基于PSO-Elman神经网络的电力系统短期负荷预测模型研究  

Research of Short-Term Load Forecasting Model Based on PSO-Elman Neural Network

作　　者：刘远龙 龚文杰 徐超 张智晟[2] 张伟[2] 王鑫[2] 赵建伟[2] 

机构地区：[1]青岛供电公司,山东青岛266002 [2]青岛大学自动化工程学院,山东青岛266071

出　　处：《青岛大学学报（工程技术版）》2010年第3期43-47,共5页Journal of Qingdao University(Engineering & Technology Edition)

基　　金：山东电力集团公司科学技术项目(2010A-41)

摘　　要：提出了基于粒子群优化算法—Elman神经网络的电力系统短期负荷预测模型,采用具有动态递归性能的Elman神经网络,可增强负荷预测模型的联想和泛化推理能力,保证负荷预测的精度。采用粒子群优化算法对Elman神经网络进行学习训练,可充分利用粒子群优化算法的全局寻优性能,克服常规学习算法易于陷入局部最优解、收敛速度慢、编程复杂等缺陷。通过对地区电网负荷系统的实例仿真证实了所提出方法的有效性,获得了较满意的预测精度,平均绝对误差和最大相对误差分别达到1.988%和4.673%。为该模型用于实际工程取得了有效的进展。In this paper,a short-term load forecasting model based on particle swarm optimization algorithm and Elman neural network is constructed.Elman neural network with dynamic recursive performance can improve association and generalization ability of forecasting model,which can ensure forecasting accuracy.Particle swarm optimization algorithm is used to train Elman neural network,and it possesses excellent global optimization performance,which can overcome some drawback of conventional training algorithm,such as local optimal solution,slow convergence rate and complex programming course.Actual load system is used in case study,and satisfactory forecasting accuracy is acquired.The simulation results show that this model is effective,and mean absolute percentage error and maximum relative error are respectively 1.988% and 4.673%.The model can attempt to be used in actual project.

关 键 词：电力系统 短期负荷预测 ELMAN神经网络 粒子群优化算法 

分 类 号：TM715[电气工程—电力系统及自动化]