Joint Smith predictor and neural network estimation scheme for compensating randomly varying time-delay in networked control system

Abstract

In this paper, we dealt with the problems of network induced delay and randomly varying time-delay (RVTD) controlled plant in networked control systems (NCS). These inherent challenges make the conventional control methods, e.g, a mathematical model of Smith predictor combined with fuzzy adaptive controller, more difficult to meet the quality requirements for NCS stability. Specifically, based on analyzing the existing techniques, we propose a novel method to efficiently compensate the RVTD for NCS. This so-called RVTD compensator for the Smith predictor using Neural network estimation scheme has not only the features of simple Smith predict structure, but also the characteristics of adaptivity, stability, and fast response. The simulation results via TrueTime Beta2.0 platform demonstrate that our design significantly improves the performance of NCS.

Jacking and Energy Consumption Control Over Network for Jack-Up Rig: Simulation and Experiment

“…Related to the time-delay problem, the response of the NCS will be lowered by the network-induced delay, causing errors in the data transmission process e −dcas [27][28]. Typically, the delay of the control signal d ca transmits to the JuR dynamic model.…”