Design of driver assistance system for air cushion vehicle with uncertainty based on model knowledge neural network

Abstract

In this paper, considering the difficult maneuverability of the air cushion vehicle (ACV), a driver assistance system (DAS) of ACV including an intuitive human-computer interface, DAS monitor and DAS controller is developed for humans. The human-computer interface is easy to be understood and used for humans. And as DAS monitor, appropriate sensors installed at handles of rudders and propellers are used to monitor driver's operational changes. For the design of DAS controller, model knowledge neural network (MKNN) method is first proposed in this paper to deal with the parameter uncertainty of ACV's complex model. Then the MKNN-based controller is designed as the DAS controller. The DAS with MKNN-based controller can assist drivers in better control operations according to their action instructions. And numerical simulations are implemented to demonstrate the effectiveness and superiority of the developed DAS with MKNN-based controller.