TFS: Towards Fuzzy Rule-Based Feedback Scheduling in Networked Control Systems

Abstract

With the pervasive applications of networked control systems (NCSs), it is increasingly important to design the network scheduling algorithm to ensure good overall system performance. Traditional static scheduling algorithms may work, but they require a-priori knowledge of the workload for processors and the network. As a result, using static scheduling may degrade the quality of control (QoC) of the control systems in the NCSs with limited bandwidth and dynamic workload. To achieve the desired overall performance of NCSs even when the available network resource for the control applications is scarce and unpredictable, we propose a dynamic time division multiple access (TDMA)-based fuzzy-controlled scheduling (TFS) scheme in this paper. Our approach uses the QoC of the control loops as continuous feedback for dynamic resource allocation. As a novel approach for network scheduling and control application co-design, the proposed scheme uses a condition-based cyclic state-switching framework to balance the network resource allocation among all the control loops. The scheme uses a fuzzy controller to provide the QoC guarantees by maintaining the overall system performance at a desired level. We investigate a simplified servo target tracking system as our case study to show the effectiveness of the proposed TFS scheme for achieving a desired level of the system performance. The simulation results confirm the advantages of TFS in comparison with the typical static and feedback scheduling algorithms.