Prediction of networked control system time-delay based on grey model and Elman neural network

In order to improve the time-delay prediction accuracy, this paper proposes a network time-delay prediction method based on grey model and Elman neural network. Grey model can weaken the randomness of the time-delay sequence, and the Elman neural network has good prediction ability for nonlinear sequence. The advantages of these two methods are combined. The final predictive time-delay is obtained by grey model and Elman neural network multiplied by their weighting coefficients. The optimal model weighting coefficients are determined by principle of minimum variance. The simulation results show that this combination prediction method has higher prediction accuracy with smaller predictive error.