基于农村发展模式分析的中长期负荷预测方法

Medium-long-term Load Forecasting Method Based on Rural Development Mode Analysis

针对农村用电方式和用电需求变化问题,本文在考虑国家政策、农村经济等影响农村发展模式及负荷变化因素的基础上,提出了一种基于农村发展模式分析的中长期负荷预测方法。首先,对电力系统大数据进行了分析,提出了K-means-Robust聚类算法与加权自适应K近邻算法,搭建了农村发展模式预测模型。然后,针对不同农村发展模式,使用基于灰色关联度分析的正则化门控循环神经网络模型预测农村中长期负荷变化曲线。最后,以某农村为例,验证了所提方法的可行性。

Aimed at the problems with rural electricity consumption mode and electricity demand change,a mediumlong-term load forecasting method based on rural development mode analysis is proposed,which takes into account fac⁃tors affecting rural development mode and load changes such as national policy and rural economy.First,the big data of power system are analyzed,the K-means-Robust clustering algorithm and the weighted adaptive K-nearest neighbor al⁃gorithm are proposed,and a rural development model prediction model is built.Then,according to different rural devel⁃opment modes,the regularized depth-gated recurrent neural network model based on the gray correlation analysis meth⁃od is used to predict the medium-long-term load change curve in rural areas.Finally,the validity of the proposed meth⁃od is verified by taking one rural area as an example.