Pedestrian risk assessment method considering pedestrian aggressivity

Abstract

The invention discloses a pedestrian risk assessment method considering pedestrian aggressivity. The pedestrian risk assessment method comprises the steps of S1, detecting whether a pedestrian passesthrough a self-vehicle driving route or not; if so, entering the step S2; S2, calling self-vehicle sensor data and positioning data to acquire pedestrian waiting time, whether pedestrians retreat or not, the number of pedestrian passing groups, the pedestrian passing speed, the acceleration and the pedestrian-vehicle distance; S3, determining a numerical value and an aggressivity index of a characteristic variable according to the data obtained in the step S2; S4, taking the aggressivity index of the characteristic variable obtained in the step S3 as input, and obtaining a pedestrian overall aggressivity value through a deep learning network based on a multilayer perceptron MLP; and S5, calculating a pedestrian risk value based on the pedestrian overall aggressivity value. By adopting themethod provided by the invention, the pedestrian aggressivity can be recognized and converted into the pedestrian risk value, so that the own vehicle can make decisions differentially, and the trafficefficiency is improved.

Description

technical field [0001] The invention relates to the field of pedestrian risk assessment for intelligent networked vehicles, in particular to a pedestrian risk assessment method considering the aggressiveness of pedestrians. Background technique [0002] The decision-making method of Intelligent Connected Vehicle (ICV) is one of the key technologies for autonomous driving. However, most of the existing decision-making methods are difficult to support high-level automatic driving, especially in crowded and complex scenes with low traffic efficiency, and one of the reasons is that there is no differentiated processing based on the individual characteristics of different objects. As a result, the dynamic adjustment ability of the vehicle according to the environmental decision-making is poor, and the degree of differentiation is insufficient. Therefore, the current ICV decision-making method needs to be optimized. [0003] A key pain point of differentiated decision-making opti...

AI Technical Summary

Problems solved by technology

[0006] Therefore, there is currently a lack of a suitable pedestrian risk assessment method, which makes subsequent decision-making difficult to achieve refinement and differentiation. It is necessary to develop a risk assessment method that can reflect the individual characteristics of pedestrians

Images