应用粒子群优化算法的短期风电功率预测.pdfVIP

第 35 卷 第 5 期 电 网 技 术 Vol. 35 No. 5 2011 年 5 月 Power System Technology May 2011 文章编号：1000-3673 （2011 ）05-0159-06 中图分类号：TM 614 文献标志码：A 学科代码：470·4054 应用粒子群优化算法的短期风电功率预测 杨志凌，刘永前 （华北电力大学，北京市昌平区 102206 ） Short-Term Wind Power Prediction With Particle Swarm Optimization YANG Zhiling, LIU Yongqian (North China Electric Power University, Changping District, Beijing 102206, China) ABSTRACT: According to the wind speed and wind direction 引起电力公司、风电企业和相关科研人员的重视。 data from numerical weather prediction, a short-term wind 风电功率预测模型主要分为统计预测模型和 power forecasting is performed by BP neural network model 物理预测模型。前者是根据风速、风向、功率的历 with particle swarm optimization (PSO). The influences of data 史数据(来自实际观测或数值气象预报) ，按照连贯 dependency on forecasting results are analyzed and the performances of BP neural network model with and without 原则和类推原则，建立风速、风向等输入量与功率 PSO are compared. Research results show that the BP neural 输出量之间的映射关系，形成统计预测模型。其主 network model with PSO possesses better performance that 要方法有持续性模型 [1-2] 、自回归平均滑动 that without PSO, and based on the testing samplings with [1-5] (auto-regressive moving average，ARMA)模型 、 better data dependency a better forecasting result can be [1-2] [3,6] [5] Markov 模型 、卡尔曼滤波法 、回归预测法 、 achieved. The higher the correlation coefficient between the 灰色预测法 [7] 、人工神经网络(artificial neural numerically predicted wind data and output powe