Network Control System With Random Delay

With the increasing complexity of control system, to meet the needs of a large number of devices communicating, mass data transmission and the development of remote control, the network control began to emerge and has become one of today's hot spot in control system research. The introduction of the network can be a good solution to the problem which the traditional point to point control can not solve.In the network environment, the communication between devices occupies the same physical medium, which will inevitably bring about the collision during information transmission and the network time delay. Excessive delay even causes the network system instability, so the research is of great practical significance.First, the impact on network performance by the network delay and the network nodes drive method is analyzed. Based on the analysis of network protocols and communication patterns, the time analysis model of network system is created, and the minimum and maximum time delay expressions generated by the data transmission is given. By analyzing the network system in the recycling process, the minimum and maximum system response time is obtained.Second, for the specificity of the network, the MATLAB/TrueTime simulation toolbox is chosen to simulate the network module and the network nodes, and to build a network simulation model in this paper. In the aspect of controller design, the Smith predictor compensator based on the time stamp is established, of which the effectiveness is verified by simulation. As the performance of compensation depends on accuracy of the recognition model excessively, fuzzy control method based on identification model gain is proposed to improve the robustness of the compensator. The results show it plays a very good correction of the identification model, and the accuracy of Smith compensator is improved. The simulation results show the good performance of the improved compensator.Again, physical platform is designed and built based on the Profibus-DP network, including the choice of control equipment and the hardware configuration, the plant selection and parameter design, and the design of monitoring system. Then the random time delay compensation algorithm is implemented by STEP7.Finally, on the basis of the successful identification of the controlled object and getting random time delay, the program code is written to achieve the control algorithm. Effectiveness of the algorithm is proved in the real system.