Multi-objective Particle Swarm Optimization Based on Angle Preference for ε-Pareto Domination

This paper proposes a multi-objective particle swarm optimization algorithm. Firstly, the external archives were obtained based on the Pareto domination. Secondly, the Pareto optimal solution of the archived set was filtered by the crowding function for the particles to be easily localized.Then, different mutations of the different sub-components of the particle population were used to increase the diversity of the solution. Finally, the preference information of the decision maker was introduced and the Pareto optimal solution was accorded with the decision maker's preference. Numerical experiments show that the final distance index of the Pareto solution is close to zero, which is close to the real Pareto boundary. The decision-making preference makes the final optimal solution not cover the entire Pareto boundary. Therefore, the search time is saved and the convergence is improved.