Improved Particle Swarm Optimization and its Application into Optimal preparing Process
Wang Ya-Lin, Wang Ning, Chunhua Yang, Gui Wei-hua · 2007
Aiming at the premature convergence problem of particle swarm optimization algorithm (PSO), an improved PSO is proposed, which is based on the effects of inertia weight on the convergence performance. In the improved PSO, the inertia weight nonlinearly decreases with iteration time increasing, and its changing curve is just like the shape of S, moreover the curvature of the curve can be adaptively adjusted according to current searching state of the particle swarm. This algorithm is applied to raw mix slurry optimal preparing of alumina process. The calculated results show that it not only has convergence property obviously prior to traditional PSO and genetic algorithm, but also can avoid the premature convergence problem effectively.