Multi-agent reinforcement learning via distributed MPC as a function approximator

Reference:
S. Mallick, F. Airaldi, A. Dabiri, and B. De Schutter, "Multi-agent reinforcement learning via distributed MPC as a function approximator," Automatica, vol. 167, p. 111803, Sept. 2024.

Abstract:
This paper presents a novel approach to multi-agent reinforcement learning (RL) for linear systems with convex polytopic constraints. Existing work on RL has demonstrated the use of model predictive control (MPC) as a function approximator for the policy and value functions. The current paper is the first work to extend this idea to the multi-agent setting. We propose the use of a distributed MPC scheme as a function approximator, with a structure allowing for distributed learning and deployment. We then show that Q-learning updates can be performed distributively without introducing nonstationarity, by reconstructing a centralized learning update. The effectiveness of the approach is demonstrated on a numerical example.