Robust distributed model predictive consensus of discrete-time multi-agent systems: a self-triggered approach

This study investigates the consensus problem of a nonlinear discrete-time multi-agent system (MAS) under bounded additive disturbances. We propose a self-triggered robust distributed model predictive control consensus algorithm. A new cost function is constructed and MAS is coupled through this function. Based on the proposed cost function, a self-triggered mechanism is adopted to reduce the communication load. Furthermore, to overcome additive disturbances, a local minimum-maximum optimization problem under the worst-case scenario is solved iteratively by the model predictive controller of each agent. Sufficient conditions are provided to guarantee the iterative feasibility of the algorithm and the consensus of the closed-loop MAS. For each agent, we provide a concrete form of compatibility constraint and a consensus error terminal region. Numerical examples are provided to illustrate the effectiveness and correctness of the proposed algorithm.