Driver´s Arms´ Time-Variant Neuromuscular Admittance During Real Car Test-Track Driving

Attempts to measure and model driver steering behavior have been so far mainly performed with driving simulators and time-invariant techniques. The goal of this paper was to quantify the driver´s arms´ time-variant admittance in real driving and to provide a range of parametrically fitted values on the estimated frequency response functions. The human arms´ neuromuscular (NMS) admittance was estimated by applying torque disturbances on the steering wheel during real car test-track driving. To capture the time-variant behavior, the admittance was estimated using a 1.28-s sliding time window. The results showed that drivers adapt their admittance while cornering, exposing a variant behavior during different corners and driving speeds. The frequency response function (FRF) of the admittance while cornering has the properties of a second-order system. During cornering, drivers have increased stiffness values, whilst in straight driving, the FRFs resemble a second-order system ( -40 dB/decade gain drop; double pole at low frequencies) for low frequencies, with a zero for frequencies above 6 Hz (on average). The FRFs during cornering were parametrically fitted to a second-order inertia-spring-damper model. The fitted parameter values can be used for NMS driver models and motivate the stability analysis of the combined closed-loop driver steering system.