论文部分内容阅读
本文将神经网络的BP算法和动态随机算法相互补充而形成一种混合算法,它克服了BP学习算法会陷入局部极小和随机算法耗时长的缺点.将这种算法用于齿轮泵出口流量预测,证明了这种算法学习速度快且能收敛于全局极小点.
In this paper, the BP neural network algorithm and the dynamic random algorithm complement each other to form a hybrid algorithm, which overcomes the BP learning algorithm will fall into the local minima and the stochastic algorithm has the disadvantage of long time. This algorithm is used for gear pump outlet flow forecast It is proved that this algorithm can learn fast and converge to the global minimum.