【作者】 李鸿亮；

【导师】 苏宏业； 褚健；

【作者基本信息】 浙江大学 ， 控制理论与控制工程， 2002， 硕士

【摘要】 预测控制策略因其预测模型、滚动优化、反馈校正三大特征符合工业过程控制的实际需要，因此自提出至今在理论和应用方面取得了长足的发展。而预测函数控制是预测控制领域中最新的研究方向之一，近年来，预测函数控制的应用已从最初的快速过程，如工业机器人的手臂控制、雷达跟踪控制等发展到慢速过程，如间歇反应过程的温度跟踪控制等，而且，在国内的应用也已呈逐步发展之趋势。本文主要从理论和应用两方面对预测控制方法进行了研究，理论方面主要是从预测函数控制的基本原理出发，研究了一般情况下的基于状态空间描述的预测函数控制策略，并通过计算机仿真验证了PFC算法比常规PID算法具有更好的鲁棒性和抗干扰性：在实际工业过程的应用上，又分为两类算法及软件的应用，预测函数控制策略及APC-PFC软件的应用主要以聚乙烯氯化过程的温度控制和青霉素发酵过程的PH值控制为主，多变量预测控制算法和APC-Hiecon软件的应用主要以扬子石化公司的液化气回收装置先进控制为主。主要研究工作集中于以下几个方面： 1．在阅读大量文献的基础上，从自动控制理论的发展历程入手，对先进控制及预测控制的的基本思想和发展状况进行了综述性描述，并结合我国流程工业的特点，切入到先进控制技术对我国流程工业发展所能发挥的作用这一主题； 2．详细讨论了基于状态空间描述的预测函数控制算法，从离散系统的结构、内模控制系统的特性出发，以阶跃函数为例，针对PFC系统的稳定性和鲁棒性进行了分析。通过仿真表明预测函数控制方法是一种计算简单、鲁棒性较强、抑制干扰能力好、控制精度高的控制策略； 3．预测控制从工业现场而来，又回到工业现场去解决实际的问题。因此任何一种预测控制方法都应该紧密地与实际工业过程结合起来，应用到实际工业过程中去。预测函数控制方法是近年来兴起的一种新型预测控制算法，本文详细介绍了预测函数控制工程化软件包APC-PFC，通过该技术在流程工业中的三个应用实例，说明这种控制方法具有跟踪快速，控制精确，无超调等优点，明显优于传统的PID控制方法； 4．介绍了多变量预测控制技术及多变量预测控制软件包APC-Hiecon，通过 浙江大学硕士学位论文 该技术在流程工业中的应用实例，说明使用该技术控制多变量过程可最大 程度地保证控制系统的稳定性和支全性，优化过程工作点，有效地提高生 产率； 5．对下一步的研究工作从理论和应用两个角度作了相应的分析与探讨。更多还原

【Abstract】 The Predictive Control strategy has been developing quickly both in the theories research and application, with its three characteristics of predictive model, receding horizon control construction and feedback correction meeting the industrial process control’s needs. The Predictive Function Control is one of the most novel direction in this area, and it applies in the fast-processes initially, such as robot’s arm control and radar tracking control, now it applies also in the slow-processes, such as fitful reaction temperature control, etc. The theory and application of the Predictive Control are studied in this paper. On the theory, the Predictive Function Control method based on the state space is discussed and the simulation results validate the PFC method’s advantages on robust and anti-jamming comparing with PID method by computer simulation. The application research includes the application of the software of Predictive Function Control (APC-PFC) and the software of the Multiple-variables Predictive Control (APC-Hiecori). The former were applied in the temperature control Chlorinating process and PH control in the process of zymolysis of penicillin, the latter were applied in the advanced control of reclaim equipment of lox in China Petroleum & Chemical Corporation Yangzi Petrochemical Co., LTD. The main works are summed up as following: The basic concepts and the development of the Advanced Control and the Predictive Control are stated with the focus on the effects of the Advanced Control on the process industry in our country after the author has studied a lot of references. The Predictive Function Control is discussed in detail. Considering the step function as an example, the stability and robustness of the PFC system are analyzed through the discrete system structure and the inter-model control system characteristics. The Predictive Function Control is proved to be a simple-computation, robust, anti-jamming and high precision control method by detail simulation study. The Predictive Control comes into being from the industry production field, and it returns to the industry field to solve the practical control problems. The Predictive Function Control is one of new predictive control methods and thecorresponding software packet ofAPC-PFC is introduced. This method is provedto be much better than PID control with the advantages of fast-tracking, highprecision, non-oversheet through three applications in the industry process.The Multiple-variables Predictive Control strategy and the correspondingsoftware packet of APC-Hiecon are introduced. The results of its application inthe process control show this method can guarantee the system’s stability andsecurity furthest, optimize the operation point of process and enhance theproductivity.The future research and development of the theory and application in this areaare also presented.更多还原