Enhanced robust output tracking of nonlinear systems with dynamic event-triggering using neural network-based method

This paper investigates the problem of robust tracking control for a class of nonlinear systems using a novel three-layer fully connected feedforward neural network controller. The weights of the hidden and output layers of this neural network controller are obtained by solving linear matrix inequalities, while the weights of the input and hidden layers are optimized using a genetic algorithm. Notably, the fitness function for training the genetic algorithm is the square of the difference between the reference signal and the controlled system output signal within the whole period. Moreover, considering external disturbances and time delays of networks, a novel Lyapunov-Krasovskii functional is constructed to derive sufficient conditions for the asymptotic stability with an \(H_{\infty }\) $H_{\infty }$ performance of the nonlinear system. Furthermore, to conserve communication resources and reduce the computational load of the neural network controller, a dynamic event-triggered scheme with a non-negative intermediate variable is implemented. Finally, the tracking effect of the nonlinear system on two types of reference signals is tested on an inverted pendulum model to illustrate and validate the effectiveness of the proposed controller.