The Research Of Neural Network Control And Application In Distillation Tower

Based on the rectification tower in pharmaceutical enterprises, the neural network adaptive control for nonlinear time-delay systems and application in the process of rectification tower is studied. In view of the nonlinear system with time-delay control process, the indirect adaptive control method based on cellular neural networks is considered, including adaptive identification method and neural network control method. The Lyapunuv- Krasovskii functional is used to compensate the time delay in system. Based on adaptive online learning algorithm, the weights of neural network through indirect adaptive neural network method to design a neural network identification algorithm, the controller was designed on the basis of the neural network identification model, and algorithms is proved with Lyapunov stability theory and the convergence of the controller, through the convergence is proved that the dynamic system is uniformly asymptotically bounded stability, through Matlab simulation experiments verify the feasibility of the algorithm. Finally for rectifying column system, design of PLC control system structures, DCS control platform, the control process of system implementation, to complete the rectification tower system control process.