Research On Networked Control System And Its Simulation And Experiment

Networked control system is one of outcomes of cross development between computer network and automation technology. It is one means of control and it can be called networked automation. In networked control system, network is not only one medium through which data of control system can be transmitted but also one part of control system. Comparing with traditional control system, networked control system can decrease line of system, reduce cost of system, make structure flexiable and make system easy to extend and maintain. It also can share the resource of system. Although networked control system has many advantages, induced delay and data dropout which is caused by network can greatly affect performance of control system. Aiming at network delay and data dropout, in this paper the following work is done:(1) Due to the condition that in networked control system tracking problem is rarely studied, stochastic and time varying networked control system can be transformed one deterministic system by setting buffer. The deterministic system can become one equivalent system by using one special mathematics transformation. The optimal tracking control law of the equivalent system is derived and the existence of optimal tracking control law is discussed.(2) For the networked control system of which network delay subjects to deterministic probability distribution, stochastic optimal control law is derived through using the related knowledge of stochastic control in this paper. The existence and stabilization of control law is discussed.(3) For the simulation and experiment of networked control system, in this paper TrueTime and SimEvents is used to simulate networked control system. Data package sending and receiving program is used to conduct network experiment. And the OPC server program is used to conduct experiment. Both simulation and experiment achieve good effect.(4) For the networked control system of which the dropout process subjects to known two-point distribution, optimal filter is derived. The filter has the function of compensation to dropout.