Intelligent power equipment health monitoring system integrated with sensor network

Abstract

The invention relates to the technical field of power equipment monitoring, and discloses an intelligent power equipment health monitoring system integrated with a sensor network, and the system comprises the steps: constructing the sensor network to collect the operation state data of power equipment; summarizing and preprocessing data from different sensors through a data concentrator, and removing noise and abnormal data to obtain preprocessed data of each sensor; carrying out fusion processing on the preprocessed data from different sensors to obtain target monitoring data; analyzing the target monitoring data, extracting equipment feature data, inputting the equipment feature data into a health state evaluation model, and evaluating the health state of the equipment through the health state evaluation model to obtain an evaluation result; when the health state evaluation result of the equipment shows that the equipment has a potential fault, starting a fault diagnosis program, and sending out corresponding early warning information according to a fault diagnosis result; potential faults of equipment can be found in time, and safe and stable operation of a power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment monitoring, and in particular to an intelligent power equipment health monitoring system integrated with a sensor network. Background Art

[0002] With the continuous development of power systems and the improvement of their intelligence, the requirements for the reliability and stability of power equipment are becoming increasingly higher. Traditional power equipment monitoring methods mainly rely on regular inspections and preventive tests, which have problems such as untimely monitoring, low accuracy, and difficulty in discovering potential faults. The development of sensor technology has made it possible to monitor power equipment in real time. By installing various sensors on power equipment, the operating status data of the equipment, such as temperature, humidity, vibration, current, voltage, etc., can be collected in real time, providing a basis for equipment health monitoring and fault diagnosis. Summary of the invention

[0003] The purpose of the present invention is to solve the above problems and designs an intelligent power equipment health monitoring system integrating a sensor network.

[0004] The present invention provides an intelligent power equipment health monitoring system integrated with a sensor network, wherein the intelligent power equipment health monitoring system integrated with a sensor network comprises a data acquisition module, a data preprocessing module, a data fusion module, a health status assessment module and a fault diagnosis module, wherein: A data acquisition module is used to arrange various types of sensors at key parts of the power equipment to build a sensor network, and the sensors collect the operating status data of the power equipment according to a preset sampling frequency, wherein the key parts of the power equipment include at least transformer windings, circuit breaker contacts and cable connectors, and the sensor network includes at least temperature sensors, humidity sensors, vibration sensors, current sensors and voltage sensors; The data preprocessing module is used to transmit the data collected by the sensor to the data concentrator through wireless communication, and the data from different sensors are aggregated and preprocessed by the data concentrator to remove noise and abnormal data to obtain the preprocessed data of each sensor; A data fusion module is used to fuse the pre-processed data from different sensors using data fusion technology to obtain target monitoring data; A health status assessment module is used to analyze the target monitoring data, extract device feature data, input the device feature data into a health status assessment model, and assess the health status of the device through the health status assessment model to obtain an assessment result; The fault diagnosis module is used to start the fault diagnosis program based on the evaluation result, diagnose the fault type and location of the equipment when the health status evaluation result of the equipment shows that the equipment has a potential fault, and issue corresponding warning information according to the fault diagnosis result.

[0005] Optionally, in a first implementation of the present invention, the data preprocessing module includes a data transmission submodule, a denoising processing submodule, an initialization submodule, an update submodule and an abnormality deletion submodule, wherein: The data transmission submodule is used to use the ZigBee protocol for wireless communication, transmit the data collected by the sensor to the data concentrator, and encrypt the transmitted data using the AES encryption algorithm; The denoising submodule is used to sort the adjacent data points of the data from different sensors, take the middle value as the filtered result, and perform denoising on the data from different sensors through median filtering; The initialization submodule is used to obtain data similarity based on the distance of the denoised data sequence information entropy, and use the data similarity to construct a cell grid, initialize the data list, calculate the shortest distance between the denoised data and the standard data set in each cell, and fill the calculation results, data similarity, and the number of data points into the data list; The updating submodule is used to judge the state of the data according to the dynamic perception vector of the data points in the neighborhood of the updated data list. If the calculated shortest distance is greater than the preset distance threshold, the data in the cell is abnormal data, otherwise it is normal data; The abnormality deletion submodule is used to delete the abnormal data after detecting the abnormal data, and obtain the pre-processed data of each sensor.

[0006] Optionally, in a second implementation of the present invention, the data fusion module includes a weight determination submodule and a coefficient adjustment submodule, wherein: The weight determination submodule is used to obtain the pre-processed data from different sensors and calculate the weight coefficient of each sensor according to the determined weight distribution strategy, wherein the weight distribution strategy is determined according to the installation location of the sensor and the criticality of the measurement object; The coefficient adjustment submodule is used to multiply the pre-processed data of each sensor by the corresponding weight coefficient, add them up to get the weighted average, and calculate the error between the fusion result and the data of each sensor. If the error is large, a genetic algorithm is used to take the accuracy of the fusion result as the objective function, and the weight coefficient is continuously adjusted to complete the data fusion processing.

[0007] Optionally, in a third implementation of the present invention, the health status assessment module includes a data analysis submodule, a data input submodule and a data output submodule, wherein: A data analysis submodule is used to analyze the target monitoring data, perform time domain feature extraction, frequency domain feature extraction and feature parameter extraction, and obtain device feature data; A data input submodule, used to input the device feature data into a health status assessment model, and construct a k-nearest neighbor graph for the device feature data using a multi-metric fusion graph construction method, wherein the health status assessment model is constructed based on a graph convolutional neural network; The data output submodule is used to input the constructed graph structure data into the dual-input residual graph convolutional network, extract data features through the graph convolution layer and the graph pooling layer, and perform feature fusion through the self-attention mechanism. Finally, the evaluation result is output after the fully connected layer and the Softmax function.

[0008] Optionally, in a fourth implementation of the present invention, the data analysis submodule includes a traversal unit, a conversion unit, an extraction unit and a sorting unit, wherein: A traversal unit, used to traverse the target monitoring data, find the maximum value and the minimum value in the target monitoring data, calculate the mean, variance, standard deviation, peak value, peak-to-peak value, and root mean square value of the target monitoring data, and obtain corresponding time domain features; A conversion unit, used to perform Fourier transform on the target monitoring data, convert the target monitoring data from the time domain to the frequency domain, perform main frequency extraction, frequency distribution, and harmonic content analysis, and obtain corresponding frequency domain features; An extraction unit, used to perform time series analysis on the target monitoring data by linear regression to extract trend characteristics of the target monitoring data; The sorting unit is used to sort the extracted time domain features, frequency domain features and feature parameters into device feature data.

[0009] Optionally, in a fifth implementation of the present invention, the data input submodule includes a calculation unit and a construction unit, wherein: A calculation unit, used for calculating the edge weight between every two sample nodes in the device feature data by using a Gaussian kernel function, cosine similarity and a Pearson correlation coefficient, wherein each monitoring variable in the device feature data is regarded as a node, and the relationship between the variables is regarded as an edge with a weight, the Gaussian kernel function measures the similarity between vectors by projecting the vectors into a high-order space, the cosine similarity measures the difference between sample nodes, and the Pearson correlation coefficient measures the linear relationship between two sample nodes; The construction unit is used to add the absolute values ​​of the three calculated correlation matrices and calculate the mean as the final adjacency matrix, and construct a k-nearest neighbor graph.

[0010] Optionally, in a sixth implementation of the present invention, a method for implementing an intelligent power equipment health monitoring system integrated with a sensor network includes the following steps: Arrange various types of sensors at key parts of the power equipment to build a sensor network. The sensors collect the operating status data of the power equipment according to a preset sampling frequency. The key parts of the power equipment include at least transformer windings, circuit breaker contacts and cable connectors. The sensor network includes at least temperature sensors, humidity sensors, vibration sensors, current sensors and voltage sensors. The data collected by the sensor is transmitted to the data concentrator through wireless communication. The data concentrator aggregates and preprocesses the data from different sensors, removes noise and abnormal data, and obtains the preprocessed data of each sensor. Data fusion technology is used to fuse the pre-processed data from different sensors to obtain target monitoring data; Analyze the target monitoring data, extract equipment feature data, input the equipment feature data into a health status assessment model, and assess the health status of the equipment through the health status assessment model to obtain an assessment result; Based on the evaluation results, when the health status evaluation results of the equipment show that the equipment has a potential fault, the fault diagnosis program is started to diagnose the fault type and location of the equipment, and corresponding warning information is issued according to the fault diagnosis results.

[0011] Optionally, in a seventh implementation of the present invention, the data collected by the sensor is transmitted to a data concentrator by wireless communication, and the data from different sensors are aggregated and preprocessed by the data concentrator to remove noise and abnormal data to obtain preprocessed data of each sensor, including: ZigBee protocol is used for wireless communication to transmit the data collected by the sensor to the data concentrator, and the transmitted data is encrypted by AES encryption algorithm; The data from different sensors are denoised by sorting adjacent data points, taking the middle value as the filtered result, and performing median filtering on the data from different sensors; According to the distance of the denoised data sequence information entropy, the data similarity is obtained, and the unit grid is constructed using the data similarity to initialize the data list. In each cell, the shortest distance between the denoised data and the standard data set is calculated, and the calculation results, data similarity, and the number of data points are filled into the data list; The updated data list is used to determine the state of the data according to the dynamic perception vector of the data points in the neighborhood. If the calculated shortest distance is greater than the preset distance threshold, the data in the cell is abnormal data, otherwise it is normal data; When abnormal data is detected, the abnormal data is deleted to obtain the pre-processed data of each sensor.

[0012] Optionally, in an eighth implementation of the present invention, the adopting of data fusion technology to fuse pre-processed data from different sensors to obtain target monitoring data includes: Obtain pre-processed data from different sensors, and calculate the weight coefficient of each sensor according to a determined weight allocation strategy, wherein the weight allocation strategy is determined according to the installation location of the sensor and the criticality of the measurement object; Multiply the preprocessed data of each sensor by the corresponding weight coefficient, add them up to get the weighted average, and calculate the error between the fusion result and the data of each sensor. If the error is large, use a genetic algorithm with the accuracy of the fusion result as the objective function to continuously adjust the weight coefficient to complete the data fusion processing.

[0013] In the technical solution provided by the present invention, real-time monitoring of power equipment is achieved through a sensor network, abnormal conditions of the equipment can be discovered in time, and the reliability and stability of the equipment can be improved; data processing and analysis technology is used to establish a health status assessment model for the equipment, which can accurately assess and predict the health status of the equipment; when a potential fault occurs in the equipment, the fault diagnosis can be carried out in time to determine the type and location of the fault, provide accurate fault information for operation and maintenance personnel, and improve fault handling efficiency; automated data collection, processing and analysis are realized, manual intervention is reduced, and monitoring efficiency and management level are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiment.The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting the invention.

[0015] Figure 1 A schematic diagram of the structure of an intelligent power equipment health monitoring system with integrated sensor network provided in an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a data preprocessing module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention 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 data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0017] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 The structural diagram of the intelligent power equipment health monitoring system integrated with the sensor network provided by the embodiment of the present invention includes a data acquisition module, a data preprocessing module, a data fusion module, a health status assessment module and a fault diagnosis module, wherein: A data acquisition module is used to arrange various types of sensors at key parts of the power equipment to build a sensor network, and the sensors collect the operating status data of the power equipment according to a preset sampling frequency, wherein the key parts of the power equipment include at least transformer windings, circuit breaker contacts and cable connectors, and the sensor network includes at least temperature sensors, humidity sensors, vibration sensors, current sensors and voltage sensors; The data preprocessing module is used to transmit the data collected by the sensor to the data concentrator through wireless communication, and the data from different sensors are aggregated and preprocessed by the data concentrator to remove noise and abnormal data to obtain the preprocessed data of each sensor; A data fusion module is used to fuse the pre-processed data from different sensors using data fusion technology to obtain target monitoring data; The health status assessment module is used to analyze the target monitoring data, extract the equipment characteristic data, input the equipment characteristic data into the health status assessment model, and evaluate the health status of the equipment through the health status assessment model to obtain the assessment result; The fault diagnosis module is used to start the fault diagnosis program based on the evaluation results, diagnose the fault type and location of the equipment when the health status evaluation results of the equipment show that the equipment has a potential fault, and issue corresponding warning information according to the fault diagnosis results.

[0018] In this embodiment, the data fusion module includes a weight determination submodule and a coefficient adjustment submodule, wherein: The weight determination submodule is used to obtain the pre-processed data from different sensors and calculate the weight coefficient of each sensor according to the determined weight distribution strategy, wherein the weight distribution strategy is determined according to the installation location of the sensor and the criticality of the measurement object; The coefficient adjustment submodule is used to multiply the pre-processed data of each sensor by the corresponding weight coefficient, add them up to get the weighted average, and calculate the error between the fusion result and the data of each sensor. If the error is large, a genetic algorithm is used to take the accuracy of the fusion result as the objective function, and the weight coefficient is continuously adjusted to complete the data fusion processing.

[0019] In this embodiment, the health status assessment module includes a data analysis submodule, a data input submodule and a data output submodule, wherein: The data analysis submodule is used to analyze the target monitoring data, extract time domain features, frequency domain features and feature parameters, and obtain device feature data; The data input submodule is used to input the device feature data into the health status assessment model, and construct a k-nearest neighbor graph for the device feature data using a multi-metric fusion graph construction method, wherein the health status assessment model is constructed based on a graph convolutional neural network; The data output submodule is used to input the constructed graph structure data into the dual-input residual graph convolutional network, extract data features through the graph convolution layer and the graph pooling layer, and perform feature fusion through the self-attention mechanism. Finally, the evaluation result is output after the fully connected layer and the Softmax function.

[0020] In this embodiment, the data analysis submodule includes a traversal unit, a conversion unit, an extraction unit and a sorting unit, wherein: A traversal unit is used to traverse the target monitoring data, find the maximum value and the minimum value in the target monitoring data, calculate the mean, variance, standard deviation, peak value, peak-to-peak value, and root mean square value of the target monitoring data, and obtain the corresponding time domain features; A conversion unit is used to perform Fourier transform on the target monitoring data, convert the target monitoring data from the time domain to the frequency domain, perform main frequency extraction, frequency distribution, and harmonic content analysis, and obtain corresponding frequency domain features; An extraction unit, used to perform time series analysis on the target monitoring data by linear regression, and extract trend characteristics of the target monitoring data; The sorting unit is used to sort the extracted time domain features, frequency domain features and feature parameters into device feature data.

[0021] In this embodiment, the data input submodule includes a calculation unit and a construction unit, wherein: A calculation unit, used for calculating the edge weight between every two sample nodes in the device feature data by using a Gaussian kernel function, cosine similarity and Pearson correlation coefficient, wherein each monitoring variable in the device feature data is regarded as a node, and the relationship between the variables is regarded as an edge with weight, the Gaussian kernel function measures the similarity between vectors by projecting the vectors into a high-order space, the cosine similarity measures the difference between the sample nodes, and the Pearson correlation coefficient measures the linear relationship between the two sample nodes; The construction unit is used to add the absolute values ​​of the three calculated correlation matrices and calculate the mean as the final adjacency matrix, and construct a k-nearest neighbor graph.

[0022] See also Figure 2 , a schematic diagram of the structure of a data preprocessing module provided in an embodiment of the present invention, the data preprocessing module includes a data transmission submodule, a denoising processing submodule, an initialization submodule, an update submodule and an abnormality deletion submodule, wherein, The data transmission submodule is used to use the ZigBee protocol for wireless communication, transmit the data collected by the sensor to the data concentrator, and encrypt the transmitted data using the AES encryption algorithm; The denoising submodule is used to sort the adjacent data points of the data from different sensors, take the middle value as the filtered result, and perform denoising on the data from different sensors through median filtering; The initialization submodule is used to obtain data similarity based on the distance of the denoised data sequence information entropy, and use the data similarity to construct a cell grid, initialize the data list, calculate the shortest distance between the denoised data and the standard data set in each cell, and fill the calculation results, data similarity, and the number of data points into the data list; The updating submodule is used to judge the state of the data according to the dynamic perception vector of the data points in the neighborhood of the updated data list. If the calculated shortest distance is greater than the preset distance threshold, the data in the cell is abnormal data, otherwise it is normal data; The abnormality deletion submodule is used to delete the abnormal data after detecting the abnormal data, and obtain the pre-processed data of each sensor.

[0023] Through the implementation of the above scheme, real-time monitoring and health assessment of power equipment can timely discover potential failures of equipment, improve equipment reliability and stability, reduce equipment maintenance costs, and ensure the safe and stable operation of the power system.

[0024] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An intelligent power equipment health monitoring system integrated with a sensor network, characterized in that: The intelligent power equipment health monitoring system integrated with the sensor network includes a data acquisition module, a data preprocessing module, a data fusion module, a health status assessment module and a fault diagnosis module, wherein: A data acquisition module is used to arrange various types of sensors at key parts of the power equipment to build a sensor network, and the sensors collect the operating status data of the power equipment according to a preset sampling frequency, wherein the key parts of the power equipment include at least transformer windings, circuit breaker contacts and cable connectors, and the sensor network includes at least temperature sensors, humidity sensors, vibration sensors, current sensors and voltage sensors; The data preprocessing module is used to transmit the data collected by the sensor to the data concentrator through wireless communication, and the data from different sensors are aggregated and preprocessed by the data concentrator to remove noise and abnormal data to obtain the preprocessed data of each sensor; A data fusion module is used to fuse the pre-processed data from different sensors using data fusion technology to obtain target monitoring data; A health status assessment module is used to analyze the target monitoring data, extract device feature data, input the device feature data into a health status assessment model, and assess the health status of the device through the health status assessment model to obtain an assessment result; The fault diagnosis module is used to start the fault diagnosis program based on the evaluation result, diagnose the fault type and location of the equipment when the health status evaluation result of the equipment shows that the equipment has a potential fault, and issue corresponding warning information according to the fault diagnosis result.

2. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 1, characterized in that: The data preprocessing module includes a data transmission submodule, a denoising processing submodule, an initialization submodule, an update submodule and an abnormality deletion submodule, wherein: The data transmission submodule is used to use the ZigBee protocol for wireless communication, transmit the data collected by the sensor to the data concentrator, and encrypt the transmitted data using the AES encryption algorithm; The denoising submodule is used to sort the adjacent data points of the data from different sensors, take the middle value as the filtered result, and perform denoising on the data from different sensors through median filtering; The initialization submodule is used to obtain data similarity based on the distance of the denoised data sequence information entropy, and use the data similarity to construct a cell grid, initialize the data list, calculate the shortest distance between the denoised data and the standard data set in each cell, and fill the calculation results, data similarity, and the number of data points into the data list; The updating submodule is used to judge the state of the data according to the dynamic perception vector of the data points in the neighborhood of the updated data list. If the calculated shortest distance is greater than the preset distance threshold, the data in the cell is abnormal data, otherwise it is normal data; The abnormality deletion submodule is used to delete the abnormal data after detecting the abnormal data, and obtain the pre-processed data of each sensor.

3. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 1, characterized in that: The data fusion module includes a weight determination submodule and a coefficient adjustment submodule, wherein: The weight determination submodule is used to obtain the pre-processed data from different sensors and calculate the weight coefficient of each sensor according to the determined weight distribution strategy, wherein the weight distribution strategy is determined according to the installation location of the sensor and the criticality of the measurement object; The coefficient adjustment submodule is used to multiply the pre-processed data of each sensor by the corresponding weight coefficient, add them up to get the weighted average, and calculate the error between the fusion result and the data of each sensor. If the error is large, a genetic algorithm is used to take the accuracy of the fusion result as the objective function, and the weight coefficient is continuously adjusted to complete the data fusion processing.

4. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 1, characterized in that: The health status assessment module includes a data analysis submodule, a data input submodule and a data output submodule, wherein: A data analysis submodule is used to analyze the target monitoring data, perform time domain feature extraction, frequency domain feature extraction and feature parameter extraction, and obtain device feature data; A data input submodule, used to input the device feature data into a health status assessment model, and construct a k-nearest neighbor graph for the device feature data using a multi-metric fusion graph construction method, wherein the health status assessment model is constructed based on a graph convolutional neural network; The data output submodule is used to input the constructed graph structure data into the dual-input residual graph convolutional network, extract data features through the graph convolution layer and the graph pooling layer, and perform feature fusion through the self-attention mechanism. Finally, the evaluation result is output after the fully connected layer and the Softmax function.

5. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 4, characterized in that: The data analysis submodule includes a traversal unit, a conversion unit, an extraction unit and a sorting unit, wherein: A traversal unit, used to traverse the target monitoring data, find the maximum value and the minimum value in the target monitoring data, calculate the mean, variance, standard deviation, peak value, peak-to-peak value, and root mean square value of the target monitoring data, and obtain corresponding time domain features; A conversion unit, used to perform Fourier transform on the target monitoring data, convert the target monitoring data from the time domain to the frequency domain, perform main frequency extraction, frequency distribution, and harmonic content analysis, and obtain corresponding frequency domain features; An extraction unit, used to perform time series analysis on the target monitoring data by linear regression to extract trend characteristics of the target monitoring data; The sorting unit is used to sort the extracted time domain features, frequency domain features and feature parameters into device feature data.

6. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 4, characterized in that: The data input submodule includes a calculation unit and a construction unit, wherein: A calculation unit, used for calculating the edge weight between every two sample nodes in the device feature data by using a Gaussian kernel function, cosine similarity and a Pearson correlation coefficient, wherein each monitoring variable in the device feature data is regarded as a node, and the relationship between the variables is regarded as an edge with a weight, the Gaussian kernel function measures the similarity between vectors by projecting the vectors into a high-order space, the cosine similarity measures the difference between sample nodes, and the Pearson correlation coefficient measures the linear relationship between two sample nodes; The construction unit is used to add the absolute values ​​of the three calculated correlation matrices and calculate the mean as the final adjacency matrix, and construct a k-nearest neighbor graph.

7. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 1, characterized in that: The method for implementing an intelligent power equipment health monitoring system integrated with a sensor network comprises the following steps: Arrange various types of sensors at key parts of the power equipment to build a sensor network. The sensors collect the operating status data of the power equipment according to a preset sampling frequency. The key parts of the power equipment include at least transformer windings, circuit breaker contacts and cable connectors. The sensor network includes at least temperature sensors, humidity sensors, vibration sensors, current sensors and voltage sensors. The data collected by the sensor is transmitted to the data concentrator through wireless communication. The data concentrator aggregates and preprocesses the data from different sensors, removes noise and abnormal data, and obtains the preprocessed data of each sensor. Data fusion technology is used to fuse the pre-processed data from different sensors to obtain target monitoring data; Analyze the target monitoring data, extract equipment feature data, input the equipment feature data into a health status assessment model, and assess the health status of the equipment through the health status assessment model to obtain an assessment result; Based on the evaluation results, when the health status evaluation results of the equipment show that the equipment has a potential fault, the fault diagnosis program is started to diagnose the fault type and location of the equipment, and corresponding warning information is issued according to the fault diagnosis results.

8. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 7, characterized in that: The data collected by the sensor is transmitted to the data concentrator by wireless communication, and the data from different sensors are aggregated and preprocessed by the data concentrator to remove noise and abnormal data to obtain the preprocessed data of each sensor, including: ZigBee protocol is used for wireless communication to transmit the data collected by the sensor to the data concentrator, and the transmitted data is encrypted by AES encryption algorithm; The data from different sensors are denoised by sorting adjacent data points, taking the middle value as the filtered result, and performing median filtering on the data from different sensors; According to the distance of the denoised data sequence information entropy, the data similarity is obtained, and the unit grid is constructed using the data similarity to initialize the data list. In each cell, the shortest distance between the denoised data and the standard data set is calculated, and the calculation results, data similarity, and the number of data points are filled into the data list; The updated data list is used to determine the state of the data according to the dynamic perception vector of the data points in the neighborhood. If the calculated shortest distance is greater than the preset distance threshold, the data in the cell is abnormal data, otherwise it is normal data; When abnormal data is detected, the abnormal data is deleted to obtain the pre-processed data of each sensor.

9. The intelligent power equipment health monitoring system integrated with a sensor network as claimed in claim 7, characterized in that: The data fusion technology is used to fuse the pre-processed data from different sensors to obtain target monitoring data, including: Obtain pre-processed data from different sensors, and calculate the weight coefficient of each sensor according to a determined weight distribution strategy, wherein the weight distribution strategy is determined according to the installation location of the sensor and the criticality of the measurement object; Multiply the preprocessed data of each sensor by the corresponding weight coefficient, add them up to get the weighted average, and calculate the error between the fusion result and the data of each sensor. If the error is large, use a genetic algorithm with the accuracy of the fusion result as the objective function to continuously adjust the weight coefficient to complete the data fusion processing.

Images