Research and Optimization of Decision-Making Mechanism for Passenger Vehicle AEB System

Abstract

As intelligent driving technology continues to advance, Automatic Emergency Braking (AEB) systems increasingly contribute to road safety. The paper investigates the decision-making process for conventional passenger AEB systems, summarizing the mainstream collision judgment techniques based on distribution, including distance-based models and Time-to-Collision (TTC) models, as well as analyzing their adaptability in urban and highway settings. This paper explores both single-stage and multi-stage deceleration approaches and provides a comparative analysis of fixed-threshold decision-making and driver-behavior models. To solve problems like the high rate of misjudgments for AEB systems in complex traffic situations, this paper proposes a refined approach by using multi-sensor fusion, deep learning methods for scene recognition, and human–machine collaboration in braking systems. The findings indicate that the future AEB systems will develop towards “high perception robustness - strong decision-making intelligence - human-machine collaborative integration” as the basis for building safer, more reliable, and personalized intelligent driving systems.