An Improved Adaptive Dynamic Particle Swarm Optimization Algorithm

In order to overcome the weakness that particle swarm optimization algorithm is likely to fall into local minimum when the complex optimization problems are solved, a new adaptive dynamic particle swarm optimization algorithm is proposed. The paper introduces the evaluation index of particle swarm premature convergence to judge the state of particle swarm in the population space, for the sake of investigates the timing of taking effect of influence function. The influence function is adaptively adopted for the aberrance and renovation of the population space when algorithm is trapped into local optimization, so as to efficiently exert the mechanism of “double evolving and promoting” in cultural particle swarm optimization algorithm. According to the premature convergence degree of population space to adaptively adjust the inertia weight of particle, it makes the population can maintain diversity of the inertia weight during evolution and keep the balance between the global convergence property and the convergence speed. Finally, the test on four benchmark functions shows that the proposed algorithm not only has strong search capability, but also convergence speed and precision of the algorithm is improved effectively.