CN-121997085-A - Salt density analysis method of electrical insulation equipment based on online conductivity algorithm

CN121997085ACN 121997085 ACN121997085 ACN 121997085ACN-121997085-A

Abstract

The invention discloses an on-line conductivity algorithm-based salt density analysis method for electric insulation equipment, which relates to the technical field of intelligent operation and maintenance and insulation state monitoring of electric equipment, and comprises the steps of collecting conductivity and environmental parameters; the method comprises the steps of constructing a hybrid parameter optimizing model, iteratively optimizing super-parameter combinations of an OPTICS clustering algorithm by utilizing a particle swarm-multi-target gray wolf algorithm, outputting optimal super-parameter combinations, configuring the optimal super-parameter combinations to an OPTICS clustering algorithm kernel to generate an optimal salt density clustering model, importing conductivity and environmental parameters to the optimal salt density clustering model, determining a core distance and an reachable distance to analyze a connected cluster structure, converting an unsupervised cluster into a supervised salt density reference value through a cluster salt density mapping mechanism, outputting a predicted salt density value, combining a multi-level risk threshold based on the predicted salt density value, and triggering abnormal alarm.

Inventors

HUANG HE
ZHONG HAI
SUN DAWEI
WANG YIRU
HAN YIPIN
Chan Keyang
XU JINGJIA
XU KENAN
SHEN WENWEN
YANG TIANQI

Assignees

国网辽宁省电力有限公司鞍山供电公司

Dates

Publication Date: 20260508
Application Date: 20251224

Claims (9)

1. An electrical insulation equipment salt density analysis method based on an online conductivity algorithm is characterized by comprising the following steps: when the insulator enters a detection period, the dissolution cup and the environment sensor array are controlled by the lifting table, and conductivity and environment parameters are synchronously collected; Constructing a hybrid parameter optimizing model, iteratively optimizing the super-parameter combination of the OPTICS clustering algorithm by utilizing a particle swarm-multi-target gray wolf algorithm, outputting an optimal super-parameter combination, configuring the optimal super-parameter combination to an OPTICS clustering algorithm kernel, and generating an optimal salt density clustering model; the conductivity and the environmental parameters are imported into an optimal salt density clustering model, the core distance and the reachable distance are determined, so that a connected cluster structure is resolved, and an unsupervised cluster is converted into a supervised salt density reference value through a cluster salt density mapping mechanism; based on the predicted salt deposit density value, an abnormal alarm or a dynamic adjustment processing strategy is triggered by combining a multi-level threshold judging mechanism.
2. The method for analyzing the salt density of the electrical insulation equipment based on the online conductivity algorithm according to claim 1, wherein the step of collecting the conductivity and the environmental parameters comprises the following steps: the lifting platform synchronously triggers and controls the dissolving cup and the environment sensor array; under the triggering of a lifting table, the dissolving cup contacts the surface of the insulator through the ascending, immersing, emulsifying and descending action sequences to obtain the conductivity of the solution; the environment sensor array measures environment parameters in real time and performs time stamp alignment with the conductivity data; the conductivity and environmental parameters are pre-treated.
3. The method of on-line conductivity algorithm based electrical insulation device salt density analysis according to claim 1, wherein the hyper-parametric combination comprises a maximum neighborhood radius and an achievable distance threshold.
4. The method for analyzing the salt density of the electrical insulation equipment based on the online conductivity algorithm according to claim 3, wherein the outputting the optimal super parameter combination comprises the following steps: Determining super-parameter combinations of an OPTICS clustering algorithm to construct a two-dimensional parameter search space; Executing global search of a particle swarm algorithm, randomly generating an initial population in a two-dimensional parameter search space, wherein the population consists of m particles, each particle has a unique position vector and a unique speed vector; introducing a multi-target wolf algorithm to perform local mining, namely sorting the population in descending order of fitness, selecting N0 particles before sorting, converting the particles into initial leading wolf position vectors, performing dynamic calibration, and if the maximum iteration number is reached, guiding the population to move by adjusting the wolf position vectors, otherwise, outputting an optimal super-parameter combination, wherein N0 is a positive integer greater than 0.
5. The method for analyzing the salt density of an electrical insulation device based on an on-line conductivity algorithm according to claim 1, wherein the determining the core distance and the reachable distance comprises: determining a core distance: Invoking the maximum neighborhood radius, traversing the sample space, and counting the neighborhood sample number of any sample point in the sample space within the maximum neighborhood radius; comparing the number of the neighborhood samples with a preset minimum point threshold, wherein if the number of the neighborhood samples is larger than or equal to the preset minimum point threshold, the sample point is judged to be a core object, otherwise, the sample point is judged to be a noise point; Based on the core object, sequentially ordering all sample points in the maximum neighborhood radius range, marking the minimum point threshold value as N1, and taking Euclidean distance between the N1 th sample point and the nearest sample point as a core distance, wherein N1 is a positive integer greater than 0; Determining the reachable distance: Based on the core distance, selecting a core object and sample points in the neighborhood of the core object, calculating Euclidean distance between the core object and the sample points, comparing the core distance of the core object with the Euclidean distance, and taking a larger value in the core object and the Euclidean distance as the reachable distance of the sample points relative to the core object point.
6. The method for analyzing the salt density of the electrical insulation equipment based on the online conductivity algorithm according to claim 5, wherein the analyzing the connected cluster structure comprises the following steps: Constructing an reachable distance graph based on the reachable distances; invoking an optimal reachable threshold of the hybrid parameter optimizing model, performing low-valley cutting on the reachable distance graph, identifying low-valley regions, marking the continuous low-valley regions as an independent connected cluster, and distributing unique cluster IDs; And dividing the salt density clusters into salt density clusters with different risk grades based on the distribution positions of the connected clusters, wherein the salt density clusters comprise low, medium and high risk salt density grade clusters.
7. The method for analyzing the salt density of the electrical insulation equipment based on the online conductivity algorithm according to claim 1, wherein the calculating the salt density cluster grade to which the new sample belongs and outputting the predicted salt density value comprises the following steps: receiving conductivity and environmental parameters acquired in real time and taking the conductivity and the environmental parameters as new samples; Starting on-line matching by using an optimal salt density clustering model, traversing all known communication cluster structures which are analyzed in the model, respectively calculating the reachable distances of a new sample relative to each communication cluster core object, comparing and analyzing all reachable distances, and selecting a communication cluster with the smallest reachable distance as a target home cluster of the new sample so as to determine the salt density cluster grade of the new sample; and calling a cluster salt density mapping mechanism, searching and extracting a historical salt density average value corresponding to a target home cluster based on a unique cluster ID of the cluster to serve as a salt density reference value, and directly assigning the salt density reference value to serve as a predicted salt density value of current electrical insulation equipment.
8. The method for analyzing the salt density of the electrical insulation equipment based on the online conductivity algorithm according to claim 1, wherein the multi-level threshold discrimination mechanism comprises: comparing the predicted salt deposit density value with a preset first safety threshold value and a preset second safety threshold value: the first-stage judgment, namely triggering an abnormal alarm signal when the predicted salt density value is larger than or equal to a first safety threshold value; second-level discrimination, namely automatically shortening the detection period of the lifting platform for controlling the dissolution cup when the predicted salt density value is larger than the first safety threshold value and smaller than the second safety threshold value; And three-stage discrimination, namely when the predicted salt density value is smaller than or equal to a first safety threshold value, judging that the equipment is in a safety operation interval, adopting a sliding average strategy to merge the current sample into a historical database, updating a reference mean value of the corresponding salt density cluster, and realizing model self-adaptive calibration.
9. The method for analyzing the salt density of the electrical insulation equipment based on the online conductivity algorithm according to claim 1, wherein the environmental parameters comprise temperature and humidity, wind speed and pollutant concentration.

Description

Salt density analysis method of electrical insulation equipment based on online conductivity algorithm Technical Field The invention relates to the technical field of intelligent operation and maintenance and insulation state monitoring of power equipment, in particular to an electric insulation equipment salt deposit density analysis method based on an online conductivity algorithm. Background The equivalent salt deposit density value on the surface of the insulator is one of important bases for judging the severity of the insulating pollution condition outside the electric porcelain, and the salt deposit density measurement work has important significance for the safe operation of the electric power system; in the conventional insulator salt density detection work, the conventional detection method has the following limitations: On one hand, a destructive detection mode is often adopted, an insulator is required to be detached for manual cleaning and laboratory measurement, so that equipment is stopped, the power supply reliability is affected, the operation and maintenance cost is increased rapidly due to high risk caused by high-altitude operation of scenes such as a power transmission line insulator, meanwhile, the traditional detection method has obvious hysteresis and inefficiency, the detection period is usually quarterly or annually, sudden pollution conditions such as sand storm and industrial pollution are difficult to capture in time, and spot sampling cannot reflect regional equipment pollution distribution differences. On the other hand, the single salt density value obtained by the traditional monitoring method is not related to environmental parameters, a pollution accumulation dynamic model cannot be built, and early warning is carried out by relying on an empirical threshold value, so that the false alarm rate is high, the single and empirical detection mode is difficult to adapt to the requirements of a power system on the dynamic and accurate sensing of the equipment state, the problems of untimely pollution early warning, insufficient pertinence of operation and maintenance decisions and the like are often caused, the equipment is caused to safely operate, the potential risk is caused, and the power grid fault probability is even increased. Therefore, there is a need to develop a salt density detection technique that is non-destructive, efficient in real time, and capable of correlating environmental parameters. Disclosure of Invention (One) solving the technical problems Aiming at the defects of the prior art, the invention provides a salt density analysis method of electric insulation equipment based on an online conductivity algorithm, which comprises the steps of controlling a dissolution cup to be immersed into an insulator through a lifting platform, collecting the environmental parameters and the conductivity of the insulator, obtaining optimal parameters through PSO-MOGWO mixing and optimizing, constructing an OPTICS cluster model, inputting the environmental parameters and the conductivity into the model, completing the division of salt density level clusters, the salt density value prediction and the risk early warning, accurately identifying the salt density state of the insulator through the model output, realizing the effective prevention and control of the operation risk of the insulation equipment outside a power grid, guaranteeing the operation reliability of the equipment, improving the accuracy and the timeliness of the evaluation of the pollution state of the insulator, and solving the problems in the background technology. (II) technical scheme In order to achieve the above purpose, the invention is realized by the following technical scheme: the application provides an electrical insulation equipment salt density analysis method based on an online conductivity algorithm, which comprises the following steps: when the insulator enters a detection period, the dissolution cup and the environment sensor array are controlled by the lifting table, and conductivity and environment parameters are synchronously collected; Constructing a hybrid parameter optimizing model, iteratively optimizing the super-parameter combination of the OPTICS clustering algorithm by utilizing a particle swarm-multi-target gray wolf algorithm, outputting an optimal super-parameter combination, configuring the optimal super-parameter combination to an OPTICS clustering algorithm kernel, and generating an optimal salt density clustering model; the conductivity and the environmental parameters are imported into an optimal salt density clustering model, the core distance and the reachable distance are determined, so that a connected cluster structure is resolved, and an unsupervised cluster is converted into a supervised salt density reference value through a cluster salt density mapping mechanism; based on the predicted salt deposit density value, an abnormal alarm or a dynamic adjustment processing