考虑灵敏度区域的多目标鲁棒性优化算法

The multi-objective robust optimization method considering sensitivity region

随机变量导致工程问题具有不确定性.设计者希望设计方案不仅能满足目标性能最优,而且希望目标性能受不确定性的影响在可接受范围之内.对此,本文提出了考虑灵敏度区域的多目标鲁棒性优化方法(multi-objective robust optimization based on sensitivity region,SR–MORO).SR–MORO可以用来解决设计变量存在不确定性时目标鲁棒性优化设计问题.该方法假定不确定性变量属于区间变量,并不需要知道随机变量的概率分布.SR–MORO采用非梯度优化方法,所以,它可以解决目标函数和约束条件不连续的情况.当参数变化幅度大,超过目标函数线性变化范围,该方法也同样适用.最后,通过实例验证了本方法的适用性.

Engineering design problems are usually uncertain because of the random variables. Designers want to design scheme to meet the goal of not only the best performance, but also want to target performance is affected by the uncertainty within the acceptable range. To solve this problem, we propose a multi-objective robust optimization method considering sensitivity region（multi-objective robust optimization based on sensitive region, SR–MORO）. SR–MORO can be used to solve optimization design problem involving uncertainty design variables. This method assumes that the uncertainty variables belong to the interval variables, so it does not need to know the probability distribution of random variables. Nongradient optimization method is used to solve the robust optimization problem, so that the approach is applicable for cases that have discontinuous objective and constraint functions with respect to uncontrollable parameters. When the parameters changed much over the linear range of the objective function, the method is also applicable. Finally, the applicability is also verified by an example.