Dynamic abuse testing of airbag system for passenger car under vehicle-road conditions and identification method for high-risk triggering items

Abstract

Determination of the triggering conditions for automobile airbags is a complex problem because vehicles often run under transient conditions due to the dynamic influence from varying road conditions. The research into malfunctions of automobile airbags is an important and challenging topic. In this work, the peak acceleration and moving window integration algorithms were used to obtain the high-risk abuse items from a test matrix that consisted of 111 items which were under multiple road environment excitation, including condition categories for typical roads, curb striking, potholes and drains, and roadblocks. A novel multi-channel data reduction method and improved clustering method based on abuse probability sorting were proposed, and based on the specific characteristic constraints for the automobile airbag malfunction test results, by construction of a false triggering probability model and discussion of the weight factors of the two algorithms, importance rankings for the high-risk false action items were realized based on the relative probability. Finally, 53 high-risk abuse items were identified from a large set of 111 test items through 10-channel acceleration sensors. After analysis and comparison of the test condition details, conclusions of high-risk items are drawn with regard to the vehicle-road excitation that affects airbag abuse. The consistency of the identification results with the Chinese national standard verifies the accuracy and effectiveness of the proposed method.