【作者】 张立群；

【导师】 唐多元；

【作者基本信息】 清华大学 ， 动力工程及工程热物理， 2004， 硕士

【摘要】 PID控制器结构简单、鲁棒性好,在实际热工过程控制中广泛采用PID控制器。过去的四十年里,出现了大量的PID控制器的设计方法,但是以往的设计方法都是基于简化后的一阶加纯延时模型提出的。基于一阶加纯延迟对象的整定公式不一定适合热工过程中的高阶对象,因此获得的控制效果常常不理想。从某种意义上,设计控制效果优良的PID控制器的难点就是准确地获得PID控制器参数的稳定域。本文利用广义Hermite-Biehler定理,提出在PID控制器参数稳定域内设计高阶对象的最优PID控制器,与此同时利用Kharitonov定理、Monte-Carlo随机实验方法和H∞混合灵敏度约束研究PID控制器的鲁棒稳定性和性能鲁棒性。首先,根据广义Hermite-Biehler定理分析了高阶对象的PID控制器参数稳定域的计算方法以及编写出通用算法程序。利用遗传寻优算法,在PID控制器参数的稳定域约束下,设计ITAE指标最优的PID控制器。同时利用Kharitonov定理讨论了该区间对象最优PID控制器的鲁棒稳定性。假设对象参数在区间范围内摄动,利用Monte-Carlo随机实验方法分析了系统过渡过程的性能鲁棒性。然后,基于上述最优PID控制器的设计方法,将其应用于过程控制中普遍存在的对象:高阶对象、非最小相位对象、不稳定对象和大延迟、大滞后对象。针对各类对象,分别讨论了基于稳定域最优PID控制器的设计及进一步改进,并进行了大量仿真实验。接着,将该方法分别应用于典型热工控制系统:锅炉再热蒸汽温度控制系统、循环流化床锅炉床压控制系统、锅炉过热蒸汽串级控制系统和电厂主汽压力串级控制系统。最后,本文探讨了基于H∞混合灵敏度约束下的最优PID控制器的设计方法,研究了满足H∞混合灵敏度约束条件的PID控制器参数域的计算方法,并设计出H∞混合灵敏度范数指标最优下的PID控制器。研究结果表明,基于稳定域最优PID控制器的设计具有良好的控制品质和鲁棒性。更多还原

【Abstract】 The PID controllers have simple structure and good robustness, and have been used by far most in the thermal process control field. In the last 40 years, a large amount of PID controller design methods have appeared, but those methods were almost developed based on First-order time-delay models. First-order time-delay models do not always suit high order plants of thermal process, so traditional PID controllers cannot provide satisfactory control. A major obstacle to designing the best PID controller, in any sense whatsoever, has been the difficulty in characterizing the entire set of stabilizing PID controllers. In this thesis, an optimized PID controller is designed of high order system, using the generalized Hermite-Biehler theorem, in stable region of PID controllers. Moreover the robust stability and performance of the PID controllers were evaluated using the Kharitonov theorem, Monte-Carlo stochastic methods and H∞ mixed sensitivity restriction.First, the stable region algorithm of PID controller for high-order plants has been studied, and a common program has been implemented, using generalized Hermite-Biehler theorem. The genetic algorithm was used to tune a PID controller with the optimal integrated time absolute error (ITAE) criterion from this stable region. The robustness stability of the optimal PID controller of interval plants are studied using Kharitonov theorem. When the parameter of plants varied in the interval range, the dynamic performance and robustness of PID control system are analyzed by Monte-Carlo stochastic methods. Second, based on above optimal design methods of PID controller, the stable region method is used to control the common plants in process control field: high order plant, non-minimum phase plant, unstable plant and large inertia plant. The design methods and further improvement of optimal PID controller for these plants are investigated respectively and a lot of simulation experiments are performed. Furthermore, this method is used in typical control systems of thermal process: the reheated temperature control system for boiler, the <WP=5>pressure control system of the circulating fluidized-bed boilers, the superheated temperature cascade control system for boiler, the main steam pressure cascade control system in generating set. Finally, a PID controller design method based on H∞ mixed sensitivity restriction is researched. The algorithm for parameter’s region of PID controller, which satisfies H∞ mixed sensitivity restriction, has been studied, and a PID controller with the optimal H∞ mixed sensitivity norm criterion has been designed. All this work has demonstrated that the proposed method can gain effectively control in thermal process control.更多还原