{"title":"基于改进k -均值算法和LSTM的配电网低压站区线损计算方法","authors":"Yuanzhu Fan","doi":"10.1109/EEI59236.2023.10212489","DOIUrl":null,"url":null,"abstract":"As renewable energy sources with random and intermittent characteristics are connected to the distribution network, the traditional theoretical calculation method of line loss is no longer applicable to the estimation of line loss in low-voltage distribution networks with DGs. A novel method of line loss calculation method based on improved K-means algorithm and LSTM for low-voltage stations area of distribution networks with DGs. Firstly, in order to effectively explore the relevance of line loss impact factors in distribution networks containing distributed power sources, the maximum mutual information coefficient is used to select line loss impact factors and establish a line loss index evaluation system. Secondly, the Cuckoo Search algorithm is introduced to optimise the k-means clustering algorithm, the clustering criterion function of weighted euclidean distance is proposed, and the adaptive step size and adaptive bird's nest elimination probability are defined with the inverse tangent function to solve the problem that the traditional clustering algorithm is too sensitive to the initial clustering centre and to improve the clustering accuracy. Finally, the clustering data is fed into an LSTM neural network, which is trained and fitted to obtain line loss estimates. The accuracy and validity of the proposed method was verified by an example analysis using a typical load day sample data of 410 active low-voltage stations area containing PV in a region.","PeriodicalId":363603,"journal":{"name":"2023 5th International Conference on Electronic Engineering and Informatics (EEI)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Line Loss Calculation Method Based on Improved K-Means Algorithm and LSTM for Low-Voltage Stations Area of Distribution Networks with DGs\",\"authors\":\"Yuanzhu Fan\",\"doi\":\"10.1109/EEI59236.2023.10212489\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"As renewable energy sources with random and intermittent characteristics are connected to the distribution network, the traditional theoretical calculation method of line loss is no longer applicable to the estimation of line loss in low-voltage distribution networks with DGs. A novel method of line loss calculation method based on improved K-means algorithm and LSTM for low-voltage stations area of distribution networks with DGs. Firstly, in order to effectively explore the relevance of line loss impact factors in distribution networks containing distributed power sources, the maximum mutual information coefficient is used to select line loss impact factors and establish a line loss index evaluation system. Secondly, the Cuckoo Search algorithm is introduced to optimise the k-means clustering algorithm, the clustering criterion function of weighted euclidean distance is proposed, and the adaptive step size and adaptive bird's nest elimination probability are defined with the inverse tangent function to solve the problem that the traditional clustering algorithm is too sensitive to the initial clustering centre and to improve the clustering accuracy. Finally, the clustering data is fed into an LSTM neural network, which is trained and fitted to obtain line loss estimates. The accuracy and validity of the proposed method was verified by an example analysis using a typical load day sample data of 410 active low-voltage stations area containing PV in a region.\",\"PeriodicalId\":363603,\"journal\":{\"name\":\"2023 5th International Conference on Electronic Engineering and Informatics (EEI)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-06-30\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2023 5th International Conference on Electronic Engineering and Informatics (EEI)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/EEI59236.2023.10212489\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 5th International Conference on Electronic Engineering and Informatics (EEI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/EEI59236.2023.10212489","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Line Loss Calculation Method Based on Improved K-Means Algorithm and LSTM for Low-Voltage Stations Area of Distribution Networks with DGs
As renewable energy sources with random and intermittent characteristics are connected to the distribution network, the traditional theoretical calculation method of line loss is no longer applicable to the estimation of line loss in low-voltage distribution networks with DGs. A novel method of line loss calculation method based on improved K-means algorithm and LSTM for low-voltage stations area of distribution networks with DGs. Firstly, in order to effectively explore the relevance of line loss impact factors in distribution networks containing distributed power sources, the maximum mutual information coefficient is used to select line loss impact factors and establish a line loss index evaluation system. Secondly, the Cuckoo Search algorithm is introduced to optimise the k-means clustering algorithm, the clustering criterion function of weighted euclidean distance is proposed, and the adaptive step size and adaptive bird's nest elimination probability are defined with the inverse tangent function to solve the problem that the traditional clustering algorithm is too sensitive to the initial clustering centre and to improve the clustering accuracy. Finally, the clustering data is fed into an LSTM neural network, which is trained and fitted to obtain line loss estimates. The accuracy and validity of the proposed method was verified by an example analysis using a typical load day sample data of 410 active low-voltage stations area containing PV in a region.