Optimization Layout Of Warning Signs Based On Road Information Pulse Model

Sharp turn and slope are not rare in mountainous highways due to the topographic condition. Warning signs will always be needed as one of the supplements to the road alignment which may not sufficiently provide the information of the dangerous section in the front. Behaviors are expected to be consistent with the environment to keep it safe. This paper took warning signs and the road alignment into the same scope to set a road information pulse model. Experiments were carried out in Tibet by the means of naturalistic driving. Warning signs were abstracted by two parameters including location and intensity. Road alignment was described by spatial geometric curvature collected by GPS and drivers' visual curvature calculated by drivers' visual lane model. Samples of vehicle passing through curves were extracted. Driving speed was analyzed as sequence data using R/S trend detective combined with Mann-Kendall catastrophe point analysis. Comparative analysis was conducted between road information pulse model and driving speed sequence. The results suggested that there exists certain type and reasonable range of warning signs when the geometric information and the visual information do not coordinate. Warning signs are expected to meet the gap of the spatial geometric curvature and the visual curvature, inducting driving behavior to be in the reasonable range of the road supply.