A Proposal to Overcome Constraints of IP Networks in Applications of Closed Loop Control

This work proposes the use of a Smith predictor with jitter transport delay adaptation to feedback control systems on which control loops are closed through an IP network. The well know Smith predictor is efficient to compensate constant transport delay but IP networks presents jitters which disturb the performance of the predictor. Therefore, together to the predictor it is implemented a new compensation way to this typical jitter of these networks. This compensation has a data structure in order to store values from the output of the predictor, in order to the right value be read to from it to compare with to the value from the system feedback. The value is chosen based in the last delay in the control loop. The proposed approach was validated via simulation and experimental results are presented in the paper.