动力总成悬置系统全局稳健设计方法

To overcome the problem that the traditional robust design method has low efficiency, a design method, which combines global optimal algorithm and parallel computation, was proposed to perform robust design of powertrain mounting system (PMS). First, the global optimal algorithm was employed to search the feasible region in design space. Then parallel computation was introduced to evaluate robustness of all feasible design points. This method can effectively improve computational efficiency and reduce possibility of repeated design. This method was used to perform robust design of one PMS. Results indicate that, with increasing of uncertainty, robust design becomes a necessary choice and its performance is better than deterministic design. As the standard deviation of modal decoupling ratio has a larger variation compared to its mean value, improvement of its robustness is implemented by decreasing of standard deviation.