摘要
提出基于类电磁机制算法的对角递归神经网络的风电功率预测模型.对角递归神经网络属于动态递归神经网络,具有较好的动态性能;类电磁机制算法模拟电磁场中带电粒子间吸引与排斥机制,可进行全局优化,具有好的收敛性能.模型采用类电磁机制算法对对角递归神经网络进行优化,可避免使神经网络训练陷入局部最小点,提高模型的预测精度.仿真结果表明,模型可有效降低预测误差,获得满意的预测精度.
Diagonal recursive neural network based wind power forecasting model using electro- magnetism-like mechanism algorithm is constructed in this paper. The diagonal recursive neural network is contained in dynamic recursive neural network, which possesses good dynamic performance. Electromagnetism-like mechanism algorithm simulates attraction and repulsion mechanism for particles in the electromagnetic field. It possesses global optimization ability and good convergence performance. Diagonal recursive neural network is optimized by electromagnetism- like mechanism algorithm, which can avoid neural network to immerse in the local minimal points and improve the forecasting precision. The testing results show that the proposed forecasting mode can reduce forecasting error and obtain satisfactory forecasting precision.
出处
《电力科学与技术学报》
CAS
2012年第4期17-21,共5页
Journal of Electric Power Science And Technology
基金
山东省优秀中青年科学家奖励计划(BS2011NJ005)
关键词
类电磁机制算法
对角递归神经网络
风电功率预测
风电场
electromagnetism-like mechanism algorithm
diagonal recursive neural networks
wind power forecasting
wind farm
