Novel Smith Predictor and BPNN for Wireless Networked Control Systems

In order to effectively restrain the impact of network delays which are brought because of bandwidth constraint, path loss, signal attenuation and ambient noise in wireless networked control systems (WNCS), aiming at time-variant network delays and imprecise Smith predictor models, a new approach is proposed that novel Smith dynamic predictor combined with back propagation neural network (BPNN) control. Because new Smith predictor does not include network delay models, therefore network delays do not need to be measured, identified or estimated on-line, it is applicable to some occasions that network delays are random, time-variant or uncertain. Based on IEEE 802.11 b/g (WLAN), the results of simulation experiment show validity of the control scheme.