Preference Incorporation in Evolutionary Multi-Objective Optimization By Logistic Function

In the field of engineering and economics, we frequently face challenges where several goals must be optimized simultaneously. Such scenarios fall under the category of multi-criteria decision-making, specifically known as multi-objective optimization problems (MOPs). Evolutionary multi-objective optimization algorithms (EMOAs) are successful applications of evolutionary computation in the field of MOPs. In order to enable traditional EMOAs to obtain the optimal solutions preferred by the decision maker (DM), preference-based evolutionary multi-objective optimization is proposed. Preference-based EMO can incorporate the preference information of the DM into EMOAs, so as to guide the population to search the region of interest (ROI) of the MOPs. This paper proposes a preference-based EMOA, called NSGA-II/LF, which incorporates the preference information of the DM into NSGA-II by logistic function. The performance of NSGA-II/LF is compared with several classical algorithms in many aspects on a large number of test problems. Through the summary and analysis of the experimental results, it is verified that the proposed algorithm can obtain high-quality optimal solutions in line with the preference of the DM when facing MOPs.