Particle Swarm Optimization with Opposition-based Disturbance

Particle swarm optimization (PSO) often traps in the local optimal solutions. In this paper, an opposition-based disturbance procedure was introduced into a basic PSO, which was abbreviated as PSOOD. For this proposed algorithm, opposition-based disturbance was implemented according to the probability when the personal best position was updated for each particle. Such procedure not only avoids the missing of cognition component in the velocity update equation, but also increases the population diversity. Numerical tests on three benchmark functions were conducted to compare the algorithm performances. The results show that PSOOD is able to escape from the local optimal solutions efficiently when solving complex optimization problems, which enhances the global search ability obviously while guaranteeing the convergence.