Location design for intelligent vehicle warning information: level 1 situational awareness modeling based on attention allocation and cognitive mechanisms

Abstract

The traditional layout of warning information positions mostly relies on experience or local experiments, lacking systematic theoretical guidance. The objective of this study is to simulate and model the human performance of drivers under different warning information position conditions through computational methods. Firstly, a comprehensive area division for the head-up display (HUD) and head-down display (HDD) interfaces was conducted. Based on the SEEV model, the N-SEEV calculation framework was introduced to achieve theoretical predictions of drivers' time to first fixation on the target under different warning information position conditions. Additionally, by innovatively integrating attention allocation and cognitive mechanisms, the QN-ACTR-SA1 computational model was developed to achieve mechanism modeling and performance prediction for level 1 situational awareness (SA1). A driving simulator experiment was conducted, recruiting 40 drivers to validate the time to first fixation on the target and the SA1 theoretical values across various scenarios. Results indicate that the lower-central HUD area delivers the most optimal overall performance in target acquisition efficiency, SA1 values, and vehicle control stability. More importantly, the proposed model can effectively estimate both time to first fixation on the target and SA1 values. This provides quantifiable design criteria for optimizing warning interface placement in intelligent vehicles, thereby helping reduce the reliance on large-scale end-user experimental validation during the early design stage.