Can Bus-based Network Control Study

As the product of interdiscipline, the networked control system with itsgood openness, flexibility and reliability has been applied widely in theindustrial field. The introduction of the network for the control system brings notonly the convenience but also the time-delay, package lost, informationdisorderly sequence. These problems will be great influence on the controlproperty, even can cause the system instable.Based on the analysis of several main influence factors of the networkedcontrol system performance and the present research of the network controlsystem, the networked control system is studied from three angles of delaycompensation, integrated modeling, improved scheduling algorithm. Becausethe full state information of controlled plant can't be got and the randomtime-delays exist in networked control systems,the predictive controller isdesigned based on grey forecasting,which can compensate the delay ofnetworked control system. For a class of the networked control systems withuncertain parameters, considering package lost as a special kind of long timedelay process and the existence of the pulse dynamic behavior, the modelingmethod of the network control system based on discrete switching of the pulse isproposed, which has regarded synthetically time-delay and packet loss problems,and will use Lyapunov function and linear matrix inequality analysis method.The sufficient condition of the system asymptotic stability is given, and therobust controller is designed for each modal switch. For the influence of thescheduling problem on the performance in the networked control system basedon CAN bus, the existing research directions of scheduling algorithm innetworked control system are summarized and an improved MTS schedulingalgorithm is proposed because of the defects of the traditional MTS schedulingalgorithm.The networked control system simulation model is set up using TRUETIMEtoolbox in the MATLAB environment and each scheme suggested in this paperis simulated and proved. Then the experimental results show the effectiveness of the proposed scheme.