Adaptive simulated annealing particle swarm optimization for catalyst protected region parameter identification

Aiming at the parameter identification problem of catalyst protected region in the process of propylene oxidation, a novel parameter identification method has been proposed for catalyst protected region using an adaptive simulated annealing particle swarm optimization (ASAPSO) algorithm. Synchronous change learning factors and linear decrease progressively inertia weights are embedded in the simulated annealing particle swarm optimization algorithm. The information exchange capacity is enhanced by the synchronous change learning factors. The overall search ability and local improved ability are balanced by the linear decrease progressively inertia weights. The proposed algorithm has some advantages in the aspect of good stability, strong information exchange capacity and fast convergence. Meanwhile, the shortcoming of local minimum valve is solved by the proposed algorithm. Simulation results show that the algorithm is feasible and accurate. The catalyst protected region of propylene oxidation from 6.35% to 11.25% is determined. Finally, the proposed ASAPSO algorithm is efficient.