Distributed generalized Nash equilibrium seeking for noncooperative games with unknown cost functions

In this article, a distributed nonmodel based generalized Nash equilibrium (GNE) seeking algorithm is proposed for class of constrained noncooperative games with unknown cost functions. the game, strategy each agent restricted by both coupled equality constraint and local inequality constraints. By virtue exact penalty method, an auxiliary function constructed The main feature depends on capability to estimate gradient information functions only values costs. This obtained extremum control (ESC). To deal constraints, Lagrange multiplier transmitted among agents some prior about Moreover, diminishing dither signal introduced in remove undesirable steady-state oscillations occurred classical ESC. As result, nonlocal convergence designed GNE game singular perturbation theory, averaging analysis Lyapunov stability theory. Numerical examples are given verify effectiveness our method.