Many-objective evolutionary algorithm based on three-way decision

Abstract

In recent years, many-objective optimization problems have been widely used. however, with the increase of the number of objectives, the difficulty of solving increases exponentially, and the imbalance between convergence and diversity becomes more serious. In view of the above problems, this paper combines the idea of three-way decision, redesigns the environment selection strategy, and proposes a many-objective optimization algorithm based on three-way decision. Firstly, the distance from the individual to the ideal point is used as an index to measure individual convergence, the minimum distance from the individual to other solutions is used as an indicator to measure individual diversity, and the individuals with good convergence and good diversity are selected separately by combining the thresholds of the three-way decision; and secondly, A dynamic threshold acquisition method is designed to further improve the performance of the algorithm; Finally, it is proved that the algorithm can effectively balance convergence and diversity through tests of different data sets, so as to verify the feasibility and effectiveness of the algorithm