摘要
建立了磨损自补偿状态下 4 5钢 /黄铜摩擦副表面复合层铜 ,锌元素分布规律的BP神经网络模型 ,该网络可较准确地预测摩擦副表层转移元素铜和锌沿表层深度分布的规律。用L -M规则训练可使网络收敛加快精度提高。本研究可为摩擦学设计的程序化计算和分析提供方便且有效的工具。
In a state of self-compensate for wear, a BP neutral net model of copper and zinc element distribution law on 45 copper /brass rubbing secondary surface is set up. This net can predict precisely the distribution law of transfer element of copper and zinc on rubbing secondary surface. With L-M regular training it can promote the speed and precision of net convergence. This study provides an effective and convenient way for programmed calculation and analysis in tribology design.
出处
《重庆工学院学报》
2000年第4期1-4,共4页
Journal of Chongqing Institute of Technology
基金
国家自然科学基金
