Optimal operation and maintenance route decision-making method and system based on risk assessment
Through a risk assessment-based method, combined with density outlier detection, Markov chain model and least squares support vector machine and other technologies, the optimal operation and maintenance route was built, which solved the problem of inefficiency of the traditional operation and maintenance model and realized operation and maintenance Resource optimization and fault prediction ensure grid stability and cost control.
Patent Information
- Application Number
- CN202411122792.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-22
AI Technical Summary
In the operation and maintenance of modern power systems, the traditional operation and maintenance model is inefficient and difficult to detect and deal with potential failures in a timely manner, resulting in high operation and maintenance costs and affecting the reliability and economicality of power supply. The existing technology lacks systematic and comprehensive solutions. It is impossible to effectively integrate and utilize multi-source operation and maintenance data, predict failures and optimize operation and maintenance routes.
Using the optimal operation and maintenance route decision-making method based on risk assessment, by obtaining resource consumption, equipment-related data and fault statistics of multiple maintenance operations, density outlier detection, Markov chain model, principal component analysis and minimum two Multiply support vector machine and other algorithms to build objective functions and constraints for optimal operation and maintenance routes, and optimize resource configuration and fault prediction.
It realizes intelligent optimization of operation and maintenance resources, reduces operation and maintenance costs, improves resource utilization efficiency, accurately evaluates the operating status of transmission lines, reduces the probability of major failures, ensures stable operation of the power grid, and optimizes the operation and maintenance paths to improve operations and maintenance The scientificity and economicality of dimensional management.
Smart Images

Figure CN119005471A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data preprocessing and identification, and in particular to a method and system for making an optimal operation and maintenance route decision based on risk assessment. Background Art
[0002] In the operation and maintenance of modern power systems, transmission lines are the arteries of power transmission, and their stable operation is crucial to ensuring the security of power supply. However, faced with the increasingly complex grid structure, diverse equipment types, and uncertainties in the natural environment and human factors, the operation and maintenance of transmission lines faces huge challenges. The traditional operation and maintenance model relies on experience judgment and regular inspections, which is not only inefficient, but also difficult to detect and effectively respond to potential faults in a timely manner, resulting in high operation and maintenance costs, while affecting the reliability and economy of power supply.
[0003] In recent years, with the rapid development of big data, artificial intelligence, and the Internet of Things, new ideas and methods have been provided for power operation and maintenance management. By integrating multi-source data collection, intelligent analysis, and optimized decision-making technology, accurate configuration of operation and maintenance resources, early warning of faults, scientific assessment of risks, and intelligent planning of operation and maintenance paths can be achieved, thereby significantly improving operation and maintenance efficiency and cost control.
[0004] Technical challenges and requirements: 1. Data complexity: Operation and maintenance data covers multiple dimensions such as equipment status, geographical environment, fault records, etc. The data volume is large and the structure is diverse. How to effectively integrate and mine the value of this data has become a major challenge. 2. Difficulty in fault prediction: The occurrence of different types of faults is random and complex. Accurately predicting faults and responding in a timely manner requires a highly intelligent prediction model. 3. Operation and maintenance route optimization: Under the premise of ensuring the stable operation of the power grid, how to scientifically plan the operation and maintenance route and achieve cost optimization is a complex multi-objective optimization problem. 4. Risk assessment system: Building a system that can not only reflect the operating status of the line but also accurately assess potential risks is crucial to improving the scientific nature of operation and maintenance decisions.
[0005] When dealing with the above problems, existing technologies often focus on optimizing a single link and lack systematic and comprehensive solutions. For example, some studies focus on the construction of fault prediction models, but ignore the optimal allocation of operation and maintenance resources; other studies focus on the planning of operation and maintenance routes, but do not fully consider the comprehensive impact of risk assessment and cost control. Therefore, there is an urgent need for an integrated and intelligent method that can comprehensively cover the entire chain of data collection, processing, analysis, and decision-making to achieve a smart upgrade of operation and maintenance control. Summary of the invention
[0006] In view of the above analysis, the embodiments of the present invention aim to provide a method and system for making an optimal operation and maintenance route decision based on risk assessment, so as to solve the problem of how to improve the scientificity of the operation and maintenance route design in combination with the actual operation status of the transmission line.
[0007] On the one hand, an embodiment of the present invention provides an optimal operation and maintenance route decision method based on risk assessment, including: obtaining basic data, wherein the basic data includes resource consumption, equipment-related data and fault statistics of multiple maintenance operations, wherein the equipment-related information includes the geographical location, environment, region, and power grid of the equipment; using a density-based outlier detection method to perform outlier detection on the basic data, and then interpolating the missing values, wherein the missing values are interpolated according to improved stratified mean interpolation; evaluating the operation status of the transmission line through a Markov chain model and using a principal component analysis algorithm to extract key influencing factors from the external factors after interpolation, predicting the transmission line operation risk level through a least squares support vector machine to identify and classify the transmission line faults, wherein the key influencing factors and the transmission line operation status are used as inputs of the least squares support vector machine; and constructing an objective function set and corresponding constraint conditions for the optimal operation and maintenance route based on the transmission line operation status and the transmission line operation risk level to obtain the optimal operation and maintenance route.
[0008] The beneficial effects of the above technical solution are as follows: Intelligent optimization of operation and maintenance resource allocation: The present invention identifies anomalies and omissions in resource consumption during operation and maintenance operations through meticulous data collection and intelligent analysis, and uses advanced algorithms to accurately complete and correct data to ensure the accuracy and completeness of the decision-making basis. On this basis, optimize resource allocation, reduce unnecessary waste, and achieve a significant reduction in operation and maintenance costs and a substantial improvement in resource utilization efficiency. Fault prediction and intelligent risk management: Combining the Markov chain model with PCA technology, the present invention can not only accurately evaluate the operating status of transmission lines, but also extract key factors that affect line safety from complex data. The constructed least squares support vector machine risk assessment system provides a scientific basis for early intervention and preventive maintenance, effectively reducing the probability of major failures and ensuring stable operation of the power grid.
[0009] Based on the further improvement of the above method, the basic data is divided into different levels according to the equipment type, operation type, or regional characteristics, including: setting a domain radius ε and a minimum number of points MinPts, wherein the domain radius ε represents the proximity relationship between points, and the minimum number of points MinPts represents the minimum number of points required to be reached in the neighborhood of a certain point; for each point in the basic data, the distance between each point and all other points is calculated; according to the calculated distance between each point and all other points, each data point in the basic data is divided into core points, boundary points and noise points, wherein when there are at least the minimum number of points MinPts in the ε neighborhood of the current point, the current point is marked as a core point; when the current point is directly or indirectly connected to the core point, but there are less than MinPts points in the ε neighborhood of the current point, the current point is marked as a boundary point; and when the current point is a data point that is neither a core point nor a boundary point, the current point is marked as a noise point as an outlier.
[0010] The beneficial effects of the above technical solution are as follows: Data collection of multiple operation and maintenance operations: Collecting the resource consumption of multiple operation and maintenance operations and relevant information such as the geographical location, environment, region and power grid of the equipment. Comprehensiveness of data collection: The collection of multiple operation and maintenance operation data ensures the comprehensiveness and richness of the data, providing a solid foundation for subsequent analysis; DBSCAN clustering algorithm for outlier detection: Based on the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm, outliers in the data are detected, which can effectively handle noise and irregular data. Stratified mean interpolation model handles missing values: The improved stratified mean interpolation model fills in the missing values in the data to ensure the integrity and accuracy of the data. Accuracy of outlier detection: The DBSCAN clustering algorithm can identify and eliminate outliers in the data, reduce the impact of abnormal data on the model, and improve data quality; Effectiveness of missing value processing: When processing missing values, the improved stratified mean interpolation model takes into account the hierarchy and correlation of the data, and the filling effect is more accurate.
[0011] Based on the further improvement of the above method, the operation status of the transmission line is evaluated by the Markov chain model and the key influencing factors are extracted from the external factors after interpolation processing by the principal component analysis algorithm. The operation risk level of the transmission line is predicted by the least squares support vector machine, which further includes: evaluating the operation status of the transmission line by the Markov chain model, wherein the Markov chain model is used to average the future state of the transmission line according to the current state of the transmission line to predict the probability of the transmission line transferring from the current state to the fault state; extracting some key factors that have the greatest impact on the operation risk from multiple external factors by the principal component analysis algorithm as the key influencing factors; predicting the operation risk level of the transmission line by the least squares support vector machine, wherein the key influencing factors are used as the input of the least squares support vector machine model to train the least squares support vector machine model.
[0012] Based on the further improvement of the above method, the operation status of the transmission line is evaluated by the Markov chain model and the key influencing factors are extracted from the external factors after interpolation by the principal component analysis algorithm, including: obtaining the influencing factors that may cause the failure of the transmission line from the basic data, and standardizing the influencing factors to form a matrix Q=[Q1,Q2,...,Q n ] T ,in,
[0013]
[0014] According to the relationship between the m influencing factors, a correlation matrix is established: R = (r ij )m×m=Q′Q, and find the eigenvalues of R λ1≥λ2≥...≥λ m ≥0, and the eigenvectors α1,α2,...,α m ; Using the following formula:
[0015]
[0016] Calculate the cumulative impact rate of the first p impact factors in, represents the total impact rate of p influencing factors on the transmission line operation failure, if When the information of m influencing factors is replaced by the p influencing factors, p<m, let H=[H1,H2,...,H p ] T , by calculating the cumulative impact rate of the p impact factors, the impact factors with small contribution rate and redundancy are eliminated; the p impact factors are weighted, and the comprehensive impact value W of the p impact factors on the transmission line when they act together is calculated, and the degree of impact of each impact factor on the transmission line and the weight G are calculated:
[0017] W=λH=[λ1λ2...λ P ]H
[0018] =(λ1α1+λ2α2+...+λ p α p ) T ×[Q1 Q2...Q P ] T ;
[0019] The weight G of the impact factor on the comprehensive score is calculated by the following formula:
[0020] G=[k1,k2,...,k p ]=(λ1α1+λ2α2+...+λ p α p ) T ;
[0021] The Markov process is defined by the following formula:
[0022] O{T t+1 =t t+1 T0=t0,T1=t1,...,T n =t n}=O{T t+1 =t t+1 T n =t n};
[0023] Among them, for the power system, the Markov chain represents the operating state of the transmission line at t+1 t+1 Only the operating state O of the transmission line at time t t The decision is independent of the operating status at any time before t; λ represents the probability of transmission line failure, μ represents the ability to repair the transmission line after failure, and a probability model of the change in the operating status of the transmission line is established by combining multiple factors that affect the failure of the transmission line:
[0024]
[0025] The Markov process in the limit state is expressed by the following formula:
[0026] PM = P;
[0027] P = [p N p1 p2 p3 p4 p5];
[0028]
[0029] Among them, p Nrepresents the probability of normal operation of the transmission line; p1, p2, p3, p4, p5, and p6 represent the probability of transmission line failure caused by six factors, respectively. The probability of transmission line failure caused by external factors is obtained by the following formula:
[0030]
[0031] The beneficial effects of the above technical solution are as follows: Markov chain model evaluates the operating status of the transmission line: The operating status of the transmission line is dynamically evaluated through the Markov chain model, which can capture the law of state change. PCA algorithm extracts key influencing factors: The principal component analysis (PCA) algorithm is used to extract key influencing factors from multidimensional data, reduce data dimensions, and simplify model complexity. Least squares support vector machine constructs risk assessment system: The least squares support vector machine (LSSVM) method is used to establish an operation risk assessment system for transmission lines and predict the probability of risk occurrence. Dynamicity of operation status assessment: The Markov chain model can reflect the changes in the operation status of the transmission line in real time, improving the timeliness and accuracy of the assessment. Simplicity of influencing factor extraction: The PCA algorithm effectively extracts key influencing factors, simplifies the calculation complexity of the model, and retains the main information. Accuracy of risk assessment: The LSSVM method shows good generalization ability and prediction accuracy in risk assessment, and can accurately predict the risk level of transmission lines.
[0032] Based on the further improvement of the above method, the objective function set for constructing the optimal operation and maintenance route further includes: expressing the operation and maintenance time objective function by the following formula:
[0033]
[0034] Among them, the time for maintenance personnel to arrive at the fault point when a fault occurs is set as t i , is the operation and maintenance time of a certain area; t mi,mi+1 From the operation and maintenance point m i To the operation and maintenance point m i+1 The time required; the economic loss objective function is expressed by the following formula:
[0035]
[0036] Among them, Pt mi Operation and maintenance point m i The duration from the fault occurrence point to the completion of power supply restoration; α i is the risk level of occurrence, P l The loss per unit time generated for the stated risk level.
[0037] Based on the further improvement of the above method, the corresponding constraint conditions are further constructed including: preventing the operation and maintenance point from being repeatedly restored, the power grid fault point m i Restoration state constraints:
[0038]
[0039] in, Operation and maintenance for any point; Operation and maintenance point m i To the fault point m i +1 path; the path from the maintenance point warehouse to the operation and maintenance point has no duplication constraints:
[0040]
[0041] in, is any load point; W Dmi Optimize the path for the operation and maintenance point to return to the warehouse point; the power grid operation and maintenance point m i Restoration status constraints; line fault recovery constraints to prevent line repeated restoration; personnel and fault point m i Resource allocation; consider the staff's operation and maintenance response and emergency speed, operation and maintenance equipment and other resource configurations to minimize recovery time:
[0042]
[0043]
[0044]
[0045] in, Any time; The resource is sufficient at any time;
[0046]
[0047] Among them, Z is the resource reserve, Z x The amount of resources carried by the staff, Maintenance materials required for all fault points; grid structure constraints, branch flow constraints and node voltage constraints in power grid constraints:
[0048] g k ∈G R ;
[0049] |P a |≤P amax ;
[0050] U min ≤U≤U max ;
[0051] Among them, gk To restore the regional grid structure; G R Total grid structure collection; P a is the power of branch a; P amax is the maximum allowable power of the branch; U is the node voltage; U min , U max are the lower and upper limits of the node voltage.
[0052] The beneficial effects of the above technical solution are as follows: Construction of a multi-objective set: Construct a multi-objective set including constraints such as cost optimization and grid operation to ensure that the decision takes into account requirements in multiple dimensions. Diversity of decision-making objectives: The construction of a multi-objective set ensures that the decision-making process comprehensively considers multiple factors such as cost and operational reliability, thereby improving the scientificity and comprehensiveness of the decision. Improved Pareto-Zebra Algorithm Solution: The improved Pareto-Zebra algorithm is used to solve multi-objective optimization problems, which can find a balance between multiple objectives and determine the optimal route. Effectiveness of the optimization algorithm: The improved Pareto-Zebra algorithm can quickly converge to the optimal solution when solving multi-objective optimization problems, ensuring the efficiency and rationality of the decision.
[0053] On the other hand, an embodiment of the present invention provides an optimal operation and maintenance route decision system based on risk assessment, including: a data acquisition module, used to acquire basic data, wherein the basic data includes resource consumption, equipment-related data and fault statistics of multiple maintenance operations, wherein the equipment-related information includes the geographical location, environment, region, and power grid of the equipment; a data processing module, used to perform outlier detection on the basic data through a density-based outlier detection method, and then interpolate the missing values, wherein the missing values are interpolated according to the improved stratified mean interpolation; a fault failure classification module, used to evaluate the operation status of the transmission line through a Markov chain model and extract key influencing factors from the external factors after interpolation using a principal component analysis algorithm, and predict the transmission line operation risk level through a least squares support vector machine to identify and classify the transmission line fault, wherein the key influencing factors and the transmission line operation status are used as inputs of the least squares support vector machine; and an operation and maintenance optimal route module, used to construct an objective function set and corresponding constraint conditions for the operation and maintenance optimal route based on the transmission line operation status and the transmission line operation risk level to obtain the operation and maintenance optimal route.
[0054] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:
[0055] 1. Intelligent optimization of operation and maintenance resource allocation: This invention uses meticulous data collection and intelligent analysis to identify anomalies and omissions in resource consumption during operation and maintenance operations, and uses advanced algorithms to accurately complete and correct data to ensure the accuracy and completeness of the decision-making basis. On this basis, it optimizes resource allocation, reduces unnecessary waste, and achieves a significant reduction in operation and maintenance costs and a substantial improvement in resource utilization efficiency.
[0056] 2. Fault prediction and intelligent risk management: Combining the Markov chain model with PCA technology, the present invention can not only accurately evaluate the operating status of the transmission line, but also extract the key factors affecting the line safety from complex data. The constructed least squares support vector machine risk assessment system provides a scientific basis for early intervention and preventive maintenance, effectively reducing the probability of major faults and ensuring the stable operation of the power grid.
[0057] 3. Scientific planning and cost control of operation and maintenance paths: The improved Pareto-Zebra algorithm is used to optimize and solve multi-objective sets (including key constraints such as cost optimization and grid operation stability), realizing intelligent planning of operation and maintenance routes. This innovative strategy not only ensures operation and maintenance efficiency, but also minimizes the interference of operation and maintenance activities on the normal operation of the grid, and improves the flexibility and economy of operation and maintenance management.
[0058] 4. Promote the intelligent transformation of operation and maintenance decision-making: Combining the above technical means, the present invention promotes the transformation of power operation and maintenance from experience-based to data-driven and intelligent decision-making. Through in-depth analysis and pattern recognition of a large amount of operation and maintenance data, it not only improves the fault response speed and processing efficiency, but also provides comprehensive and accurate decision-making support information for the management, accelerates the modernization process of the power operation and maintenance management system, and lays a solid foundation for building a smart grid.
[0059] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] The drawings are only for the purpose of illustrating particular embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like components throughout the drawings.
[0061] Figure 1 is a flow chart of an optimal operation and maintenance route decision method based on risk assessment according to an embodiment of the present invention;
[0062] Figure 2A simplified flow chart of an optimal operation and maintenance route decision method based on risk assessment according to an embodiment of the present invention;
[0063] Figure 3 4 is a block diagram of an optimal operation and maintenance route decision system based on risk assessment according to an embodiment of the present invention. DETAILED DESCRIPTION
[0064] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.
[0065] refer to Figure 2 A specific embodiment of the present invention discloses a method for determining an optimal operation and maintenance route based on risk assessment, comprising: in step S201, obtaining basic data, wherein the basic data includes resource consumption, equipment-related data and fault statistics of multiple maintenance operations, wherein the equipment-related information includes the geographical location, environment, region and power grid of the equipment; in step S202, using a density-based outlier detection method to detect outliers on the basic data, and then interpolating the missing values, wherein the missing values are interpolated according to improved stratified mean interpolation; in step S203, evaluating the operation status of the transmission line through a Markov chain model and extracting key influencing factors from the external factors after interpolation through a principal component analysis algorithm, predicting the operation risk level of the transmission line through a least squares support vector machine to identify and classify the transmission line faults, wherein the key influencing factors and the operation status of the transmission line are used as inputs of the least squares support vector machine; and in step S204, constructing an objective function set and corresponding constraint conditions of the optimal operation and maintenance route based on the operation status of the transmission line and the operation risk level of the transmission line to obtain the optimal operation and maintenance route.
[0066] Compared with the prior art, in the optimal operation and maintenance route decision method based on risk assessment provided in this embodiment, multiple operation and maintenance operation data are collected: the resource consumption of multiple operation and maintenance operations and relevant information such as the geographical location, environment, region and power grid of the equipment are collected. Comprehensiveness of data collection: the collection of multiple operation and maintenance operation data ensures the comprehensiveness and richness of the data, and provides a solid foundation for subsequent analysis; DBSCAN clustering algorithm is used for outlier detection: based on the DBSCAN algorithm to detect outliers in the data, it can effectively handle noise and irregular data. The stratified mean interpolation model handles missing values: the improved stratified mean interpolation model fills in the missing values in the data to ensure the integrity and accuracy of the data. Accuracy of outlier detection: the DBSCAN clustering algorithm can identify and eliminate outliers in the data, reduce the impact of abnormal data on the model, and improve data quality; effectiveness of missing value processing: the improved stratified mean interpolation model takes into account the hierarchy and correlation of the data when processing missing values, and the filling effect is more accurate.
[0067] In the following, reference Figure 2 , each step of the optimal operation and maintenance route decision-making method based on risk assessment according to an embodiment of the present invention is described in detail.
[0068] In step S201, basic data is obtained, wherein the basic data includes resource consumption of multiple maintenance operations, equipment-related data and fault statistics data, and the equipment-related information includes the geographical location, environment, region, and power grid of the equipment.
[0069] The resource consumption of multiple maintenance operations includes manpower consumption, material consumption, time cost, financial cost and energy consumption. Among them, manpower consumption is used to record the number of people involved in each operation and maintenance task, working hours, and professional skill requirements; material consumption includes the number and cost of tools, spare parts, and consumables used; time cost includes the total time from response time to task completion, distinguishing between preparation time, travel time, and actual operation time; financial cost includes direct expenditures of labor, material costs, transportation costs, and indirect costs of equipment rental and outsourcing service fees; energy consumption: fuel and electricity consumption of vehicles used in operation and maintenance operations.
[0070] Equipment-related data includes equipment location and environmental data as well as regional and power grid-related data. (1) Equipment location and environmental information: Longitude and latitude coordinates: Accurately locate the location of the equipment to facilitate geographic information system (GIS) analysis. Topography: Mountainous areas, plains, water areas, etc., which affect the difficulty and strategy of operation and maintenance. Climate conditions: Temperature, humidity, wind speed, precipitation, etc., which have a direct impact on equipment performance and maintenance frequency. Environmental pollution level: Air pollution, salt spray corrosion, sandstorms, etc., which affect the aging speed of equipment. Transportation convenience: Road accessibility and distance to the nearest maintenance station affect response speed and transportation costs. (2) Regional and power grid-related information: Power grid structure: The location of the equipment in the power grid, the relationship between superiors and subordinates, and the contribution to the stability of the power grid. Load conditions: The characteristics of the power load in the equipment service area, peak and valley periods. Historical fault records: Fault types, frequencies, and cause analysis of the same region or connected equipment. Power grid upgrade plan: Future power grid transformation or expansion plans and potential impacts on existing equipment.
[0071] Fault statistics data include fault type, fault occurrence time, fault duration, power loss and fault repair process, among which fault types include insulation damage, overload, short circuit and animal touch; fault occurrence time includes the date and time of the fault to determine whether there is a seasonal pattern in the occurrence of the fault; fault duration refers to the power outage time caused by the fault; power loss refers to the estimation of direct power loss and indirect economic loss caused by the fault; fault repair process refers to the time, cost, measures taken and whether parts are replaced for fault repair.
[0072] In step S202, a density-based outlier detection method is used to perform outlier detection on the basic data, and then the missing values are interpolated, wherein the missing values are interpolated according to improved stratified mean interpolation.
[0073] Specifically, the method of using a density-based outlier detection method to detect outliers on the basic data and then interpolating the missing values further includes: in a stratified identification step, the basic data are divided into different levels according to equipment type, operation type, or regional characteristics; in a mean calculation step within the level, for each record containing missing values, the level to which it belongs is determined, and the missing value is filled with the mean calculated within the level; in a stratified interpolation step, when there are missing value records in the basic data, the missing value is filled with the mean within the level according to the level to which the missing value belongs; in a weighted or complex interpolation step, different weights are assigned to different levels or weighted average or multiple interpolation is used to reflect the internal structure of the basic data, considering the variability within the level and the sample size; and in an iteration and verification step, the stratification and statistics are re-evaluated based on the interpolated basic data, and iteration is performed again until a predetermined convergence criterion or an upper limit on the number of iterations is met.
[0074] Specifically, the basic data is divided into different levels according to the equipment type, operation type, or regional characteristics, including: setting the domain radius ε and the minimum number of points MinPts, wherein the domain radius ε represents the proximity relationship between points, and the minimum number of points MinPts represents the minimum number of points required to be reached in the neighborhood of a certain point; for each point in the basic data, the distance between each point and all other points is calculated; according to the calculated distance between each point and all other points, each data point in the basic data is divided into core points, boundary points and noise points, wherein when there are at least the minimum number of points MinPts in the ε neighborhood of the current point, the current point is marked as a core point; when the current point is directly or indirectly connected to the core point, but there are less than MinPts points in the ε neighborhood of the current point, the current point is marked as a boundary point; and when the current point is a data point that is neither a core point nor a boundary point, the current point is marked as a noise point as an outlier.
[0075] In step S203, the operating status of the transmission line is evaluated by a Markov chain model and the key influencing factors are extracted from the external factors after interpolation using a principal component analysis algorithm. The transmission line operating risk level is predicted by a least squares support vector machine to identify and classify transmission line faults, wherein the key influencing factors and the operating status of the transmission line are used as inputs of the least squares support vector machine.
[0076] Specifically, the operation status of the transmission line is evaluated by a Markov chain model and the key influencing factors are extracted from the external factors after interpolation processing by a principal component analysis algorithm. The operation risk level of the transmission line is predicted by a least squares support vector machine, which further includes: evaluating the operation status of the transmission line by a Markov chain model, wherein the Markov chain model is used to average the future state of the transmission line according to the current state of the transmission line to predict the probability of the transmission line transferring from the current state to a fault state; extracting some key factors that have the greatest impact on the operation risk from multiple external factors by a principal component analysis algorithm as key influencing factors; predicting the operation risk level of the transmission line by a least squares support vector machine, wherein the key influencing factors are used as the input of the least squares support vector machine model to train the least squares support vector machine model.
[0077] The Markov chain model is used to evaluate the operation status of the transmission line and the principal component analysis algorithm is used to extract the key influencing factors from the external factors after interpolation. The influencing factors that may cause transmission line failures are obtained from the basic data, and the influencing factors are standardized to form a matrix Q = [Q1, Q2, ..., Q n ] T ,in,
[0078]
[0079] According to the relationship between m influencing factors, a correlation matrix is established: R = (r ij )m×m=Q′Q, and find the eigenvalues of R λ1≥λ2≥...≥λ m ≥0, and the eigenvectors α1,α2,...,α m ; Use the following formula to calculate the cumulative impact rate of the first p impact factors
[0080]
[0081] in, It represents the total impact rate of p influencing factors on the transmission line operation failure. If When the information of m influencing factors is replaced by p influencing factors, p<m, let H=[H1,H2,...,H p ] T , by calculating the cumulative impact rate of p influencing factors, we can eliminate the influencing factors with small contribution rate and redundancy;
[0082] The p influencing factors are weighted, and the comprehensive impact value W of the p influencing factors acting together on the transmission line is calculated. The degree of influence and weight G of each influencing factor on the transmission line are also calculated:
[0083] W=λH=[λ1 λ2 ... λ P ]H
[0084] =(λ1α1+λ2α2+...+λ p α p ) T ×[Q1 Q2 ... Q P ] T ;
[0085] The weight G of the impact factor on the comprehensive score is calculated by the following formula:
[0086] G=[k1,k2,...,k p ]=(λ1α1+λ2α2+...+λ p α p ) T ;
[0087] The Markov process is defined by the following formula:
[0088] O{T t+1 =t t+1 |T0=t0,T1=t1,...,T n =t n}=O{T t+1 =t t+1|T n =t n};
[0089] Among them, for the power system, the Markov chain represents the operating state of the transmission line at t+1 t+1 Only the operating state O of the transmission line at time t t The decision is independent of the operating status at any time before t; λ represents the probability of transmission line failure, μ represents the ability to repair the transmission line after failure, and a probability model of the change in the operating status of the transmission line is established by combining multiple factors that affect the failure of the transmission line:
[0090]
[0091] The Markov process in the limit state is expressed by the following formula:
[0092] PM = P;
[0093] P = [p N p1 p2 p3 p4 p5];
[0094]
[0095] Among them, p N represents the probability of normal operation of the transmission line; p1, p2, p3, p4, p5, and p6 represent the probability of transmission line failure caused by six factors, respectively. The probability of transmission line failure caused by external factors is obtained by the following formula:
[0096]
[0097] In step S204, an objective function set and corresponding constraint conditions of an optimal operation and maintenance route are constructed based on the operation status of the transmission line and the operation risk level of the transmission line to obtain the optimal operation and maintenance route.
[0098] The objective function set for constructing the optimal operation and maintenance route further includes: the operation and maintenance time objective function is expressed by the following formula:
[0099]
[0100] Among them, the time for maintenance personnel to arrive at the fault point when a fault occurs is set to t i , is the operation and maintenance time of a certain area; t mi,mi+1 From the operation and maintenance point m i To the operation and maintenance point m i+1 the time required;
[0101] The economic loss objective function is expressed by the following formula:
[0102]
[0103] Among them, Pt mi Operation and maintenance point m i The duration from the fault occurrence point to the completion of power supply restoration; α i is the risk level of occurrence, P l The loss per unit time generated by the risk level.
[0104] The corresponding constraints are further constructed to prevent the operation and maintenance points from being restored repeatedly, and the grid fault point m i Restoration state constraints:
[0105]
[0106] in, Operation and maintenance for any point; Operation and maintenance point m i To the fault point m i +1 for the path;
[0107] The path from the maintenance point warehouse to the operation and maintenance point must not be repeated:
[0108]
[0109] in, is any load point; W Dmi Optimize the path from the operation and maintenance point back to the warehouse point;
[0110] Power grid operation and maintenance point m i Restoration status constraints; line fault recovery constraints to prevent line repeated restoration; personnel and fault point m i Resource allocation; consider the staff's operation and maintenance response and emergency speed, operation and maintenance equipment and other resource configurations to minimize recovery time:
[0111]
[0112]
[0113]
[0114] in, Any time; The resource is sufficient at any time;
[0115]
[0116] Among them, Z is the resource reserve, Z x The amount of resources carried by the staff, Repair materials needed for all failure points;
[0117] Grid structure constraints, branch flow constraints and node voltage constraints in power grid constraints
[0118] g k ∈G R ;
[0119] |P a |≤P amax ;
[0120] U min ≤U≤U max ;
[0121] Among them, g k To restore the regional grid structure; G R Total grid structure collection; P a is the power of branch a; P amax is the maximum allowable power of the branch; U is the node voltage; U min , U max are the lower and upper limits of the node voltage.
[0122] 1. The relationship between Markov chain model and least squares support vector machine.
[0123] The relationship between the Markov chain model and the least squares support vector machine in the transmission line operation status evaluation and risk prediction can be understood as a hierarchical analysis process. First, the Markov chain model is used to evaluate the operation status of the transmission line. This step determines the various possible operation states of the line and evaluates its change trend by analyzing the historical data of the transmission line and the probability of state transition.
[0124] After completing the status evaluation, the least squares support vector machine uses the operating status data obtained from the Markov chain model and the key influencing factors extracted from external factors to predict the operating risk of the transmission line. The LSSVM model learns and predicts the operating risk level of the transmission line based on the input influencing factors and status data. The prediction results of LSSVM provide an accurate basis for the risk classification of transmission lines and help further identify potential fault risks. Therefore, the Markov chain model provides a preliminary assessment of the status of the transmission line, while the LSSVM further refines these assessments and provides specific risk level predictions. Together, these two models complete a comprehensive analysis of the operating status of the transmission line.
[0125] 2. The relationship between this step and the objective function step.
[0126] In the above steps, the operating status and risk level of the transmission line are effectively evaluated through the joint application of the Markov chain model and LSSVM. These evaluation results provide important input information for the objective function of constructing the optimal operation and maintenance route. Specifically, the risk level predicted by LSSVM can directly affect the priority of operation and maintenance decisions and route planning.
[0127] The construction of the objective function set and constraint conditions is to develop the optimal action route in actual operation and maintenance to minimize risks, reduce maintenance costs and improve efficiency. These objective functions may include: reducing the time and cost of line maintenance, reducing the probability of failure, optimizing resource allocation, etc. The higher the predicted risk level of the line, the earlier and more prioritized maintenance is usually required. Therefore, the output of LSSVM provides a key optimization indicator for the objective function, ensuring that the planning of the operation and maintenance route can effectively deal with transmission lines of different risk levels, thereby achieving the reasonable allocation and optimization of operation and maintenance resources.
[0128] 3. The relationship between basic data and external factors.
[0129] Basic data and external factors play different but complementary roles in the operation status and risk assessment of transmission lines. Basic data usually include the operation records, equipment conditions, historical fault data, etc. of the transmission line, reflecting the operation status of the transmission line itself. These data are the main basis for the Markov chain model to conduct status evaluation.
[0130] External factors refer to external conditions that affect the operation of transmission lines, such as meteorological data (temperature, humidity, wind speed, etc.), geographical environment (topography, geology, etc.), and external power demand. These external factors can directly or indirectly affect the operation status of transmission lines. Through principal component analysis (PCA), the key factors that have the greatest impact on the operation of transmission lines are extracted from external factors and used as input for LSSVM to help predict the operation risk level.
[0131] Basic data provides basic information about the internal operation of transmission lines, while external factors provide complementary information about environmental and external conditions. The combination of the two can provide a more comprehensive understanding and prediction of the operating status and risks of transmission lines. The analysis of basic data (such as through Markov chain models) and the analysis of external factors (such as through principal component analysis and LSSVM) together constitute a complete framework for transmission line risk assessment.
[0132] In the following, reference Figure 1 The optimal operation and maintenance route decision system based on risk assessment according to an embodiment of the present invention is described in detail by way of a specific example.
[0133] In step S101, resource consumption of multiple operation and maintenance operations and relevant information such as the geographical location, environment, region, power grid, etc. of the corresponding equipment are collected, and relevant information such as the time when different types of faults occur and the scale of power loss are counted.
[0134] (1) Resource consumption of operation and maintenance tasks. Human consumption: record the number of people involved in each operation and maintenance task, working hours, professional skills requirements, etc. Material consumption: including the number and cost of tools, spare parts, and consumable materials used. Time cost: the total time from response time to task completion, distinguishing between preparation time, travel time, and actual operation time. Financial cost: direct expenses such as labor costs, material costs, transportation costs, and indirect costs such as equipment rental, outsourcing service fees, etc. Energy consumption: fuel and electricity consumption of vehicles used in operation and maintenance operations.
[0135] (2) Equipment geographic location and environmental information. Longitude and latitude coordinates: Accurately locate the equipment to facilitate geographic information system (GIS) analysis. Topography: Mountainous areas, plains, water areas, etc., affect the difficulty and strategy of operation and maintenance. Climate conditions: Temperature, humidity, wind speed, precipitation, etc. have a direct impact on equipment performance and maintenance frequency. Environmental pollution level: Air pollution, salt spray corrosion, sandstorms, etc. affect the aging speed of equipment. Transportation convenience: Road accessibility and distance to the nearest maintenance station affect response speed and transportation costs.
[0136] (3) Regional and power grid related information. Power grid structure: the location of the equipment in the power grid, the relationship between superiors and subordinates, and the contribution to the stability of the power grid. Load conditions: the characteristics of the power load in the equipment service area, peak and valley periods. Historical fault records: fault types, frequencies, and cause analysis of the same region or connected equipment. Power grid upgrade plan: future power grid transformation or expansion plans, and potential impacts on existing equipment.
[0137] (4) Fault statistics. Fault type: such as insulation damage, overload, short circuit, animal contact, etc. Occurrence time: date and time of the fault, whether there is a seasonal pattern. Duration: duration of power outage caused by the fault and the impact on users. Power loss: estimate the direct power loss and indirect economic loss caused by the fault. Repair process: time and cost of repair, measures taken, whether parts need to be replaced.
[0138] In step S102, outlier detection and processing are performed based on the data, the DBSCAN clustering algorithm is used to carry out the outlier detection method, and the missing values are processed according to the improved stratified mean interpolation model.
[0139] (1) In the outlier detection method, the density-based clustering algorithm improves the defect of the distance-based outlier algorithm that gives the same weight to all objects. It is suitable for the case where the sample data is unevenly distributed. It can make good outlier judgments for the case where there are different densities in the data, reducing the limitations of use and performing probabilistic evaluation on the judgment of whether it is an outlier. In addition, clusters in reality often have various shapes. Traditional clustering methods such as Kmeans clustering can only process spherical clusters, which obviously cannot meet the needs of clusters of other irregular shapes. Therefore, a DBSCAN clustering algorithm that meets the needs of arbitrary cluster shapes is proposed.
[0140] DBSCAN is a density-based outlier detection method. It belongs to unsupervised learning, is resistant to noise, can process clusters of any shape and size, and does not require the number of clusters to be divided in advance. Therefore, for the collected multi-source data, the internal structure of the data is unknown, and the data type that cannot be predicted before cleaning, this method has strong operability for such data.
[0141] The principle of abnormal data identification is as follows: Define density and neighborhood: First, set two key parameters, namely the neighborhood radius (ε, epsilon) and the minimum number of points (MinPts). ε defines the proximity relationship between points, while MinPts defines the minimum number of points required in the neighborhood of a point to determine the density state of the point. Classification points: DBSCAN classifies each data point into one of three types by checking its neighborhood: Core points: There are at least MinPts points (including the point itself) in the ε neighborhood. Boundary points: Not in the ε neighborhood of any core point, but can be connected to the core point through a series of core points. Noise points (outliers): Neither core points nor part of the neighborhood of any core point.
[0142] (2) Clustering and anomaly detection: The algorithm starts from any unvisited point. If the point is a core point, a cluster is expanded around it, including all points in its neighborhood, and the neighborhood points of these points are recursively added until no new points can be added. In the process, points that are not included in any cluster are considered outliers.
[0143] Determine the parameters: ε (epsilon): neighborhood radius, which determines the maximum distance between points that are considered "neighboring". MinPts: minimum number of points, for a point to become a core point, its ε neighborhood must contain at least the number of points (including the point itself). Calculate the distance: For each point in the data set, calculate the distance between it and all other points. Euclidean distance or other distance metrics are usually used. Identify core points and boundary points: If a point has at least MinPts points in its ε neighborhood, the point is marked as a core point. Boundary points are those points that are directly or indirectly connected to core points, but have less than MinPts points in their own ε neighborhood. Mark noise points: Any data point that is neither a core point nor a boundary point is considered a noise point, that is, an outlier. Output results: Normal data points will be classified into one or more clusters (as core points or boundary points), while noise points will be identified separately and regarded as outliers.
[0144] (3) The stratified mean interpolation method is a further refinement of the single interpolation method. It selects appropriate attributes according to specific data to divide the population. If missing data appears in a layer, the missing data in that layer are used to fill in the missing data. The improved stratified mean interpolation model is a more advanced missing value processing technology developed on the basis of the traditional mean interpolation method. This model mainly targets the hierarchical structure in the data set (such as stratification by time, geographic location, category, etc.), and uses this hierarchical information to guide the missing value filling process to improve the accuracy and effectiveness of interpolation.
[0145] Layer identification: First, identify the natural hierarchical structure in the data set, such as dividing the data into different layers or subsets based on characteristics such as equipment type, job type, region, etc. This is done because data points within the same layer may have stronger similarities and correlations.
[0146] Calculate the mean within the layer: Within each layer, calculate the mean (or median, mode, etc., depending on the specific situation) of the records containing complete observations. The statistic reflects the central tendency of the data in this layer and is used as a benchmark for interpolation.
[0147] Hierarchical interpolation: For each record with missing values, first determine the level to which it belongs, and then use the mean (or other statistics) calculated within the level to fill in the missing values. Compared with global mean interpolation, this method takes into account the structural differences in the data and can provide more reasonable estimates.
[0148] Consider weights or complex imputation strategies: In some cases, the improved model will further consider the variability within the level, sample size or other relevant factors, assign different weights to different levels or records, or adopt more complex imputation strategies, such as weighted averaging, multiple imputation, etc., to better reflect the inherent structure of the data.
[0149] Iteration and validation: To improve the imputation results, the process is sometimes performed iteratively, i.e., a preliminary imputation is performed first, then the stratification and statistics are reassessed based on the imputed data, and if necessary, the imputation is performed again until the predetermined convergence criteria or the upper limit of the number of iterations is met.
[0150] Let n' be the number of non-missing samples in the population, the total number of samples is n, and use Represents the sample population mean estimate, and the calculation formula is as follows
[0151]
[0152] Among them, a i is an indicator variable, when a i When it is 0, it means data is missing. i 1 means that the data is not missing. is the mean of the non-missing samples, the estimated value of the population mean is the interpolation value of the missing data, and the variance is assumed to be The variance after data interpolation is s 2 , the variance changes before and after interpolation are as follows:
[0153]
[0154] The above display This shows that the mean interpolation method can reduce sample variance and improve overall data quality, but there will be a situation where all missing data are interpolated by the same data, resulting in a distorted sample distribution. The stratified mean interpolation method is a further refinement of the single interpolation method. The population is divided according to the specific data, and if there is missing data in the layer, the missing data in the layer is used to fill in the missing data. This method is better than the above method, but it still cannot avoid the problem of distorted sample distribution.
[0155] Assume the total number of samples is n, H is the total number of layers, n h It refers to the total number of samples in the hth layer. The mean of the non-missing data in the hth layer is Estimation of the population mean The calculation formula is as follows:
[0156]
[0157] After imputation, the change in the overall variance is as follows, where n1h is the number of non-missing data in the hth layer:
[0158]
[0159] In step S103, the operation status of the transmission line is evaluated by using a Markov chain model, the key influencing factors are extracted using a PCA algorithm, and a transmission line operation risk assessment system is constructed using a least squares support vector machine method.
[0160] (1) Fault risk level judgment: The operational risk of a transmission line can be described by the probability of an adverse event occurring and the degree of harm to the line after the adverse event occurs. R represents the amount of risk that an adverse event may cause to the transmission line; p represents the probability that an adverse event may cause a failure in the transmission line; and h represents the degree of damage to the transmission line caused by an adverse event. Among them, p can also represent the probability of an adverse event occurring p1 and the probability of an adverse event causing a transmission line failure p2; h can be expressed by the loss of the line's transmission capacity Y and the impact coefficient S, as shown in the following formula:
[0161] R = f(p,h) = p·h;
[0162] p=p1·p2;
[0163] h = Y·S;
[0164] Since there are many risk factors that affect the operation of transmission lines, the main consideration is the impact of external environmental factors on the lines. Different environmental factors appear at different times, and have different risks and impacts on the operation of transmission lines. Therefore, the risk period of external environmental factors and the risk weight caused by environmental factors are introduced into the expression of the degree of damage to the transmission line caused by adverse events, that is, S i =a i ·ω i a is the influence coefficient of the risk that the line may encounter due to environmental influence factor i; i represents the risk period of environmental factor i; ω i is the weight of environmental factor i. By reorganizing the above formula, we can get:
[0165]
[0166] Because the loss of transmission capacity of the transmission line cannot be quantified, the value of p2 multiplied by Y is taken as the transmission capacity loss, that is, y = p2Y, so formula (4) can be converted to:
[0167]
[0168] Combining the Regulations on Safety Risk Management of Power Grid Operation and the value range of R, the risks and risk levels of transmission line operation are obtained, as shown in the following table.
[0169]
[0170] When conducting risk analysis, the complexity and diversity of influencing factors will increase the difficulty and complexity of the analysis. In order to avoid repeated calculation of related influencing factors, redundant influencing factors should be eliminated, key influencing factors should be extracted, and the difficulty of risk assessment should be reduced.
[0171] The first step is to use PCA to extract key influencing factors. Key influencing factors are representative and can describe most of the information among many influencing factors. By extracting key influencing factors, the risk assessment process of transmission line operation can be simplified.
[0172] Collect the influencing factors that may cause transmission line failure, standardize the influencing factors, and form a matrix Q = [Q1, Q2, ..., Q n ] T ,or:
[0173]
[0174] According to the relationship between multiple influencing factors, a correlation matrix is established: R = (r ij )m×m=Q′Q, and find the eigenvalues of R λ1≥λ2≥...≥λ m ≥0, and the eigenvectors α1,α2,...,α m .
[0175] Determine the number of influencing factors. Calculate the cumulative impact rate of the first p influencing factors using the following formula:
[0176]
[0177] use The total impact rate of the p impact factor on the transmission line operation failure can be described. Then use these p influencing factors to represent the information of m influencing factors (p<m), and let H=[H1,H2,...,H p ] T , by calculating the cumulative impact rate of the first p influencing factors, the influencing factors with small contribution rate and redundancy are eliminated.
[0178] Then calculate the comprehensive impact value of the influencing factors. Weight these p key influencing factors and calculate the comprehensive impact value W of the transmission line when multiple influencing factors act together; calculate the degree of influence and weight G of each influencing factor on the transmission line.
[0179] W=λH=[λ1λ2...λ P ]H
[0180] =(λ1α1+λ2α2+...+λ p α p )T ×[Q1 Q2...Q P ] T ;
[0181] From the above formula, we can get the weight of the impact factor on the comprehensive score:
[0182] G=[k1,k2,...,k p ]=(λ1α1+λ2α2+...+λ p α p ) T ;
[0183] The weight G of the impact factor is normalized. The larger the weight of the impact factor, the more representative the impact factor is. Therefore, when G>0.5, the corresponding impact factor is selected as the key impact factor.
[0184] Transmission line failure is a random event that can be described by the Markov chain model. For the power system, the Markov chain represents the operating state of the transmission line at t+1. t+1 Only the operating state O of the transmission line at time t t The decision is independent of the running state at any time before t. By using this characteristic of Markov, we can reduce the dependence on the historical state, reduce the amount of calculation, and improve the operation efficiency. The Markov process is defined as:
[0185] O{T t+1 =t t+1 |T0=t0,T1=t1,...,T n =t n}=O{T t+1 =t t+1 |T n =t n};
[0186] λ represents the probability of transmission line failure, μ represents the ability to repair the transmission line after failure, and combining the many factors that affect the failure of transmission lines, a probability model for changes in the operating status of transmission lines is established.
[0187]
[0188] Among them, λ1 represents the occurrence rate of transmission line failures due to environmental factors (such as extreme weather); λ2 represents the occurrence rate of transmission line failures due to equipment aging factors (such as equipment disrepair); λ3 represents the occurrence rate of transmission line failures due to improper maintenance (such as too long maintenance cycle or poor maintenance quality); λ4 represents the occurrence rate of transmission line failures due to overload operation (such as load exceeding the rated value); λ5 represents the occurrence rate of transmission line failures due to human errors (such as operating errors); λ6 represents the occurrence rate of transmission line failures due to other unknown reasons (such as unknown fault causes); μ1 represents the occurrence rate of transmission line failures due to environmental factors after the failure. μ1: indicates the ability to repair the transmission line after a fault due to equipment upgrading (such as the ability to recover after replacing aging equipment); μ2: indicates the ability to repair the transmission line after a fault due to maintenance improvement (such as the ability to recover after improving maintenance quality); μ3: indicates the ability to repair the transmission line after a fault due to load adjustment (such as the ability to reduce the load to restore normal operation); μ5: indicates the ability to repair the transmission line after a fault due to correcting human errors (such as the ability to recover after correcting operating errors); μ6: indicates the ability to repair the transmission line after a fault due to other repair measures (such as other unknown repair means).
[0189] The Markov process has the following equation in the limit state:
[0190] PM = P;
[0191] P = [p N p1 p2 p3 p4 p5];
[0192]
[0193] Use p N represents the probability of normal operation of the transmission line; p1, p2, p3, p4, p5, and p6 represent the probability of line failure caused by 6 factors. The probability of transmission line failure caused by external factors is:
[0194]
[0195] There are many factors that affect the normal operation of transmission lines, and the main consideration is the impact of external factors on the operation of transmission lines. External factors include bird damage, lightning strikes, strong winds, external forces and many other factors. It is necessary to establish a probability model that includes the above-mentioned external factors:
[0196]
[0197] According to the research, P lim (l) = 1.4P max(l), where the severity of damage after a line failure is:
[0198]
[0199] The purpose of constructing a transmission line operation risk assessment model is to find the normal operation risk level of the transmission line corresponding to the key influencing factors. The least squares support vector machine algorithm is used to identify and classify transmission line faults.
[0200] Since the key influencing factors of line operation risk are nonlinear characteristics, functions are needed to map and transform the nonlinear key influencing factors of training data and convert the nonlinear influencing factors into linear equations.
[0201]
[0202] For ω,ξ i ,b,α i Find the partial derivatives and use KKT to remove ω and ξ i , we get the following linear equation:
[0203]
[0204] Ω=y k y h (φ(x k )) T φ(x h )=y k y h K(x k ,x h );
[0205] This formula is an expression of the expected value of the product of probability density functions, Ω represents the product of probability density functions, Ω is a model parameter, usually a vector; ξ i represents the regularization term, which is used to prevent overfitting; b represents the bias term; α i represents the weight of the i-th base learner.
[0206]
[0207] This formula is the definition of the Gaussian kernel function. The Gaussian kernel function is a commonly used kernel function that can map data in a low-dimensional space to a high-dimensional space, thereby achieving nonlinear classification and regression tasks.
[0208] Where k, h = 1, 2, ..., n; x k ,x h ,y h ,y k is the training sample vector; α is the iThe vector formed; Y = [y1, y2, ..., y n ], 1 v =[1,1,...,1]; K(x k ,x h ) is the kernel function; I is the unit vector.
[0209] Secondly, the classification equation is used to classify the line operation faults.
[0210]
[0211] Among them, φ is the nonlinear mapping function vector, x i and i is the training sample data. In this way, the nonlinear key influencing factors are converted into linear data for solution. The transformed linear transmission line operation risk key influencing factors are input into the least squares support vector machine risk assessment model established above, and the operation risk faced by the transmission line can be obtained.
[0212] First, identify and extract the key influencing factors of transmission line operation, which may be diverse, including but not limited to equipment status, environmental conditions, operating behavior, etc. These factors often present complex nonlinear relationships.
[0213] Then, for the nonlinear influencing factors identified in the first step, specific mathematical tools (such as nonlinear mapping functions) are used to convert them into linear forms. This is because many traditional statistical and machine learning methods assume that the data is linear, so it is necessary to convert nonlinear problems into linear problems for subsequent analysis and modeling. In short, the above two steps are a continuous process. The first step is to determine the research object, and the second step is to pre-process the research object to adapt to subsequent analysis and modeling. Only after completing these two steps can the least squares support vector machine or other similar models be used for risk assessment.
[0214] Since γ and σ 2 Directly affects the accuracy of the final transmission line fault classification. Through continuous iterative updates, until the optimal γ and σ are found 2 .
[0215] (2) Calculation of unit failure time loss by fault: First, classify the same fault types for further classification. Use the clustering method to classify the same type of faults, and then combine the historical fault data to calculate the average loss cost per unit failure time by weighted average.
[0216] 1) Classification indicators. Classification indicators and basis:
[0217] Fault phenomenon and cause: preliminary classification is carried out according to the fault manifestation (such as abnormal current, voltage fluctuation, insulation reduction, etc.) and the root cause of the fault (such as equipment aging, external force damage, operating error, etc.). This is the most basic classification basis, which is directly related to the nature of the fault and the way to deal with it.
[0218] Fault impact scope and consequences: Classify according to the scope of service interruption caused by the fault (such as the number of users affected, the size of the blackout area), the scale of power loss, and the potential impact on the safe and stable operation of the power grid. Faults with a wide impact and serious consequences may require more urgent and detailed handling.
[0219] Historical data and frequency: Use historical fault records to count the frequency, repetitiveness, and seasonality of faults. Frequently occurring fault types may require special attention and be classified as key objects.
[0220] Cost and repair difficulty: Classify according to the time required for fault repair, cost (including direct cost and indirect cost) and complexity of repair. This involves many factors such as personnel skills, spare parts acquisition, and technical difficulty.
[0221] 2) Based on fuzzy C-means clustering. Fuzzy C-means (FCM) is a clustering algorithm based on objective function partitioning. It can assign a weight to each object and each cluster to indicate the degree to which the object belongs to the cluster. It can avoid hard clustering and provide more flexible clustering results.
[0222] The main principle of FCM is to maximize the similarity between objects classified into the same cluster and minimize the similarity between objects in different clusters. By calculating the membership degree of each data sample to be clustered with each cluster center, it is divided into clusters with obvious separation of characteristics to achieve data sample clustering. A weight is assigned to each sample and each cluster, and the degree to which the sample belongs to the cluster is judged by the cluster distance. Through multiple iterations, the distance within each cluster sample is minimized and the distance between clusters is maximized.
[0223] Suppose the set of m samples to be clustered is X = {x1, x2, ..., x m}, where each sample x i There are n characteristic indicators, let it be x i ={x i1 ,x i2 ,...,x in If we want to divide X into C categories, each subset can be represented as C = {c1, c2, ..., c k}Then the clustering loss function is as follows:
[0224]
[0225] Among them, J is the value function of FCM; b is the weighting coefficient, which is used to adjust the influence of the membership index in the optimization target, and its value is generally 2; u ij is the membership degree of the i-th sample to the j-th cluster; x i is the i-th sample, with dimension n; c j is the cluster center of the jth cluster, with a dimension of n; ||*|| represents the distance between the sample and the cluster center. To solve this optimization problem, the Lagrange multiplier method is used to introduce the equality constraint into the objective function, and the conditional extreme value is converted into the unconditional extreme value, resulting in the following formula:
[0226]
[0227] Where K is the number of clusters predefined in cluster analysis, that is, we hope to divide the data set into K different clusters. Purpose: In clustering algorithms (such as K-means algorithm), K is an important parameter that determines how many clusters the data set will be divided into. Each cluster has a corresponding cluster center μk. For example, suppose we have a data set that needs to be divided into 3 clusters. At this time, set K=3. In K-means clustering, there are 3 cluster centers, which represent the center point position of each cluster. The clustering process will assign each sample point to the closest cluster center to form 3 clusters. N is the total number of samples in the data set. N represents the total number of samples that need to be clustered. In algorithm calculations, it is a basic parameter that determines the scale and complexity of the calculation. Importance: The size of N directly affects the execution time and memory consumption of the clustering algorithm. Large-scale data sets (that is, larger N) require more efficient algorithms and more computing resources. Where: λ i Lagrange multiplier for the ith sample.
[0228] C j and membership Find partial derivatives:
[0229]
[0230] The fuzzy C-means clustering algorithm calculates the membership degree and cluster center through continuous iteration until it reaches the optimal value. The expressions of membership degree and cluster center can be obtained from the following formulas, as shown in the following formulas:
[0231]
[0232]
[0233] According to the above two formulas, the membership coefficient and cluster center are iterated until the convergence condition of FCM is reached. The convergence condition is:
[0234]
[0235] Where T is the number of iterations and ε is the error threshold.
[0236] 3) Calculation of average loss: Use the weighted average method to calculate the unit failure time loss scale of similar equipment.
[0237] In step S104, a multi-objective set of optimal operation and maintenance routes is constructed, and a set of constraint conditions such as optimal cost and power grid operation is proposed.
[0238] Objective function:
[0239] (1) Time recovery. Considering the “abundance” in elastic response, on the basis of path optimization, it is also necessary to consider the problem of operation and maintenance resource allocation, that is, the degree of response of the resources and personnel allocation of the operation and maintenance reserve to the work. Preventive planning in advance during the operation and maintenance stage can shorten the repair time when an emergency occurs, so the time for maintenance personnel to arrive at the fault point when a fault occurs is set as t i , which is also the optimal path time, then all operation and maintenance time T is:
[0240]
[0241] Where: is the operation and maintenance time of a certain area; t mi,mi+1 From point a to m i To point b m i+1 The time required. m usually represents the dimension of the sample data, that is, the number of features of each data point. In cluster analysis or machine learning, the data set usually consists of multiple features, and m represents the dimension of the feature vector of each sample data. i represents the number of a specific sample in the data set. i varies from 1 to n in the data set, corresponding to the 1st to nth samples respectively. n is the total number of samples in the data set, indicating the total number of data points that need to be analyzed or processed. t can represent time or the number of iterations, depending on the context of the problem. In optimization problems, t usually represents the number of iterations or the time step.
[0242] (2) Economic loss. The economic loss can be regarded as F, which is expressed as the weighted sum of the failure losses per unit time after failures occur at different operation and maintenance points. The objective function is:
[0243]
[0244] Among them, Pt mi Operation and maintenance point m i The duration from the fault occurrence point to the completion of power supply restoration; α i is the level of risk, P l The loss per unit time caused by this level of risk.
[0245] Constraints:
[0246] When setting the constraints of the operation and maintenance response, a mixed integer nonlinear programming algorithm is used. The state can be defined as [o, 1]. If the fault has been restored, it is 1, and the restoration state of all lines is represented by 1. The optimal path for the maintenance personnel of the power supply company to reach the operation and maintenance point from the warehouse is W Dmi , reaching the state of all operation and maintenance points is 1. Constraints are proposed from the aspects of operation and maintenance work response and power grid.
[0247] Prevent the operation and maintenance point from being restored repeatedly, and the power grid fault point m i Recovery state; the path from the maintenance point warehouse to the operation and maintenance point is not repeated; the power grid operation and maintenance point m i Restoration status; To prevent line duplication, line fault recovery; Personnel and fault point m i resource allocation; in order to minimize the recovery time, it is also necessary to consider the configuration of resources such as the operation and maintenance response and emergency speed of the staff, and the operation and maintenance equipment.
[0248]
[0249] in, Operation and maintenance for any point; W mi,m+1 Operation and maintenance point m i To the fault point m i +1 for the path.
[0250]
[0251] in, is any load point; W Dmi Optimize the path from the operation and maintenance point back to the warehouse point.
[0252]
[0253]
[0254]
[0255] in, Any time; The resource is sufficient at any time.
[0256]
[0257] Among them, Z is the resource reserve, Z x The amount of resources carried by the staff, Repair materials required for all fault points.
[0258] Grid constraints: Considering that my country's distribution network structure is a closed-loop design and open-loop operation, it is a radial structure during distribution network maintenance and fault recovery. The safe operation of the power grid must be guaranteed during the power supply recovery process. Therefore, the grid constraints are
[0259] g k ∈G R ;
[0260] |P a |≤P amax ;
[0261] U min ≤U≤U max ;
[0262] Among them, g k To restore the regional grid structure; G R Total grid structure collection; P a is the power of branch a; P amax is the maximum allowable power of the branch; U is the node voltage; U min , U max are the lower and upper limits of node voltage. The above constraints are grid structure constraints, branch flow constraints and node voltage constraints.
[0263] In step S105, the improved Pareto-Zebra algorithm is applied to solve the objective function and finally obtain the optimal route.
[0264] Although the Zebra Optimization Algorithm (ZOA) is mainly designed to solve single-objective optimization problems, it can be adjusted and extended to cope with multi-objective and multi-constraint optimization scenarios. The following is the general process of applying ZOA to solve multi-objective and multi-constraint problems:
[0265] Multi-objective optimization extension: (1) Objective function conversion: First, the multi-objective problem needs to be converted into a single-objective problem. Common methods include Pareto dominance, weighted sum method, ε-constraint method, etc. in multi-objective optimization. The most commonly used method is to calculate the non-dominated rank and crowding distance of each solution to maintain the diversity of the solution set. (2) Fitness evaluation: For multi-objective problems, fitness evaluation is no longer a single objective minimization or maximization, but an evaluation of the position and distribution of solutions on the Pareto front. This usually involves calculating the non-dominated rank and crowding distance of the solution. (3) Update rule adjustment: In ZOA, the update rule of the zebra (solution) needs to be modified to ensure that the algorithm can explore and maintain solutions on the Pareto front. It may be necessary to consider multi-objective characteristics in the selection of leader zebras, information sharing, and other links.
[0266] Handling constraints: (1) Penalty function method: Add a penalty term to the objective function. When a solution violates a constraint, its fitness value will increase significantly, thereby preventing these solutions from being selected during the optimization process. (2) Repair operation: Repair the solution that violates the constraint and adjust it to the nearest feasible solution that satisfies the constraint, and then further optimize it. (3) Embedded constraint handling mechanism: Directly consider the constraints during the algorithm's iteration process. For example, a mechanism can be designed so that only solutions that satisfy all constraints participate in information sharing and position updates.
[0267] The overall process is as follows:
[0268] Initialization: Generate an initial population of zebras, each zebra represents a potential solution, while ensuring that the initial solution covers the solution space as much as possible and takes into account the constraints.
[0269] Fitness evaluation: Each zebra is evaluated according to the above multi-objective evaluation method.
[0270] Multi-objective optimization mechanism: perform non-dominated sorting, determine the Pareto rank of each solution, and maintain diversity based on crowding distance.
[0271] Constraint processing: Apply the above constraint processing techniques to ensure that all solution sets after operations satisfy the constraints.
[0272] Update and explore: Update the position of zebras based on the fitness value after multi-objective optimization and the zebra behavior model (such as the selection of leader zebras, group movement, etc.).
[0273] Termination condition check: Repeat the above process until a predetermined number of iterations or other stopping criteria (such as convergence index) is reached.
[0274] Solution set extraction: The solution set finally obtained on the Pareto front is the solution to the multi-objective optimization problem.
[0275] Zebra Optimization Algorithm (ZOA) is an optimization algorithm based on zebra behavior. The algorithm simulates zebra behavior to find the optimal solution. It has the characteristics of strong optimization ability and fast convergence speed.
[0276] The specific steps are as follows:
[0277] 1.1 Initialization.
[0278] Similar to other optimization algorithms, the population is randomly initialized in the optimization space:
[0279] x i,j =lb j +r·(ub j -lb j );
[0280] where x i,j For individuals, j To find the upper bound, lb j To find the optimal lower bound, r is a random number between [0,1].
[0281] 1.2 Stage 1: Foraging Behavior
[0282] In the first phase, population members are updated based on simulations of zebra behavior as they search for forage. Zebras' primary diet consists primarily of grasses and sedges, but they may also eat buds, fruits, bark, roots, and leaves if their preferred foods are scarce. Depending on the quality and availability of vegetation, zebras may spend 60-80% of their time feeding. Among zebras, there is a species called the plains zebra that is a pioneer herbivore that, by devouring the canopy of upper and less nutritious grasses, provides for other species that require shorter and more nutritious grasses. In ZOA, the best members of the population are considered pioneer zebras and lead other population members toward their position in the search space.
[0283] Therefore, (3) and (4) can be used to mathematically model the zebra's position update during the foraging phase.
[0284]
[0285]
[0286] in, It represents the value of the jth feature of the i-th sample in the new state, after one iteration update. i,j Represents the value of the jth feature of the ith sample in the current state. r is a scaling factor or step size, often used to adjust the update amplitude. PZ j Target value or expected value, representing the ideal or target state of the jth feature. I Identity matrix or identifier, depending on the context. Usually used in matrix operations. X i Represents the feature vector of the i-th sample. Represents the new feature vector of the i-th sample, the state after the update. Represents the objective function value of the i-th sample in the new state. i Represents the objective function value of the i-th sample in the current state. Where r is a random number between [0,1], and I is a random value belonging to the set {1,2}.
[0287] 1.3 Stage 2: Defense strategies against predators
[0288] In the second stage, the defense strategies of zebras against predator attacks are simulated to update the positions of ZOA population members in the search space. The main predator of zebras is lions; however, they are threatened by cheetahs, leopards, wild dogs, brown hyenas, and spotted hyenas. Crocodiles are another predator of zebras when they are close to the water. The defense strategies of zebras vary depending on the predator. The defense strategy of zebras against lion attacks is to escape with zigzag and random side-turning movements. Zebras are more aggressive against smaller predators, such as hyenas and dogs, and they do this by gathering together to confuse and intimidate hunters. In the ZOA design, one of the following two situations is assumed to occur with equal probability:
[0289] (i) A lion attacks a zebra, so the zebra chooses an escape strategy;
[0290] (ii) If other predators attack the zebra, the zebra will choose an attack strategy.
[0291] In the first strategy, when the zebra is attacked by a lion, the zebra will hide near its location to avoid the lion's attack. Therefore, mathematically, this strategy can be modeled using pattern S1 in (5). In the second strategy, when other predators attack one of the zebras, the other zebras in the group will move closer to the attacked zebra and try to intimidate and confuse the predator by building a defensive structure. This strategy of the zebra is mathematically modeled using pattern S2 in (5). When updating the position of the zebra, if the zebra has a better value for the objective function in the new position, the new position is accepted. This update condition is modeled using (6).
[0292]
[0293]
[0294] in, represents the new value of the jth feature of the i-th sample under the second updating method (P2). i,j Represents the current value of the jth feature of the ith sample. S1 and S2 are two different scaling factors or weights used to adjust the update of feature values. R is a random factor or weight used to introduce randomness or variation to prevent falling into a local optimal solution. r is another scaling factor or step size, usually used to adjust the amplitude of the update. t is the current number of iterations or time steps. T is the maximum number of iterations or total time steps. P s A probability value used to decide whether to apply a specific update strategy. j Target value or expected value, representing the ideal or target state of the jjjth feature. I Identity matrix or identifier, depending on the context. Usually used in matrix operations. X i Represents the feature vector of the i-th sample. Represents the new feature vector of the i-th sample, the state after the update. Represents the objective function value of the i-th sample in the new state. i Represents the objective function value of the i-th sample in the current state.
[0295] Where t is the number of iterations, T is the maximum number of iterations, and R is a constant of 0.01. Ps is the probability of switching between the two strategies. Its value is a random number between [0,1]. AZ is the state of the attacked zebra.
[0296] Although directly converting multi-objective problems into single-objective problems is not a strategy directly adopted by ZOA, it is helpful to understand the adaptability of ZOA by understanding common multi-objective optimization processing methods. ZOA mainly handles multi-objective problems indirectly by maintaining the Pareto front and diversity of the solution set. However, if we want to mention converting multi-objective problems into single-objective problems, we use the weighted method to express it:
[0297] f(x)=w1f1(x)+w2f2(x)+...+w n f n (x);
[0298] Among them, f i () is the i-th objective function, w i is the corresponding weight, and satisfies all w i ≥0,∑w i = 1. However, ZOA prefers to directly deal with multi-objective problems and evaluates the quality of solutions through non-dominated ranks and crowding distances.
[0299] In ZOA, fitness evaluation focuses on the non-dominated rank and crowding distance of the solution. The non-dominated rank is defined as how many other solutions a solution is dominated by. A solution that is not dominated by any other solution belongs to the first level (non-dominated rank is 1). The crowding distance is calculated to measure the distribution of solutions on the Pareto front. The formula is as follows: For each target j, calculate the crowding distance on the target dimension (denoted as CD j ):
[0300]
[0301] in, represents the value of the kth solution at the jth objective, P is the set of solutions on the current Pareto frontier, and the solution sets are sorted from small to large according to each objective. The crowding distance of the frontier solution is set to infinity or a sufficiently large value to ensure that it will not be easily eliminated.
[0302] Update rule adjustment: In ZOA, when updating the position of the zebra (solution), it is necessary to ensure that the algorithm can effectively explore and maintain the diversity of the Pareto frontier. This usually includes the following steps:
[0303] Select leader zebra: Select leader zebra based on non-dominated rank and crowding distance. Solutions with low non-dominated rank (i.e. better) and large crowding distance are preferred as leader zebras because these solutions represent key points on the Pareto front.
[0304] Information sharing and speed update: The zebra updates its speed v according to its position relationship with the leader zebra and possible random disturbances. k and position x k The speed update can be considered as follows:
[0305]
[0306] Among them, w is the inertia weight, c1, c2 are learning factors, r1, r2 are random numbers, is the historical optimal position of zebra k, L k is the position of the leading zebra. This step ensures a balance between exploration and exploitation:
[0307]
[0308] Boundary checking and constraint processing: Ensure that the updated solution meets all constraints, such as cost budget, grid operation stability, etc. Constraints can be processed through penalty functions, linear scaling or other methods to ensure the validity of the solution.
[0309] Repeat steps 2 to 4 until a predetermined number of iterations or convergence criterion is reached.
[0310] Convergence judgment: Determine whether the algorithm has converged based on the changes in the solution set (such as the stability of the Pareto front and the fluctuation range of the fitness value).
[0311] Result output: Output the solution set on the current Pareto frontier, which is the optimal solution set for the multi-objective problem. These solutions represent the optimal trade-off between the cost of different operation and maintenance routes and the stability of power grid operation under given constraints.
[0312] refer to Figure 3Another specific embodiment of the present invention discloses an optimal operation and maintenance route decision system based on risk assessment, including: a data acquisition module 301, used to acquire basic data, wherein the basic data includes resource consumption of multiple maintenance operations, equipment-related data and fault statistics, wherein the equipment-related information includes the geographical location, environment, region, and power grid of the equipment; a data processing module 302, used to perform outlier detection on the basic data by using a density-based outlier detection method, and then interpolate the missing values, wherein the missing values are interpolated according to the improved stratified mean interpolation; fault A failure classification module 303 is used to evaluate the operation status of the transmission line through a Markov chain model and extract key influencing factors from the external factors after interpolation processing using a principal component analysis algorithm, and predict the operation risk level of the transmission line through a least squares support vector machine to identify and classify transmission line faults, wherein the key influencing factors and the operation status of the transmission line are used as inputs of the least squares support vector machine; and an operation and maintenance optimal route module 304 is used to construct an objective function set and corresponding constraint conditions for the operation and maintenance optimal route based on the operation status of the transmission line and the operation risk level of the transmission line to obtain the operation and maintenance optimal route.
[0313] Specifically, the data processing module includes: a stratified identification module, which is used to divide the basic data into different levels according to equipment type, operation type, or regional characteristics; an intra-level mean calculation module, which is used to determine the level to which each record containing missing values belongs, and use the mean calculated within the level to fill the missing values; a stratified interpolation module, which is used to fill the missing values with the mean within the level according to the level to which the missing values belong when there are missing value records in the basic data; a weighted or complex interpolation module, which is used to consider the variability within the level and the sample size, assign different weights to different levels, or use weighted average or multiple interpolation to reflect the inherent structure of the basic data; and an iteration and verification module, which is used to re-evaluate the stratification and statistics based on the interpolated basic data, and iterate again until the predetermined convergence criteria or the upper limit of the number of iterations is met.
[0314] Hierarchical recognition refers to dividing data into different levels or subsets based on the natural hierarchical structure in the data set, such as equipment type, operation type, region and other characteristics. This method can more accurately capture the internal characteristics of the data through stratification and improve the accuracy of data analysis and processing. The following is a specific example to illustrate the detailed process of this step.
[0315] Suppose we have an operation and maintenance dataset for a large power grid, which contains the following main features: equipment type, such as transformers, transmission lines, circuit breakers, etc.; operation type, such as routine inspection, emergency repair, preventive maintenance, etc.; region, such as city A, city B, rural C, etc.; environmental characteristics, such as temperature, humidity, geographic location, etc.
[0316] For example, when stratified by equipment type, it includes transformers, which contain all operation and maintenance data about transformers; transmission lines, which contain all operation and maintenance data about transmission lines; and circuit breakers, which contain all operation and maintenance data about circuit breakers.
[0317] For example, when stratified by job type, it includes routine inspection, which contains the operation and maintenance data of all routine inspections; emergency repair, which contains the operation and maintenance data of all emergency repairs; preventive maintenance, which contains the operation and maintenance data of all preventive maintenance.
[0318] For example, when stratified by region, it includes city A, which contains the operation and maintenance data of all areas in city A; city B, which contains the operation and maintenance data of all areas in city B; and rural C, which contains the operation and maintenance data of all areas in rural C.
[0319] In the outlier detection method, the density-based clustering algorithm improves the defect of the distance-based outlier algorithm that gives the same weight to all objects. It is suitable for the case of uneven distribution of sample data. It can make good outlier judgments for the case of different densities in the data, reducing the limitations of use and performing probability evaluation for the judgment of whether it is an outlier. In addition, clusters in reality often have various shapes. Traditional clustering methods such as Kmeans clustering can only process spherical clusters, which obviously cannot meet the needs of clusters of other irregular shapes. Therefore, a DBSCAN clustering algorithm that meets the needs of arbitrary cluster shapes is proposed.
[0320] The stratified mean interpolation method is a further refinement of the single interpolation method. It selects appropriate attributes according to specific data to divide the population. If missing data appears in a layer, the missing data in that layer is used to fill it. The improved stratified mean interpolation model is a more advanced missing value processing technology developed on the basis of the traditional mean interpolation method. This model mainly targets the hierarchical structure in the data set (such as stratification by time, geographic location, category and other characteristics), and uses these hierarchical information to guide the missing value filling process to improve the accuracy and effectiveness of interpolation.
[0321] The system also includes other modules. The method in this application corresponds to the system. In order to avoid redundancy, the description of other modules is omitted here.
[0322] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.
[0323] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.
Claims
1. A method for determining the optimal operation and maintenance route based on risk assessment, characterized in that: include: Obtaining basic data, wherein the basic data includes resource consumption of multiple maintenance operations, equipment-related data and fault statistics, wherein the equipment-related information includes the geographical location, environment, region, and power grid of the equipment; Performing outlier detection on the basic data using a density-based outlier detection method, and then interpolating the missing values, wherein the missing values are interpolated according to an improved stratified mean interpolation; Evaluating the transmission line operation status by using a Markov chain model and extracting key influencing factors from the external factors after interpolation using a principal component analysis algorithm, predicting the transmission line operation risk level by using a least squares support vector machine to identify and classify the transmission line fault, wherein the key influencing factors and the transmission line operation status are used as inputs to the least squares support vector machine; and Based on the operation status of the transmission line and the operation risk level of the transmission line, an objective function set of an optimal operation and maintenance route and corresponding constraint conditions are constructed to obtain the optimal operation and maintenance route.
2. The optimal operation and maintenance route decision method based on risk assessment according to claim 1 is characterized in that: The resource consumption of the multiple maintenance operations includes manpower consumption, material consumption, time cost, financial cost and energy consumption, wherein the manpower consumption is used to record the number of people involved in each operation and maintenance task, working hours, and professional skill requirements; the material consumption includes the number and cost of tools, spare parts, and consumable materials used; the time cost includes the total time from response time to task completion, distinguishing between preparation time, road time, and actual operation time; the financial cost includes direct labor costs, material costs, transportation costs, and indirect costs of equipment rental and outsourcing service fees; the energy consumption includes fuel and electricity consumption of vehicles used in the operation and maintenance operations; The fault statistics data include fault type, fault occurrence time, fault duration, power loss and fault repair process, among which the fault types include insulation damage, overload, short circuit and animal touch; the fault occurrence time includes the date and time of the fault to determine whether there is a seasonal pattern in the occurrence of the fault; the fault duration refers to the power outage time caused by the fault; the power loss refers to the estimation of the direct power loss and indirect economic loss caused by the fault; the fault repair process refers to the time, cost, measures taken and whether parts are replaced for fault repair.
3. The optimal operation and maintenance route decision method based on risk assessment according to claim 1 is characterized in that: Using a density-based outlier detection method to perform outlier detection on the basic data, and then interpolating the missing values further includes: In the hierarchical identification step, the basic data is divided into different levels according to equipment type, operation type, or regional characteristics; In the intra-level mean calculation step, for each record containing missing values, the level to which it belongs is determined, and the missing value is filled using the mean calculated within the level; In the stratified interpolation step, when there are missing value records in the basic data, the missing values are filled using the mean value within the stratum according to the stratum to which the missing values belong; In weighted or complex imputation steps, consider intra-stratum variability, sample size, assign different weights to different strata or use weighted averages or multiple imputations to reflect the inherent structure of the underlying data; and In the iteration and validation step, the stratification and statistics are re-evaluated based on the imputed underlying data, and iterations are performed again until the predetermined convergence criteria or the upper limit of the number of iterations is met.
4. The optimal operation and maintenance route decision method based on risk assessment according to claim 3 is characterized in that: The basic data is divided into different levels according to equipment type, operation type, or regional characteristics, including: Set a domain radius ε and a minimum number of points MinPts, wherein the domain radius ε represents the proximity relationship between points, and the minimum number of points MinPts represents the minimum number of points that need to be reached in the neighborhood of a certain point; For each point in the base data, calculate the distance between each point and all other points; According to the calculated distance between each point and all other points, each data point in the basic data is divided into core points, boundary points and noise points, wherein when there are at least a minimum number of points MinPts in the ε neighborhood of the current point, the current point is marked as a core point; when the current point is directly or indirectly connected to the core point, but there are less than MinPts points in the ε neighborhood of the current point, the current point is marked as a boundary point; and when the current point is neither a core point nor a boundary point, the current point is marked as a noise point as an outlier.
5. The optimal operation and maintenance route decision method based on risk assessment according to claim 1 is characterized in that: The Markov chain model is used to evaluate the operation status of the transmission line and the principal component analysis algorithm is used to extract the key influencing factors from the external factors after interpolation. The least squares support vector machine is used to predict the operation risk level of the transmission line, which further includes: Evaluating the operating state of the transmission line by using the Markov chain model, wherein the Markov chain model is used to average the future state of the transmission line according to the current state of the transmission line to predict the probability of the transmission line transferring from the current state to a fault state; Extracting some key factors that have the greatest impact on operational risk from multiple external factors through the principal component analysis algorithm as the key influencing factors; The transmission line operation risk level is predicted by the least squares support vector machine, wherein the least squares support vector machine model is trained by using key influencing factors as inputs of the least squares support vector machine model.
6. The optimal operation and maintenance route decision method based on risk assessment according to claim 5 is characterized in that: The Markov chain model is used to evaluate the operation status of the transmission line and the principal component analysis algorithm is used to extract the key influencing factors from the external factors after interpolation processing, including: Obtain the influencing factors that may cause the transmission line failure from the basic data, and standardize the influencing factors to form a matrix Q=[Q1, Q2, ..., Q n ] T ,in, According to the relationship between the m influencing factors, a correlation matrix is established: R = (r ij )m×m=Q′Q, and find the eigenvalues of R λ1≥λ2≥...≥λ m ≥0, and the eigenvectors α1,α2,...,α m ; The cumulative impact rate of the first p influencing factors is calculated using the following formula: in, represents the total impact rate of p influencing factors on the transmission line operation failure, if When the information of m influencing factors is replaced by the p influencing factors, p<m, let H=[H1,H2,...,H p ] T , by calculating the cumulative impact rate of the p impact factors, the impact factors with small contribution rate and redundancy are eliminated; The p influencing factors are weighted to calculate the comprehensive influencing value W of the p influencing factors acting together on the transmission line, and the degree of influence of each influencing factor on the transmission line and the weight G are calculated: W=λH=[λ1 λ2...λ P ]H =(λ1α1+λ2α2+...+λ p a p ) T ×[Q1 Q2...Q P ] T ; The weight G of the impact factor on the comprehensive score is calculated by the following formula: G=[k1,k2,...,k p ]=(λ1α1+λ2α2+...+λ p a p ) T ; The Markov process is defined by the following formula: O{T t+1 =t t+1 |T0=t0,T1=t1,...,T n =t n }=O{T t+1 =t t+1 |T n =t n }; Among them, for the power system, the Markov chain represents the operating state of the transmission line at t+1 t+1 Only the operating state O of the transmission line at time t t The decision is independent of the operating status at any time before t; λ represents the probability of transmission line failure, μ represents the ability to repair the transmission line after failure, and a probability model of the change in the operating status of the transmission line is established by combining multiple factors that affect the failure of the transmission line: The Markov process in the limit state is expressed by the following formula: PM = P; P=[p N p1 p2 p3 p4 p5]; Among them, p N represents the probability of normal operation of the transmission line; p1, p2, p3, p4, p5, and p6 represent the probability of transmission line failure caused by six factors, respectively. The probability of transmission line failure caused by external factors is obtained by the following formula:
7. The optimal operation and maintenance route decision method based on risk assessment according to claim 1 is characterized in that: The objective function set for constructing the optimal operation and maintenance route further includes: The operation and maintenance time objective function is expressed by the following formula: Among them, the time for maintenance personnel to arrive at the fault point when a fault occurs is set to t i , is the operation and maintenance time of a certain area; t mi,mi+1 From the operation and maintenance point m i To the operation and maintenance point m i+1 the time required; The economic loss objective function is expressed by the following formula: Among them, Pt mi Operation and maintenance point m i The duration from the fault occurrence point to the completion of power supply restoration; α i is the risk level, P l The loss per unit time generated for the stated risk level.
8. The optimal operation and maintenance route decision method based on risk assessment according to claim 7 is characterized in that: Constructing the corresponding constraints further includes: Prevent the operation and maintenance point from being restored repeatedly, and the power grid fault point m i Restoration state constraints: in, Operation and maintenance for any point; Operation and maintenance point m i To the fault point m i +1 for the path; The path from the maintenance point warehouse to the operation and maintenance point must not be repeated: in, is any load point; W Dmi Optimize the path from the operation and maintenance point back to the warehouse point; Power grid operation and maintenance point m i Restoration status constraints; line fault recovery constraints to prevent line repeated restoration; personnel and fault point m i Resource allocation; consider the staff's operation and maintenance response and emergency speed, operation and maintenance equipment and other resource configurations to minimize recovery time: in, Any time; The resource is sufficient at any time; Among them, Z is the resource reserve, Z x The amount of resources carried by the staff, Repair materials needed for all failure points; Grid structure constraints, branch flow constraints and node voltage constraints in power grid constraints: g k ∈G R ; |P a |≤P amax ; IN min ≤U≤U max ; Among them, g k To restore the regional grid structure; G R Total grid structure collection; P a is the power of branch a; P amax is the maximum allowable power of the branch; U is the node voltage; U min , U max are the lower and upper limits of the node voltage.
9. An optimal operation and maintenance route decision system based on risk assessment, characterized in that: include: A data acquisition module, used to acquire basic data, wherein the basic data includes resource consumption of multiple maintenance operations, equipment-related data and fault statistics, wherein the equipment-related information includes the geographical location, environment, region, and power grid of the equipment; A data processing module, used for performing outlier detection on the basic data by using a density-based outlier detection method, and then interpolating the missing values, wherein the missing values are interpolated according to improved stratified mean interpolation; a fault failure classification module, for evaluating the operation status of the transmission line through a Markov chain model and extracting key influencing factors from the external factors after interpolation processing using a principal component analysis algorithm, and predicting the transmission line operation risk level through a least squares support vector machine to identify and classify the transmission line fault, wherein the key influencing factors and the operation status of the transmission line are used as inputs of the least squares support vector machine; and The operation and maintenance optimal route module is used to construct an objective function set and corresponding constraint conditions of the operation and maintenance optimal route based on the operation status of the transmission line and the operation risk level of the transmission line to obtain the operation and maintenance optimal route.
10. The optimal operation and maintenance route decision system based on risk assessment according to claim 9 is characterized in that: The data processing module comprises: A hierarchical identification module, for dividing the basic data into different hierarchies according to equipment type, operation type, or regional characteristics; A layer mean calculation module, for determining the layer to which each record containing missing values belongs, and using the mean calculated within the layer to fill the missing values; A hierarchical interpolation module is used to fill in missing values using the mean value within the hierarchy according to the hierarchy to which the missing values belong when there are missing value records in the basic data; Weighted or complex imputation modules that take into account intra-stratum variability, sample size, assigning different weights to different strata or using weighted averages or multiple imputation to reflect the inherent structure of the underlying data; and The iteration and verification module is used to re-evaluate the stratification and statistics based on the basic data after imputation, and iterate again until the predetermined convergence standard or the upper limit of the iteration number is met.
Citation Information
Cited By
Distributed multi-center machine learning intelligent decision-making method
CN120069100A
Equipment state monitoring and anomaly analysis strategy method and system based on PMS
CN120561837A
Pms-based device state monitoring and abnormality analysis strategy method and system
CN120561837B
Management decision-making method and system for road maintenance
CN120894184A
Resource scheduling system and method for wheel set maintenance workshop based on Internet of Things
CN120952747A
