In this paper, a longitudinal vehicle controller is proposed for an active driver safety system and its validation via Hardware-In-Loop Simulation (HILS) is presented. Since the desired speed and distance are chosen arbitrarily by a driver, there are usually discontinuities or discrete jumps between the desired and current vehicle state right after the desired input is changed. To minimize performance degradation resulting from this discrete jump, one of nonlinear control techniques, Dynamic Surface Control (DSC) with an input-shaping filter, is applied for both velocity and distance control. Furthermore, while much cost and effort are in general required for experimental validation of the longitudinal controller, its validation via HILS is performed with a better efficiency. Finally, various switching scenarios including discrete desired inputs in terms of velocity and distance are considered and the corresponding performances of the controller are shown in the HILS.
Adaptive neural network dynamic surface control for musculoskeletal robots
