Application of PSO to multimodel self-tuning dynamic matrix control

作者：Yue Heng; Zhang Hai-Jun; Chai Tian-You

来源：Journal of Northeastern University, 2008, 29(2): 153-156+165.

• Dynamic matrix
• Multimodel
• Particle swarm optimization (PSO)
• Predictive control
• Self tuning
• Transient response

To obtain the fixed models in multimodel control problem, the PSO (particle swarm optimization) algorithm is applied to multimodel self-tuning dynamic matrix control. For a class of SISO discrete time systems with jumping model parameters, the PSO algorithm is used to optimize the self-tuning model parameters online and manage dynamically the fixed models in accordance to the similarity between models, so that the number of models may be handled and system burden can be reduced effectively. Model switching tactics is used to select the controller best to describe the actual controlled system. Simulation results show that the method proposed can improve the transient response greatly in comparison with conventional control algorithms and demonstrate its efficiency and feasibility.

• 出版日期2008
• 单位东北大学