{"title":"基于SSA非降噪处理的电力负荷预测","authors":"Yindong Jin, He Xiao, Chengui Fu","doi":"10.1145/3569966.3569999","DOIUrl":null,"url":null,"abstract":"Aiming at the problems of high randomness and low prediction accuracy of power load data, a power load prediction model is formed by integrating Singular Spectrum Analysis(SSA) and a Gated Recurrent Unit(GRU) network with a feature mapping layer added, which can effectively improve the power load prediction accuracy. The method takes historical load data as input, uses nonlinear time series processing technology SSA to extract features reflecting complex dynamic changes of load, constructs the extracted feature vector into a time series form as the input of the FL-GRU network, and superimposes the prediction results of each subsequence. get the final prediction result. To avoid the loss of effective information in the data during the noise reduction process, the method performs non-noise reduction processing. Experiments are carried out with a household power load data set in the UK and a data set provided by ISO New England, the method achieved 98.86% and 97.31% prediction accuracy on both datasets.","PeriodicalId":145580,"journal":{"name":"Proceedings of the 5th International Conference on Computer Science and Software Engineering","volume":"118 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Power load forecasting based on SSA non-noise reduction processing\",\"authors\":\"Yindong Jin, He Xiao, Chengui Fu\",\"doi\":\"10.1145/3569966.3569999\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Aiming at the problems of high randomness and low prediction accuracy of power load data, a power load prediction model is formed by integrating Singular Spectrum Analysis(SSA) and a Gated Recurrent Unit(GRU) network with a feature mapping layer added, which can effectively improve the power load prediction accuracy. The method takes historical load data as input, uses nonlinear time series processing technology SSA to extract features reflecting complex dynamic changes of load, constructs the extracted feature vector into a time series form as the input of the FL-GRU network, and superimposes the prediction results of each subsequence. get the final prediction result. To avoid the loss of effective information in the data during the noise reduction process, the method performs non-noise reduction processing. Experiments are carried out with a household power load data set in the UK and a data set provided by ISO New England, the method achieved 98.86% and 97.31% prediction accuracy on both datasets.\",\"PeriodicalId\":145580,\"journal\":{\"name\":\"Proceedings of the 5th International Conference on Computer Science and Software Engineering\",\"volume\":\"118 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-10-21\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 5th International Conference on Computer Science and Software Engineering\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3569966.3569999\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 5th International Conference on Computer Science and Software Engineering","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3569966.3569999","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Power load forecasting based on SSA non-noise reduction processing
Aiming at the problems of high randomness and low prediction accuracy of power load data, a power load prediction model is formed by integrating Singular Spectrum Analysis(SSA) and a Gated Recurrent Unit(GRU) network with a feature mapping layer added, which can effectively improve the power load prediction accuracy. The method takes historical load data as input, uses nonlinear time series processing technology SSA to extract features reflecting complex dynamic changes of load, constructs the extracted feature vector into a time series form as the input of the FL-GRU network, and superimposes the prediction results of each subsequence. get the final prediction result. To avoid the loss of effective information in the data during the noise reduction process, the method performs non-noise reduction processing. Experiments are carried out with a household power load data set in the UK and a data set provided by ISO New England, the method achieved 98.86% and 97.31% prediction accuracy on both datasets.