论文部分内容阅读
首次将响应面方法应用到车轮踏面优化中。分别基于多项式响应面车轮踏面优化方法和高斯径向基函数响应面车轮踏面优化方法,用C++语言编写优化软件模块,实现在C++环境中调用多体动力学软件ADAMS/Rail进行轨道车辆系统的动力学仿真分析;将动力学仿真分析得到的数据反馈给车轮踏面优化软件模块,完成整个优化设计的循环过程。分析比较这2种响应面方法在车轮踏面优化中应用的结果表明,2种响应面方法非常适合于车轮踏面优化,而且收敛速度快;从磨耗指数的优化来看,用2种响应面方法对200 km.h-1客车车轮踏面进行优化后,新车轮踏面较原车轮踏面的磨耗指数降低52%左右,说明对磨耗指数的优化非常有效。在优化计算时间方面,多项式响应面车轮踏面优化方法为50 566 s,高斯径向基函数响应面车轮踏面优化方法为16 449 s;在1次试验设计的试验次数方面,多项式响应面车轮踏面优化方法为77次,而高斯径向基函数响应面车轮踏面优化方法只有25次。在车轮踏面优化中高斯径向基函数响应面方法优于多项式响应面方法。
For the first time, the response surface method is applied to wheel tread optimization. Based on polynomial response surface method of wheel tread optimization and Gaussian radial basis function response surface wheel tread optimization method, C + + language optimization software modules were written to achieve the C + + environment to call multi-body dynamics software ADAMS / Rail rail vehicle system dynamics Learn simulation analysis; the dynamic simulation analysis of the data back to the wheel tread optimization software module to complete the cycle of the optimal design cycle. The results of the analysis and comparison of the two response surface methods applied to the wheel tread optimization show that the two response surface methods are very suitable for the wheel tread optimization and the convergence speed is fast. From the optimization of wear index, two response surface methods 200 km.h-1 After optimizing the wheel tread of the bus, the wear index of the new wheel tread reduced by 52% compared with that of the original wheel tread, which shows that the optimization of the wear index is very effective. In terms of optimization calculation time, the optimization method of polynomial response surface wheel tread is 50 566 s, Gaussian radial basis function response surface wheel tread optimization method is 16 449 s. In terms of the number of tests for one experimental design, the polynomial response surface wheel tread optimization The method is 77 times, while Gaussian radial basis function response surface wheel tread optimization method is only 25 times. Gaussian radial basis function response surface method is better than polynomial response surface method in wheel tread optimization.