基于PSO-OMP优化的WD-ASD超短期负荷预测

发布时间：2018-04-26 14:40

本文选题：超短期负荷预测 + 原子稀疏分解　；参考：《电工电能新技术》2017年12期

【摘要】：为提高负荷预测精度,降低电力系统规划决策的保守性,本文提出了一种基于小波-原子稀疏分解(WD-ASD)的超短期负荷预测模型。该模型使用模糊聚类算法提取相似日为历史数据,采用小波分解(WD)作为前置环节,以基于原子表达式的自预测和基于最小二乘支持向量机(LSSVM)的残余分量预测为基础构建原子稀疏分解(ASD)预测模型,分别对负荷的高低频分量进行预测,并将结果相加得到最终预测值。其中ASD分解过程由正弦原子库自适应匹配分解完成,并将粒子群算法(PSO)和正交匹配追踪(OMP)算法相结合以增强原子稀疏分解能力。实际负荷数据算例验证了所提方法的自适应性、快速性及有效性。
[Abstract]:In order to improve the accuracy of load forecasting and reduce the conservatism of power system planning decision, this paper presents an ultra-short-term load forecasting model based on wavelet atomic sparse decomposition (WD-ASD). The model uses fuzzy clustering algorithm to extract similar date as historical data and wavelet decomposition (WD) as the leading link. Based on the self-prediction based on atomic expression and the residual component prediction based on least squares support vector machine (LSSVM), the atomic sparse decomposition (ASD) prediction model is constructed to predict the high and low frequency components of the load, respectively. The final prediction value is obtained by adding the results together. The process of ASD decomposition is accomplished by sine atom library adaptive matching decomposition, and particle swarm optimization (PSO) algorithm is combined with orthogonal matching tracking (OMP) algorithm to enhance the ability of atomic sparse decomposition. An example of actual load data shows that the proposed method is adaptive, fast and effective.
【作者单位】：电力电子节能与传动控制河北省重点实验室燕山大学;华北电力大学电气与电子工程学院;北京亦庄国际开发建设有限公司;
【基金】：河北省高等学校科学技术研究项目(QN2016064)
【分类号】：TM715

【相似文献】