Study of population diversity of multiobjective evolutionary algorithm based on immune and entropy principles

A key problem in a multiobjective evolutionary system is how to take measures to preserve diversity in the population. The mechanism of natural immune system and entropy principle are applied in a multiobjective evolutionary process to solve this problem and a strategy of preserving diversity in the population of a multiobjective evolutionary algorithm based on immune and entropy principles is introduced. The detailed design method is shown. Finally, we describe the computer simulation of implementing several two-objective flow shop scheduling problems and compare the computing results of the new method with the multiobjective genetic algorithm. Experimental results show that this strategy can effectively preserve population diversity and it has good search performance.