Dynamic Multiobjective Optimization Using Evolutionary Algorithm with Kalman Filter

摘要: Abstract Multiobjective optimization is a challenging task, especially in changing environment. The study on dynamic multiobjective so far very limited. Benchmark problems, appropriate performance metrics, as well efficient algorithms are required to further the research this field. In paper, Kalman Filter prediction-based evolutionary algorithm proposed solve problems. This prediction model uses historical information predict for future generations and thus, direct search towards Pareto optimal solutions. A scoring scheme then devised enhance by hybridizing with random re-initialization method. models tested analysis of experiment results presented. It shown that capable improving performances, compared using method alone. also suggests additional features could be added improvements much more field still needed.