Transient Performance Control in Neural Network Tracking Systems

Training stable neural network controllers for closed-loop tracking control systems remains a challenging task. Previously, the authors proposed a method to ensure the stability of a neural network control systems by finding for the Lyapunov function using mixed integer programming. However, in addition to stability, the transient performance is also crucial for control systems. In this paper, we propose an approach related to the introduction of additional requirements to the Lyapunov function, which allows to control the parameters of the transient process.