Design of networked control system based on wireless sensor networks

Wireless sensor networks (WSNs) are emerging as a new generation of sensor networks. Serving as the backbone of control applications, WSNs will enable an unprecedented degree of distributed and mobile control. In the paper we analyze and design a networked control system where the communication between the sensors, the controller, and the actuator takes place over WSNs. However, WSNs also introduce random delays and packets loss in the feedback loop. We present a state predictive and state estimation based on the extended Kalman filter for the networked control system. The model predictive control with both random delays and packet loss in the feedback loop is introduced. This paper is motivated by many of applications for the mobile controllers during the setup of an experimental NCS. The simulations show that our experimental system which takes the delays and loss into account indeed can perform well in the real applications.