Analysis of autonomous vehicle safety constraints based on systems-theoretic process analysis

Abstract

This article starts with the safety challenges faced by autonomous vehicles, sorts out the development stages of autonomous vehicles and the cases of safety threats. It combines the safety standards of connected vehicles through case analysis, and uses STPA to analyze unsafe controls. Based on the four dimensions of laws, regulations, hardware, software, and network, this paper analyzes the safety constraints of autonomous driving technology, and proposes corresponding safety constraint strategies. The analysis results show that the improvement of the automobile life cycle safety regulation system, the establishment of a new type of anti-deep learning algorithms, and the construction of big data control links by OEMs can effectively reduce the safety risks of autonomous vehicles, which is worth further experimental research.