Design Of The Networked Control Systems And Research On Predictive Control

Networked control system is a closed loop control system composed of a controller, sensors, actuators, etc, in which the object can be controlled from a long distance and the data communicates via a network channel among controller, sensor and actuator. The networked control system has a wide range of application prospects in other fields such as industrial control. Due to the use of network, the design of the network control system is more complex than traditional control systems. How to connect the controlled object to the network and the delay, packet loss which network brings and other issues are the subject to be solved. Based on this, a predictive control algorithm is designed in this thesis to fix the delay impact caused by the network on the system and to improve control performance.Firstly, a DC motor speed control system based on the internet is designed in this thesis, and the system consists of the server, controller and controlled object subsystem. The server is responsible for connecting the controller to a network, managing users and other functions. The controller as a user can be connected to the networked control system via server and develops their own control algorithms loaded into the system. The development of the controlled object subsystem is based on embedded system, using ARM9 as the controller and the motor as the controlled object. ARM9 is responsible for completing the network communication, AD conversion and other functions to control the motor.In-depth analysis and research on the delay existing in the network control system is done in this thesis, and a predictive control algorithm is used to compensate for the network delay. In this thesis, dynamic matrix control and generalized predictive control are designed based on the network and simulations are made by using Matlab software to prove their implementation and effectiveness. Finally, algorithm procedures can be written and DLL files are generated and loaded into the network control system, so the motor can be controlled. Experimental results show that predictive control algorithm designed in this thesis can effectively eliminate the impact of network delay and brings good control effect.