摘要
基于BP人工神经网络的L M算法 ,建立了磨合磨损的分形参数预测模型。将该模型用于销 盘磨合磨损试验 ,对最佳分形维数进行了准确预测。该模型收敛速度快、误差小 。
Based on L-M algorithm of BP artificial neural network, the fractal parametric prediction model of running in and wear was established. Let the model be used in the pin-disc running in and wear test, an accurate prediction was carried out upon the optimum fractal dimension. This model has a quick converging speed and a smaller error. The result of output has a very good coincidence with the result of test.
出处
《机械设计》
CSCD
北大核心
2004年第11期43-44,共2页
Journal of Machine Design
关键词
神经网络
L-M算法
磨合磨损
最佳分形维数
neural network
L-M algorithm
running in and wear
optimum fractal dimemsion
