A new neuromuscular driver model for steering system development

The evaluation of steering systems can be enhanced by replacing the test driver by a driver model capable of mimicking the driver’s needs, wants and limitations. This model can subsequently be used to find the optimal steering system design by using computerized optimization routines, or by developing a steering system that adapts to the driver while driving. Also vehicle stability control systems (VSC) can be improved by driver centered design, providing artificial force cues via the steering wheel that aid in controlling the vehicle. A suitable driver model needs to incorporate a neuromuscular (NMS) structure. The validated model of de Vlugt (2004) is used as a basis. Since this model is identified for isometric conditions only, two main adaptations are needed before this model can be used for driver simulation during more extreme maneuvers. First, the driver’s reflexes are given a moving reference point determined by the desired steering wheel position. Second, the muscle forces needed to generate the desired steering wheel angle are calculated by a so called inverse internal model. This model reflects the driver’s learnt dynamics of his arms and the system to be controlled. The new driver model shows to be capable of steering large angles and simultaneously modeling the physical limitations of the driver. Optimization techniques were used to find the optimal driver model parameters, depending on his driving mood and driving task. The same optimization techniques were used to find the optimal steering ratio, where the optimal ratio showed to changes significantly depending on the mood and task. This indicates there is room for improving steering systems as they currently mainly depend on vehicle velocity.