节点文献

时滞过程的闭环辨识与先进控制策略的仿真与应用研究

Simulation and Application Study on Closed-loop Identification and Advanced Control Strategy for Time-delay Process

【作者】 肖术骏

【导师】 朱学峰;

【作者基本信息】 华南理工大学 , 控制理论与控制工程, 2010, 硕士

【摘要】 时滞现象普遍存在于工业生产过程中,本文所研究的两个对象,治疗膀胱癌的热疗仪实验装置和水厂加药絮凝过程都是具有明显纯滞后的过程,尤其是水厂的加药絮凝过程,其纯滞后时间与时间常数之比甚至超过了1。时滞的存在会对控制系统产生不利的影响,特别是对于大时滞过程,常规的PID控制器很难取得理想的控制效果,甚至会导致系统的不稳定。因此,有必要针对大时滞过程设计先进的控制策略,而了解过程的对象模型对于控制策略的设计是有极大帮助的。本文首先提出了一种新的基于PI控制的闭环辨识方法,该方法是假定对象为二阶加时滞(SOPTD)过程,用PI控制器使得闭环响应呈欠阻尼状态,根据欠阻尼的闭环响应曲线,首先得到闭环传递函数,然后通过等效法和比较系数法将PI控制器从闭环传递函数中分离出来,反推出对象的模型。将此辨识方法用于热疗仪实验装置中,取得较好的辨识效果。然后针对水厂加药絮凝过程,设计了基于内模控制的PID控制器(IMC_PID),比较了IMC_PID控制器与Lopez法和Z-N法整定的PID控制器的控制效果,仿真结果表明IMC_PID具有结构简单,利于在线整定的优点。之后在IMC_PID的基础上进行了二自由度IMC_PID的仿真研究,表明了二自由度IMC_PID能够较好的兼顾快速性和鲁棒性。接下来采用改进的差分进化算法,以一种能抗超调的MITAE指标为优化目标函数,进行PID参数的优化,结果表明MITAE指标达到了预期的目标。本文最后采用了美国通控集团博软分部的一款基于无模型自适应控制技术的控制器(MFA)对水厂的加药絮凝过程进行了前馈-反馈控制系统的设计,并在佛山某水厂的中试装置中得到应用,取得了良好的控制效果。

【Abstract】 Lag is commonly found in industrial processes. This paper studies two objects, the hyperthermia instrument experimental device used for the treatment of bladder cancer and the dosing flocculation process of the water plant are both processes with large time-delay. Especially for the dosing flocculation process, the ratio of the pure delay-time and the time constant is even more than one. The existence of time-delay would have a negative impact on the control system. Especially for large time-delay process, the conventional PID controller is difficult to achieve the desired response, or even badly lead to instability. Therefore, it is necessary for the large time-delay process design of advanced control technologies. However the advanced control technologies are based on the process model as the basis for design.First, this paper presents a new closed-loop identification based on PI control method, which assumes that the process for the second-order plus time delay (SOPTD) model. It makes the closed-loop response was underdamped with the PI controller, then the PI closed-loop transfer function with SOPDT model can be approximatively obtained through the closed-loop underdamped step response. Then the open-loop transfer function of the process can be obtained by the equivalent transforming of block diagram to separate the proportional-integral controller from the closed-loop transfer function. At last, the SOPDT model of the process can be attained by coefficient of comparison. This identification method is used in the hyperthermia instrument experimental device to obtain a better identification results.Then, a PID controller based on the internal model control(IMC_PID) is designed for the dosing flocculation process. And it compares the IMC_PID controller with controller tuning by Lopez and Z-N law. Simulation results show that the IMC_PID controller has the advantages of simple structure and easily tuning on-line. And then the simulation studies of a two degrees of freedom IMC_PID show that the two degrees of freedom IMC_PID can be better able to balance the speediness and robustness.Next, it uses an improved differential evolution algorithm to optimize the PID parameter with the objective function of an anti-overshoot MITAE index. The results prove that MITAE index achieve the desired target.At last, it designs a feedforward-feedback control system based on the model free adaptive(MFA) controller of the Cybosoft. The system has been applied in the dosing flocculation process of a pilot plant in a water plant in Foshan. And it achieved a good control effect.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络