A novel dynamic reference point model for preference-based evolutionary multiobjective optimization

Abstract In the field of preference-based evolutionary multiobjective optimization, optimization algorithms are required to search for Pareto optimal solutions preferred by decision maker (DM). The reference point is a type techniques that effectively describe preferences DM. So far, either static or interactive with process. However, existing do not cover all application scenarios. A novel case, i.e., changes over time due environment change, has been considered. This paper focuses on problems dynamic First, we propose change model simulate preference DM time. Then, algorithm framework clonal selection (ĝa-NSCSA) and genetic (ĝa-NSGA-II) designed solve such kind problems. addition, in terms practical applications, experiments portfolio tested. Experimental results benchmark applications show ĝa-NSCSA exhibits better performance among compared algorithms.