CN-121642944-B - Real-time monitoring and early warning method and system for distribution network equipment state

CN121642944BCN 121642944 BCN121642944 BCN 121642944BCN-121642944-B

Abstract

The invention relates to the technical field of distribution network equipment state monitoring and discloses a distribution network equipment state real-time monitoring and early warning method and system, wherein the method comprises the steps of establishing a source load balance analysis model for describing the energy supply and energy consumption balance relation of distribution equipment to output a power supply frequency fluctuation value of distribution network equipment when the distribution network equipment is in unbalance of supply and demand; the method comprises the steps of judging the duration time of the power distribution equipment in a supply-demand unbalanced state and outputting a performance evaluation value of the power distribution network equipment according to a power supply frequency fluctuation value, combining the performance evaluation value with a load safety early warning level to output a power distribution equipment early warning scheme, outputting real-time operation data of the power distribution equipment to output a load safety result of the power distribution equipment in a real-time state and in a future period, matching the load safety result and executing the power distribution equipment monitoring early warning scheme. The invention can identify the load security risk possibly occurring in the future period in advance before the current state is not deteriorated, and improves the operation security and reliability of the distribution network equipment.

Inventors

DONG YULONG
LV YIBO
CHEN LIANG
WANG JUAN
DENG FULI
Cheng Miaohai
SI JINGJING
JIANG GANNING
Ma Huaigang
WU ZITING

Assignees

国网甘肃省电力公司兰州供电公司

Dates

Publication Date: 20260508
Application Date: 20260205

Claims (8)

1. The real-time monitoring and early warning method for the state of the distribution network equipment is characterized by comprising the following steps of: Analyzing the load bearing capacity of the distribution network equipment according to the monitoring standard, and establishing a load safety early warning level in an uncertain scene based on the load bearing capacity; Acquiring historical operation data comprising the operation load and the power supply capacity of the power distribution equipment, and generating an operation topological graph by taking the power supply line connection between the power distribution equipment as an edge and the historical operation data as nodes; Analyzing an operation topological graph, acquiring a balance relation between energy supply and energy consumption in the power transmission process, and establishing a source load balance analysis model based on the balance relation by utilizing an elastic network fitting technology; Extracting different energy supply and energy consumption data from a source load balance analysis model based on a variation inference technology as sampling points, and simulating a power supply frequency fluctuation state of power distribution equipment under the condition of power supply fluctuation; extracting the offset amplitude of the fluctuation change of the power supply frequency according to the historical operation data, defining a supply-demand unbalance state judgment standard based on the offset amplitude, and analyzing the amplitude of the fluctuation value of the power supply frequency through the judgment standard; Analyzing duration time and fault frequency of power supply frequency fluctuation of the power distribution equipment based on the amplitude change trend, and judging that the power distribution equipment is in a supply-demand unbalanced state when the duration time and the fault frequency exceed a threshold value; adopting a window division technology to generate a sliding window structure to perform dependency analysis on duration and fault frequency, and generating a model training data set by utilizing an incremental learning mode according to a dependency analysis result; Training a back propagation neural network model by taking a model training data set as input, outputting a predicted value of the duration time and the fault frequency of the power distribution equipment in a supply-demand unbalanced state by using the back propagation neural network model, and outputting a performance evaluation value of the power distribution equipment; combining the performance evaluation value with the load safety early warning level to output a power distribution equipment early warning scheme; And simulating the future operation state of the power distribution equipment by combining the real-time operation data of the power distribution equipment with the frequency calculation technology, outputting a load safety result of the power distribution equipment in the real-time state and in a future period according to a simulation result, and matching and executing a power distribution equipment monitoring and early warning scheme based on the load safety result.
2. The method for real-time monitoring and early warning of distribution network equipment state according to claim 1, wherein the analyzing the operation topological graph, obtaining a balance relation between energy supply and energy consumption in a power transmission process, and establishing a source load balance analysis model based on the balance relation by using an elastic network fitting technology comprises: Converting the operation topological graph into a multi-relation graph based on element paths, and analyzing the multi-relation graph by using a power supply quality relation selection graph convolution layer to acquire aggregation neighbor information so as to model global correlation among the element paths; Screening the aggregated neighbor information to distinguish the importance of each power supply quality relation, and reserving the power supply quality relation with the importance meeting a threshold value as the energy supply and demand balance change condition of the power distribution equipment under different running conditions; the running time and the operation and maintenance information of the power distribution equipment are used as independent variable indexes, the energy supply and demand balance change condition is used as an independent variable index, and a characteristic index pool for constructing a source load balance analysis model is obtained; and determining an index combination for constructing a source load balance analysis model based on characteristic index pool screening influence factor characteristic indexes, and constructing the source load balance analysis model by utilizing the index combination and elastic network fitting technology.
3. The method for real-time monitoring and early warning of a distribution network device state according to claim 2, wherein the converting the operation topological graph into a multi-relation graph based on element paths, analyzing the multi-relation graph by using a power supply quality relation selection graph convolution layer to obtain aggregation neighbor information, and modeling global correlation among the element paths comprises: Defining an operation topological graph as a heterogeneous graph based on a power transmission relation between power distribution equipment and power supply nodes, defining a meta-path set according to service characteristics of the power distribution equipment, and acquiring a neighbor set of the power supply nodes under each meta-path by using the meta-path set so as to convert the heterogeneous graph into a multi-relation graph based on the meta-paths; projecting the power supply node into a characteristic subspace of a graph rolling network to obtain an influence factor index, distributing the influence factor index for each element path, calculating neighbor information of the power supply node under each element path through a mean value aggregation function, and extracting global power supply quality information of each element path by utilizing a reading function; the method comprises the steps of calculating the correlation degree between neighbor information and global power supply quality information, and eliminating neighbor information which is related to the global power supply quality information and is lower than a threshold value based on the correlation degree to obtain global neighbor information; and projecting the global neighbor information to the weight coefficient of the same feature space computing element path, and generating node representation with weight fusion by combining power supply node information through an activation function to obtain the aggregated neighbor information.
4. The method for real-time monitoring and early warning of distribution network equipment state according to claim 3, wherein the step of determining an index combination for constructing a source load balance analysis model based on characteristic index pool screening influence factor characteristic indexes, and the step of constructing the source load balance analysis model by using the index combination and elastic network fitting technology comprises the following steps: respectively quantifying the association degree between each variable index and dependent variable index in the characteristic index pool by adopting a Pearson linear correlation coefficient, a Kendell rank correlation coefficient and a Spilot rank correlation coefficient; Comparing the association degree with the selected standard, screening key characteristic indexes influencing source load balance, determining effective index combinations for constructing a source load balance analysis model, and judging dimension differences and dispersion degree differences of the effective index combinations; Preprocessing the effective index combination according to the dimension difference and the dispersion degree difference, and constructing a source load balance analysis model by using a ridge regression algorithm and a lasso regression algorithm based on the preprocessed effective index combination; And calculating the correlation coefficient elastic network estimated value of each variable index to the dependent variable index in the source load balance analysis model by solving the optimization problem with quadratic programming, and obtaining a source load balance analysis result.
5. The method for real-time monitoring and early warning of a distribution network device state according to claim 1, wherein the step of simulating the future operation state of the distribution device by combining the real-time operation data of the distribution device with a frequency calculation technology, outputting a load safety result of the distribution device in the real-time state and in a future period according to a simulation result, and matching and executing a distribution device monitoring and early warning scheme based on the load safety result comprises the steps of: Acquiring real-time operation data of the power distribution equipment, simulating the operation state probability of the power distribution equipment according to the real-time operation data by adopting a frequency dynamic calculation technology, and generating a random scene of future operation of the power distribution equipment; simulating a random scene of the power distribution equipment based on a Monte Carlo simulation technology, simulating state performance of the power distribution equipment under different loads and environmental fluctuation factors, and evaluating performance evaluation values of the power distribution equipment; And outputting load safety results of the distribution equipment in a real-time state and in a future period according to the real-time operation data and the performance evaluation value, matching a monitoring and early warning scheme of the distribution equipment, and executing early warning processing of the distribution network equipment based on the matching results.
6. The method for real-time monitoring and early warning of a distribution network device state according to claim 5, wherein the steps of obtaining real-time operation data of the distribution device, simulating the operation state probability of the distribution device according to the real-time operation data by using a frequency dynamic calculation technology, and generating a random scene of future operation of the distribution device comprise: The method comprises the steps that real-time operation data of the power distribution equipment in a target time period are collected through a sensor, and denoising and normalization processing are carried out on the real-time operation data through a moving average filtering technology so as to complete data preprocessing; based on real-time operation data and specification and structural parameters of the power distribution equipment, a dynamic modeling mode is adopted to construct a digital twin body comprising the structure, the operation parameters, the environment variables and the operation logic of the power distribution equipment; Mapping processing of the digital twin body and the physical entity of the power distribution equipment is carried out, so that a mature digital twin body is obtained, and the running state probability of the power distribution equipment is simulated by combining the running characteristic dividing state set of the power distribution equipment; And constructing a transition probability matrix of each operation state by adopting a Markov chain based on the operation state probability, and generating a random operation scene of the power distribution equipment in a future prediction period by utilizing the transition probability matrix.
7. The method for real-time monitoring and early warning of distribution network equipment state according to claim 6, wherein the performing mapping processing of the digital twin body and the physical entity of the distribution equipment to obtain a mature digital twin body, and dividing the state set by combining the operation characteristics of the distribution equipment to simulate the operation state probability of the distribution equipment comprises: executing state mapping and behavior simulation mapping, completing real-time synchronization of a physical entity of the power distribution equipment to the digital twin, and adding the re-etching capability of the real-time running state of the digital twin to obtain a mature digital twin; And dividing the state set based on the operation characteristics of the power distribution equipment, dividing time windows, counting the occurrence frequency of real-time operation data corresponding to each operation state in the mature digital twin body in each time window, and determining the operation state probability of the power distribution equipment by utilizing the frequency.
8. A distribution network equipment state real-time monitoring and early warning system for realizing the distribution network equipment state real-time monitoring and early warning method according to any one of claims 1-7, which is characterized in that the system comprises: the standard making module is used for analyzing the load bearing capacity of the distribution network equipment according to the monitoring standard and establishing a load safety early warning level under an uncertain scene based on the load bearing capacity; the frequency fluctuation analysis module is used for establishing a source load balance analysis model for describing the energy supply and energy consumption balance relation of the distribution equipment based on historical operation data, randomly sampling the source load balance analysis model and outputting a power supply frequency fluctuation value of the distribution network equipment when the supply and demand are unbalanced; the performance evaluation module is used for judging the duration time and the fault frequency of the distribution equipment in the supply-demand unbalance state according to the power supply frequency fluctuation value, iteratively training the counter propagation neural network model and outputting a performance evaluation value of the distribution network equipment; The early warning scheme output module is used for combining the performance evaluation value with the load safety early warning level and outputting an early warning scheme of the power distribution equipment; And the real-time monitoring output module is used for simulating the future operation state of the power distribution equipment by combining the real-time operation data of the power distribution equipment with the frequency calculation technology, outputting the load safety result of the power distribution equipment in the real-time state and in the future period according to the simulation result, and matching and executing a power distribution equipment monitoring and early warning scheme based on the load safety result.

Description

Real-time monitoring and early warning method and system for distribution network equipment state Technical Field The invention relates to the technical field of distribution network equipment state monitoring, in particular to a distribution network equipment state real-time monitoring and early warning method and system. Background Distribution network equipment refers to various equipment used for electric energy transmission, distribution and scheduling in a distribution network, and mainly comprises a distribution transformer, a switching device, a circuit breaker, a load switch, a distribution cable, line equipment, a grounding device and the like, wherein the main function of the distribution equipment is to transmit high-voltage electric power from a transmission network to a user side and ensure that the electric power can be safely and stably distributed to each user. The real-time monitoring and early warning of the state of the distribution network equipment is realized by carrying out real-time monitoring on various equipment in the distribution network, combining data analysis and a prediction model, finding potential fault risks or abnormal states of the equipment in time and sending early warning signals in advance, so that early identification and processing of equipment faults are realized, and power interruption or other potential safety hazards caused by the equipment faults are avoided. However, the existing real-time monitoring and early warning scheme for the state of the distribution network equipment still uses a fixed threshold value and a single working condition assumption as a core, mainly depends on whether voltage, current or load is out of limit to trigger an alarm, and lacks systematic description of uncertain operation scenes and long-term operation pressure of the equipment, so that fine and prospective risk management and control are difficult to realize. For the problems in the related art, no effective solution has been proposed at present. Disclosure of Invention Aiming at the problems in the related art, the invention provides a real-time monitoring and early warning method and a real-time monitoring and early warning system for the state of distribution network equipment, so as to overcome the technical problems existing in the prior related art. For this purpose, the invention adopts the following specific technical scheme: in a first aspect, the present invention provides a method for monitoring and early warning a state of a distribution network device in real time, including: Analyzing the load bearing capacity of the distribution network equipment according to the monitoring standard, and establishing a load safety early warning level in an uncertain scene based on the load bearing capacity; Establishing a source load balance analysis model for describing the energy supply and energy consumption balance relation of the distribution equipment based on historical operation data, randomly sampling the source load balance analysis model, and outputting a power supply frequency fluctuation value of the distribution network equipment when the supply and demand are unbalanced; Judging the duration time and the fault frequency of the distribution equipment in a supply-demand unbalanced state according to the power supply frequency fluctuation value, iteratively training a counter propagation neural network model, and outputting a performance evaluation value of the distribution network equipment; combining the performance evaluation value with the load safety early warning level to output a power distribution equipment early warning scheme; And simulating the future operation state of the power distribution equipment by combining the real-time operation data of the power distribution equipment with the frequency calculation technology, outputting a load safety result of the power distribution equipment in the real-time state and in a future period according to a simulation result, and matching and executing a power distribution equipment monitoring and early warning scheme based on the load safety result. Preferably, the method for establishing a source load balance analysis model for describing the energy supply and energy consumption balance relation of the distribution equipment based on historical operation data, randomly sampling the source load balance analysis model, and outputting the power supply frequency fluctuation value of the distribution network equipment when the supply and demand are unbalanced comprises the following steps: Acquiring historical operation data comprising the operation load and the power supply capacity of the power distribution equipment, and generating an operation topological graph by taking the power supply line connection between the power distribution equipment as an edge and the historical operation data as nodes; Analyzing an operation topological graph, acquiring a balance relation between energy supply and energy consumption in