基于智能电表的住宅短期电力负载预测  

Prediction of short-term electric loads based on smart meter

作　　者：杨继革 严俊 陈丽春 贺乐华 余圣彬 YANG Ji-ge;YAN Jun;CHEN Li-chun;HE Le-hua;YU Sheng-bin(Quzhou Power Supply Company,State Grid Zhejiang Electric Power Co.Ltd.,Quzhou 324400,China;Longyou Power Supply Co.Ltd.,State Grid Zhejiang Electric Power Co.Ltd.,Quzhou 324400,China;Jiangshan Power Supply Co.Ltd.,State Grid Zhejiang Electric Power Co.Ltd.,Quzhou 324400,China)

机构地区：[1]国网浙江省电力有限公司衢州供电公司,浙江衢州324400 [2]国网浙江省电力有限公司龙游县供电有限公司,浙江衢州324400 [3]国网浙江省电力有限公司江山市供电有限公司,浙江衢州324400

出　　处：《沈阳工业大学学报》2022年第3期255-258,共4页Journal of Shenyang University of Technology

基　　金：国家自然科学基金项目(61033004);国网浙江省衢州电力有限公司科技项目(5211QZ17001Z).

摘　　要：为了解决居民用电短期负荷预测率低的问题,提出了一种基于用户智能电表实时测量数据的数学建模方法.利用谱分析设计了成型滤波器来评估高斯噪声,再结合卡尔曼滤波对不同采样周期的监测数据进行负荷预测和精度评价.结果表明,提高采样精度、获取更多的实时测量数据可以明显改善负荷预测的准确性,但相应地也会带来更高的计算成本.在避免高计算负荷的同时达到预期的预测精度需要限制用于预测的数据量,而通过选择合理的测量采样率,可以获取满意的折衷方案.In order to solve the problem for inefficient prediction of short-term loads for residential electricity consumption,a mathematical modeling method based on real-time measured data obtained from users’smart meters was proposed.A shaping filter was designed by the spectrum analysis so as to evaluate Gaussian noise.In combination with Kalman filtering,the load prediction and accuracy evaluation of monitored data were carried out during different sampling periods.The results show that the accuracy of load prediction can be significantly improved by improving the sampling accuracy and obtaining more real-time measured data,but the corresponding calculation cost will be higher.It is necessary to limit the data amount for prediction,so as to avoid high calculation load and achieve the expected prediction accuracy.By reasonably selecting the measured sampling rate,a satisfactory compromise scheme can be obtained.

关 键 词：居民用电 短期负载 智能电表 成型滤波 高斯噪声 卡尔曼滤波 采样周期 

分 类 号：TM715[电气工程—电力系统及自动化]