Rational option of optimum parameters of robust design of hybrid of PMOO with sequential uniform algorithm

In this paper, reasonable regulation of optimum parameters in sequential uniform design (algorithm) is developed by means of probabilistic multi–objective optimization (PMOO) in term of total preferable probability, which aims to conduct rational option in the subsequent deep optimization with a series of provisional ″optimum status″ candidates. The provisional ″optimum statuses″ were produced in the sequential uniform algorithm of subsequent deep optimization in each step, which is in turn used to form a ″special point set″ in this study, the total preferable probability is evaluated for the ″special point set″. The final optimum status is with the highest total preferable probability of the ″special point set″ comparatively. Besides, under condition of ″target value being the best″, both discrepancy of average value \(\overline{Y}\) $\overline{Y}$ of a response from its target value Y₀ (ε =| \(\overline{Y}\) $\overline{Y}$ −Y₀|) and averaged deviation γ of the actual response value Y from the target value Y₀ are taken as the dual individual response objectives for robust design simultaneously. Two examples are given to illuminate the proposed procedure.