| With the rapid development of power technology,the demand for electrical equipment by electric power customers is increasing,showing the characteristics of diverse types and large-scale quantities.Whenever a power customer performs a switching operation on the electrical equipment that it is equipped with,the corresponding operation mode also changes,showing a variability.Different operating modes will generate different harmonic eigenvalues,and the calculation of the threshold is an important basis for determining whether the harmonic data is abnormal.Although all countries have put forward corresponding standards for the threshold calculation of harmonics,these standards use the threshold of a certain(for example,the most valued)operation mode as the basis for judgment,and the abnormal accident cases that occur during the internship process are reflected.This judgment is based on the inability to accurately discriminate whether the harmonic data is abnormal data.Therefore,the power quality detection level of power customers needs to be further improved to adapt to the variability of power customers’ operation modes.In order to solve the problem of low accuracy of harmonic data anomaly detection under different operation modes,a harmonic data anomaly monitoring module is developed based on the existing power intelligent operation and maintenance system of the company,which is suitable for harmonic data anomaly detection under different operation modes of power customers.The specific research content of this paper is divided into the following points:Firstly,a normal cloud model is established according to the harmonic data under normal operation conditions.The fluctuation range of harmonic data under normal operation mode is measured by the entropy of the cloud model.The abnormal threshold of harmonic data is determined according to the’3En outer boundary of the membership curve of the cloud model.By comparing the harmonic data to be detected with the harmonic abnormal threshold,the abnormal detection of harmonic data can be realized,and the harmonic data anomaly detection method in a single operation mode can be obtained,and the program is written based on the Matlab platform.Based on the harmonic data anomaly detection method in single operation mode,the method of harmonic data anomaly detection under different operation modes is realized.Secondly,the relationship between the harmonic threshold determined by the cloud model and the capacity of the harmonic data samples is analyzed,which leads to the question of how much the harmonic data sample size should be taken during the modeling process.Aiming at this problem,this paper improves on the traditional sample capacity determination method,and proposes two methods for determining the data size of harmonic samples.The calculation formula of harmonic sample data capacity based on confidence is obtained.Thirdly,referring to the idea of regional merging in image processing,the process of harmonic data anomaly detection under different operating modes is optimized.The method of combining harmonic thresholds under different operating modes is proposed,and the program is written based on Matlab platform.Fourth,verify the method proposed in this paper in the actual engineering environment.The verification content includes:(1)Discrimination of harmonic data anomaly detection algorithm proposed in this paper under different operation modes;(2)The effectiveness of the method for determining the data size of two harmonic samples designed in different operating modes;(3)The accuracy of the harmonic anomaly threshold combining method proposed in this paper under different operating modes.Finally,based on the previous method of harmonic current data anomaly detection and its optimization scheme proposed in different modes,combined with the existing electric power intelligent operation and maintenance monitoring system of the internship company,the harmonic data abnormality monitoring module of the power customer is added.Through this module,the power customer can monitor the change of harmonic data in different operation modes in real time,and master the operation status of the electrical equipment under different operation modes,so that the power customers can take corresponding control measures. |