基于改进PSO-LSSVM的短期电力负荷预测  

Short-Term Power Load Forecasting Based on Improved PSO-LSSVM

作　　者：马小津[1] 朱博[2] 戴琳[1] 张伟[1] 陈熙[1] 

机构地区：[1]合肥通用机械研究院,安徽合肥230088 [2]安徽省电力设计院,安徽合肥230022

出　　处：《自动化技术与应用》2016年第3期5-9,19,共6页Techniques of Automation and Applications

摘　　要：短期电力负荷预测是电力系统安全调度、经济运行的重要依据,随着电力系统的市场化,负荷预测的精度直接影响到电力系统运行的可靠性、经济性和供电质量。LSSVM不仅保持了SVM的优点,同时降低了计算复杂性,加快求解速度,为短期电力负荷预测提供了一个新的研究方向。本文将最小二乘支持向量机(LSSVM)用于短期电力负荷预测,提出基于LSSVM的短期电力负荷预测模型,同时建立改进粒子群模型对LSSVM进行参数优化,并以浙江台州某地区的历史负荷数据和气象数据为例进行验证,实例验证表明,改进PSO-LSSVM模型的预测效果明显提高。Short-term power load forecasting is an important basis for the safe dispatch of power system and economic operation, With the marketization of power system, the precision of load forecasting directly affects the reliability, the economy and the quality of power supply of power system operation. LSSVM not only keeps the advantages of SVM, but also reduces the complexity of calculation, speed up the computation, provides a new research direction for short-term load forecasting of power system. This article uses LSSVM to short-term power load forecasting, and proposes the model of short-term power load forecasting based on the LSSVM. Simultaneously, it establishes the improvement PSO model to optimize the LSSVM parameter. Taking the historical load data and meteorological data of Zhejiang Taizhou some areas for example, it indicates that the forecast effect of improved PSO-LSSVM model enhances distinctly.

关 键 词：最小二乘支持向量机(LSSVM) 短期电力负荷 预测 粒子群(PSO) 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]