Distributed data traffic scheduling with awareness of dynamics state in cyber physical systems for distributed voltage control in smart micro-grid

Abstract

The design of communication infrastructure in cyber physical system (CPS) is of vital importance for proper operations of physical dynamics. In this paper, the scheduling of packet transmission over wireless communication networks in CPS is studied for the purpose of system state stabilization. To measure the effective information amount of different packets, the concept of virtual queue is defined based on the Delay-tolerant Kalman filtering. Virtual queue based back pressure scheduling algorithm is proposed for the distributed data traffic scheduling. The impact of packet losses on virtual queue lengths is also studied for the formulation of distributed scheduling strategy. The scheduling method is then applied in the context of voltage control in micro-grid and is demonstrated to achieve better performance than general scheduling algorithm without the awareness of dynamics state. Further numerical simulations of data traffic scheduling considering packet losses are carried out at networked system with data packet dropout governed by a Markov process to demonstrate algorithm validity.