Event-Triggered Optimal Safety Control for Nonlinear Safety-Critical Systems with Disturbance

In this paper, the event-triggered optimal safety control method is proposed to solve the zero-sum game problem of nonlinear safety-critical systems with disturbance. First, we transform the safety-critical system with safety constraints into an equivalent system without safety constraints by using the barrier function. Then, for relieving the computation pressure and saving communication cost, the event-triggered mechanism is introduced, and a safe event-triggered condition is presented, meanwhile, the Zeno behavior is excluded. In addition, only a critic neural network (NN) is used to implement the proposed method. During the learning process, the past data and current data are used to relax the persistence of excitation (PE) condition. According to the Lyapunov theory, it can be proved that the states and the weight estimation error of critic NN are uniformly ultimately bounded (UUB). Finally, a simulation example shows the effectiveness of the proposed method.