基于FOA-Elman神经网络的光伏电站短期出力预测模型  

Short-Term Photovoltaic Power Forecasting Based on Elman Neural Network with Fruit Fly Optimization Algorithm

作　　者：韩伟[1] 王宏华[1] 杜炜[2] 

机构地区：[1]河海大学能源与电气学院,南京211100 [2]国网电力科学研究院,南京210003

出　　处：《电测与仪表》2014年第12期120-124,128,共6页Electrical Measurement & Instrumentation

基　　金：江苏省研究生培养创新工程(CXZZ120228)

摘　　要：提出了基于果蝇优化算法(FOA)-Elman神经网络的光伏电站出力短期预测模型,采用具有动态递归性能的Elman神经网络,可增强光伏电站出力预测模型的联想和泛化推理能力,保证出力预测的精度。引入人体舒适度,减少输入向量个数;通过FOA对Elman神经网络进行学习训练,可充分利用FOA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。最后,与常规Elman模型进行对比仿真实验,结果表明所提出预测模型的正确性和有效性。The model based on Elman neural network （NN） with fruit fly optimization algorithm （FOA） is proposedto forecast the short-term photovohaic （PV） power. Using dynamic recurrent Elman NN, the reasoning andgeneralization capacity of PV power forecasting model is enhanced, and forecasting accuracy is ensured. The humanbody amenity is introduced to reduce the number of input vectors. The FOA is used to train the Elman NN, whichcan make full use of the global optimization performance of FOA and overcome the defects such as local optimalsolution, slow convergence speed and complex programming. Finally, in comparison with the simulation results ofElman NN, the numerical results verify the effectiveness and correctness of the proposed mode.

关 键 词：光伏电站 出力预测 ELMAN神经网络 FOA算法 

分 类 号：TM615[电气工程—电力系统及自动化]