本文已被：浏览 1577次 下载 699次
Received:March 09, 2017
Received:March 09, 2017
中文摘要: 精确的电力负荷预测有利于保障电网运行的安全性、稳定性、高效性及经济性.为提高预测精度，采用了一种PSO改进T-S（Takagi-Sugeno）模糊神经网络方法.分析了数据预处理对改善输入量的重要性，讨论了可以让学习率和平滑因子动态调节的改进T-S模糊神经网络算法，从而使PSO找到最优参数，然后结合历史负荷数据、相关影响因素进行预测，以表明改进的T-S模糊神经网络在短期电力负荷中具有更高的控制精度.
中文关键词: 预测精度 PSO优化T-S算法 模糊模型 异常数据
Abstract:Accurate power load forecasting is beneficial to ensure the safety,stability,efficiency and economy of grid operation.In order to improve the prediction accuracy,a PSO improved T-S (Takagi-Sugeno) fuzzy neural network method is applied.The importance of data preprocessing to improving the input is analyzed and an improved T-S fuzzy neural network algorithm that can make the learning rate and smoothing factor adjust dynamically is discussed.Therefore,it can make PSO find the optimal parameters.And then,combined with historical load data and related influencing factors,it is concluded that the improved T-S fuzzy neural network has higher control precision in short-term power load.
文章编号：20174004 中图分类号： 文献标志码：
基金项目:大型电气传动系统与装备技术国家重点实验室开放基金课题（SKLLDJ032016021）.
Reference text: