摘要 |
<p>Disclosed is a method for forecasting residential quarter short-term load, characterized by comprising: step 1: reading history sample data, and performing data screening on the history sample data; step 2: obtaining a load influencing factor coefficient set; step 3: constructing a training sample set and a forecasting sample set for neural network load forecasting; step 4: performing influencing factor separating processing on the training sample set to obtain a trained neural network; step 5: performing influencing factor separating processing on the forecasting sample set, and using a result obtained after processing as input of the trained neural network, to obtain a corresponding output result; and step 6: performing influencing factor adding processing on the output result to obtain load forecasting data of a week in the future. The present invention solves problems that residential quarter load data is small, has large fluctuation, and has lots of influencing factors, and improves the accuracy and operation speed of residential quarter load forecasting.</p> |
申请人 |
STATE GRID CORPORATION OF CHINA;JIANGSU ELECTRIC POWER COMPANY;JIANGSU ELECTRIC POWER COMPANY YANGZHOU POWER COMPANY;JIANGSU FANGTIAN POWER TECHNOLOGY CO., LTD. |
发明人 |
LIU, ZHONG;JIANG, YIQUAN;ZHANG, CHUNYAN;LI, PEIPEI;LI, XINJIA;FAN, YILONG;WANG, GONGQIN;WU, GANG;ZHENG, FEI |