摘要
针对传统设计方法难以解决滑块刚度和轻量化之间的矛盾问题,建立起基于BP神经网络和多种群遗传算法的滑块尺寸参数优化设计模型。采用正交试验设计方法安排试验方案,以有限元计算数据作为训练和测试样本,获得了滑块尺寸参数组合对其刚度的高精度映射关系。在保证滑块刚度指标的前提下,以滑块质量最小化为优化目标,并采用遗传算法求解,实现了滑块轻量化设计。
A slide dimension optimization model is established on the basis of BP neural networks and genetic algorithms to solve the contradictory problem between the the stiffness and light weight of the slide that can not be solved with conventional methods. A test plan is prepared with orthogonal design method. The data that are calculated with finite element method are trained and tested to obtain the high-precision mapping relationship between the combined slide dimensions and slide stiffness. The light weight design of the slide is realized by using genetic algorithms to solve the equations under the precondition of ensuring the required stiffness of the slide and the optimization target of minimizing the slide weight.
出处
《一重技术》
2014年第6期1-5,共5页
CFHI Technology
