Identification and validation of a driver steering control model incorporating human sensory dynamics

Abstract

Most existing models of driver steering control do not consider the driver's sensory dynamics, despite many aspects of human sensory perception having been researched extensively. The authors recently reported the development of a driver model that incorporates sensory transfer functions, noise and delays. The present paper reports the experimental identification and validation of this model. An experiment was carried out with five test subjects in a driving simulator, aiming to replicate a real-world driving scenario with no motion scaling. The results of this experiment are used to identify parameter values for the driver model, and the model is found to describe the results of the experiment well. Predicted steering angles match the linear component of measured results with an average 'variance accounted for' of 98% using separate parameter sets for each trial, and 93% with a single fixed parameter set. The identified parameter values are compared with results from the literature and are found to be physically plausible, supporting the hypothesis that driver steering control can be predicted using models of human perception and control mechanisms. ARTICLE HISTORY

Measurement and Modeling of the Effect of Sensory Conflicts on Driver Steering Control

“…Assuming a linear driver model is a reasonable fit to the control behaviour of the drivers, and that all drivers were adopting a similar control strategy, the time-domain steering signals can be averaged over the five drivers to give a set of average responses for each trial. Comparison of parameter values identified for individual drivers and averaged signals in the full motion experiment showed that this approach is reasonable [3]. The averaged signals have less noise in the measurements than the individual signals, therefore this data is used to identify parameter values for a typical driver.…”

“…lowing drivers to preview future target values, or a straight line which moved over the road, removing preview so that visual delays could be measured more easily (see Figure 4). The first experiment (from [3]) used unity gains on all motion to measure driver performance without sensory conflicts. Details of the analysis and results of this 'full motion' experiment are reported in [3].…”

“…The first experiment (from [3]) used unity gains on all motion to measure driver performance without sensory conflicts. Details of the analysis and results of this 'full motion' experiment are reported in [3]. In the next two experiments the lateral and yaw motion were scaled by constant scaling factors; one experiment focussed on trials where the driver could not preview the upcoming target path and the other involved only trials with preview.…”

“…In previous work, a new model of driver steering control was developed incorporating models of the driver's sensory systems and assuming optimal integration of noisy sensory measurements [2]. An identification procedure was developed to find values of the physical parameters of the model which matched results from an experiment carried out in a driving simulator [3]. The model was able to represent the experimental results well, and the parameter values were found to be physically plausible when compared with estimates found from a review of the literature on sensory dynamics [1].…”

“…The experiment carried out in [3] was carefully designed so that the lateral and yaw motion applied to the driver in the moving-base simulator matched the motion of the simulated vehicle at full scale. Therefore the information perceived by the driver's visual and vestibular systems was coherent, without any sensory conflicts.…”