A Collision Severity Prediction Method Oriented to Occupant Injury Risk

Abstract

This application relates to a collision severity prediction method. By constructing a first learning model and inputting vehicle data and collision accident scene feature data into the first learning model to obtain a collision acceleration curve, the road traffic conditions and The vehicle's own data, the actual calculation of the vehicle's collision acceleration. By constructing the second learning model and inputting the occupant characteristic data, restraint system characteristic data and collision acceleration curve into the second learning model, the collision dynamic curves of different human body parts can be obtained, which can comprehensively consider the occupant characteristics and restraint system characteristics, etc. The characteristics of the collision scene greatly improve the reliability of occupant collision severity prediction. Finally, the quantification of crash severity is achieved through scalar transformation of the crash dynamics curves of different body parts. The collision severity prediction method involved in the present application can realize rapid and accurate occupant collision severity prediction under the premise of considering comprehensive factors.

Description

technical field [0001] The present application relates to the technical field of vehicle accident analysis, in particular to a collision severity prediction method oriented to occupant injury risk. Background technique [0002] At present, the path planning algorithm for vehicle driving has been relatively mature in path search and selection. Commonly used path search algorithms include Dijkstra's algorithm and A * Algorithms and more. Usually, the path selected by planning should meet the two basic requirements of feasibility and safety, so as to ensure that the selected path can not only be realized in control, but also ensure driving safety. When there is a safe path in the alternative path, the collision probability of this path is zero, regardless of whether the collision severity calculation is performed, the collision severity is zero, and it is meaningless to calculate the collision severity at this time. However, when there is no safe path among the alternative p...

AI Technical Summary

Problems solved by technology

[0007] Based on this, it is necessary to provide a prediction method of collision severity for the problem that the occupant collision process is oversimplified and the accuracy is low in the traditional prediction method of collision severity

Images