改进BP算法的轮胎侧向力神经网络模型

用改进的误差反向传播 (BP)算法建立了轮胎侧向力的数学模型 ,研究了模型精度与网络结构、训练方法之间的关系。结果表明 ,该算法克服了传统BP算法网络收敛速度慢及易陷入局部极小点的两个缺点 ,能较精确地对轮胎侧向力进行建模

Neural Network Model withImproved Backward Error Propagation Algorithm for Tyres

An improved Backward Error Propagation Algorithm is used to build neural network model of tyre sideforce. The relations between the model precision and the network structure as well as the training method are researched. The results show which the algorithm has overcome two shortcomings which the traditional BP Algorithm has, one is the low network weakening speed, the other is to fall into the partial minimun value easily, and the model of tyre sideforce can be precisely build.