CMAC-PID control and new smith predictor for networked control systems

The cerebellar model articulation controller (CMAC) neural network is a practical tool for improving existing nonlinear control systems, and it can effectively reduce tracking error of the control system. In order to availably restrain the impact of the network delay for the networked control systems (NCS), a novel approach is proposed that new Smith predictor combined with the CMAC-PID control for the NCS. This new Smith predictor comes true to hide the predictor models of the network delays into real network data transmission processes, therefore the network delays don't need to be measured, identified or estimated on-line. Simultaneously, this new Smith predictor doesn't include the prediction model of the controlled plant, thus it doesn't need to know the exact mathematical model of the controlled plant beforehand. Therefore it is applicable to some occasions that network delays are the random, time-variant or uncertain. Based on CSMA/CD (Ethernet), the results of the simulation experiment show the validity of the control scheme.