A cooperative game approach to synthesizing a sequential distributed model predictive controller

This paper presents a distributed model predictive control algorithm for a discrete linear time-invariant system that involves
a cooperative game approach. It is assumed that this system consists of several subsystems which are connected through
inputs. The subsystems are updated one by one in a sequential manner. Once a subsystem is completed its optimization, its
optimal input is shared with the next interconnected subsystem. The process of synthesizing the optimal control for each
subsystem involves the computation of disagreement points for each subsystem by employing a feasible cooperation approach.
Meanwhile, in the negotiation process, the subsystem can validate whether or not it should cooperate at each time step. The
feasibility of the proposed approach at each iteration and time step is proven. Furthermore, the stability of the proposed
method is investigated by using the convexity of the cost function. Finally, the proposed method is applied to a canal irrigation
system.