CN-121998629-A - Intelligent operation and maintenance management method and system for direct-current charging pile

CN121998629ACN 121998629 ACN121998629 ACN 121998629ACN-121998629-A

Abstract

The invention relates to an intelligent operation and maintenance management method and system for a direct current charging pile, and belongs to the technical field of operation and maintenance of charging piles. The method comprises the steps of firstly, counting road data and fault rate around a direct current charging pile, establishing an environment index, then collecting a charging direct current voltage and direct current signal, obtaining direct current voltage and direct current time sequence data through analog-to-digital conversion and an EEMD algorithm, calculating mutation, strength, fluctuation and trend indexes to construct a reliability index, simultaneously collecting electric power and electric energy quality parameters in real time, carrying out feature aggregation and pooling through a graph isomorphic neural network, extracting electric energy features by combining the convolutional neural network, inputting a multi-layer perceptron after being fused through an attention mechanism to obtain an error prediction value, further establishing a performance index by combining a true value, finally integrating the environment, the reliability and the performance index to construct an operation and maintenance management index, and formulating an operation and maintenance strategy according to the operation and maintenance management index.

Inventors

HUANG TIANFU
WU XIANG
ZHANG YING
WU ZHIWU
LIN TONGYAO
WANG CHUNGUANG
Tu Yanzhao
HUANG HANBIN
YU HONGHUI
HE WENZHI

Assignees

国网福建省电力有限公司营销服务中心
国网福建省电力有限公司

Dates

Publication Date: 20260508
Application Date: 20260411

Claims (10)

1. The intelligent operation and maintenance management method for the direct-current charging pile is characterized by comprising the following steps of: Calculating the road condition and the direct current charging pile fault rate of the direct current charging pile position in the area to be managed, and calculating the direct current charging pile environment index, including the road index and the fault index, based on the road condition and the direct current charging pile fault rate; Acquiring direct-current voltage and direct-current signals in the charging process of the direct-current charging pile, and acquiring time sequence data corresponding to the direct-current voltage and the direct-current signals by utilizing an analog-to-digital conversion and EEMD algorithm; The method comprises the steps of collecting power and electric energy quality parameters of a direct-current charging pile in real time, extracting feature vectors of the power quality parameters by using a graph isomorphic neural network, extracting feature vectors of the electric energy quality parameters by using a convolution neural network, updating the feature vectors of the power quality parameters and the electric energy quality parameters by using an attention mechanism to obtain a feature vector set, taking the feature vector set as input of a multi-layer perceptron with which is pre-trained, obtaining a metering error predicted value of the direct-current charging pile in a region to be managed, and constructing an operation performance index of the direct-current charging pile based on the metering error predicted value and a metering error true value; Based on the environment, the operation reliability and the operation performance index of the direct-current charging pile, calculating an operation and maintenance management index of the direct-current charging pile, and formulating a corresponding operation and maintenance management strategy based on the operation and maintenance management index.
2. The intelligent operation and maintenance management method of a direct current charging pile according to claim 1, wherein calculating the direct current charging pile environmental index based on the road condition and the direct current charging pile fault rate is specifically as follows: the road condition includes a road class, a traffic density, a single lane width, and a number of unidirectional lanes, and a road index is calculated based on the road condition, expressed as: ; in the formula, The road index is represented by a number of road indices, Represents a natural constant of the natural product, The density of the traffic flow is indicated, The road class is indicated as such, Indicating the number of one-way lanes, Representing a single lane width; Calculating a fault index based on the fault rate of the direct current charging pile, and expressing the fault index as the following formula: ; in the formula, Indicating the failure index(s) of the device, The number of faults of the direct current charging pile is represented, The total number of the installation of the direct current charging piles is represented, The time of the statistics is represented by, Indicating the total fault times of the direct current charging pile in the statistical time, Indicating the index of the fault times of the direct current charging pile, Indicating that the direct current charging pile occurs Time of minor fault repair; constructing an environment index of the direct current charging pile according to the calculated road index and the fault index, and expressing the environment index as follows by a formula: ; in the formula, And the environmental index of the direct current charging pile is represented.
3. The intelligent operation and maintenance management method of a direct current charging pile according to claim 1, wherein the calculation of the operation reliability index of the direct current charging pile according to the time sequence data is specifically as follows: The mutation index is expressed as: ; in the formula, The mutation index is indicated as such, Indicating that the time sequence data corresponding to the direct current voltage or the direct current signal is in the frequency band The total duration of the abrupt change to smooth operation is generated, Time sequence data representing direct current voltage or direct current signal in frequency band Time for the inner stable operation to generate no mutation; The intensity index is expressed as: ; in the formula, The intensity index is represented by a number of values, Representing the calculation of the maximum value to be taken, The representation takes the minimum value to calculate, Represented at the sampling point A direct voltage or a direct current value at the voltage sensor, Representing the average value of direct current voltage or direct current in the running process of the direct current charging pile; the fluctuation index is expressed as: ; in the formula, The index of the fluctuation is represented as, The number of maximum sampling points is indicated, Representing the slave sampling points To the sampling point Performing accumulation calculation; dividing the time sequence data according to different preset lengths to obtain corresponding subsequences, calculating trend indexes based on the subsequences, and expressing the trend indexes as follows by a formula: ; in the formula, The trend index is represented by a trend index, A linear regression operation function is represented as such, The preset length is indicated to be the same as the preset length, Representing the difference between the maximum and minimum values of the sub-sequences, The standard deviation of the subsequence is indicated, The total number of sub-sequences is indicated, Representing a sub-sequence index; Respectively calculating mutation indexes, intensity indexes, fluctuation indexes and trend indexes of the direct current voltage and direct current signals, and respectively calculating corresponding operation reliability indexes based on calculation results, wherein the operation reliability indexes of the direct current voltage signals are expressed as follows by formulas: ; in the formula, An operation reliability index representing the direct current voltage signal, Representation fetch The operation of the index is performed, Representing an arctangent operation; The operational reliability index calculation formula of the direct current signal is consistent with the operational reliability index of the direct current voltage signal, and is expressed as ; Constructing a comprehensive operation reliability index of the direct-current charging pile according to the operation reliability index of the direct-current voltage and direct-current signals Expressed as: 。
4. The intelligent operation and maintenance management method of a direct current charging pile according to claim 1, wherein the feature vector for extracting the power quality parameter by using the graph isomorphic neural network is specifically as follows: The power quality parameters comprise direct-current voltage amplitude, power grid frequency, temperature, direct-current voltage harmonic components and three-phase direct-current voltage unbalance; Carrying out normalization processing on the power quality parameters, and mapping each normalized power quality parameter to a corresponding node in a preset graph structure as an initial feature vector of the node, wherein the preset graph structure comprises a plurality of nodes, and each node represents one type of power quality parameter; Dynamically calculating the behavior similarity between any two nodes in the graph structure based on the data sequence of each power quality parameter; judging the connection relation among the nodes according to a preset similarity threshold value, determining node pairs with behavior similarity meeting preset conditions as being adjacent to each other, and constructing a graph topological structure describing the association relation among the power quality parameters according to a judging result; Inputting the constructed graph topological structure into a graph isomorphic neural network, enabling each node to aggregate the characteristic information of adjacent nodes through iterative computation of the graph isomorphic neural network, and updating and generating a new characteristic vector containing neighborhood information by combining with the initial characteristic vector of the node; And carrying out average pooling operation on the new feature vectors of the nodes output by each layer of the graph isomorphic neural network, integrating the features obtained after the average pooling operation, and outputting a feature vector set of the power quality parameters.
5. The intelligent operation and maintenance management method for the direct current charging pile according to claim 4, wherein the feature vector for extracting the electric energy quality parameter by using the convolutional neural network is specifically as follows: the direct current voltage and direct current characteristics in the electric energy quality parameters are extracted by utilizing a convolutional neural network, and the characteristics are expressed as follows in a formula: ; ; in the formula, Indicating that a CNN convolution operation is to be performed, The extracted dc voltage characteristic vector is represented, The extracted dc current characteristic vector is represented, The dc voltage is indicated as being a function of the dc voltage, Representing a direct current; combining the DC voltage characteristic vector and the DC current characteristic vector to form an electric energy parameter characteristic vector set 。
6. The intelligent operation and maintenance management method for the direct current charging pile according to claim 5, wherein updating the feature vectors of the power quality parameter and the power quality parameter by the attention mechanism is specifically as follows: Based on the influence matrix, updating each characteristic vector in the characteristic vector set of the power quality parameter and each characteristic vector in the characteristic vector set of the power quality parameter respectively, and expressing the characteristic vector as a formula: ; ; ; in the formula, Representing the updated power quality parameter feature vector, Representing the updated power quality parameter feature vector, Representing the influence matrix of the image, The weight matrix is represented by a matrix of weights, Representing the transpose of the matrix, Representing the hyperbolic tangent activation function, The first of the feature vector sets representing the power quality parameter The number of feature vectors is chosen to be the same, A set of eigenvectors representing the power quality parameters, The first of the feature vector sets representing the quality parameters of the electrical energy The number of feature vectors is chosen to be the same, A set of eigenvectors representing the quality of the electrical energy, A weight matrix representing the eigenvectors of the power quality parameter, A weight matrix representing the eigenvectors of the power quality parameter, Representing the topology of the graph.
7. The intelligent operation and maintenance management method of a direct current charging pile according to claim 1, wherein the specific step of constructing the operation performance index comprises: And taking the characteristic vector set as the input of the multi-layer perceptron which is pre-trained, obtaining a metering error predicted value of the direct current charging pile of the area to be managed, and expressing the metering error predicted value as follows by a formula: ; in the formula, The predicted value of the metering error is indicated, A multi-layer perceptron is shown, Representing the updated power quality parameter feature vector, Representing the updated electric energy quality parameter characteristic vector; based on the predicted value of the metering error and the true value of the metering error, the running performance index of the direct current charging pile is constructed, and the running performance index is expressed as follows by a formula: ; ; in the formula, The running performance index is indicated to be the same, The probability is represented by a probability that, Representing the standard deviation of the true value of the metrology error, Representing the variance of the true value of the metrology error, Representing a sample Is used for the measurement of the error prediction value of (a), Representing a sample Is used to measure the true value of the error of the (a), The index of the sample is represented and, Representing the true value of the metering error, Indicating the acquisition of electric energy, Representing the actual power.
8. The intelligent operation and maintenance management method of a direct current charging pile according to claim 1, wherein the operation and maintenance management index of the direct current charging pile is calculated based on the direct current charging pile environment, the operation reliability and the operation performance index, and is expressed as the following formula: ; in the formula, 、 And Respectively corresponding weight parameters of the preset direct current charging pile environment, the running reliability and the running performance index, Representing the operation and maintenance management index of the direct current charging pile, Indicating the environmental index of the DC charging pile The values after the normalization are used for the calculation, Indicating an operational reliability index The values after the normalization are used for the calculation, Indicating the running performance index Normalized values.
9. The intelligent operation and maintenance management system for the direct-current charging pile is characterized by comprising the following modules: The environment index construction module is used for counting road conditions and direct current charging pile fault rates of the direct current charging pile positions in the area to be managed, and calculating direct current charging pile environment indexes including road indexes and fault indexes based on the road conditions and the direct current charging pile fault rates; The system comprises an operation reliability index construction module, a direct current charging pile operation reliability index calculation module and a direct current charging pile operation reliability index calculation module, wherein the operation reliability index construction module is used for acquiring direct current voltage and direct current signals in the charging process of the direct current charging pile and acquiring time sequence data corresponding to the direct current voltage and the direct current signals by utilizing an analog-to-digital conversion and EEMD algorithm; The system comprises an operation performance index construction module, a characteristic vector collection and a measurement error prediction value, wherein the operation performance index construction module is used for acquiring electric power and electric energy quality parameters of the direct current charging pile in real time, extracting characteristic vectors of the electric power quality parameters by using a graph isomorphic neural network, extracting characteristic vectors of the electric energy quality parameters by using a convolution neural network, updating the characteristic vectors of the electric power quality parameters and the electric energy quality parameters by using an attention mechanism to obtain a characteristic vector collection, taking the characteristic vector collection as input of a multi-layer perceptron which is pre-trained, obtaining a measurement error prediction value of the direct current charging pile in a region to be managed, and constructing the operation performance index of the direct current charging pile based on the measurement error prediction value and a measurement error true value; And the operation and maintenance management module is used for calculating the operation and maintenance management index of the direct-current charging pile based on the direct-current charging pile environment, the operation reliability and the operation performance index, and formulating a corresponding operation and maintenance management strategy based on the operation and maintenance management index.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1 to 8 when the program is executed by the processor.

Description

Intelligent operation and maintenance management method and system for direct-current charging pile Technical Field The invention relates to an intelligent operation and maintenance management method and system for a direct current charging pile, and belongs to the technical field of operation and maintenance of charging piles. Background Along with the continuous expansion of the electric automobile market, the direct current charging pile is used as a key device for quick charging, and the operation and maintenance management of the direct current charging pile becomes a key link for ensuring the stability and the safety of charging service. Because the direct current charging pile runs in a complex and changeable environment for a long time, challenges such as equipment aging and fault frequency are faced, effective operation and maintenance management is important for guaranteeing the continuity and reliability of charging service, and a series of challenges exist in the operation and maintenance of a traditional direct current charging pile network. First, conventional operation and maintenance often requires a lot of human resources and time investment, which is certainly a huge burden for a large-scale direct current charging pile network. Second, traditional operation and maintenance is slow in response to problems and usually requires human intervention to resolve. The low-efficiency operation and maintenance mode can cause instability and service interruption of the direct-current charging pile network, and great inconvenience is brought to the charging experience of users. The Chinese patent application with the publication number of CN117952592A discloses an intelligent management method of a direct current charging pile, which comprises the steps of acquiring historical operation data of the direct current charging pile, corresponding historical environment data and historical operation data according to time sequence, evaluating the health state of the direct current charging pile based on the historical operation data of the direct current charging pile, the corresponding historical environment data and the historical operation data, acquiring real-time operation data and real-time environment data of the direct current charging pile, inputting a health state evaluation result, the real-time operation data and the real-time environment data of the direct current charging pile into a pre-trained fault prediction model for carrying out fault risk prediction, and managing and maintaining the direct current charging pile according to a fault risk prediction result. In view of the foregoing, there is a need for a method and system for implementing intelligent operation and maintenance management of a dc charging pile without requiring a complicated iterative process. Disclosure of Invention In order to solve the problems in the prior art, the invention provides an intelligent operation and maintenance management method and system for a direct current charging pile. The technical scheme of the invention is as follows: On the one hand, the invention provides an intelligent operation and maintenance management method for a direct current charging pile, which comprises the following steps: Calculating the road condition and the direct current charging pile fault rate of the direct current charging pile position in the area to be managed, and calculating the direct current charging pile environment index, including the road index and the fault index, based on the road condition and the direct current charging pile fault rate; Acquiring direct-current voltage and direct-current signals in the charging process of the direct-current charging pile, and acquiring time sequence data corresponding to the direct-current voltage and the direct-current signals by utilizing an analog-to-digital conversion and EEMD algorithm; The method comprises the steps of collecting power and electric energy quality parameters of a direct-current charging pile in real time, extracting feature vectors of the power quality parameters by using a graph isomorphic neural network, extracting feature vectors of the electric energy quality parameters by using a convolution neural network, updating the feature vectors of the power quality parameters and the electric energy quality parameters by using an attention mechanism to obtain a feature vector set, taking the feature vector set as input of a multi-layer perceptron with which is pre-trained, obtaining a metering error predicted value of the direct-current charging pile in a region to be managed, and constructing an operation performance index of the direct-current charging pile based on the metering error predicted value and a metering error true value; Based on the environment, the operation reliability and the operation performance index of the direct-current charging pile, calculating an operation and maintenance management index of the direct-current charging pile, and formula