In this paper, the modeling and stability analysis of time-delay compensated network control system, the robust control method based on time-delay compensation and the improvement of time-delay compensated network control system are studied. The neural network strategy is used to optimize the controller, combined with predictive control and time-delay compensation control to improve the system performance. Optimize the network communication, improve the communication rate and reliability, reduce the probability of data loss and error, and enhance the anti-interference performance. After optimization, the network communication time is shortened by more than 30%, the probability of data loss and error is reduced by more than 50%, and the system stability and reliability are significantly improved.