MPSO: Modified particle swarm optimization and its applications

Abstract

Abstract Particle swarm optimization (PSO) is a population based meta-heuristic search algorithm that has been widely applied to a variety of problems since its advent. In PSO, the inertial weight not only has a crucial effect on its convergence, but also plays an important role in balancing exploration and exploitation during the evolution. However, PSO is easily trapped into the local optima and premature convergence appears when applied to complex multimodal problems. To address these issues, we present a modified particle swarm optimization with chaos-based initialization and robust update mechanisms. On the one side, the Logistic map is utilized to generate uniformly distributed particles to improve the quality of the initial population. On the other side, the sigmoid-like inertia weight is formulated to make the PSO adaptively adopt the inertia weight between linearly decreasing and nonlinearly decreasing strategies in order to achieve better tradeoff between the exploration and exploitation. During this process, a maximal focus distance is formulated to measure the particle's aggregation degree. At the same time, the wavelet mutation is applied for the particles whose fitness value is less than that of the average so as to enhance the swarm diversity. In addition, an auxiliary velocity-position update mechanism is exclusively applied to the global best particle that can effectively guarantee the convergence of MPSO. Extensive experiments on CEC′13/15 test suites and in the task of standard image segmentation validate the effectiveness and efficiency of the MPSO algorithm proposed in this paper.