Moving horizon estimation based distributed model predictive control of networked systems with communication delays

This work discusses about output feedback cooperative distributed model predictive control of networked systems in the present of disturbances and time-varying communication delays. The networked system is composed of several interconnected subsystems coupled via states. The subsystems share their control input sequences through a delayed communication network. A buffer based distributed moving horizon estimation strategy is designed that estimates the local states of each subsystem and those states which are coupled between subsystems and their actual values are not accessible due to communication delays. The boundedness of the estimation error and robust exponential stability of the closed loop system are established. The theoretical results are verified by applying the proposed method on two interconnected continuous stirred tank reactors.