Incentive Stochastic Stackelberg Games With Measurement Feedback Control

Abstract

In this paper, incentive Stackelberg games are investigated for a class of stochastic linear systems. Unlike the existing ordinary Stackelberg games, output feedback control is developed in such a way that state variables cannot be measured directly, especially for networked control systems. First, the incentive Stackelberg strategy is designed by the leader to achieve team-optimal solution. Second, the follower achieves its equilibrium ensuring the incentive Stackelberg strategy by solving some cross-coupled stochastic Riccati differential equations. Finally, the optimal estimator (conditional expectation) for the leader and follower is derived.