偏好聚合不精确法在稳健设计中的应用

Improving Robust Design with Preference Aggregation Methods

在稳健设计中,最好的设计是通过求解多标准优化问题获得。通常使用的权重和方法,可能丢失Pareto点,同时权重的人为选取具有盲目性,不能获得满意的设计。为此,采用不相关点计算权重,运用全面的偏好聚合方法获得满意的Pareto分布。应用于一个和两个性能变量的湿式多盘制动器稳健设计中,表明该方法通过简单的计算就可得到整个Pareto沿,同时避免了权重的人为盲目选取。

The putative best design is obtained by solving a multi-criteria optimization problem in robust design. The problem solved, the usually used weighted sum method may lose Pareto points, and the artificial selection of the weight is ambiguous, which may fail to obtain a desired design. So that, adopting an indifference point calculating weight, a more comprehensive preference aggregation method was implemented for the robust design. Using this method for wet multi-disc brake robust design with one and two performance variables respectively, the results indicate that this method can get allover Pareto with a simple calculation, and it can avoid the artificial blindly selection of the weight at the same time.