An agent-based power transmission and distribution equipment operating status monitoring system

Abstract

The present invention discloses an intelligent-based power transmission and distribution equipment operating status monitoring system, which relates to the field of data analysis technology. The system includes a data acquisition module, a data feature module, a fault prediction module, and a model update module; the data acquisition module is used to acquire historical data; the data feature module is used to integrate historical data, pre-process the data, and extract features; the fault prediction module is used to construct a classification model based on operating parameters and operating status, construct a time series model to predict operating parameters, and combine the classification and time series models to predict the operating status of operating parameters; the model update module is used to construct a regression model based on the corresponding equipment and operating parameters of different equipment and their corresponding classification model parameters, predict the parameters of the new equipment classification model based on the regression model and the new equipment parameters, and continuously iterate and update the new equipment classification model. The present invention can effectively improve the situation in which the existing technology is difficult to quickly adjust the model to achieve prediction when new equipment is connected.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an intelligent-agent-based power transmission and distribution equipment operating status monitoring system. Background Art

[0002] As a key component of the power system, the operating status of power transmission and distribution equipment directly impacts the reliability and stability of power supply. As power systems continue to expand, the number and complexity of equipment continue to rise. Traditional methods relying on manual inspections are unable to meet the requirements for efficient and safe operation of modern power systems. To achieve real-time and accurate monitoring of the operating status of power transmission and distribution equipment, building a scientific monitoring system leveraging intelligent agents and data processing technologies is becoming an inevitable trend. By acquiring equipment operating data, analyzing characteristics, and predicting faults, potential problems can be promptly identified, ensuring stable power system operation, which is of great significance to the intelligent development of the power industry.

[0003] However, existing technologies for monitoring the operating status of power transmission and distribution equipment lack a dynamic model update mechanism. This prevents timely model optimization when new equipment is connected or operating conditions change, impacting the adaptability and effectiveness of the monitoring system. This solution, through the collaborative work of multiple modules and the integration of various data processing and model building methods, establishes a dynamic update mechanism that significantly improves fault prediction accuracy and enhances the system's adaptability to new equipment, effectively addressing the shortcomings of existing technologies. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent agent-based power transmission and distribution equipment operating status monitoring system to solve the problems raised in the prior art.

[0005] To achieve the above-mentioned object, the present invention provides the following technical solutions: an agent-based power transmission and distribution equipment operating status monitoring system, the system comprising a data acquisition module, a data feature module, a fault prediction module and a model update module;

[0006] The data acquisition module is used to obtain historical power transmission and distribution equipment replacement information, operating status data of each device, and real-time acquisition of operating status data;

[0007] The data feature module is used to integrate historical data, pre-process the acquired equipment operation status data and perform feature extraction on the pre-processed data;

[0008] The fault prediction module is used to build a classification model based on the operating parameters and operating status, build a time series model to predict future operating parameters, and combine the classification and time series models to predict the operating status of future operating parameters;

[0009] The model updating module is used to build a regression model by combining the device parameters and operating parameters corresponding to different devices and their corresponding classification model parameters, predict the new device classification model parameters based on the regression model and the new device parameters, and continuously iteratively update the new device classification model;

[0010] The output end of the data acquisition module is connected to the input end of the data feature module and the model update module; the output end of the data feature module is connected to the input end of the fault prediction module; and the model update module is connected to the fault prediction module.

[0011] The operating status data of each device includes operating parameter data of different power transmission and distribution equipment under normal operating conditions and operating parameters when a fault occurs as well as the corresponding fault type.

[0012] The data acquisition module includes a device parameter unit, a normal data unit, an abnormal data unit and a new data acquisition unit;

[0013] The device parameter unit is used to obtain device parameters of power transmission and distribution equipment in historical data;

[0014] The normal data unit is used to obtain operating parameters of different power transmission and distribution equipment in normal operating conditions in historical data;

[0015] The abnormal data unit is used to obtain the operating parameters and fault types of different power transmission and distribution equipment when faults occur in historical data;

[0016] The new data acquisition unit is used to collect the operating parameters of the power transmission and distribution equipment in real time;

[0017] The output end of the device parameter unit is connected to the input end of the model update module and the data feature module; the output end of the normal data unit is connected to the input end of the data feature module and the model update module; the output end of the abnormal data unit is connected to the input end of the data feature module and the model update module; the output end of the new data acquisition unit is connected to the input end of the data feature module.

[0018] The equipment parameters of the power transmission and distribution equipment are obtained through the equipment product manual; the operating parameters are obtained by arranging sensors on the power transmission and distribution equipment; the fault type is manually calibrated;

[0019] For a certain power transmission and distribution equipment, it corresponds to a unique set of equipment parameters and multiple sets of operating parameters with different timestamps; the data feature module includes a preprocessing unit and a feature extraction unit;

[0020] The pre-processing unit is used to integrate historical data and pre-process the operating parameter data in the historical data;

[0021] The feature extraction unit is used to extract features reflecting the operating status of the device from the preprocessed data;

[0022] The output end of the preprocessing unit is connected to the input end of the feature extraction unit; the output end of the feature extraction unit is connected to the input end of the fault prediction module.

[0023] The specific method of integrating historical data is as follows:

[0024] Divide by different power transmission and distribution equipment; for a certain power transmission and distribution equipment, integrate its operating status data during operation in chronological order, including operating parameters, operating status, and corresponding timestamps; the format is expressed as: [a certain power transmission and distribution equipment: corresponding equipment parameters, operating parameters, operating status: normal / abnormal category 1 / abnormal category 2 / ..., timestamp];

[0025] Integrate historical data corresponding to all power transmission and distribution equipment;

[0026] The specific method of preprocessing the operating parameter data in the historical data is:

[0027] For the operating parameters of power transmission and distribution equipment, normalization processing is performed and the data is mapped to the [0,1] interval;

[0028] Replace the categorical running status with a numerical value;

[0029] The specific method of extracting the features reflecting the equipment operation status from the pre-processed data is as follows:

[0030] Transform domain feature extraction: Perform discrete cosine transform and wavelet transform on the normalized operating parameter data to convert the data from the original domain to other transform domains;

[0031] Statistical feature extraction: Calculate various statistics of normalized operating parameter data, including mean, variance, skewness, and kurtosis;

[0032] All the extracted features are combined into a feature matrix, and principal component analysis is performed on it; by calculating the covariance matrix, the principal component direction of the data is found, and the original high-dimensional features are projected onto the principal component direction to obtain the feature matrix after dimensionality reduction.

[0033] The fault prediction module includes a time series model unit, a classification model unit and a fault prediction unit;

[0034] The time series model unit is used to construct a time series model to predict the newly collected operating parameters of the power transmission and distribution equipment;

[0035] The classification model unit is used to construct a classification model based on the extracted operating parameter features in combination with their corresponding operating states;

[0036] The fault prediction unit is used to predict the operating status of the operating parameters predicted by the time series model according to the classification model;

[0037] The output end of the timing model unit is connected to the input end of the classification model unit; the output end of the classification model unit is connected to the input end of the fault prediction unit and the model updating module.

[0038] The specific method of constructing the time series model to predict the newly collected operating parameters of the power transmission and distribution equipment is as follows:

[0039] Performing normalization processing on the newly collected operating parameter data sequence to obtain a normalized operating parameter data sequence;

[0040] Build an LSTM time series model: Use the normalized operating parameter data sequence as the model input; the model predicts the operating parameters for a period of time in the future by learning the time dependency in the operating parameter data sequence, and outputs the predicted operating parameter data sequence;

[0041] The specific method of constructing a classification model based on the extracted operating parameter features and their corresponding operating states is as follows:

[0042] For a certain power transmission and distribution equipment, the extracted operating parameter feature matrix and the corresponding operating status are integrated in chronological order; the integrated operating parameter feature matrix and the corresponding operating status are used as input to construct an LSTM classification model;

[0043] During the model training process, the back propagation algorithm is used, with the cross entropy loss function between the manually calibrated true value and the predicted value as the optimization target. The model parameters are continuously adjusted to obtain the classification model of the power transmission and distribution equipment.

[0044] Build corresponding classification models for each type of power transmission and distribution equipment;

[0045] The specific method of predicting the operating status of the operating parameters predicted by the time series model based on the classification model is as follows:

[0046] Determine the device to which the newly collected operating parameter data sequence belongs, and select the classification model corresponding to the device to classify and predict the operating parameter data sequence predicted by the LSTM time series model; when the predicted classification is abnormal, feedback the predicted abnormality and the type of abnormality to the staff.

[0047] The model updating module includes a new parameter acquiring unit, a data analyzing unit, a parameter analyzing unit and a model updating unit;

[0048] The new parameter acquisition unit is used to acquire the device parameters of the new device when replacing a new device;

[0049] The data analysis unit is used to calculate the data statistics of each operating parameter at different time stamps;

[0050] The parameter analysis unit is used to analyze the correlation between different equipment parameters and their impact on the equipment operating status;

[0051] The model updating unit is used to construct a regression model by combining data analysis and parameter analysis with different classification model parameters; when replacing a new device, the classification model parameters are adjusted by combining the new device parameters and the regression model;

[0052] The output end of the new parameter acquisition unit is connected to the input end of the parameter analysis unit; the output end of the data analysis unit is connected to the input end of the model updating unit; the output end of the parameter analysis unit is connected to the input end of the model updating unit; and the model updating unit is connected to the fault prediction module.

[0053] The data statistics specifically include mean, variance, maximum and minimum values;

[0054] The specific method of analyzing the correlation between different equipment parameters and their influence on the equipment operation state is as follows: screening out the parameter combination that influences the equipment operation state by calculating the Pearson correlation coefficient;

[0055] The specific method of combining data analysis and parameter analysis with different classification model parameters to construct a regression model is as follows: for different devices and their corresponding classification models, an SVR regression model is constructed with device parameter combinations and data statistics as independent variables and the device corresponding classification model parameters as dependent variables;

[0056] When replacing new equipment, obtain the new equipment parameters (including equipment parameters and operating parameters), predict the corresponding classification model parameters of the new equipment based on the regression model, and update the model; and continuously collect the new equipment operating parameter data and corresponding status to iteratively train the model to update it.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention can more accurately extract the characteristics of the equipment operating status through the comprehensive processing of historical data and real-time data, and the combined use of multiple models, thereby improving the accuracy of fault prediction, timely discovering potential equipment failures, reducing the occurrence of power outages, and ensuring the reliability of power supply; the existence of the model update module of the present invention enables the system to quickly adapt to the access of new equipment without rebuilding the entire monitoring system or manually adjusting a large number of parameters. The classification model can be updated according to the new equipment parameters and regression model, which improves the flexibility and scalability of the system and facilitates effective monitoring when new power transmission and distribution equipment is continuously introduced into the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 The figure is a flow chart of an intelligent-based power transmission and distribution equipment operating status monitoring system according to the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0060] Example: Figure 1 As shown, the present invention provides a technical solution, a power transmission and distribution equipment operation status monitoring system based on intelligent agent, the system includes a data acquisition module, a data feature module, a fault prediction module and a model update module;

[0061] The data acquisition module is used to obtain historical power transmission and distribution equipment replacement information, operating status data of each device, and real-time acquisition of operating status data;

[0062] The data feature module is used to integrate historical data, pre-process the acquired equipment operation status data and perform feature extraction on the pre-processed data;

[0063] The fault prediction module is used to build a classification model based on the operating parameters and operating status, build a time series model to predict future operating parameters, and combine the classification and time series models to predict the operating status of future operating parameters;

[0064] The model updating module is used to build a regression model by combining the device parameters and operating parameters corresponding to different devices and their corresponding classification model parameters, predict the new device classification model parameters based on the regression model and the new device parameters, and continuously iteratively update the new device classification model;

[0065] The output end of the data acquisition module is connected to the input end of the data feature module and the model update module; the output end of the data feature module is connected to the input end of the fault prediction module; and the model update module is connected to the fault prediction module.

[0066] The operating status data of each device includes operating parameter data of different power transmission and distribution equipment under normal operating conditions and operating parameters when a fault occurs as well as the corresponding fault type.

[0067] The data acquisition module includes a device parameter unit, a normal data unit, an abnormal data unit and a new data acquisition unit;

[0068] The device parameter unit is used to obtain device parameters of power transmission and distribution equipment in historical data;

[0069] The normal data unit is used to obtain operating parameters of different power transmission and distribution equipment in normal operating conditions in historical data;

[0070] The abnormal data unit is used to obtain the operating parameters and fault types of different power transmission and distribution equipment when faults occur in historical data;

[0071] The new data acquisition unit is used to collect the operating parameters of the power transmission and distribution equipment in real time;

[0072] The output end of the device parameter unit is connected to the input end of the model update module and the data feature module; the output end of the normal data unit is connected to the input end of the data feature module and the model update module; the output end of the abnormal data unit is connected to the input end of the data feature module and the model update module; the output end of the new data acquisition unit is connected to the input end of the data feature module.

[0073] The equipment parameters of the power transmission and distribution equipment are obtained through the equipment product manual; the operating parameters are obtained by arranging sensors on the power transmission and distribution equipment; the fault type is manually calibrated;

[0074] For a certain power transmission and distribution equipment, it corresponds to a unique set of equipment parameters and multiple sets of operating parameters with different timestamps;

[0075] The data feature module includes a preprocessing unit and a feature extraction unit;

[0076] The pre-processing unit is used to integrate historical data and pre-process the operating parameter data in the historical data;

[0077] The feature extraction unit is used to extract features reflecting the operating status of the device from the preprocessed data;

[0078] The output end of the preprocessing unit is connected to the input end of the feature extraction unit; the output end of the feature extraction unit is connected to the input end of the fault prediction module.

[0079] The specific method of integrating historical data is as follows:

[0080] Divide by different power transmission and distribution equipment; for a certain power transmission and distribution equipment, integrate its operating status data during operation in chronological order, including operating parameters, operating status, and corresponding timestamps; the format is expressed as: [a certain power transmission and distribution equipment: corresponding equipment parameters, operating parameters, operating status: normal / abnormal category 1 / abnormal category 2 / ..., timestamp];

[0081] Integrate historical data corresponding to all power transmission and distribution equipment;

[0082] The specific method of preprocessing the operating parameter data in the historical data is:

[0083] For the operating parameters of power transmission and distribution equipment, normalization processing is performed and the data is mapped to the [0,1] interval;

[0084] Replace the categorical running status with a numerical value;

[0085] The specific method of extracting the features reflecting the equipment operation status from the pre-processed data is as follows:

[0086] Transform domain feature extraction: Perform discrete cosine transform and wavelet transform on the normalized operating parameter data to convert the data from the original domain to other transform domains;

[0087] Statistical feature extraction: Calculate various statistics of normalized operating parameter data, including mean, variance, skewness, and kurtosis;

[0088] All the extracted features are combined into a feature matrix, and principal component analysis is performed on it; by calculating the covariance matrix, the principal component direction of the data is found, and the original high-dimensional features are projected onto the principal component direction to obtain the feature matrix after dimensionality reduction.

[0089] The fault prediction module includes a time series model unit, a classification model unit and a fault prediction unit;

[0090] The time series model unit is used to construct a time series model to predict the newly collected operating parameters of the power transmission and distribution equipment;

[0091] The classification model unit is used to construct a classification model based on the extracted operating parameter features in combination with their corresponding operating states;

[0092] The fault prediction unit is used to predict the operating status of the operating parameters predicted by the time series model according to the classification model;

[0093] The output end of the timing model unit is connected to the input end of the classification model unit; the output end of the classification model unit is connected to the input end of the fault prediction unit and the model updating module.

[0094] The specific method of constructing the time series model to predict the newly collected operating parameters of the power transmission and distribution equipment is as follows:

[0095] Performing normalization processing on the newly collected operating parameter data sequence to obtain a normalized operating parameter data sequence;

[0096] Build an LSTM time series model: Use the normalized operating parameter data sequence as the model input; the model predicts the operating parameters for a period of time in the future by learning the time dependency in the operating parameter data sequence, and outputs the predicted operating parameter data sequence;

[0097] The specific method of constructing a classification model based on the extracted operating parameter features and their corresponding operating states is as follows:

[0098] For a certain power transmission and distribution equipment, the extracted operating parameter feature matrix and the corresponding operating status are integrated in chronological order; the integrated operating parameter feature matrix and the corresponding operating status are used as input to construct an LSTM classification model;

[0099] During the model training process, the back propagation algorithm is used, with the cross entropy loss function between the manually calibrated true value and the predicted value as the optimization target. The model parameters are continuously adjusted to obtain the classification model of the power transmission and distribution equipment.

[0100] Build corresponding classification models for each type of power transmission and distribution equipment;

[0101] The specific method of predicting the operating status of the operating parameters predicted by the time series model based on the classification model is as follows:

[0102] Determine the device to which the newly collected operating parameter data sequence belongs, and select the classification model corresponding to the device to classify and predict the operating parameter data sequence predicted by the LSTM time series model; when the predicted classification is abnormal, feedback the predicted abnormality and the type of abnormality to the staff.

[0103] The model updating module includes a new parameter acquiring unit, a data analyzing unit, a parameter analyzing unit and a model updating unit;

[0104] The new parameter acquisition unit is used to acquire the device parameters of the new device when replacing a new device;

[0105] The data analysis unit is used to calculate the data statistics of each operating parameter at different time stamps;

[0106] The parameter analysis unit is used to analyze the correlation between different equipment parameters and their impact on the equipment operating status;

[0107] The model updating unit is used to construct a regression model by combining data analysis and parameter analysis with different classification model parameters; when replacing a new device, the classification model parameters are adjusted by combining the new device parameters and the regression model;

[0108] The output end of the new parameter acquisition unit is connected to the input end of the parameter analysis unit; the output end of the data analysis unit is connected to the input end of the model updating unit; the output end of the parameter analysis unit is connected to the input end of the model updating unit; and the model updating unit is connected to the fault prediction module.

[0109] The data statistics specifically include mean, variance, maximum and minimum values;

[0110] The specific method of analyzing the correlation between different equipment parameters and their influence on the equipment operation state is as follows: screening out the parameter combination that influences the equipment operation state by calculating the Pearson correlation coefficient;

[0111] The specific method of combining data analysis and parameter analysis with different classification model parameters to construct a regression model is as follows: for different devices and their corresponding classification models, an SVR regression model is constructed with device parameter combinations and data statistics as independent variables and the device corresponding classification model parameters as dependent variables;

[0112] When replacing new equipment, obtain the new equipment parameters, predict the corresponding classification model parameters of the new equipment based on the regression model, and update the model; and continuously collect the new equipment operating parameter data and corresponding status to iteratively train the model to update it.

[0113] In this embodiment, an agent-based power transmission and distribution equipment operating status monitoring system is applied in a regional power grid.

[0114] High-precision voltage sensors, current transformers, and temperature sensors are strategically placed on the surfaces and core areas of key power transmission and distribution equipment, such as transformers and circuit breakers, based on the equipment's structure and operating characteristics. These sensors collect real-time operating parameters such as voltage fluctuations, current levels, and temperature changes during equipment operation at frequencies in seconds or even milliseconds. Simultaneously, by consulting equipment product manuals, technical documentation, and other materials, equipment parameters such as rated voltage, rated current, capacity, and model are obtained. Furthermore, historical databases are deeply mined to comprehensively summarize operating parameter data for each device during normal operation, as well as operating parameters and fault types during faults. For example, data on sudden temperature rise and abnormal current in transformers caused by winding short circuits provides a rich data foundation for subsequent analysis.

[0115] After acquiring the data, the data is categorized by device type, such as transformers, circuit breakers, and disconnectors. Furthermore, the operating status data for the same device at different times is integrated chronologically to form a complete data chain. For operating parameters, the minimum-maximum normalization method is used to map the data to the [0, 1] interval, eliminating dimensional differences between parameters and facilitating subsequent analysis. For categorical data such as fault types, one-hot encoding or label encoding is used to convert them into numerical data. Next, the normalized operating parameters are subjected to discrete cosine transforms and wavelet transforms, converting the data from the time domain to the frequency domain or other transform domains to explore the data's characteristics across different dimensions. Statistical measures such as mean, variance, skewness, and kurtosis are calculated to characterize the data's distribution from a statistical perspective. All extracted features are combined to construct a feature matrix. Principal component analysis (PCA) is then used to calculate the covariance matrix to identify the principal component directions of the data. The original high-dimensional features are projected onto these principal component directions, achieving data dimensionality reduction, removing redundant information, and retaining key features.

[0116] The processed data is then used for fault prediction. A time series model is constructed using a long short-term memory (LSTM) network. The newly collected and normalized operating parameter data series is first divided into training and test sets. The LSTM model is trained, with appropriate parameters such as the number of hidden layer neurons, learning rate, and number of iterations. By learning the temporal dependencies within the historical operating parameter data series, the model can predict operating parameters for the next hour, two hours, or even longer periods. The extracted features and corresponding operating status data are then integrated chronologically. The integrated feature matrix and operating status labels are used as input to construct an LSTM classification model. During classification model training, a backpropagation algorithm is employed, using the cross-entropy loss function between manually calibrated actual operating status and model predictions as the optimization objective. Model parameters are continuously adjusted to improve classification accuracy. After training, the classification model is used to perform classification predictions on the operating parameter data series predicted by the time series model. If an operating status is determined to be abnormal, such as a transformer temperature predicted to exceed a safety threshold within the next two hours, the system immediately issues a high-temperature fault warning to the operation and maintenance personnel, detailing the type of abnormality and the potential impact range.

[0117] When new equipment, such as a new circuit breaker, is connected to the grid, the system initiates a model update mechanism. First, comprehensive equipment parameters are collected for the new equipment, including but not limited to rated voltage, rated current, mechanical lifespan, and electrical lifespan. Furthermore, after the new equipment is put into operation, operating parameter data is continuously collected, and statistics such as the mean, variance, maximum, and minimum values ​​of each operating parameter at different timestamps are calculated. The Pearson correlation coefficient is then calculated to analyze the correlations between different equipment parameters and identify parameter combinations that significantly impact the equipment's operating status. Based on this data, a support vector regression (SVR) model is constructed, using the equipment parameter combinations and data statistics as independent variables and the corresponding classification model parameters of existing equipment as dependent variables. The trained regression model is then used to predict the classification model parameters based on the new equipment's parameters, completing the initialization of the classification model for the new equipment. During the subsequent operation of the new equipment, new operating parameter data and corresponding status are continuously collected, and the classification model is iteratively trained to continuously optimize the model parameters, ensuring that the system can consistently accurately monitor the operating status of the new equipment and promptly identify potential fault hazards.

[0118] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

Claims

1. An agent-based power transmission and distribution equipment operating status monitoring system, characterized by: The system includes a data acquisition module, a data feature module, a fault prediction module and a model update module; The data acquisition module is used to obtain historical power transmission and distribution equipment replacement information, operating status data of each device, and real-time acquisition of operating status data; The data feature module is used to integrate historical data, pre-process the acquired equipment operation status data and perform feature extraction on the pre-processed data; The fault prediction module is used to build a classification model based on the operating parameters and operating status, build a time series model to predict future operating parameters, and combine the classification and time series models to predict the operating status of future operating parameters; The model updating module is used to build a regression model by combining the device parameters and operating parameters corresponding to different devices and their corresponding classification model parameters, predict the new device classification model parameters based on the regression model and the new device parameters, and continuously iteratively update the new device classification model; The output end of the data acquisition module is connected to the input end of the data feature module and the model update module; the output end of the data feature module is connected to the input end of the fault prediction module; the model update module is connected to the fault prediction module; The fault prediction module includes a time series model unit, a classification model unit and a fault prediction unit; The time series model unit is used to construct a time series model to predict the newly collected operating parameters of the power transmission and distribution equipment; The classification model unit is used to construct a classification model based on the extracted operating parameter features in combination with their corresponding operating states; The fault prediction unit is used to predict the operating status of the operating parameters predicted by the time series model according to the classification model; The specific method of constructing the time series model to predict the newly collected operating parameters of the power transmission and distribution equipment is as follows: Performing normalization processing on the newly collected operating parameter data sequence to obtain a normalized operating parameter data sequence; Build an LSTM time series model: use the normalized operating parameter data sequence as the model input; The model predicts the operating parameters in the future by learning the time dependency in the operating parameter data sequence and outputs the predicted operating parameter data sequence; The specific method of constructing a classification model based on the extracted operating parameter features and their corresponding operating states is as follows: For a certain power transmission and distribution equipment, the extracted operating parameter feature matrix and the corresponding operating status are integrated in chronological order; the integrated operating parameter feature matrix and the corresponding operating status are used as input to construct an LSTM classification model; During the model training process, the back propagation algorithm is used, with the cross entropy loss function between the manually calibrated true value and the predicted value as the optimization target. The model parameters are continuously adjusted to obtain the classification model of the power transmission and distribution equipment. Build corresponding classification models for each type of power transmission and distribution equipment; The specific method of predicting the operating status of the operating parameters predicted by the time series model based on the classification model is as follows: Determine the device to which the newly collected operating parameter data series belongs, and select the classification model corresponding to the device to perform classification prediction on the operating parameter data series predicted by the LSTM time series model; When the prediction is classified as an anomaly, the predicted anomaly and the type of anomaly are fed back to the staff.

2. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 1, characterized in that: The operating status data of each device includes operating parameter data of different power transmission and distribution equipment under normal operating conditions and operating parameters when a fault occurs as well as the corresponding fault type.

3. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 2, characterized in that: The data acquisition module includes a device parameter unit, a normal data unit, an abnormal data unit and a new data acquisition unit; The device parameter unit is used to obtain device parameters of power transmission and distribution equipment in historical data; The normal data unit is used to obtain operating parameters of different power transmission and distribution equipment in normal operating conditions in historical data; The abnormal data unit is used to obtain the operating parameters and fault types of different power transmission and distribution equipment when faults occur in historical data; The new data acquisition unit is used to collect the operating parameters of the power transmission and distribution equipment in real time; The output end of the device parameter unit is connected to the input end of the model update module and the data feature module; the output end of the normal data unit is connected to the input end of the data feature module and the model update module; the output end of the abnormal data unit is connected to the input end of the data feature module and the model update module; the output end of the new data acquisition unit is connected to the input end of the data feature module.

4. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 3, characterized in that: The equipment parameters of the power transmission and distribution equipment are obtained through the equipment product manual; the operating parameters are obtained by arranging sensors on the power transmission and distribution equipment; the fault type is manually calibrated; For a certain power transmission and distribution equipment, it corresponds to a unique set of equipment parameters and multiple sets of operating parameters with different timestamps.

5. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 4, characterized in that: The data feature module includes a preprocessing unit and a feature extraction unit; The pre-processing unit is used to integrate historical data and pre-process the operating parameter data in the historical data; The feature extraction unit is used to extract features reflecting the operating status of the device from the preprocessed data; The output end of the preprocessing unit is connected to the input end of the feature extraction unit; the output end of the feature extraction unit is connected to the input end of the fault prediction module.

6. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 5, characterized in that: The specific method of integrating historical data is as follows: Divide by different power transmission and distribution equipment; for a certain power transmission and distribution equipment, integrate its operating status data during operation in chronological order, including operating parameters, operating status, and corresponding timestamps; the format is expressed as: [a certain power transmission and distribution equipment: corresponding equipment parameters, operating parameters, operating status: normal / abnormal category 1 / abnormal category 2 / ..., timestamp]; Integrate historical data corresponding to all power transmission and distribution equipment; The specific method of preprocessing the operating parameter data in the historical data is: For the operating parameters of power transmission and distribution equipment, normalization processing is performed and the data is mapped to the [0,1] interval; Replace the categorical running status with a numerical value; The specific method of extracting the features reflecting the equipment operation status from the pre-processed data is as follows: Transform domain feature extraction: Perform discrete cosine transform and wavelet transform on the normalized operating parameter data to convert the data from the original domain to other transform domains; Statistical feature extraction: Calculate various statistics of normalized operating parameter data, including mean, variance, skewness, and kurtosis; All the extracted features are combined into a feature matrix, and principal component analysis is performed on it; by calculating the covariance matrix, the principal component direction of the data is found, and the original high-dimensional features are projected onto the principal component direction to obtain the feature matrix after dimensionality reduction.

7. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 6, characterized in that: The output end of the timing model unit is connected to the input end of the classification model unit; the output end of the classification model unit is connected to the input end of the fault prediction unit and the model updating module.

8. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 7, characterized in that: The model updating module includes a new parameter acquiring unit, a data analyzing unit, a parameter analyzing unit and a model updating unit; The new parameter acquisition unit is used to acquire the device parameters of the new device when replacing a new device; The data analysis unit is used to calculate the data statistics of each operating parameter at different time stamps; The parameter analysis unit is used to analyze the correlation between different equipment parameters and their impact on the equipment operating status; The model updating unit is used to construct a regression model by combining data analysis and parameter analysis with different classification model parameters; when replacing a new device, the classification model parameters are adjusted by combining the new device parameters and the regression model; The output end of the new parameter acquisition unit is connected to the input end of the parameter analysis unit; the output end of the data analysis unit is connected to the input end of the model updating unit; The output end of the parameter analysis unit is connected to the input end of the model updating unit; and the model updating unit is connected to the fault prediction module.

9. The agent-based power transmission and distribution equipment operating status monitoring system according to claim 8, characterized in that: The data statistics specifically include mean, variance, maximum and minimum values; The specific method of analyzing the correlation between different equipment parameters and their influence on the equipment operation state is as follows: screening out the parameter combination that influences the equipment operation state by calculating the Pearson correlation coefficient; The specific method of combining data analysis and parameter analysis with different classification model parameters to construct a regression model is as follows: for different devices and their corresponding classification models, an SVR regression model is constructed with device parameter combinations and data statistics as independent variables and the device corresponding classification model parameters as dependent variables; When replacing new equipment, obtain the new equipment parameters, predict the corresponding classification model parameters of the new equipment based on the regression model, and update the model; and continuously collect the new equipment operating parameter data and corresponding status to iteratively train the model to update it.

Images