Snowplow Steering Guidance with Gain Stabilization

Abstract

Acquisition and utilization of lateral guidance information is crucial for steering a vehicle. With practice, drivers can successfully perform the steering function using visual perception and hand-eye coordination. However, this seemingly simple task becomes difficult when the visual information loses its clarity. Driving a snowplow during whiteout conditions is one such example. In order to improve the safety and efficiency of snow removal operations, a supplemental guidance display was proposed and successfully implemented in a California Department of Transportation (Caltrans) snowplow. The guidance information was calculated based on the magnetic markers embedded in the roadway. The crucial step to this success is a transformation of the guidance display problem into a robust driver-in-the-loop control problem. This transformation considers the ‘display’ law as part of the overall driver-steering-control algorithm. Two key ‘assumptions’ for this design are (1) the ‘display’ law should be designed in such a way that drivers can use ‘proportional’ control gain alone to satisfy the performance and stability requirements of the steering tasks, and (2) the driver steering model can be described as a combination of gain and dynamic delay under the display law in (1). This paper describes the validation process for the ‘gain’ stabilization design concept as well as the associated driver steering model using the initial snowplow test data.