Modeling And Control Of Wireless Networked Control System

Wireless Networked control systems (WNCSs) are control systems in which the closed con-trol loop is connected by wireless communication networks. WNCSs are the development and in-tegration of computer science, wireless communication theory and control technology. Compared with traditional control systems, which are connected by point-to-point cables, the WNCSs have the advantages of low cost, easy maintenance and so on. However, some inevitable disadvantages arisen with the use of wireless communication network, such as network-induced delays, packet dropouts and node competition. It is well known that network-induced delays and packet dropouts are the source of instability, and can degrade the control performance. Consequently, considering the characteristic of WNCSs, it has become a hot issue that how to solve these problems in the WNCSs by using the existing control theory.The main purpose of our research is to study the influence of network-induced delays, multi-packets transmission policy and packet dropouts on the modeling, analysis and stability of WNCSs. Thus, in our research, a kind of ZigBee-based WNCS is used, and the research results are sum-marised as follows:(1) Based on the protocols of ZigBee network, the components of network-induced delay are introduced and the main reasons causing packet dropouts are discussed. Considering the system with both single hop transmission channel and multi-hops transmission channels, the distribution of network-induced delay and the probability of packet dropout are analyzed, respectively.(2) Considering the characteristic of ZigBee technology, a novel stochastic uncertain switch-ing model is provided to describe the WNCSs with network-induced delays and packet dropouts. In such model, the network-induced delays can be longer or less than one sampling period. This mod-el is suitable for both single-hop and multi-hops WNCSs, and the uncertainties are satisfying norm bounded condition. By using multi-Lyapunov functional, under arbitrary switching situation, a sufficient condition is presented for the exponentially mean square stability of the closed-loop WNCS, and the corresponding controller is designed. (3) For WNCSs with both sensor-to-controller and controller-to-actuator channels, a stochas-tic model to describe packet dropout process is given. The modeling approach for WNCSs with consecutive packet dropouts and time-varying network-induced delays are discussed, the closed-loop system is modeled as a kind of stochastic switching system. Based on the characteristic of packet dropouts, such as the probability of packet dropout and the number of consecutive packet dropouts, a novel Lyapunov functional is provided, and a packet dropout probability dependent condition to ensure the whole system to be mean square stable is obtained.(4) In ZigBee-based WNCSs, if a packet exceeding the maximum allowed packet size, the multiple packets transmission policy needs to be adopted. Thus, a novel stochastic model is used to describe the multi channels WNCS subject to packet dropouts and network-induced delays under multi packets transmission policy, furthermore, a sufficient condition is given to ensure the closed-loop WNCS to be exponentially mean square stable.(5) In practical applications of ZigBee-based WNCS, due to the physical limitation on actu-ators, the saturation on actuator is inevitable. For ZigBee-based WNCSs with actuator saturation, the saturation nonlinearity is considered as the nonlinearity satisfying sector bounded condition, and the whole system is modeled as nonlinear stochastic system. Based on the characteristic of packet dropout and the sector bounded condition, a novel Lyapunov functional is established, and the corresponding state feedback controller is designed, furthermore, the stabilization problem is discussed.(6) The controller design method of ZigBee-based WNCS is used in networked bilateral tele-operation systems. The information exchange between the master and slave manipulators via in-dependent communication channels. For such systems, the mismatched-model terms are treated as disturbance, then, the model of closed-loop error dynamics are derived. Furthermore, H∞con-trollers are designed for both master and slave manipulators to ensure the exponentially mean square stability of the whole systems with a given disturbance attenuation levelγ.