Enhanced Pareto multi-objective collaborative optimization strategy

In order to improve the convergence performance of standard collaborative optimization strategy and extend its multi-objective optimization compatibility, by adopting Pareto multi-objective genetic algorithm in the system level optimization, an enhanced collaborative optimization using Pareto multi-objective genetic algorithm (ECO-PMGA) is proposed. A sequential ranking method considering the crowed degree is developed to ensure the Pareto optimality and even distribution of non-inferior solutions. The interdisciplinary consistency constraints of 2-norm format are employed to improve the efficiency of discipline level optimizations in ECO-PMGA. The numerical stability and capability of searching Pareto non-inferior solution set are validated through two typical optimization problems. The results indicate that the convergence of system level optimization and numerical stability of ECO-PMGA are fairly enhanced, moreover, the ECO-PMGA shows a good performance in achieving Pareto optimal set. Accordingly, the proposed ECO-PMGA is practical and valuable for multi-objective optimization problems for complex and coupled systems.