Design And Analysis Of Communication Logic For Networked Control Systems

Networked Control Systems(NCS) belongs to distributed control systems which the sensors, controllers and actuators are connected by a shared communication network. This type of feedback control systems where the control loops are closed through a network is more efficient, flexible, and reliable with less cost. However, network also brings some new problems, such as network delay, dropouts, bandwidth contraints, etc. The remote controller can't get enough information when there are many nodes on the network while the bandwidth is limited, and that will result degradation of system performance, even leads to instability. Dynamicly allocating the bandwidth usage and reducing the amount of the transmitted data is effective to solve the problem of bandwidth constraints, but no enough attention has been paid.Communication Logic is an effective key method for the bandwidth constraints problem in Networked Control Systems. First we introduce the principle idea of communication logic and summarize the recent research paper on it. The work is focused on linear time invariant processes and a new communication logic, which is built on a State Dependent Poisson Process, is proposed and combined into a new type of smart sensor, to determine when the sensor should transmite data to remote controllor. In order to analyze the influence of Communication Logic as well as the whole system, we set up the model of the states-feedback networked control systems with data dropouts described through Bounulli Process, and study the mean square asymptotically stability of systems by using the stochastic theory about Poisson Process with time varying density. Sufficient conditions are derived through mathematic deduction. The stability for output-feedback control systems with Communication Logic is also discussed later. Finally, theoretical results of NCS with the new communication logic are validated by Matlab/Simulink tools. The stochastic communication logic based on states of systems is compared with periodic commnunication logic through simulation and the result shows that the former can achieve better system performance.It turns out that the communication logic, which is based on State Dependent Poisson Process, can determine the appropriate time for sensors to send the data and reduce the usage of network bandwidth, meanwhile assure the stability and performence for Networked Control Systemes.