Parameter approximate dynamic optimization for PSO systems

— This paper presents a novel swarm approximate dynamic programming method (swarm-ADP) for parameter optimization of PSO systems, from the perspective of optimal control. Based on the stability results of a simplified PSO and the swarm-ADP algorithm, parameter dynamic optimization and computation is studied in detail for a deterministic PSO system and a stochastic PSO system, respectively. Furthermore, numerical simulations based on several benchmarks optimization are performed. The results show the validity of the proposed parameter optimization method for PSO systems.

• Type: Conference
• Year: 2009
• Where: CDC
• Authors: Qi Kang, Lei Wang, Derong Liu, Qidi Wu
• Keywords: CDC 2009, Control Systems, Parameter Dynamic Optimization, Parameter Optimization, Pso Systems