Power transmission equipment information data maintenance system

Through the integrated environment perception, data fusion, analytical diagnostics and adaptive execution modules, manifold learning and collaborative filtration technology, personalized maintenance strategies and automated execution of power transmission systems are realized, solving maintenance inflexibility and failures in the existing technology. The problem of risk improves the stability and efficiency of the system.

CN119047864APending Publication Date: 2024-11-29GONGYI POWER SUPPLY CO OF STATE GRID HENAN ELECTRIC POWER CO +1

Patent Information

Application Number
CN202410977419.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing power transmission system maintenance technology lacks flexibility and makes it difficult to optimize personalized maintenance strategies, resulting in uneconomical maintenance and an increased risk of failure.

Method used

Adopting environment perception module, data fusion processing module, analysis and diagnosis module, collaborative decision-making module and adaptive execution module, through manifold learning algorithms and collaborative filtering technology, the equipment status is monitored and analyzed in real time, personalized maintenance strategies are generated, and automated devices are used. Perform maintenance tasks.

Benefits of technology

Improve the accuracy and efficiency of maintenance decisions, reduce equipment downtime, and enhance the stability and safety of the transmission system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119047864A_ABST
    Figure CN119047864A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of power system maintenance, in particular to a power transmission equipment information data maintenance system which comprises an environment sensing module, a data fusion processing module, an analysis and diagnosis module, a collaborative decision module and a self-adaptive execution module. Wherein the environment sensing module is used for monitoring and collecting meteorological and physical condition parameters of a power transmission environment; the data fusion processing module is used for carrying out fusion processing on the data acquired by the environment sensing module and the operation data of the power transmission equipment; the analysis and diagnosis module is used for carrying out equipment health state evaluation, fault prediction and maintenance demand prediction; the collaborative decision-making module is used for optimizing a maintenance decision and realizing a personalized and precise maintenance strategy; and the self-adaptive execution module is used for automatically adjusting the priority of maintenance operation and resource allocation. According to the invention, by integrating an advanced data analysis technology and an automatic execution module, the accuracy and efficiency of maintenance decision making are remarkably improved, so that high reliability and economic operation of a power transmission system are ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power system maintenance, and in particular to a power transmission equipment information data maintenance system. Background Art

[0002] In the operation and maintenance of modern power transmission systems, it is crucial to monitor the health status of equipment in real time and predict potential failures. With the advancement of technology, more and more automated and intelligent equipment are introduced to improve the efficiency of detection and maintenance. However, since power transmission equipment usually operates in a complex and changeable environment, involving a large amount of data collection, processing and analysis, it makes it a challenge to accurately assess equipment status and maintenance needs. In addition, traditional maintenance strategies often rely on periodic inspections or fixed rules, which may not only lead to inefficient use of resources, but also increase the risk of system failure due to untimely response.

[0003] In the existing technology, although various sensors and monitoring technologies are used to collect equipment data, the analysis and processing of these data usually need to be transmitted to a remote center, which not only increases the processing delay, but also may affect the accuracy and timeliness of decisions due to data transmission problems. In addition, the existing maintenance decision-making system often lacks sufficient flexibility to adapt to the personalized needs of various equipment, making it difficult to optimize maintenance strategies, resulting in maintenance activities that are neither economical nor efficient. Therefore, developing a system that can optimize maintenance decisions, rationalize resource allocation, and automatically perform maintenance tasks through an intelligent system has become a necessary condition for improving the reliability and efficiency of the power transmission system. Summary of the invention

[0004] Based on the above purpose, the present invention provides a power transmission equipment information data maintenance system.

[0005] The power transmission equipment information data maintenance system includes an environment perception module, a data fusion processing module, an analysis and diagnosis module, a collaborative decision-making module, and an adaptive execution module; wherein:

[0006] Environmental perception module: used to monitor and collect meteorological and physical condition parameters of the power transmission environment in real time, including temperature, humidity, wind speed, wire slack and tilt angle;

[0007] Data fusion processing module: It integrates the data collected by the environmental perception module with the operation data of the power transmission equipment, and uses the edge computing algorithm to perform data preprocessing and preliminary analysis to reduce the amount of data transmission and improve the processing speed;

[0008] Analysis and diagnosis module: Apply manifold learning algorithm to extract deep features from fused data, and perform equipment health status assessment, fault prediction and maintenance demand prediction;

[0009] Collaborative decision-making module: Based on the output of the analysis and diagnosis module and combined with historical maintenance data, the collaborative filtering algorithm optimizes maintenance decisions to achieve personalized and precise maintenance strategies;

[0010] Adaptive execution module: According to the maintenance strategy of the collaborative decision-making module, it automatically adjusts the priority and resource allocation of maintenance operations, and executes maintenance tasks through preset automation devices.

[0011] Further, the environment perception module includes a temperature sensor unit, a humidity sensor unit, a wind speed sensor unit, a wire slack sensor unit and a tilt angle sensor unit; wherein:

[0012] Temperature sensor unit: The temperature of the power transmission environment is measured using thermocouples and resistance temperature detectors, where the thermocouple generates an electromotive force proportional to the temperature through the thermoelectric effect, and the resistance temperature detector detects temperature changes through changes in the resistance of the material;

[0013] Humidity sensor unit: A capacitive humidity sensor and a resistive humidity sensor are used to measure the humidity of the power transmission environment. The capacitive humidity sensor detects humidity through the change in capacitance caused by humidity changes, while the resistive humidity sensor measures humidity through the change in resistance caused by humidity.

[0014] Wind speed sensor unit: Ultrasonic wind speed sensor and thermal wind speed sensor are used to measure the wind speed of the power transmission environment. The ultrasonic wind speed sensor determines the wind speed by measuring the propagation speed of ultrasonic waves in the air, while the thermal wind speed sensor calculates the wind speed by measuring the cooling effect of air flow on the heating element;

[0015] Wire slack sensor unit: uses a fiber Bragg grating sensor and a laser distance sensor to measure the slack of the wire. The fiber Bragg grating sensor measures the strain of the wire by detecting the change in the wavelength of light in the optical fiber, while the laser distance sensor determines the displacement of the wire by measuring the laser reflection time.

[0016] Tilt angle sensor unit: An accelerometer and a gyroscope are used to measure the tilt angle of the transmission tower. The accelerometer measures the tilt angle by detecting the change in acceleration caused by gravity, and the gyroscope determines the tilt angle by measuring the angular velocity.

[0017] Furthermore, the data fusion processing module includes a data receiving unit, a data standardization unit, a data fusion unit and an edge computing processing unit; wherein:

[0018] Data receiving unit: used to receive real-time monitoring data from the environment sensing module, including temperature, humidity, wind speed, wire slack and tilt angle data, and operation data of the power transmission equipment, including current, voltage and power;

[0019] Data standardization unit: converts various received data into a unified format and makes adjustments to ensure consistent data quality and provide standardized input for subsequent analysis;

[0020] Data fusion unit: It uses data fusion algorithm to combine environmental data and equipment operation data, and improves the comprehensive representation ability of data by identifying the correlation and interdependence between data;

[0021] Edge computing processing unit: Use edge computing algorithms to preprocess and analyze data near the data source. The edge computing processing unit uses local computing resources to reduce dependence on the central processing system and performs calculations at the data generation site, thereby reducing data transmission requirements and response time.

[0022] Furthermore, the data fusion unit includes:

[0023] Data synchronization: The data fusion unit first synchronizes the environmental data and equipment operation data to ensure the correspondence of data from different sensors at the same time point; and uses the timestamp alignment method to sort and match data from different sources according to timestamps;

[0024] Data cleaning: Clean the synchronized data to remove noise and outliers to ensure data accuracy, and use the z-score method to detect and remove outliers in the data;

[0025] Data standardization: standardize data from different sources to make them have the same dimension and range; and use the Min-Max normalization method to normalize the data to the [0,1] interval;

[0026] Feature extraction: Extract features from the standardized data to form feature vectors. Specifically, principal component analysis is used to extract features, reduce data dimensions, and retain predetermined information.

[0027] Feature fusion: The feature vectors of environmental data and equipment operation data are fused to form a comprehensive feature vector. Specifically, the weighted average method is used for feature fusion. The formula is: F = αE + βO, where F is the fused feature vector, E is the feature vector of environmental data, O is the feature vector of equipment operation data, α and β are weighting coefficients, satisfying α + β = 1;

[0028] Feature enhancement: Enhance the features of the fused feature vector to improve the comprehensive representation ability of the data. Specifically, the sparse coding method is used to enhance the expression ability of the feature vector. The formula is:

[0029] Where D is the dictionary matrix, which represents the pre-trained feature basis; s is the sparse coding coefficient vector, which represents the representation of the fused feature vector on the feature basis; F is the fused feature vector; λ is the regularization parameter, which is used to balance the reconstruction error and sparsity.

[0030] Furthermore, the edge computing processing unit:

[0031] Receiving fused data: The edge computing processing unit first receives the fused feature vector output by the data fusion unit;

[0032] Anomaly detection: Anomaly detection is performed on the received fused feature vector to identify potential anomalies or fault signals. Specifically, a statistical anomaly detection method is used to calculate the score of each feature vector and compare it with the threshold. The formula is: Among them, v represents each eigenvalue in the eigenvector, μ v and σ v Respectively represent the mean and standard deviation of the feature in historical data, and S represents the standardized score;

[0033] Data compression: Use data compression technology to reduce the amount of data in the feature vector to facilitate efficient transmission; specifically, use the principal component analysis method to extract the main components to reduce the data dimension while retaining the preset important information: Z = X p W p Among them, X p represents the processed eigenvector data matrix, W p is the PCA transformation matrix, and Z is the data after dimensionality reduction;

[0034] Rapid response analysis: Rapid response analysis is performed based on the reduced-dimensional data to predict and respond to potential maintenance needs. Specifically, the decision tree algorithm is used to classify the data to quickly determine the operation or maintenance strategy. The formula is: R = DecisionTree (Z), where R represents the decision result, which is used to indicate the specific response or maintenance operation;

[0035] Optimize data transmission: Based on the results of edge computing processing, determine the data that needs to be transmitted to the analysis and diagnosis module to optimize the overall system performance and response speed; specifically, determine the priority of its transmission by calculating the information entropy of the data. The information entropy of the data measures the amount of information contained in the data. The greater the amount of information, the greater its contribution to analysis and diagnosis. The specific formula is: Among them, H(X) is the information entropy of data set X, x i is the i-th element in the data set, p(x i ) is the probability distribution function of the element, indicating that among all the data x iThe probability of occurrence; then, according to the information entropy of each data block, a transmission threshold T is set. Only when the information entropy of the data block is higher than this threshold, the data block will be transmitted to the analysis and diagnosis module. The judgment formula is: Wherein, Transmit(X) represents the decision function of whether data block X is transmitted.

[0036] Furthermore, the analysis and diagnosis module includes a feature extraction unit, a health status assessment unit, a fault prediction unit and a maintenance demand prediction unit; wherein:

[0037] Feature extraction unit: used to receive optimized data from the edge computing processing unit, and extract low-dimensional embedding structure from the data based on the manifold learning algorithm of local linear embedding to capture the intrinsic geometric features in complex data. The manifold learning algorithm can reduce the dimension of the data while retaining the characteristics of the data and generate high-quality feature vectors, providing a basis for subsequent analysis;

[0038] Health status assessment unit: Use the feature vector output by the feature extraction unit to conduct real-time assessment of the current health status of the power transmission equipment. Specifically, the pre-established equipment health model is used to assess the operating status of the equipment, including the degree of wear and performance degradation of the equipment, to determine whether the equipment is in normal working condition.

[0039] Fault prediction unit: Based on the data provided by the feature extraction unit, the system uses time series analysis and machine learning algorithms to predict the future operating status of the equipment, identify potential failure modes, and predict the probability and time of failure.

[0040] Maintenance demand prediction unit: Based on the results of the health status assessment unit and the fault prediction unit, it analyzes the maintenance needs of the equipment. By comprehensively considering the health status and failure risks of the equipment, it generates specific maintenance plans and suggestions, including when to perform maintenance and the resources required, to ensure the continuous and efficient operation of the equipment.

[0041] Furthermore, the feature extraction unit includes:

[0042] Data reception: receiving optimized data from the edge computing processing unit;

[0043] Construct an adjacency graph: Use the k-nearest neighbor algorithm to construct an adjacency graph of data points. i Only its k nearest neighbors x j Through edge connections, it lays the foundation for weight optimization of local linear embedding algorithm;

[0044] Weight calculation: In the local linear embedding algorithm, each data point attempts to reconstruct itself through a linear combination of its k nearest neighbors to maintain the relationship between data points in the local neighborhood, by calculating the reconstruction weight w corresponding to the neighborsij , so that the reconstruction error is minimized. The specific calculation formula is:

[0045] Among them, ∈(W) represents the reconstruction error, N(i) represents the set of neighbor nodes of point i, and x i is the high-dimensional representation of the i-th data point, w ij is the weight of the i-th data point when reconstructed through the j-th neighbor;

[0046] Optimize reconstruction weights: For each data point, optimize the reconstruction weights of its neighbors to minimize the overall reconstruction error and satisfy the constraint that the sum of the weights is 1. The formula is: ∑ j∈N(i) w ij =1foralli, where ∑ j∈N(i) w ij represents the sum of the weights of all neighbors, which must be 1 to ensure that the reconstructed linear combination accurately reflects the original data points;

[0047] Mapping to low-dimensional space: Mapping the data from high-dimensional space to low-dimensional space using the optimized reconstruction weights is done by solving the following low-dimensional embedding optimization problem: Among them, Φ(Y) represents the reconstruction error in the low-dimensional space, y i is the low-dimensional representation of the i-th data point, w ij is the neighbor reconstruction weight, ∑ j∈N(i) w ij y j represents the reconstruction of the i-th data point through its neighbors in the low-dimensional space.

[0048] Furthermore, the maintenance demand prediction unit includes:

[0049] Data collection: The maintenance demand prediction unit first receives the equipment health status data from the health status assessment unit, including the current wear and tear of the equipment and performance degradation; at the same time, it receives the fault prediction results from the fault prediction unit, including potential fault modes, probability of fault occurrence and time estimate;

[0050] Comprehensive analysis: Combine health status data and fault prediction results to conduct a comprehensive analysis to evaluate the overall operating status and risk level of the equipment. Specifically, Bayesian network is used for risk assessment to calculate the health status and fault probability of the equipment at different time nodes. The Bayesian network model formula is expressed as: Among them, P(X) represents the joint probability distribution of the entire system, X i Represents the state variable of the device, Parents(X i ) indicates the influence of X i The parent node set of

[0051] Priority assessment: Based on the comprehensive analysis results, the maintenance needs of the equipment are prioritized to determine the equipment that needs immediate maintenance. The maintenance tasks are sorted using a multi-attribute decision-making method, taking into account factors such as failure risk, health status, equipment importance, and maintenance resource availability. The multi-attribute decision-making formula is expressed as: Among them, S i is the priority score of device i, w j is the weight of attribute j, v ij is the score of device i on attribute j;

[0052] Maintenance strategy generation: Based on the priority evaluation results, a specific maintenance strategy is generated, including maintenance time, maintenance content, and required resources. The dynamic programming method is used to optimize the maintenance plan to minimize downtime and cost while maximizing the maintenance effect. The dynamic programming optimization formula is expressed as: V(t) = max a [R(t,a)+∑ t′ P(t′|t,a)V(t′)], where V(t) represents the optimal value at time point t, R(t,a) represents the immediate benefit of taking action a at time point t, and P(t′|t,a) represents the transition probability of taking action a from time point t to t′;

[0053] Recommendation generation: Generate maintenance recommendations based on maintenance strategies, including specific operating steps and required materials and tools. Maintenance recommendations include preventive maintenance and emergency maintenance to ensure the stability and reliability of equipment under different operating conditions.

[0054] Furthermore, the collaborative decision module specifically includes a data receiving unit, a data preprocessing unit, a similarity calculation unit, a collaborative filtering unit and a decision generation unit; wherein:

[0055] Data receiving unit: receives output data from the analysis and diagnosis module, including equipment health status assessment and fault prediction results; and also receives historical maintenance data, including maintenance records, fault history, and equipment performance change information;

[0056] Data preprocessing unit: preprocesses the received data, including data cleaning, standardization and feature extraction, to ensure data quality and consistency;

[0057] Similarity calculation unit: Calculates the similarity between the current device and similar devices in historical data, so as to make maintenance decisions through collaborative filtering algorithm. Specifically, the cosine similarity calculation method is used, and the formula is:

[0058] Among them, A and B represent the feature vectors of the current device and the historical device respectively, and n is the number of features;

[0059] Collaborative filtering unit: Based on the similarity calculation results, a collaborative filtering algorithm is used to generate maintenance strategy recommendations for the current equipment. The collaborative filtering algorithm is divided into user-based collaborative filtering and item-based collaborative filtering, which correspond to recommendations between equipment and between maintenance operations respectively.

[0060] Decision-making unit: used to comprehensively consider the health status of the current equipment, fault prediction results and maintenance experience of similar equipment to formulate a specific maintenance operation plan.

[0061] Furthermore, the adaptive execution module includes a strategy receiving unit, a priority adjustment unit, a resource allocation unit, an execution control unit and an automation device control unit; wherein:

[0062] Task receiving unit: used to receive the maintenance strategy from the collaborative decision-making module, including specific maintenance operations, priority sorting and resource requirements;

[0063] Priority adjustment unit: used to analyze the priority of each maintenance task according to the received maintenance strategy to ensure that the most urgent and important tasks are executed first;

[0064] Resource Allocation Unit: Allocate required maintenance resources, including manpower, equipment, and materials, based on the task priorities provided by the Priority Adjustment Unit;

[0065] Execution control unit: Assigns maintenance tasks to corresponding automation devices or maintenance personnel and provides operation instructions and workflow. The execution control unit will also monitor the execution progress of maintenance tasks to ensure that the tasks are carried out as planned;

[0066] Automation device control unit: includes intelligent robots, automatic detection equipment and remote operating systems, which are used to perform corresponding maintenance tasks and to carry out specific operations such as automation device equipment inspection, fault repair and performance tuning according to the instructions of the execution control unit.

[0067] Beneficial effects of the present invention:

[0068] The present invention significantly improves the efficiency of data processing and the accuracy of maintenance decisions by integrating advanced data processing and analysis technologies. By utilizing manifold learning algorithms and collaborative filtering technologies, the system can deeply analyze large amounts of data collected from complex environments and extract key features from them, making equipment health status assessment and fault prediction more accurate. In addition, through an automated decision support module, the system can not only generate personalized maintenance plans, but also dynamically adjust maintenance strategies according to actual conditions and optimize resource allocation, thereby improving the efficiency and effectiveness of overall maintenance.

[0069] The present invention, through preset automation devices, such as intelligent robots and automatic detection equipment, the system can automatically perform complex maintenance tasks, reduce human intervention, and reduce the risk of operational errors. This automated maintenance execution not only speeds up the maintenance response time, but also greatly reduces the downtime of equipment, and improves the operating stability and safety of the power transmission system. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0071] Figure 1 A schematic diagram of a power transmission equipment information data maintenance system according to an embodiment of the present invention;

[0072] Figure 2 Schematic diagram of an analysis and diagnosis module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0073] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0074] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. The words "include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0075] like Figure 1-Figure 2 As shown, the power transmission equipment information data maintenance system includes an environment perception module, a data fusion processing module, an analysis and diagnosis module, a collaborative decision-making module and an adaptive execution module; wherein:

[0076] Environmental perception module: used to monitor and collect meteorological and physical condition parameters of the power transmission environment in real time, including temperature, humidity, wind speed, wire slack and tilt angle, so as to capture the impact of the environment on the equipment;

[0077] Data fusion processing module: It integrates the data collected by the environmental perception module with the operation data of the power transmission equipment, and uses the edge computing algorithm to perform data preprocessing and preliminary analysis to reduce the amount of data transmission and improve the processing speed;

[0078] Analysis and diagnosis module: Apply manifold learning algorithm to extract deep features from fused data, and perform equipment health status assessment, fault prediction and maintenance demand prediction. Manifold learning can discover the low-dimensional embedded structure of data, thereby capturing the inherent geometric features in complex data and improving the accuracy and efficiency of analysis;

[0079] Collaborative decision-making module: Based on the output of the analysis and diagnosis module and combined with historical maintenance data, the collaborative filtering algorithm optimizes maintenance decisions to achieve personalized and precise maintenance strategies;

[0080] Adaptive execution module: According to the maintenance strategy of the collaborative decision-making module, it automatically adjusts the priority and resource allocation of maintenance operations, and executes maintenance tasks through preset automation devices.

[0081] The environment sensing module includes a temperature sensor unit, a humidity sensor unit, a wind speed sensor unit, a wire slack sensor unit and a tilt angle sensor unit; wherein:

[0082] Temperature sensor unit: The temperature of the power transmission environment is measured using thermocouples and resistance temperature detectors (RTDs). The thermocouples generate an electromotive force proportional to the temperature through the thermoelectric effect, while the resistance temperature detectors detect temperature changes through changes in the resistance of the material.

[0083] Humidity sensor unit: A capacitive humidity sensor and a resistive humidity sensor are used to measure the humidity of the power transmission environment. The capacitive humidity sensor detects humidity through the change in capacitance caused by humidity changes, while the resistive humidity sensor measures humidity through the change in resistance caused by humidity.

[0084] Wind speed sensor unit: Ultrasonic wind speed sensor and thermal wind speed sensor are used to measure the wind speed of the power transmission environment. The ultrasonic wind speed sensor determines the wind speed by measuring the propagation speed of ultrasonic waves in the air, while the thermal wind speed sensor calculates the wind speed by measuring the cooling effect of air flow on the heating element;

[0085] Wire slack sensor unit: uses a fiber Bragg grating (FBG) sensor and a laser distance sensor to measure the slack of the wire. The fiber Bragg grating sensor measures the strain of the wire by detecting the change in the wavelength of light in the optical fiber, while the laser distance sensor determines the displacement of the wire by measuring the laser reflection time.

[0086] Tilt angle sensor unit: An accelerometer and a gyroscope are used to measure the tilt angle of the transmission tower. The accelerometer measures the tilt angle by detecting the acceleration change caused by gravity, and the gyroscope determines the tilt angle by measuring the angular velocity. The above-mentioned environmental perception module monitors the meteorological and physical conditions of the transmission environment in real time through these high-precision sensor units to ensure the accuracy and reliability of the data, and provide high-quality input data for subsequent data fusion processing and analysis and diagnosis.

[0087] The data fusion processing module includes a data receiving unit, a data standardization unit, a data fusion unit and an edge computing processing unit; wherein:

[0088] Data receiving unit: used to receive real-time monitoring data from the environment sensing module, including temperature, humidity, wind speed, wire slack and tilt angle data, and operation data of the power transmission equipment, including current, voltage and power;

[0089] Data standardization unit: converts various received data into a unified format and makes adjustments to ensure consistent data quality and provide standardized input for subsequent analysis;

[0090] Data fusion unit: It uses data fusion algorithm to combine environmental data and equipment operation data, and improves the comprehensive representation ability of data by identifying the correlation and interdependence between data;

[0091] Edge computing processing unit: Edge computing algorithms are used near the data source to preprocess and perform preliminary analysis of data. The edge computing processing unit uses local computing resources to reduce dependence on the central processing system, and performs calculations at the data generation location, thereby reducing data transmission requirements and response time. The above-mentioned data fusion processing module adopts edge computing algorithms and efficient data fusion algorithms to perform fast and effective data processing at the data generation location, which not only reduces the data processing burden of the central server, but also shortens the data processing time, enabling the entire system to respond and process real-time data of the transmission equipment more quickly.

[0092] The data fusion unit includes:

[0093] Data synchronization: The data fusion unit first synchronizes the environmental data and equipment operation data to ensure the correspondence of data from different sensors at the same time point; and uses the timestamp alignment method to sort and match data from different sources according to timestamps;

[0094] Data cleaning: Clean the synchronized data to remove noise and outliers to ensure data accuracy, and use the z-score method to detect and remove outliers. The formula is: Where x is the value of the measured data point, μ is the average value of the same type of data, σ is the standard deviation of the same type of data, and z represents the standardized score, which is used to identify data points that deviate greatly from the average value;

[0095] Data standardization: Standardize data from different sources to make them have the same dimension and range; and use the Min-Max standardization method to normalize the data to the [0,1] interval. The formula is:

[0096] Among them, x is the value of the original data point, x min and x max are the minimum and maximum values ​​of the same type of data, respectively, and x′ is the normalized data value;

[0097] Feature extraction: Extract features from the standardized data to form feature vectors. Specifically, principal component analysis (PCA) is used for feature extraction to reduce the data dimension while retaining the predetermined information. The formula is: Z = XW, where X is the standardized data matrix, W is the feature vector matrix, and Z is the feature matrix after dimensionality reduction.

[0098] Feature fusion: The feature vectors of environmental data and equipment operation data are fused to form a comprehensive feature vector. Specifically, the weighted average method is used for feature fusion. The formula is: F = αE + βO, where F is the fused feature vector, E is the feature vector of environmental data, O is the feature vector of equipment operation data, α and β are weighting coefficients, satisfying α + β = 1;

[0099] Feature enhancement: Enhance the features of the fused feature vector to improve the comprehensive representation ability of the data. Specifically, the sparse coding method is used to enhance the expression ability of the feature vector. The formula is:

[0100] Among them, D is the dictionary matrix, which represents the pre-trained feature basis; s is the sparse coding coefficient vector, which represents the representation of the fused feature vector on the feature basis; F is the fused feature vector; λ is the regularization parameter, which is used to balance the reconstruction error and sparsity; through the above steps, the data fusion unit can efficiently fuse the environmental data and equipment operation data to generate a feature vector with high comprehensive representation ability, providing high-quality data input for subsequent analysis and diagnosis.

[0101] Edge computing processing unit:

[0102] Receiving fused data: The edge computing processing unit first receives the fused feature vectors output by the data fusion unit. These vectors contain comprehensive information of environmental perception data and device operation data.

[0103] Anomaly detection: Anomaly detection is performed on the received fused feature vector to identify potential anomalies or fault signals. Specifically, a statistical anomaly detection method is used to calculate the score of each feature vector and compare it with the threshold. The formula is: Among them, v represents each eigenvalue in the eigenvector, μ v and σ v Respectively represent the mean and standard deviation of the feature in historical data, and S represents the standardized score;

[0104] Data compression: Use data compression technology to reduce the amount of data in the feature vector for efficient transmission; specifically use the principal component analysis (PCA) method to extract the main components to reduce the data dimension while retaining the preset important information: Z = X p W p Among them, X p represents the processed eigenvector data matrix, W p is the PCA transformation matrix, and Z is the data after dimensionality reduction;

[0105] Rapid response analysis: Rapid response analysis is performed based on the reduced-dimensional data to predict and respond to potential maintenance needs. Specifically, the decision tree algorithm is used to classify the data to quickly determine the operation or maintenance strategy. The formula is: R = DecisionTree (Z), where R represents the decision result, which is used to indicate the specific response or maintenance operation;

[0106] Optimize data transmission: Based on the results of edge computing processing, determine the data that needs to be transmitted to the analysis and diagnosis module to optimize the overall system performance and response speed; specifically, determine the priority of its transmission by calculating the information entropy of the data. The information entropy of the data measures the amount of information contained in the data. The greater the amount of information, the greater its contribution to analysis and diagnosis. The specific formula is: Among them, H(X) is the information entropy of data set X, x i is the i-th element in the data set, p(x i ) is the probability distribution function of the element, indicating that among all the data x i The probability of occurrence; then, according to the information entropy of each data block, a transmission threshold T is set. Only when the information entropy of the data block is higher than this threshold, the data block will be transmitted to the analysis and diagnosis module. The judgment formula is: Where Transmit(X) represents the decision function of whether data block X is transmitted;

[0107] Through the above steps, the edge computing processing unit not only preprocesses and preliminarily analyzes the data, but also effectively screens the data through the calculation of information entropy, ensuring that only data with large amounts of information and high contributions to the analysis and diagnosis module are processed and transmitted first. This method significantly improves the efficiency of data processing and transmission, reduces the network burden, and optimizes the response time and performance of the entire system.

[0108] The analysis and diagnosis module includes a feature extraction unit, a health status assessment unit, a fault prediction unit, and a maintenance demand prediction unit; wherein:

[0109] Feature extraction unit: used to receive optimized data from the edge computing processing unit, and extract low-dimensional embedding structure from the data based on the manifold learning algorithm of local linear embedding to capture the intrinsic geometric features in complex data. The manifold learning algorithm can reduce the dimension of the data while retaining the characteristics of the data and generate high-quality feature vectors, providing a basis for subsequent analysis;

[0110] Health status assessment unit: Use the feature vector output by the feature extraction unit to conduct real-time assessment of the current health status of the power transmission equipment. Specifically, the pre-established equipment health model is used to assess the operating status of the equipment, including the degree of wear and performance degradation of the equipment, to determine whether the equipment is in normal working condition.

[0111] Fault prediction unit: Based on the data provided by the feature extraction unit, the system uses time series analysis and machine learning algorithms to predict the future operating status of the equipment, identify potential failure modes, predict the probability and time of failure, and provide early warnings to ensure that the maintenance team can take timely measures.

[0112] Maintenance demand prediction unit: Based on the results of the health status assessment unit and the fault prediction unit, it analyzes the maintenance needs of the equipment, and generates specific maintenance plans and suggestions by comprehensively considering the health status and fault risks of the equipment, including when to perform maintenance and the resources required, to ensure the continuous and efficient operation of the equipment; the analysis and diagnosis module, through the organic combination of the above units, realizes the extraction of deep-level features from the optimization data, and conducts comprehensive health status assessment, fault prediction and maintenance demand prediction for the equipment, providing a scientific basis and technical support for the efficient and reliable operation of the transmission equipment.

[0113] The feature extraction unit includes:

[0114] Data reception: Receive optimized data from the edge computing processing unit, which has been preliminarily screened and compressed to contain only key features for efficient further analysis;

[0115] Construct an adjacency graph: Use the k-nearest neighbor algorithm to construct an adjacency graph of data points. iOnly its k nearest neighbors x j Through edge connections, it lays the foundation for weight optimization of local linear embedding algorithm;

[0116] Weight calculation: In the local linear embedding algorithm, each data point attempts to reconstruct itself through a linear combination of its k nearest neighbors to maintain the relationship between data points in the local neighborhood, by calculating the reconstruction weight w corresponding to the neighbors ij , so that the reconstruction error is minimized. The specific calculation formula is:

[0117] Among them, ∈(W) represents the reconstruction error, N(i) represents the set of neighbor nodes of point i, and x i is the high-dimensional representation of the i-th data point, w ij is the weight of the i-th data point when reconstructed through the j-th neighbor;

[0118] Optimize reconstruction weights: For each data point, optimize the reconstruction weights of its neighbors to minimize the overall reconstruction error and satisfy the constraint that the sum of the weights is 1. The formula is: ∑ j∈N(i) w ij =1foralli, where ∑ j∈N(i) w ij represents the sum of the weights of all neighbors, which must be 1 to ensure that the reconstructed linear combination accurately reflects the original data points;

[0119] Mapping to low-dimensional space: Mapping the data from high-dimensional space to low-dimensional space using the optimized reconstruction weights is done by solving the following low-dimensional embedding optimization problem: Among them, Φ(Y) represents the reconstruction error in the low-dimensional space, y i is the low-dimensional representation of the i-th data point, w ij is the neighbor reconstruction weight, ∑ j∈N(i) w ij y j Represents the reconstruction of the i-th data point through its neighbors in the low-dimensional space; the application of the local linear embedding algorithm allows the feature extraction unit to reduce its dimension while maintaining the local features of the data, which not only improves the efficiency of data processing, but also enhances the prediction accuracy and efficiency of the subsequent analysis module. This method enables the transmission equipment information data maintenance system to more accurately evaluate the equipment status and predict potential faults, thereby effectively guiding maintenance decisions and operations.

[0120] The maintenance demand forecasting unit includes:

[0121] Data collection: The maintenance demand prediction unit first receives the equipment health status data from the health status assessment unit, including the current wear and tear of the equipment and performance degradation; at the same time, it receives the fault prediction results from the fault prediction unit, including potential fault modes, probability of fault occurrence and time estimate;

[0122] Comprehensive analysis: Combine health status data and fault prediction results to conduct a comprehensive analysis to evaluate the overall operation status and risk level of the equipment. Specifically, Bayesian Network is used for risk assessment to calculate the health status and fault probability of the equipment at different time points. The Bayesian Network model formula is expressed as: Among them, P(X) represents the joint probability distribution of the entire system, X i Represents the state variable of the device, Parents(X i ) indicates the influence of X i The parent node set of

[0123] Priority assessment: Based on the comprehensive analysis results, the maintenance needs of the equipment are prioritized to determine the equipment that needs immediate maintenance. The multi-attribute decision-making method (MADM) is used to sort the maintenance tasks, taking into account factors such as failure risk, health status, equipment importance, and maintenance resource availability. The multi-attribute decision-making formula is expressed as: Among them, S i is the priority score of device i, w j is the weight of attribute j, v ij is the score of device i on attribute j;

[0124] Maintenance strategy generation: Based on the priority evaluation results, a specific maintenance strategy is generated, including maintenance time, maintenance content, and required resources. The dynamic programming method is used to optimize the maintenance plan to minimize downtime and cost while maximizing the maintenance effect. The dynamic programming optimization formula is expressed as: V(t) = max a [R(t,a)+∑ t′ P(t′|t,a)V(t′)], where V(t) represents the optimal value at time point t, R(t,a) represents the immediate benefit of taking action a at time point t, and P(t′|t,a) represents the transition probability of taking action a from time point t to t′;

[0125] Recommendation generation: Generate maintenance recommendations based on maintenance strategies, including specific operating steps and required materials and tools. The maintenance recommendations include preventive maintenance and emergency maintenance to ensure the stability and reliability of equipment under different operating conditions. By comprehensively analyzing the health status and failure risks of equipment, scientifically evaluate maintenance needs, and generate optimized maintenance plans and recommendations, which not only improves the pertinence and effectiveness of maintenance work, but also reduces the unplanned downtime and maintenance costs of equipment, ensuring efficient, safe and reliable operation of transmission equipment.

[0126] The collaborative decision-making module specifically includes a data receiving unit, a data preprocessing unit, a similarity calculation unit, a collaborative filtering unit and a decision generation unit; wherein:

[0127] Data receiving unit: receives output data from the analysis and diagnosis module, including equipment health status assessment and fault prediction results; and also receives historical maintenance data, including maintenance records, fault history, and equipment performance change information;

[0128] Data preprocessing unit: preprocesses the received data, including data cleaning, standardization and feature extraction, to ensure data quality and consistency. Data cleaning is achieved by removing outliers and noise, standardization is achieved by converting data to the same dimension, and feature extraction is achieved through dimensionality reduction and feature selection techniques;

[0129] Similarity calculation unit: Calculates the similarity between the current device and similar devices in historical data, so as to make maintenance decisions through collaborative filtering algorithm. Specifically, the cosine similarity calculation method is used, and the formula is:

[0130] Among them, A and B represent the feature vectors of the current device and the historical device respectively, and n is the number of features;

[0131] Collaborative filtering unit: Based on the similarity calculation results, a collaborative filtering algorithm is used to generate maintenance strategy recommendations for the current equipment. The collaborative filtering algorithm is divided into user-based collaborative filtering and item-based collaborative filtering, which correspond to recommendations between equipment and between maintenance operations respectively.

[0132] Decision generation unit: used to comprehensively consider the health status of the current equipment, fault prediction results and maintenance experience of similar equipment, and formulate specific maintenance operation plans; through the organic combination of the above units, the collaborative decision-making module can generate personalized and precise maintenance strategies based on the output of the analysis and diagnosis module and historical maintenance data through collaborative filtering algorithms, thereby optimizing equipment maintenance decisions.

[0133] The adaptive execution module includes a strategy receiving unit, a priority adjustment unit, a resource allocation unit, an execution control unit and an automation device control unit; wherein:

[0134] Task receiving unit: used to receive the maintenance strategy from the collaborative decision-making module, including specific maintenance operations, priority sorting and resource requirements;

[0135] Priority adjustment unit: used to analyze the priority of each maintenance task according to the received maintenance strategy, to ensure that the most urgent and important tasks are executed first. Priority adjustment is based on the health status of the equipment, the risk of failure and the overall operation of the system to ensure the timeliness and accuracy of maintenance decisions;

[0136] Resource Allocation Unit: Allocate required maintenance resources, including manpower, equipment, and materials, based on the task priorities provided by the Priority Adjustment Unit;

[0137] Execution control unit: Assigns maintenance tasks to corresponding automation devices or maintenance personnel and provides operation instructions and workflow. The execution control unit will also monitor the execution progress of maintenance tasks to ensure that the tasks are carried out as planned;

[0138] Automation device control unit: includes intelligent robots, automatic detection equipment and remote operating systems, which are used to execute corresponding maintenance tasks and perform specific operations such as automation device inspection, fault repair and performance tuning according to the instructions of the execution control unit; through the organic combination of the above units, the adaptive execution module can automatically adjust the priority and resource allocation of maintenance operations according to the maintenance strategy of the collaborative decision-making module, and efficiently execute maintenance tasks through preset automation devices.

[0139] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A power transmission equipment information data maintenance system, characterized in that: It includes environment perception module, data fusion processing module, analysis and diagnosis module, collaborative decision-making module and adaptive execution module; among which: Environmental perception module: used to monitor and collect meteorological and physical condition parameters of the power transmission environment in real time, including temperature, humidity, wind speed, wire slack and tilt angle; Data fusion processing module: It integrates the data collected by the environmental perception module with the operation data of the power transmission equipment, and uses the edge computing algorithm to perform data preprocessing and preliminary analysis to reduce the amount of data transmission and improve the processing speed; Analysis and diagnosis module: Apply manifold learning algorithm to extract deep features from fused data, and perform equipment health status assessment, fault prediction and maintenance demand prediction; Collaborative decision-making module: Based on the output of the analysis and diagnosis module and combined with historical maintenance data, the collaborative filtering algorithm optimizes maintenance decisions to achieve personalized and precise maintenance strategies; Adaptive execution module: According to the maintenance strategy of the collaborative decision-making module, it automatically adjusts the priority and resource allocation of maintenance operations, and executes maintenance tasks through preset automation devices.

2. The power transmission equipment information data maintenance system according to claim 1, characterized in that: The environment sensing module includes a temperature sensor unit, a humidity sensor unit, a wind speed sensor unit, a wire slack sensor unit and a tilt angle sensor unit; wherein: Temperature sensor unit: The temperature of the power transmission environment is measured using thermocouples and resistance temperature detectors, where the thermocouple generates an electromotive force proportional to the temperature through the thermoelectric effect, and the resistance temperature detector detects temperature changes through changes in the resistance of the material; Humidity sensor unit: A capacitive humidity sensor and a resistive humidity sensor are used to measure the humidity of the power transmission environment. The capacitive humidity sensor detects humidity through the change in capacitance caused by humidity changes, while the resistive humidity sensor measures humidity through the change in resistance caused by humidity. Wind speed sensor unit: Ultrasonic wind speed sensor and thermal wind speed sensor are used to measure the wind speed of the power transmission environment. The ultrasonic wind speed sensor determines the wind speed by measuring the propagation speed of ultrasonic waves in the air, while the thermal wind speed sensor calculates the wind speed by measuring the cooling effect of air flow on the heating element; Wire slack sensor unit: uses a fiber Bragg grating sensor and a laser distance sensor to measure the slack of the wire. The fiber Bragg grating sensor measures the strain of the wire by detecting the change in the wavelength of light in the optical fiber, while the laser distance sensor determines the displacement of the wire by measuring the laser reflection time. Tilt angle sensor unit: An accelerometer and a gyroscope are used to measure the tilt angle of the transmission tower. The accelerometer measures the tilt angle by detecting the change in acceleration caused by gravity, and the gyroscope determines the tilt angle by measuring the angular velocity.

3. The power transmission equipment information data maintenance system according to claim 1, characterized in that: The data fusion processing module includes a data receiving unit, a data standardization unit, a data fusion unit and an edge computing processing unit; wherein: Data receiving unit: used to receive real-time monitoring data from the environment sensing module, including temperature, humidity, wind speed, wire slack and tilt angle data, and operation data of the power transmission equipment, including current, voltage and power; Data standardization unit: converts various received data into a unified format and makes adjustments to ensure consistent data quality and provide standardized input for subsequent analysis; Data fusion unit: It uses data fusion algorithm to combine environmental data and equipment operation data, and improves the comprehensive representation ability of data by identifying the correlation and interdependence between data; Edge computing processing unit: Use edge computing algorithms to preprocess and analyze data near the data source. The edge computing processing unit uses local computing resources to reduce dependence on the central processing system and performs calculations at the data generation site, thereby reducing data transmission requirements and response time.

4. The power transmission equipment information data maintenance system according to claim 3, characterized in that: The data fusion unit comprises: Data synchronization: The data fusion unit first synchronizes the environmental data and equipment operation data to ensure the correspondence of data from different sensors at the same time point; and uses the timestamp alignment method to sort and match data from different sources according to timestamps; Data cleaning: Clean the synchronized data to remove noise and outliers to ensure data accuracy, and use the z-score method to detect and remove outliers in the data; Data standardization: standardize data from different sources to make them have the same dimension and range; and use the Min-Max normalization method to normalize the data to the [0,1] interval; Feature extraction: Extract features from the standardized data to form feature vectors. Specifically, principal component analysis is used to extract features, reduce data dimensions, and retain predetermined information. Feature fusion: The feature vectors of environmental data and equipment operation data are fused to form a comprehensive feature vector. Specifically, the weighted average method is used for feature fusion. The formula is: F = αE + βO, where F is the fused feature vector, E is the feature vector of environmental data, O is the feature vector of equipment operation data, α and β are weighting coefficients, satisfying α + β = 1; Feature enhancement: Enhance the features of the fused feature vector to improve the comprehensive representation ability of the data. Specifically, the sparse coding method is used to enhance the expression ability of the feature vector. The formula is: Where D is the dictionary matrix, which represents the pre-trained feature basis; s is the sparse coding coefficient vector, which represents the representation of the fused feature vector on the feature basis; F is the fused feature vector; λ is the regularization parameter, which is used to balance the reconstruction error and sparsity.

5. The power transmission equipment information data maintenance system according to claim 4, characterized in that: The edge computing processing unit: Receiving fused data: The edge computing processing unit first receives the fused feature vector output by the data fusion unit; Anomaly detection: Anomaly detection is performed on the received fused feature vector to identify potential anomalies or fault signals. Specifically, a statistical anomaly detection method is used to calculate the score of each feature vector and compare it with the threshold. The formula is: Among them, v represents each eigenvalue in the eigenvector, μ v and σ v Respectively represent the mean and standard deviation of the feature in historical data, and S represents the standardized score; Data compression: Use data compression technology to reduce the amount of data in the feature vector to facilitate efficient transmission; specifically, use the principal component analysis method to extract the main components to reduce the data dimension while retaining the preset important information: Z = X p W p Among them, X p represents the processed eigenvector data matrix, W p is the PCA transformation matrix, and Z is the data after dimensionality reduction; Rapid response analysis: Rapid response analysis is performed based on the reduced-dimensional data to predict and respond to potential maintenance needs. Specifically, the decision tree algorithm is used to classify the data to quickly determine the operation or maintenance strategy. The formula is: R = DecisionTree (Z), where R represents the decision result, which is used to indicate the specific response or maintenance operation; Optimize data transmission: Based on the results of edge computing processing, determine the data that needs to be transmitted to the analysis and diagnosis module to optimize the overall system performance and response speed; specifically, determine the priority of its transmission by calculating the information entropy of the data. The information entropy of the data measures the amount of information contained in the data. The greater the amount of information, the greater its contribution to analysis and diagnosis. The specific formula is: Among them, H(X) is the information entropy of data set X, x i is the i-th element in the data set, p(x i ) is the probability distribution function of the element, indicating that among all the data x i The probability of occurrence; then, according to the information entropy of each data block, a transmission threshold T is set. Only when the information entropy of the data block is higher than this threshold, the data block will be transmitted to the analysis and diagnosis module. The judgment formula is: Wherein, Transmit(X) represents the decision function of whether data block X is transmitted.

6. The power transmission equipment information data maintenance system according to claim 1, characterized in that: The analysis and diagnosis module includes a feature extraction unit, a health status assessment unit, a fault prediction unit and a maintenance demand prediction unit; wherein: Feature extraction unit: used to receive optimized data from the edge computing processing unit, and extract low-dimensional embedding structure from the data based on the manifold learning algorithm of local linear embedding to capture the intrinsic geometric features in complex data. The manifold learning algorithm can reduce the dimension of the data while retaining the characteristics of the data and generate high-quality feature vectors, providing a basis for subsequent analysis; Health status assessment unit: Use the feature vector output by the feature extraction unit to conduct real-time assessment of the current health status of the power transmission equipment. Specifically, the pre-established equipment health model is used to assess the operating status of the equipment, including the degree of wear and performance degradation of the equipment, to determine whether the equipment is in normal working condition. Fault prediction unit: Based on the data provided by the feature extraction unit, the system uses time series analysis and machine learning algorithms to predict the future operating status of the equipment, identify potential failure modes, and predict the probability and time of failure. Maintenance demand prediction unit: Based on the results of the health status assessment unit and the fault prediction unit, it analyzes the maintenance needs of the equipment. By comprehensively considering the health status and failure risks of the equipment, it generates specific maintenance plans and suggestions, including when to perform maintenance and the resources required, to ensure the continuous and efficient operation of the equipment.

7. The power transmission equipment information data maintenance system according to claim 6, characterized in that: The feature extraction unit comprises: Data reception: receiving optimized data from the edge computing processing unit; Construct an adjacency graph: Use the k-nearest neighbor algorithm to construct an adjacency graph of data points. i Only its k nearest neighbors x j Through edge connections, it lays the foundation for weight optimization of local linear embedding algorithm; Weight calculation: In the local linear embedding algorithm, each data point attempts to reconstruct itself through a linear combination of its k nearest neighbors to maintain the relationship between data points in the local neighborhood, by calculating the reconstruction weight w corresponding to the neighbors ij , so that the reconstruction error is minimized. The specific calculation formula is: Among them, ∈(W) represents the reconstruction error, N(i) represents the set of neighbor nodes of point i, and x i is the high-dimensional representation of the i-th data point, w ij is the weight of the i-th data point when reconstructed through the j-th neighbor; Optimize reconstruction weights: For each data point, optimize the reconstruction weights of its neighbors to minimize the overall reconstruction error and satisfy the constraint that the sum of the weights is 1. The formula is: ∑ j∈N(i) w ij =1foralli, where ∑ j∈N(i) w ij represents the sum of the weights of all neighbors, which must be 1 to ensure that the reconstructed linear combination accurately reflects the original data points; Mapping to low-dimensional space: Mapping the data from high-dimensional space to low-dimensional space using the optimized reconstruction weights is done by solving the following low-dimensional embedding optimization problem: Among them, Φ(Y) represents the reconstruction error in the low-dimensional space, y i is the low-dimensional representation of the i-th data point, w ij is the neighbor reconstruction weight, ∑ j∈N(i) w ij y j represents the reconstruction of the i-th data point through its neighbors in the low-dimensional space.

8. The power transmission equipment information data maintenance system according to claim 7, characterized in that: The maintenance demand prediction unit comprises: Data collection: The maintenance demand prediction unit first receives the equipment health status data from the health status assessment unit, including the current wear and tear of the equipment and performance degradation; at the same time, it receives the fault prediction results from the fault prediction unit, including potential fault modes, probability of fault occurrence and time estimate; Comprehensive analysis: Combine health status data and fault prediction results to conduct a comprehensive analysis to evaluate the overall operating status and risk level of the equipment. Specifically, Bayesian network is used for risk assessment to calculate the health status and fault probability of the equipment at different time nodes. The Bayesian network model formula is expressed as: Among them, P(X) represents the joint probability distribution of the entire system, X i Represents the state variable of the device, Parents(X i ) indicates the influence of X i The parent node set of Priority assessment: Based on the comprehensive analysis results, the maintenance needs of the equipment are prioritized to determine the equipment that needs immediate maintenance. The maintenance tasks are sorted using a multi-attribute decision-making method, taking into account factors such as failure risk, health status, equipment importance, and maintenance resource availability. The multi-attribute decision-making formula is expressed as: Among them, S i is the priority score of device i, w j is the weight of attribute j, v ij is the score of device i on attribute j; Maintenance strategy generation: Based on the priority evaluation results, a specific maintenance strategy is generated, including maintenance time, maintenance content, and required resources. The dynamic programming method is used to optimize the maintenance plan to minimize downtime and cost while maximizing the maintenance effect. The dynamic programming optimization formula is expressed as: V(t) = max a [R(t,a)+∑ t′ P(t′|t,a)V(t′)], where V(t) represents the optimal value at time point t, R(t,a) represents the immediate benefit of taking action a at time point t, and P(t′|t,a) represents the transition probability of taking action a from time point t to t′; Recommendation generation: Generate maintenance recommendations based on maintenance strategies, including specific operating steps and required materials and tools. Maintenance recommendations include preventive maintenance and emergency maintenance to ensure the stability and reliability of equipment under different operating conditions.

9. The power transmission equipment information data maintenance system according to claim 1, characterized in that: The collaborative decision module specifically includes a data receiving unit, a data preprocessing unit, a similarity calculation unit, a collaborative filtering unit and a decision generation unit; wherein: Data receiving unit: receives output data from the analysis and diagnosis module, including equipment health status assessment and fault prediction results; and also receives historical maintenance data, including maintenance records, fault history, and equipment performance change information; Data preprocessing unit: preprocesses the received data, including data cleaning, standardization and feature extraction, to ensure data quality and consistency; Similarity calculation unit: Calculates the similarity between the current device and similar devices in historical data, so as to make maintenance decisions through collaborative filtering algorithm. Specifically, the cosine similarity calculation method is used, and the formula is: Among them, A and B represent the feature vectors of the current device and the historical device respectively, and n is the number of features; Collaborative filtering unit: Based on the similarity calculation results, a collaborative filtering algorithm is used to generate maintenance strategy recommendations for the current equipment. The collaborative filtering algorithm is divided into user-based collaborative filtering and item-based collaborative filtering, which correspond to recommendations between equipment and between maintenance operations respectively. Decision-making unit: used to comprehensively consider the health status of the current equipment, fault prediction results and maintenance experience of similar equipment to formulate a specific maintenance operation plan.

10. The power transmission equipment information data maintenance system according to claim 1, characterized in that: The adaptive execution module includes a strategy receiving unit, a priority adjustment unit, a resource allocation unit, an execution control unit and an automation device control unit; wherein: Task receiving unit: used to receive the maintenance strategy from the collaborative decision-making module, including specific maintenance operations, priority sorting and resource requirements; Priority adjustment unit: used to analyze the priority of each maintenance task according to the received maintenance strategy to ensure that the most urgent and important tasks are executed first; Resource Allocation Unit: Allocate required maintenance resources, including manpower, equipment, and materials, based on the task priorities provided by the Priority Adjustment Unit; Execution control unit: Assigns maintenance tasks to corresponding automation devices or maintenance personnel and provides operation instructions and workflow. The execution control unit will also monitor the execution progress of maintenance tasks to ensure that the tasks are carried out as planned; Automation device control unit: includes intelligent robots, automatic detection equipment and remote operating systems, which are used to perform corresponding maintenance tasks and to carry out specific operations such as automation device equipment inspection, fault repair and performance tuning according to the instructions of the control unit.

Citation Information

Patent Citations

  • Power consumption behavior analysis method and system based on multi-feature fusion and improved spectral clustering

    CN114065819A

  • Intelligent operation and maintenance system and method for digital twin substation

    CN118172040A

  • Composition for Skin-lightening and Improving Wrinkle Using an Extract of Maerua edulis

    KR102229943B1

Cited By

  • Predictive maintenance strategy optimization method for transformer area equipment maintenance

    CN119359287A

  • Motor health assessment method based on multi-dimensional parameters

    CN119469285A

  • Fire-fighting equipment maintenance management system and method based on Internet of Things

    CN119809603A

  • New energy garbage truck maintenance management system and method

    CN120031536A

  • Equipment operation and maintenance intelligent decision support system based on large model reasoning

    CN120105294A