Improving the quality of service in network-based control systems

The use of data networks in control loops has received much attention recently due to its flexibility and economical advantages. In addition, mutual network usage has raised new challenges such as delay and data loss. This paper aims to reduce undesired effects of network by reducing the required traffic of the network. An estimation framework for network control system is introduced, in which estimations of local Kalman filter is sent to remote estimator based on the logic decided by a novel fuzzy communication logic. In order to do so, there exist two estimators, a remote estimator which estimates the states of the plant and its local copy that gives the same output. The output of the local estimator is compared with the real states of the system, if the states of the system are estimated with small error, there is no need to send data, hence, the probability of sending data is decreased using a fuzzy decision system. In order to optimize this fuzzy system, a particle swarm optimization (PSO) algorithm is used. The proposed method is applied to control a pair of overhead crane systems with non-linear dynamics. Since the two overhead cranes need to work synchronously and their synchronization is performed over a network, the control of this system lies within the scope of the proposed controller. Simulation results show that the communication load is reduced and the purposed fuzzy communication logic is able to control the non-linear dynamical systems over a network with a sufficient performance.