Improved generalized predictive control using grey prediction for internet-based Autonomous Underwater Vehicle

Time delays are random and uncertain, which degrade the performance of an Internet-based Autonomous Underwater Vehicle (AUV) and even destabilize the system. In this paper, an improved generalized predictive control (JGPC) with predictive time delays is proposed. A grey model GM (1, 1) with limited number of delay data is established to predict delays in next control periods based on characteristics of network delays. Meanwhile the predicted delays and online identification parameters are used to regulate the system model. Then on the basis of the predicted time delays, a local controller JGPC aimed at calculating series of predictive control laws is designed to realize compensation of time delays in the system. The Internet-based AUV can operate periodically by remote actuator with register. Finally, an underwater manipulator depth control is used to validate the approach. The simulation results demonstrate the efficacy of this approach compared to the JGPC controller with constant delay parameter. The desired set-point tracking and good performance of the system with variable delays and packet losses are achieved.