论文部分内容阅读
基于优化技术，提出一种改进型ＢＰ算法。该算法收敛速度快，学习精度高，训练周期短，可适用于神经网络的离线及在线学习，是一种实用性很强的快速ＢＰ算法。并用神经网络对电力系统稳定器控制序列的学习仿真证明了该算法的优越性
Based on the optimization technique, an improved BP algorithm is proposed. The algorithm has the advantages of fast convergence rate, high learning accuracy and short training period. It can be applied to offline and online learning of neural networks. It is a fast and practical BP algorithm. The learning simulation of power system stabilizer control sequence with neural network proved the superiority of the algorithm