A Multi-objective Particle Swarm Optimization Algorithm Based on Reverse Learning

Abstract

In order to solve the contradiction between population diversity and convergence in particle swarm optimization algorithm, in this paper, a particle swarm optimization algorithm with reverse learning is proposed. On this basis, the values of learning factor and constraint factor parameters are modified, and the linear decreasing weight strategy was adopted. By modifying the learning factor and the constraint factor, the algorithm improves the particle optimization ability. It balances the global search and local search of the particle, and the convergence speed is improved by using the inertia weight. When it is detected that the algorithm falls into the local optimal region, the position information of these poor particles is used to guide some particles to reverse learning at a faster flight speed, and the particles are quickly pulled out of the local optimal region. The reverse learning process can not only improve the diversity of particle population, but also ensure the global detection ability of the algorithm. Experimental results show that, compared with the basic MOPSO algorithm, this algorithm has fast convergence speed and high solution accuracy in function optimization.