A New Particle Swarm Optimization for Solving Constrained Optimization Problems

This paper proposes a new particle swarm optimization(PSO) for solving the constrained optimization problems.Based upon an acceptable assumption that any feasible solution is better than any infeasible solution,a new mechanism for constraints handling is incorporated in the standard PSO to transform the constrained optimization problem into an unconstrained optimization problem.In addition to the mechanism of constraints handling,a mutation strategy to increase population diversity is added to the proposed algorithm,which can enhance the probability of leading the particle swarm escape from local optimums,and then improve the convergence speed and solution quality.The experimental results compared with genetic algorithm and a standard PSO show that the proposed algorithm is a feasible algorithm for solving constrained optimization problems.