摘要
用BP人工神经网络建立了重载丝杆螺母副的摩擦学特性与载荷之间关系的磨损自补偿数学模型。结果表明 :用L -M规则进行神经网络学习训练可使网络收敛快 ,误差小。网络输出结果与实验结果比较有极好吻合性。该神经网络可为工程设计人员在摩擦学设计时提供有效的计算工具。
The BP neural network used the wear-self-compensation mathematics model of the relationship of tribology characteristic with load of the heavy load screw-nut pairs was established.The network lessens quicker and the error less trained with the L-M rule .The output of the network were precise by comparison with the test results.The network model could provide calculable measure for the tribology design engineers.
出处
《机械设计与制造》
北大核心
2000年第6期42-43,共2页
Machinery Design & Manufacture
基金
国家自然科学基金资助项目
