Pedestrian-Vehicle Collision Avoidance Strategy for NLOS Conditions

Pedestrian safety is one of the most important goals of transportation systems. A set of different collision avoidance architectures and algorithms using a wide range of sensors and communication technologies were proposed. But the important issue is the accuracy of position beaconing. Also, the strategies for pedestrian-vehicle collision avoidance in scenarios with non line-of-sight (NLOS) pedestrians still need to develop. In the paper is presented the review of existing systems for collision prediction based on different types of sensors and communication technologies. We propose the enhanced collision avoidance structure for NLOS pedestrian based on the low-cost inertial navigation system for increased accuracy of the positioning and communications over IoT. This paper provides the analysis of existing collision avoidance algorithms. We propose our algorithm which has got a set of enhances. We consider the danger zones and propose our novel collision avoidance strategy which depends on the zone. Also, we have improved the probable collision point calculation by taking into account the approaching speed and direction of the movement.