Multi-Objective Optimization Using Evolutionary Algorithms
Multi-Objective Optimization Using Evolutionary Algorithms
A constraint-based approach to feasibility assessment in preliminary design
Artificial Intelligence for Engineering Design, Analysis and Manufacturing
A fast and elitist multiobjective genetic algorithm: NSGA-II
IEEE Transactions on Evolutionary Computation
| Hi-index | 0.00 |
In this paper, couple of multi-objective optimization (MOO) techniques is implemented on a concept vehicle design problem. The Pareto Set Pursuing (PSP) method, proposed by Shan and Wang [1] is compared with a commercial version of the NSGA-II algorithm. PSP uses a sequential surrogate model generation and progressive importance sampling approach as compared to the NSGA-II, which uses exact function evaluations. This comparative study is initially carried out with the aid of a continuous analytical test problem followed by a mixed discrete continuous, vehicle design problem. Based on these studies, it was found that PSP performed reasonably well as compared to the NSGA-II and offered considerable savings on the number of function evaluations. Additionally, it also provided an evenly spread Pareto frontier. For the concept vehicle design problem, the performance of PSP was comparable with NSGA-II in terms of the accuracy of Pareto frontier and better in terms of the spread of the Pareto frontier.