Work

Improved Event-triggered Approximate Optimal Control for Nonlinear Nonzero-sum Games Using Reinforcement Learning