Control Strategy Research Based On Model-free Adaptive Control Theory In Network Control System

Network Control System is a kind of new distributed control system combinednetwork communication technology and control system together. Compared with thepoint to point traditional control system, it has higher scalability, more convenience andhigh economic efficiency. However, the introduction of network technology causessome new problems to NCS. If network control system wants to run efficiently, there aretwo problems need to be solve. First is reducing the bad influence from network, and thesecond is optimizing the control method in NCS. Therefore, it has important theoreticaland practical significance to study NCS control strategy. According to the bad influencefrom network environment and the shortcomings of existing control methods, someresearch were done in this paper, specific contents are as follows:(1) A reasonable and practical network control system simulation platform is built byusing Truetime toolkit, this platform supports control strategy study for NCS andprovides two interface which can study network key factors and network controllerfor NCS. It provides a flexible, efficient and reliable support to analysis NCS networkfactors influence and control algorithm performance.(2) Research the NCS control strategies from the perspective of the networkenvironment factors, mainly study the packet loss rate, the impact of network delay,the sampling period, the data transfer rate these factors on the NCS network, andanalysis the relationship between these factors and control system stability, furtherverify the impact of these factors on the NCS in the simulation platform by multiplesimulations.(3) Further study the control method of NCS solving the hard modeling problem andtime-varying intensity problem, proposing an optimized control algorithm based control strategy, works are as follows:The model-free adaptive control algorithm (MFAC, Model Free AdaptiveControl) is introduced into the network control system to avoid modeling problem.Proposing PID-MFAC control algorithm which can reduce the steady-state erroreffectively.Using multi-innovation theory to optimize pseudo-gradient estimation andcontrol law in MFAC, Optimizing model free adaptive control algorithm based onmulti-innovation theory which has a rapid convergence, and using stochasticapproximation theory to prove the convergence. Finally, the simulation results showthe effectiveness of the algorithm.According to error feedback principle, proposing model-free adaptive controlalgorithm with error change rate. The algorithm can track system error immediately,and enhance the system’s interference capability. Finally, simulate these proposedalgorithms on Truetime to verify the effectiveness of the proposed algorithm.