Intelligent Scheduling Strategy with Bandwidth Feedback in Networked Control Systems

This paper proposes an intelligent scheduling strat-egy with bandwidth feedback. A network scheduler is designed to adjust the sampling period of each networked control loop based on its real-time control error. At a current scheduling period, the increment of the bandwidth feedback occupancy ratio is estimated by the dynamic control error of each networked control loop, and then a new bandwidth occupancy ratio within the next scheduling period is assigned to each loop based on the bandwidth constraint condition. After that, the sampling period is dynamically changed to adapt the quality of control and quality of service. To analyze the system stability, this variable sampling period networked control system is modeled as a dynamic interval discrete system with uncertain parameters. Its asymptotic stability is proved by the Lyapunov function and linear matrix inequality. Experimental results show that the scheduling strategy has the bandwidth feedback advantage, and the variable sampling period has the successive approximation ability for the time-varying reference input networked control system.