CN-121561539-B - Terminal meter reliability dynamic assessment method, system, equipment and medium based on multi-source data characteristics

CN121561539BCN 121561539 BCN121561539 BCN 121561539BCN-121561539-B

Abstract

The invention discloses a terminal meter reliability dynamic evaluation method, a system, equipment and a medium based on multi-source data characteristics, and relates to the technical field of terminal meter fault prediction. And introducing a mixed feature screening mechanism to carry out mixed feature screening to obtain a key feature set of the terminal meter. And (5) carrying out weight distribution by adopting a Bayesian algorithm to obtain comprehensive weights. And constructing a comprehensive reliability scoring model, and outputting the comprehensive reliability score of the terminal meter. And dynamically setting a risk threshold according to the environmental parameters, and outputting the comprehensive risk level of the terminal meter. The reliability dynamic evaluation of the terminal meter under the multi-source data fusion is realized, the accuracy of the reliability evaluation is improved through dynamic feature screening and self-adaptive weight adjustment mechanisms, and fault risks and operation and maintenance weak links under different scenes are effectively identified.

Inventors

LIU QINGCHAN
YANG JINGXU
JIANG BO
Wu Xianle
Tao Yunxu
YIN YUJUN
LI TENGBIN
Qi Jiaqi
Jin Qihao
ZHENG FENGYI
ZHANG YIMING
LI SHENG

Assignees

云南电网有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260121

Claims (8)

1. A dynamic evaluation method for reliability of a terminal meter based on multi-source data features is characterized by comprising the steps of, Acquiring multi-source data of a metering terminal, integrating the multi-source data to form a structural feature matrix, and preprocessing the structural feature matrix; Introducing a mixed feature screening mechanism to perform mixed feature screening on the pretreated structured feature matrix to obtain a key feature set of a terminal meter; carrying out weight distribution on the key feature set of the terminal meter by adopting a Bayesian algorithm to obtain comprehensive weights; constructing a comprehensive reliability scoring model according to the comprehensive weight, and outputting a comprehensive reliability score of the terminal meter; dynamically setting a risk threshold according to the environmental parameters, and outputting the comprehensive risk level of the terminal meter; the step of introducing a mixed feature screening mechanism to perform mixed feature screening on the preprocessed structured feature matrix to obtain a key feature set of a terminal meter, comprising the following steps: the mixed feature screening comprises a key static feature screening and a key dynamic feature screening; the key static features are screened by constructing a random forest model, and the expression is: , Wherein, the A static feature representing the importance to be assessed, Representing static features The reduction in Gini coefficient that occurs when a single decision tree node splits, Representing static features The sum of the reduction of Gini coefficients in all decision trees, Representing the sum of the reduction of Gini coefficients for all static features in all decision trees, Representing static features Is a global importance score of (2); the correlation between the dynamic characteristics and the fault state is quantified through a mutual information method, and key dynamic characteristics are screened out, wherein the expression is: , Wherein, the For the dynamic characteristics to be evaluated, For fault conditions, p (X) is the edge probability distribution of the dynamic feature X, i.e. the probability that the value of the dynamic feature X is X when the fault condition is not considered, p (Y) is the edge probability distribution of the fault Y, i.e. the prior probability that the value of the fault condition Y is Y when the dynamic feature value is not considered, For the probability that the dynamic feature X and the fault state Y occur simultaneously, For the probability that the dynamic feature X occurs independently of the fault state Y, The mutual information value of the dynamic characteristic X and the fault state Y is obtained; Integrating the key static features and the key dynamic features to obtain a key feature set of the terminal meter; The method for distributing weights of the key feature sets of the terminal meter by adopting the Bayesian algorithm to obtain comprehensive weights comprises the following steps: Inputting a key feature set of a terminal meter, and constructing a Bayesian linear regression model, wherein the expression is as follows: , where y is the reliability score of the terminal, The weight of the ith key feature in the key feature set, the key features include a key static feature and a key dynamic feature, To value the i-th key feature in the key feature set, As a noise term of the sound, the noise, As the variance of the noise is the value of the variance of the noise, As a total number of key features in the set of key features, Representing normal distribution, i being a key feature index in a key feature set; sampling 1000 times by a Markov chain Monte Carlo method, and obtaining posterior distribution mean value of key feature weights as a reference weight; according to the reference weight, a dynamic adjustment factor under the influence of a photovoltaic scene is introduced to generate a comprehensive weight, and the expression is as follows: , , Wherein beta (t) is a dynamic adjustment factor, For the integrated weight of the ith key feature in the key feature set at time t, The photovoltaic fluctuation rate at the time t, For the reference weight of the i-th key feature in the set of key features, And adjusting the weight of the ith key feature in the key feature set at the moment t.
2. The method for dynamically evaluating reliability of a terminal meter based on multi-source data features of claim 1, wherein the steps of obtaining multi-source data of a metering terminal, integrating the multi-source data to form a structured feature matrix comprise: Installing an inductor in a metering terminal to sense environmental characteristics, searching a history file to obtain static characteristics, and obtaining dynamic characteristics according to equipment feedback; all data are aligned according to the device ID and the time stamp to generate a structural feature matrix Wherein The dimension of time is represented as such, For the number of devices to be used, As a number of features, Is the real number domain.
3. The method for dynamically evaluating reliability of a terminal meter based on multi-source data features of claim 2, wherein the preprocessing of the structured feature matrix comprises: aiming at abnormal values in the structural feature matrix, a sliding window mean method is adopted to inhibit burst noise, and the expression is as follows: , Wherein, the Is that The smoothed value of the moment in time, For the point in time Is used to determine the original data value of (a), Representing temporary index variables that are summed for traversal within a sliding window.
4. The method for dynamically evaluating reliability of a terminal meter based on multi-source data features of claim 3, wherein constructing a comprehensive reliability scoring model according to comprehensive weights and outputting comprehensive reliability scores of the terminal meter comprises: the comprehensive reliability scoring model is composed of the sum of weighted characteristic functions, and the expression is: , wherein R is the comprehensive reliability score of the terminal meter, Representing a feature normalization function.
5. The method for dynamically evaluating reliability of a terminal meter based on multi-source data features of claim 4, wherein dynamically setting a risk threshold according to environmental parameters and outputting a comprehensive risk level of the terminal meter comprises: The risk threshold expression is: , Wherein, the Representation of The risk threshold value for the moment in time, As an index of the environmental risk conditions, As a total number of environmental risk condition types, For the conditional triggering of the indication function, Is an adjustment factor for environmental risk conditions.
6. The dynamic terminal meter reliability assessment system based on the multi-source data features is characterized by comprising an acquisition and preprocessing module, a feature screening module, a weight distribution module, an output module and a risk level judging module, wherein the dynamic terminal meter reliability assessment method based on the multi-source data features is applied to any one of claims 1-5; The acquisition and preprocessing module is used for acquiring multi-source data of the metering terminal, integrating the multi-source data to form a structural feature matrix, and preprocessing the structural feature matrix; the feature screening module is used for introducing a mixed feature screening mechanism to perform mixed feature screening on the preprocessed structured feature matrix to obtain a key feature set of the terminal meter; the weight distribution module is used for distributing weights of key feature sets of the terminal meter by adopting a Bayesian algorithm to obtain comprehensive weights; The output module is used for constructing a comprehensive reliability scoring model according to the comprehensive weight and outputting comprehensive reliability scores of the terminal meter; and the risk level judging module dynamically sets a risk threshold according to the environmental parameters and outputs the comprehensive risk level of the terminal meter.
7. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of a terminal meter reliability dynamic assessment method based on multi-source data features as claimed in any one of claims 1 to 5.
8. A computer readable storage medium having stored thereon a computer program, characterized in that the computer program when executed by a processor implements the steps of a terminal meter reliability dynamic evaluation method based on multi-source data features as claimed in any one of claims 1 to 5.

Description

Terminal meter reliability dynamic assessment method, system, equipment and medium based on multi-source data characteristics Technical Field The invention relates to the technical field of terminal meter fault prediction, in particular to a terminal meter reliability dynamic evaluation method, a system, equipment and a medium based on multi-source data characteristics. Background The metering terminal meter is used as core equipment of the intelligent power grid and bears key functions such as electric energy metering, data acquisition and load control, and the operation reliability of the metering terminal meter is directly related to the operation efficiency of the power system, the service quality of users and the economic benefits of power enterprises. However, the problem of operation reliability of the metering terminal meter is increasingly remarkable, especially in the scenes of high penetration of new energy and complex environments (such as coastal high salt fog and industrial high pollution), equipment faults frequently occur, so that metering errors are increased, data are lost and even safety accidents are caused, and stable operation of a power system and user electricity utilization experience are seriously influenced. The existing research relies on partial electric quantity data to evaluate the running state of the electric energy metering device, multi-source data (equipment body state, environmental factors) and the like are not fused, and evaluation indexes are required to be updated further. The index system established by the other type of evaluation method is complete, but the method only aims at the running state of the electric energy meter to be unfolded and evaluated, the index weight is solidified, the change of the risk factors in different scenes cannot be dynamically reflected, the comprehensive reliability evaluation method of the metering terminal meter is lacked, and further the operation and maintenance resource waste and unnecessary potential safety hazards are led. By comprehensively considering the problems, the invention provides a terminal meter reliability dynamic evaluation method based on multi-source data characteristics. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the method solves the technical problems of how to realize the dynamic evaluation of the reliability of the metering terminal meter by fusing multi-source data characteristics under the conditions of high penetration of new energy and complex environment, and meanwhile, the method can be used for carrying out the risk factor change analysis under different conditions according to a dynamic adjustment mechanism of weight so as to identify potential fault risks and operation and maintenance weak links. In order to solve the technical problems, the invention provides a terminal meter reliability dynamic evaluation method based on multi-source data characteristics, which comprises the following steps, Acquiring multi-source data of a metering terminal, integrating the multi-source data to form a structural feature matrix, and preprocessing the structural feature matrix; Introducing a mixed feature screening mechanism to perform mixed feature screening on the pretreated structured feature matrix to obtain a key feature set of a terminal meter; carrying out weight distribution on the key feature set of the terminal meter by adopting a Bayesian algorithm to obtain comprehensive weights; constructing a comprehensive reliability scoring model according to the comprehensive weight, and outputting a comprehensive reliability score of the terminal meter; and dynamically setting a risk threshold according to the environmental parameters, and outputting the comprehensive risk level of the terminal meter. The invention relates to a terminal meter reliability dynamic evaluation method based on multi-source data characteristics, which comprises the following steps of obtaining multi-source data of a metering terminal, integrating the multi-source data to form a structural characteristic matrix, and the method comprises the following steps: Installing an inductor in a metering terminal to sense environmental characteristics, searching a history file to obtain static characteristics, and obtaining dynamic characteristics according to equipment feedback; all data are aligned according to the device ID and the time stamp to generate a structural feature matrix WhereinThe dimension of time is represented as such,For the number of devices to be used,As a number of features,Is the real number domain. As a preferable scheme of the terminal meter reliability dynamic evaluation method based on the multi-source data characteristics, the method for preprocessing the structural feature matrix comprises the following steps: aiming at abnormal values in the structural feature matrix, a sliding window mean method is adopted to inhibit burst