the data acquisition layer is responsible for acquisition and preprocessing of multi-source heterogeneous data of the substation, the feature extraction layer realizes automatic learning of deep features, the knowledge transfer layer completes intelligent reuse of operation experience knowledge of the substation, and the decision output layer gives a final identification result and evaluation suggestion.

[0010] As a preferred scheme of the substation switching misoperation key factor adaptive identification method, the construction of the hierarchical semantic feature extraction model comprises multi-modal feature extraction for text data of the substation operation ticket, and a formula of the multi-modal feature extraction is:

[0011] S=D(X_text)

[0012] D(X_text)=σ(W_t·E(X_text)+b_t)

[0013] Wherein, X_text represents input substation operation ticket text data, E(·) represents a word embedding mapping function, which is used to convert operation text into a dense vector representation, W_t represents a weight matrix of text feature extraction, b_t represents a text feature extraction bias vector, and sigma represents a nonlinear activation function.

[0014] As a preferred scheme of the substation switching misoperation key factor adaptive identification method, the improved LSTM-based time sequence feature extraction network comprises device state feature extraction for substation equipment state data, and a formula of the device state feature extraction is:

[0015] H(t)=LSTM(X_device(t),H(t-1))

[0016] f(t) = sigma(W_d * [H(t-1), X_device(t)] + b_d)

[0017] Wherein, X_device(t) represents the substation device state input at t time, H(t) represents the device state hidden state vector, which is dynamically updated through the LSTM unit, and W_d and b_d represent the weight matrix and bias vector of the device state feature extraction, respectively.

[0018] As a preferred scheme of the substation switching misoperation key factor adaptive identification method, wherein: the designed dynamic fusion method based on attention mechanism includes a calculated feature fusion mechanism, and the formula of the calculated feature fusion mechanism is:

[0019]

[0020] Z = sigma(W_f * [alpha_text; alpha_device; alpha_human] + b_f)

[0021] Wherein, alpha_k represents the attention weight of each feature of the substation, which is calculated through the interaction of the query matrix Q_k, the key matrix K_k and the value matrix V_k, and Z represents the fused feature representation, which realizes the deep integration of the features through the learnable parameter matrix W_f and the bias vector b_f.

[0022] As a preferred scheme of the substation switching misoperation key factor adaptive identification method, wherein: the proposed substation operation knowledge distillation framework includes calculating a substation operation knowledge distillation framework according to a knowledge transfer framework, and the formula of the calculated substation operation knowledge distillation framework is:

[0023] T(x) = softmax(g(f(x)) / tau)

[0024] L = lambda_1 * L_CE(y, S(x)) + lambda_2 * L_KL(T(x), S(x)) + lambda_3 * L_reg

[0025] Wherein, T(x) represents the knowledge representation extracted by the substation expert model, tau is a temperature parameter, the loss function includes cross-entropy loss L_CE, knowledge distillation loss L_KL and regularization term L_reg, and the weights lambda_1, lambda_2 and lambda_3 are balanced.

[0026] As a preferred scheme of the substation switching misoperation key factor adaptive identification method, wherein: the identification of the substation switching includes calculating a gradient accumulation-based quantitative evaluation method according to the factor evaluation, and the formula of the calculated gradient accumulation-based quantitative evaluation method is:

[0027]

[0028] S(f_i) = I(f_i) / summation I(f_j)

[0029] Wherein, I(f_i) represents the influence degree of substation misoperation factor f_i, which is obtained by calculating the gradient accumulation of the loss function on the factor; S(f_i) is the normalized importance score.

[0030] In the second aspect, the embodiment of the present application provides a substation switching misoperation key factor adaptive identification system, which comprises: a construction module, which constructs a system architecture, processes the system architecture using a hierarchical structure system, and constructs a hierarchical semantic feature extraction model for the text data of the substation operation ticket; a calculation module, which proposes an improved LSTM-based time sequence feature extraction network for substation equipment state data, and designs a dynamic fusion method based on attention mechanism; an identification module, which proposes a substation operation knowledge distillation framework, and identifies the substation switching based on the quantitative evaluation of gradient accumulation.

[0031] In the third aspect, the embodiment of the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements any step of the substation switching misoperation key factor adaptive identification method described above when executing the computer program.

[0032] In the fourth aspect, the embodiment of the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement any step of the substation switching misoperation key factor adaptive identification method described above.

[0033] The present application has the following advantages: the present application cooperatively analyzes multi-source data such as operation ticket information, equipment state, and environmental parameters, proposes a multi-modal feature fusion processing mechanism based on deep learning, realizes unified representation and analysis of various data, applies knowledge distillation technology to the field of substation switching operation, designs a teacher-student network framework, compresses complex expert experience models into lightweight practical models, proposes a soft label generation method based on temperature scaling, effectively preserves fine-grained information of expert decision-making, and realizes real-time adaptive identification of risk factors. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor. Among them:

[0035] Figure 1 A flow chart of a substation switching misoperation key factor adaptive identification method and system is provided for an embodiment of the present application.

[0036] Figure 2 An identification schematic diagram of a substation switching misoperation key factor adaptive identification method and system is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0037] In order to make the above objectives, features and advantages of the present application more apparent, specific embodiments of the present application will be described in detail below with the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work should fall within the protection scope of the present application.

[0038] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the concept of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0039] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0040] The present application is described in detail in conjunction with the schematic diagram. In the detailed description of the embodiments of the present application, the cross-sectional view of the device structure is locally enlarged without the general proportion for the convenience of description, and the schematic diagram is only an example, which should not limit the scope of protection of the present application herein. In addition, the three-dimensional spatial dimensions of length, width and depth should be included in actual manufacturing.

[0041] Meanwhile, in the description of the present application, it should be noted that the terms "up, down, in and out" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first, second or third" are only for the purpose of description, and cannot be understood as indicating or implying relative importance.

[0042] Unless otherwise defined, the terms "mounting, connecting, associating" in the present application should be interpreted broadly, for example: it can be fixed connection, detachable connection or integral connection; it can also be mechanical connection, electrical connection or direct connection, it can also be indirectly connected through intermediate medium, or it can be the internal communication of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0043] Embodiment 1

[0044] Reference Figure 1 And Figure 2 As the first embodiment of the present application, the embodiment provides a substation switching misoperation key factor adaptive identification method, comprising:

[0045] S1: Constructing a system architecture, adopting a hierarchical structure system to process the system architecture, constructing a hierarchical semantic feature extraction model for the text data of the substation operation ticket.

[0046] Among them, the construction of system architecture includes adopting a hierarchical structure system;

[0047] The structure system includes data acquisition layer, feature extraction layer, knowledge transfer layer and decision output layer, the data acquisition layer is responsible for the acquisition and preprocessing of the multi-source heterogeneous data of the substation, the feature extraction layer realizes the automatic learning of deep features, the knowledge transfer layer completes the intelligent reuse of the operation experience knowledge of the substation, and the decision output layer gives the final identification result and evaluation suggestion.

[0048] S1.1: Constructing a hierarchical semantic feature extraction model includes multi-modal feature extraction for the text data of the substation operation ticket, and the formula of the multi-modal feature extraction is:

[0049] S=D(X_text)

[0050] D(X_text)=σ(W_t·E(X_text)+b_t)

[0051] Wherein, X_text represents the input substation operation ticket text data, E(·) represents a word embedding mapping function, which is used to convert the operation text into a dense vector representation, W_t represents a weight matrix of text feature extraction, b_t represents a text feature extraction bias vector, and σ represents a nonlinear activation function.

[0052] S2: For the substation equipment state data, an improved LSTM-based time series feature extraction network is proposed, and a dynamic fusion method based on attention mechanism is designed.

[0053] Wherein, the time sequence feature extraction network based on improved LSTM includes device state feature extraction for substation device state data, and the formula for device state feature extraction is:

[0054] H(t) = LSTM(X_device(t), H(t-1))

[0055] f(t) = σ(W_d·[H(t-1), X_device(t)] + b_d)

[0056] Wherein, X_device(t) represents the substation device state input at time t, H(t) represents the device state hidden state vector, which is dynamically updated through the LSTM unit, and W_d and b_d represent the weight matrix and bias vector of the device state feature extraction, respectively.

[0057] S2.1: Design a dynamic fusion method based on attention mechanism, which includes calculating the feature fusion mechanism, and the formula for calculating the feature fusion mechanism is:

[0058]

[0059] Z = σ(W_f·[α_text; α_device; α_human] + b_f)

[0060] Wherein, α_k represents the attention weight of each type of feature in the substation, which is calculated by the interaction of the query matrix Q_k, the key matrix K_k and the value matrix V_k, and Z represents the fused feature representation, which realizes the deep integration of the features through the learnable parameter matrix W_f and the bias vector b_f.

[0061] S3: Propose a substation operation knowledge distillation framework, which identifies the substation switching based on gradient accumulation quantitative evaluation.

[0062] Wherein, the substation operation knowledge distillation framework includes calculating the substation operation knowledge distillation framework according to the knowledge transfer framework, and the formula for calculating the substation operation knowledge distillation framework is:

[0063] T(x) = softmax(g(f(x)) / τ)

[0064] L = λ_1·L_CE(y, S(x)) + λ_2·L_KL(T(x), S(x)) + λ_3·L_reg

[0065] Wherein, T(x) represents the knowledge representation extracted by the substation expert model, τ is the temperature parameter, the loss function includes cross entropy loss L_CE, knowledge distillation loss L_KL and regularization term L_reg, which are balanced by weight coefficients λ_1, λ_2 and λ_3.

[0066] S3.1: The identification of the substation switching operation includes a quantitative evaluation method based on gradient accumulation according to factor evaluation calculation, and the formula of the quantitative evaluation method based on gradient accumulation is:

[0067]

[0068] S(f_i)=I(f_i) / ∑I(f_j)

[0069] Wherein, I(f_i) represents the influence degree of the substation misoperation factor f_i, which is obtained by calculating the gradient accumulation of the loss function on the factor; S(f_i) is the normalized importance score.

[0070] In a preferred embodiment, a substation switching operation identification system based on multi-modal and knowledge distillation, the system includes a construction module that constructs a system architecture, processes the system architecture using a hierarchical structure, and constructs a hierarchical semantic feature extraction model for text data of substation operation tickets; a calculation module that proposes a time series feature extraction network based on an improved LSTM for substation equipment state data, and designs a dynamic fusion method based on an attention mechanism; an identification module that proposes a substation operation knowledge distillation framework, and identifies the substation switching operation based on a quantitative evaluation of gradient accumulation.

[0071] The above-mentioned unit modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to call and execute the operations of the above-mentioned modules by the processor.

[0072] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved through WIFI, operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device. It can also be an external keyboard, touchpad or mouse, etc.

[0073] In summary, the present application cooperatively analyzes multi-source data such as operation ticket information, equipment state, and environmental parameters, proposes a multi-modal feature fusion processing mechanism based on deep learning, realizes unified representation and analysis of various data, applies the knowledge distillation technology to the substation switching operation field, designs a teacher-student network framework, compresses the complex expert experience model into a lightweight practical model, proposes a temperature scaling-based soft label generation method, effectively preserves the fine-grained information of expert decision-making, and realizes real-time adaptive identification of risk factors.

[0074] Embodiment 2

[0075] With reference to Figure 1 and Figure 2 For the second embodiment of the present application, the embodiment provides a substation switching misoperation key factor adaptive identification method. In order to verify the beneficial effects of the present application, a simulation experiment is carried out for scientific demonstration.

[0076] In order to verify the effectiveness of the proposed substation switching identification method based on multi-modal and knowledge distillation, detailed experimental research is carried out. In the experiment, an actual operation data set from a large substation is used, which contains a total of 12,000 switching operation records in the past year, each record contains multi-modal information such as operation ticket text data and equipment state data. First, through the constructed hierarchical system architecture, the data is preprocessed and preliminarily screened to ensure the quality of the data. Then, in the feature extraction stage, the improved LSTM model is used to process the time series data of the equipment state, and the deep neural network based on word embedding is used to extract the semantic features of the operation ticket text. Subsequently, the attention mechanism is designed to dynamically fuse the text and equipment state features, realizing the deep fusion of multi-modal data. On this basis, through the knowledge distillation framework, the operation knowledge is successfully extracted from the pre-trained expert model and applied to the target model, significantly improving the generalization ability and recognition accuracy of the model. Finally, the experimental results show that compared with the traditional method, the method of the present application improves the recognition accuracy by about 15%, reaching more than 95%, and the false positive rate is reduced by about 10%, proving the effectiveness and superiority of the method in practical application. The experimental data of the present application is shown in Table 1 below:

[0077] Table 1 Experimental data table of the present application

[0078] Data type Description Specific value Total number of operation records Number of switching operation records used in the experiment 12,000 times Time range Time span of the records Past year Feature extraction Device state data processing model Improved LSTM model Feature extraction Operation ticket text data processing model Deep neural network based on word embedding Fusion method Multi-modal data fusion mechanism Attention mechanism Knowledge transfer Operation knowledge transfer framework Knowledge distillation framework Experimental results Recognition accuracy improvement compared to traditional methods Improved by about 15% Experimental results Final recognition accuracy Reached more than 95% Experimental results False positive rate reduction compared to traditional methods Reduced by about 10%

[0079] The experiment is based on 12,000 substation switching operation records in the past year, uses an improved LSTM model to process device state data, uses a word embedding-based deep neural network to extract operation ticket text features, and realizes deep fusion of multi-modal data through an attention mechanism, and after operation knowledge transfer combined with a knowledge distillation framework, the experimental results show that compared with traditional methods, the recognition accuracy of the method is improved by about 15%, reaching more than 95%, and the false positive rate is reduced by about 10%. The comparison between the present application and the prior art is shown in Table 2 as follows:

[0080] Table 2 Comparison table of the present application and the prior art

[0081]

[0082] The technical scheme of the present application significantly improves the recognition accuracy of substation switching operation to more than 95% and reduces the false positive rate to less than 10% through multi-modal data fusion, improved LSTM model and word embedding-based deep neural network for feature extraction, and dynamic attention mechanism and knowledge distillation framework. Compared with the prior art, the present application has stronger comprehensive advantages in data processing, feature extraction, feature fusion and knowledge transfer, effectively improving the safety and reliability of substation operation.

[0083] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A method for adaptive identification of key factors of substation switching misoperation, characterized in that: Comprising, The system architecture is constructed, the hierarchical structure system is adopted to process the system architecture, and a hierarchical semantic feature extraction model is constructed for text data of substation operation tickets; For substation equipment state data, an improved LSTM-based time series feature extraction network is proposed, and a dynamic fusion method based on attention mechanism is designed; A substation operation knowledge distillation framework is proposed, and a gradient accumulation-based quantitative evaluation is used to identify substation switching; The substation operation knowledge distillation framework includes calculating the substation operation knowledge distillation framework according to the knowledge transfer framework, and the formula for calculating the substation operation knowledge distillation framework is: T(x) = softmax(g(f(x)) / tau) L = lambda_1 * L_CE(y, S(x)) + lambda_2 * L_KL(T(x), S(x)) + lambda_3 * L_reg Where T(x) represents the knowledge representation extracted by the substation expert model, tau is the temperature parameter, the loss function includes cross-entropy loss L_CE, knowledge distillation loss L_KL and regularization term L_reg, and the weights lambda_1, lambda_2 and lambda_3 are balanced; The identification of substation switching includes calculating the gradient accumulation-based quantitative evaluation method according to the factor evaluation, and the formula for calculating the gradient accumulation-based quantitative evaluation method is: S(f_i) = I(f_i) / sum I(f_j) Where I(f_i) represents the influence degree of the substation misoperation factor f_i, which is obtained by calculating the gradient accumulation of the loss function; S(f_i) is the normalized importance score.

2. The substation switching misoperation key factor self-adaptive identification method of claim 1, wherein: The system architecture includes a hierarchical structure system; The structure system includes a data acquisition layer, a feature extraction layer, a knowledge transfer layer and a decision output layer, the data acquisition layer is responsible for the acquisition and preprocessing of substation multi-source heterogeneous data, the feature extraction layer realizes the automatic learning of deep features, the knowledge transfer layer completes the intelligent reuse of substation operation experience knowledge, and the decision output layer gives the final identification result and evaluation suggestion.

3. The substation switching misoperation key factor self-adaptive identification method of claim 2, wherein: The hierarchical semantic feature extraction model includes multi-modal feature extraction for text data of substation operation tickets, and the formula for multi-modal feature extraction is: S = D(X_text) D(X_text) = sigma(W_t * E(X_text) + b_t) Where X_text represents the input substation operation ticket text data, E(·) represents a word embedding mapping function, which is used to convert the operation text into a dense vector representation, W_t represents a text feature extraction weight matrix, b_t represents a text feature extraction bias vector, and sigma represents a nonlinear activation function.

4. The substation switching misoperation key factor self-adaptive identification method of claim 3, wherein: The improved LSTM-based time series feature extraction network includes device state feature extraction for substation equipment state data, and the formula for device state feature extraction is: H(t) = LSTM(X_device(t), H(t-1)) f(t) = sigma(W_d * [H(t-1), X_device(t)] + b_d) Wherein, X_device(t) represents the substation equipment state input at t time, H(t) represents the equipment state hidden state vector, which is dynamically updated through the LSTM unit, and W_d and b_d represent the weight matrix and bias vector of the equipment state feature extraction, respectively.

5. The substation switching misoperation key factor self-adaptive identification method of claim 4, wherein: The design of the dynamic fusion method based on the attention mechanism includes a calculated feature fusion mechanism, and a formula of the calculated feature fusion mechanism is: a_k = softmax(W_k · Q_k(K_k) T V_k Z=σ(W_f·[α_text;α_device;α_human]+b_f) Wherein, α_k represents the attention weight of each feature of the substation, which is calculated through the interaction of the query matrix Q_k, the key matrix K_k and the value matrix V_k, Z represents the fused feature representation, and the deep integration of the features is realized through the learnable parameter matrix W_f and the bias vector b_f.

6. A substation switching misoperation key factor adaptive identification system based on the substation switching misoperation key factor adaptive identification method of any one of claims 1 to 5, characterized in that: It includes, The construction module constructs the system architecture, processes the system architecture by using a hierarchical structure system, constructs a hierarchical semantic feature extraction model for text data of substation operation tickets, and constructs a hierarchical semantic feature extraction model for text data of substation operation tickets. The calculation module proposes a time sequence feature extraction network based on an improved LSTM for substation equipment state data, and designs a dynamic fusion method based on an attention mechanism. The recognition module proposes a substation operation knowledge distillation framework, and performs recognition on substation switching based on quantitative evaluation of gradient accumulation. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to realize the steps of the substation switching misoperation key factor adaptive recognition method in any one of claims 1-5.

8. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the substation switching misoperation key factor adaptive recognition method in any one of claims 1-5.

Images