Network Control System Based On Generalized Predictive Control And Double Smith Predictor

Entering the 21st century, with the rapid development of the Internet, the advantage of NCS is growing and showing a booming trend in the industrial production. However, Network delay exists in the network control system of forward channel and feedback channels, it would not only reduce the dynamic character of its existence, but also cause the system to be uncontrolled. When a charged object itself with time delay, it often in large delay and nonlinear and time-varying. Accurate selection and determination of the parameters of the controller have a certain degree of difficulty. So, compensation for network control system with time-delay estimation under the attention of scholars at home and abroad.Due to the Smith Predictor compensation model exactly matches the time delay of the forward channel of the NCS can be removed from the closed-loop circuits out of the closed-loop circuits, delay feedback pathways total elimination from the system without the feedback pathway network scheduling, it has been widely used in the delay compensation. However, the traditional Smith Predictive compensation required precise mathematical model. When the model does not match completely or completely mismatch, it will be difficult to have time delay compensation effect. It is difficult to obtain accurate parameters of the model. This article will address this problem research on predictive compensation of time delay in NCS.(1) Opening first the structure of NCS introduced, the composition of delay, the causes of time delay and the network delay of the compensation strategies researched by status at home and abroad. Based on these, the ideas and methods to solve the problem of time delay compensation advanced.(2) With network delay and delay of controlled object itself, for traditional Smith Predictor compensation needs to be accurate model of the controlled object, the first-order inertia filtering joined to the forward of the primary feedback channel. To resolve object with Predictor model parameter mismatch problem. And the new control strategy is introduced into the conventional PID control strategy and adaptive Fuzzy PID control. In the Matlab/Simulink platform, the mode designed. Simulation results show that the system can get better dynamic performance and adaptation.(3) Generalized predictive control algorithm in NCS introduced, because it needs Diophantine equation in solving arithmetic, and involving a large number of matrix operation. So stepwise GPC strategy and buffer queue selection mechanism introduced.(4) Controlled object is the most representative in the process of industrial control second-order inertia big lag link, Fuzzy - PID control and the improved GPC as control strategies, then based on TrueTime1.5 toolbox in Matlab/Simulink environment, for the system with network time delay and without delay, Smith estimated compensation model match exactly and completely mismatch when comparing the simulation analysis. And experimental results show that the time delay estimated compensation scheme in the system stability and controllability of effectiveness.