论文部分内容阅读
介绍了多层前馈网的基本算法及其在镦粗压力计算中的应用.针对其基本学习算法(反向误差传播算法,BP算法)存在的收敛速度慢和易陷入局部最小值的问题,提出了基于PID调节的BP算法,由此建立起镦粗压力计算的神经网络知识库,经过测试集考核,证明该模型算法正确,能有效地提高模型的判别精度,是一种有效的和有前途的算法.
The basic algorithm of multilayer feedforward network and its application in upsetting pressure calculation are introduced. Aiming at the problems of slow convergence speed and easy falling into local minimum of its basic learning algorithm (BP algorithm), a BP algorithm based on PID adjustment is proposed to establish the neural network knowledge of upsetting pressure calculation It is an effective and promising algorithm for the library to pass the test set examination to prove that the model algorithm is correct and can effectively improve the discrimination accuracy of the model.