基于PSO神经网络的解耦控制及其在精馏塔上的应用研究

发布时间：2018-02-24 02:40

本文关键词： 精馏塔解耦控制 BP神经网络粒子群 RBF神经网络　出处：《浙江理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：精馏塔是石油化工行业中常用的化工设备，主要用于多种混合石化产品的分离过程。作为一种典型的多变量耦合系统，其控制性能的好坏将直接影响精馏生产过程的能耗及产品质量。因此，研究有效的解耦控制方法，研制高精度精馏塔控制系统，一直以来都得到了化工过程控制领域的高度关注。本文在分析精馏塔耦合特性的基础上，研究了基于神经网络PID的解耦控制方法，研制了基于PLC的精馏塔解耦控制系统，并实验验证了方法的有效性。完成的主要工作如下： (1)在综述国内外研究现状的基础上，介绍了精馏塔生产工艺与控制要求，分析了所具有的非线性、大时滞、多变量、强耦合等特点，提出了系统的总体控制方案。 (2)针对精馏塔塔底与塔顶温度耦合严重以及传统解耦控制方法的不足，提出了一种基于混沌粒子群算法的神经网络PID控制方法。用混沌粒子群算法来替代神经网络PID原先的反向传递学习算法，调节PIDNN各个神经元之间的权值，以达到快速解耦的控制效果。仿真结果表明，所提出的方法与原有的BP算法相比具有更加优秀的动态和稳态性能。 (3)进一步分析了精馏塔所具有的非线性、大惯性、强耦合等特性，提出一种基于动态RBF神经网络的单神经元PID解耦控制算法。构建动态RBF神经网络用于辨识耦合系统模型，将辨识所得到的Jcobian矩阵信息用于单神经元PID控制器参数在线整定，从而完成对精馏塔系统的解耦控制。仿真结果表明，所提出的算法与传统的基于RBF神经网络的PID解耦控制相比，控制精度提高，鲁棒性增强。 (4)以实验室乙酸乙酯生产线精馏塔设备为对象，以西门子s7-300PLC为下位机控制器，以北京亚控科技有限公司的组态王软件(6.53)为上位机监控平台，研制了精馏塔智能解耦控制系统。完成了硬件系统控制柜的设计及调试，制作了上位机组态界面，，编写了基于step7软件的解耦控制算法。进行了精馏塔温度控制的实验研究，实际运行结果表明，本文提出的解耦控制方法具有动态性能好、控制精度高、鲁棒性强等特点，明显提高了精馏塔解耦控制系统的温度控制精度，具有较高的实用价值。
[Abstract]:Distillation column is a chemical equipment commonly used in petrochemical industry. It is mainly used in the separation process of mixed petrochemical products. Its control performance will directly affect the energy consumption and product quality of distillation production process. Therefore, an effective decoupling control method is studied to develop the control system of high precision distillation column. In this paper, the decoupling control method based on neural network PID is studied, and a distillation tower decoupling control system based on PLC is developed. The effectiveness of the method is verified by experiments. The main work accomplished is as follows:. 1) on the basis of summarizing the present research situation at home and abroad, this paper introduces the production process and control requirements of distillation column, analyzes the characteristics of nonlinearity, large time delay, multivariable and strong coupling, and puts forward the overall control scheme of the system. 2) aiming at the serious temperature coupling between bottom and top of distillation column and the deficiency of traditional decoupling control method, A neural network PID control method based on chaotic particle swarm optimization (PSO) is proposed, which replaces the original reverse transfer learning algorithm of neural network PID and adjusts the weights of each neuron in PIDNN. The simulation results show that the proposed method has better dynamic and steady-state performance than the original BP algorithm. In this paper, the nonlinear, large inertia and strong coupling characteristics of distillation column are further analyzed, and a single neuron PID decoupling control algorithm based on dynamic RBF neural network is proposed. The dynamic RBF neural network is used to identify the coupled system model. The Jcobian matrix information obtained by identification is used to set the parameters of single neuron PID controller online, and the decoupling control of distillation column system is completed. The simulation results show that, Compared with the traditional PID decoupling control based on RBF neural network, the proposed algorithm improves the control precision and robustness. Taking the distillation tower equipment of ethyl acetate production line in laboratory as the object, Siemens S7-300 PLC as the lower computer controller, and the Kingview software of Beijing Asia Control Technology Co., Ltd. The intelligent decoupling control system of distillation column is developed, the design and debugging of the hardware control cabinet is completed, the configuration interface of the upper computer is made, the decoupling control algorithm based on step7 software is compiled, and the experimental research on the temperature control of the distillation column is carried out. The actual operation results show that the decoupling control method presented in this paper has the advantages of good dynamic performance, high control precision and strong robustness, and it obviously improves the temperature control accuracy of the decoupling control system of the distillation tower, and has a higher practical value.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TE962;TP273

【参考文献】