Controller design for networked control systems based on neural network prediction

To the networked control systems in which network-induced delay is time-vary and less than one sample time, a neural network predictive control method is presented to compensate the random delays in this paper. Genetic algorithm is used to optimize parameters. Neural network is introduced to do prediction and optimization calculation. And the uncertainty of the delay is transformed to parametrical uncertainties. Simulation result shows that this control method can evidently improve the control quality with increasing time delay, and has stronger adaptability.