论文部分内容阅读
基于概率鲁棒方法,针对具有实参数不确定性的多输入多输出被控对象,提出一种分散PID控制器设计方法。根据被控对象模型的参数摄动状态,计算闭环系统满足性能设计要求的概率作为优化算法的目标函数,利用遗传算法对分散PID控制器参数进行优化,用Monte Carlo实验对控制系统进行鲁棒性检验。对5个多变量化工过程进行了仿真试验,并与基于标称参数的设计方法进行比较。仿真结果表明,基于概率鲁棒的分散PID控制器设计方法对模型参数不确定性具有较好的鲁棒性,在被控对象存在一定的不确定性时,系统能以最大的概率满足设计要求。
Based on the probabilistic robust method, a decentralized PID controller design method is proposed for multi-input multi-output controlled objects with real parameter uncertainties. According to the parameter perturbation state of the controlled object model, the probability that the closed-loop system satisfies the performance design requirement is calculated as the objective function of the optimization algorithm. The parameters of the decentralized PID controller are optimized by genetic algorithm. The Monte Carlo experiment is used to control the robustness of the control system test. Five multivariate chemical processes were simulated and compared with design methods based on nominal parameters. The simulation results show that the decentralized PID controller design method based on probabilistic robustness has good robustness to model parameter uncertainty. When the controlled object has some uncertainties, the system can meet the design requirements with the maximum probability .