摘要
针对多变量系统控制中的耦合问题,提出了一种基于扩张状态观测器(ESO)的动态解耦方法。该方法将系统输入变量间的耦合作用、被控对象参数时变和外界干扰视为一个总的扰动,用ESO估计该总扰动并反馈到控制器进行补偿,从而实现动态解耦;对解耦后的每个子系统,分别设计出了基于误差最小二乘指标的神经元自适应PID(NAPID)控制器。该方法简化了解耦过程,放松了对系统模型的要求,计算量小、鲁棒性强。最后用该法对蒸馏塔进行控制仿真,仿真时使用混沌优化方法对ESO的参数进行了离线优化,并给出了与模糊PID解耦控制方法对比的仿真结果,结果证明本方法是可行的。
For the problem of coupling in MIMO system, this paper proposed a dynamic decoupling control based on ESO. To each of the subsystems of a MIMO system, external disturbances, parameters time-varying and disturbances from other subsystem were all looked as a total disturbance. ESO could estimate the total disturbance,and fed back the estimated value to controller to compensate the disturbance, and achieved decoupling. Then, for each deeoupled SISO system, schemed out a neuron adaptive PID （NAPID） controller based on error least square. By this method, simplified the decoupling, reduced the requirement of the plant model. The novel method could also achieve less calculation and better robustness. Finally, placed the method in distillation column model control. In simulation, got the parameters of ESO by chaos optimization method and gave the simulation result. The result shows that the method is available.
出处
《计算机应用研究》
CSCD
北大核心
2009年第11期4176-4178,共3页
Application Research of Computers
基金
国家"863"计划资助项目(2007AA01Z170)
关键词
多变量系统
扩张状态观测器
动态解耦控制
神经元自适应
混沌优化
multivariable system
ESO （ extended state observer）
dynamic decoupling control
neuron adaptive
chaos optimization
