Research on Networked Learning Control System for Unknown Controlled Object Model

Networked learning control system is a kind of feedback control system which the learning loops are closed through network. Network-induced delay has a great adverse effect on the control based on networked learning. For the controlled object of unknown mathematical model, it is proposed in this paper that the network-induced delay compensated strategy, which is based on the cubic spline rolling optimization multi-step predictive algorithm, and corresponding networked learning strategy. The control simulation is studied for complex controlled object of unknown mathematical model in Token network. The simulation results show that the adverse effect can be better avoided which is caused by the network-induced delay on the complex controlled object, and good rapidity and stability can be achieved by applying the complex control strategy.