An integral Nash equilibrium control scheme for a class of multi-agent linear systems

TL;DR

This paper introduces an integral Nash equilibrium seeking control law for multi-agent linear systems, achieving convergence to Nash equilibria with less perturbation, improving efficiency over existing methods.

Contribution

It develops a novel integral control scheme that guarantees convergence to Nash equilibria in multi-agent linear systems without requiring parameter estimation, unlike prior extremum seeking algorithms.

Findings

Convergence to Nash equilibrium in full-information case.

Convergence to neighborhood of Nash equilibrium with limited information.

Reduced perturbation frequencies and amplitudes needed for similar convergence speed.

Abstract

We propose an integral Nash equilibrium seeking control (I-NESC) law which steers the multi-agent system composed of a special class of linear agents to the neighborhood of the Nash equilibrium in noncooperative strongly monotone games. First, we prove that there exist parameters of the integral controller such that the system converges to the Nash equilibrium in the full-information case, in other words, without the parameter estimation scheme used in extremum seeking algorithms. Then we prove that there exist parameters of the I-NESC such that the system converges to the neighborhood of the Nash equilibrium in the limited information case where parameter estimation is used. We provide a simulation example that demonstrates that smaller perturbation frequencies and amplitudes are needed to attain similar convergence speed as the existing state-of-the-art algorithm.