基于改进的神经网络的电力系统负荷预报

Power System Load Forecast Based on Improved Neural Network

用共轭梯度算法改进BP神经网络,使得网络的性能得到改善,缩短了网络训练时间,提高了预报精度,并将其应用于电力系统负荷预报,通过对东北电网实测数据的仿真计算表明,该方法是可行的。

Conjugate gradient algorithm is used to improve BP neural network, that can improve performance ofnetwork, shorten training time, enhance forecasting precision. The improved network is used in power load fore-casting of northeast of China, the calculation results show that this approach is feasible.