基于GA参数优化整定的PI型广义预测控制

摘要

Because of the proportion and integral factor' s parameter of PI-type generalized predictive contrul is harder tuning, it propose a algorithm based on GA parameter optimization and tuning for PI-type generalized predictive control (GAPIGPC).PI-type generalized predictive control (PIGPC) use prediction technology overcome timevarying and time-delay, and use PI (PID) control technology strengthen system robustness, greatly reducing the impact of model mismatch; optimization design of parameters is an combinatorial optimization problem, and genetic algorithm is robust algorithm for complex system optimization calculation.Combinatorial parameter tuning with optimization provides an effective means.Simulation results show that this algorithm has better robustness and immunity.%针对PI型广义预测控制比例因子和积分因子参数整定困难,提出了基于遗传算法参数优化整定的PI型广义预测控制(GAPIGPC).PI型广义预测控制(PIGPC)采用预测技术克服时变时滞,用PI(PID)控制技术加强系统的鲁棒性,极大程度地减少了模型失配的影响.由于其参数的优化设计是一个组合优化问题,而遗传算法是一类可用于复杂系统优化计算的鲁棒性算法,并为组合参数整定及优化提供了一个有效的手段.仿真实例表明,该算法具有较好的鲁棒性和抗干扰性.