Distributed Multi-Step Model Predictive Control for Consensus

TL;DR

This paper develops a distributed multi-step MPC approach for multi-agent consensus under constraints, introducing a lexicographic tie-breaking rule to ensure strict convergence conditions without shrinking feasible sets.

Contribution

It proposes a novel secondary criterion in distributed MPC that guarantees strict consensus convergence by maximizing interiority measures, addressing a key geometric challenge.

Findings

Guarantees asymptotic consensus when interior feasible terminal states exist.

Provides explicit horizon conditions for single- and double-integrator agents.

Demonstrates monotone diameter decay in numerical simulations.

Abstract

This paper studies consensus of discrete-time multi-agent systems under time-varying directed communication, state and input constraints using a distributed multi-step model predictive control (MPC) framework. Consensus is recast as stabilization of the agreement set, and a geometric viewpoint based on convex-hull invariance and strict interiority is adopted. Building on an existing geometric necessary and sufficient condition for agreement, we show that enforcing terminal inclusion in local neighbor convex hulls guarantees hull invariance but does not, in general, imply the strict relative-interior property required for convergence. An explicit counterexample demonstrates that strictness cannot be deduced from feasibility and contraction constraints alone. To resolve this issue without shrinking feasible sets or altering primary performance objectives, a lexicographic tie-breaking mechanism is introduced. Among optimal (or near-optimal) MPC solutions, the proposed secondary criterion selects trajectories maximizing an interiority measure with respect to the neighbor hull. It is shown that whenever an interior feasible terminal state exists, this selection rule enforces the strictness condition required for asymptotic consensus. Explicit horizon conditions are derived for single- and double-integrator agents with bounded inputs, ensuring feasibility and automatic existence of interior feasible terminal points. The resulting scheme provides a distributed and implementable route to consensus via finite-step set-Lyapunov contraction. Numerical simulations with distributed inter-process communication illustrate monotone diameter decay and report per-agent computational complexity.