A Novel Feedback Scheduling Approach for Resource-Constrained Network Control System

In networked control system (NCS), increasing network workload would make network resource scare and yield unaccepted QoC of subsystems. Deadband scheduling is proved a better scheduling method than traditional static schedule to solve the problem by reducing network traffic to improve overall system´s QoC. However, fixed deadband range in primal deadband scheduling couldn´t be dynamically adjusted with respect to network utilization and control performance variation of each subtask In the context of control and network scheduling co-design, model based deadband feedback scheduling approach (MBDFS) proposed in this paper is applied to enhance QoC of the overall system under overloaded condition. In this approach, a task that deadband feedback control should be carried on is dynamically selected and the deadband range is adjusted based on a specified mathematic model which takes into account dynamic changes of subtask´s own QoC as well as network utilization. Simulation results indicate the scheduling approaches proposed can effectively improve overall system´s QoC in the circumstance of narrow network bandwidth and overload network condition.