Research on Driver KSS Rating Prediction Model Based on EU 2021/1341 DDAW

In order to establish a scientific driver KSS rating prediction model, screen validation test drivers meeting popular requirements, and carry out DDAW system testing and certification, this paper first collected physiological data such as eye movement, EEG and behavioral characteristics from drivers with different KSS sleepiness rating under fixed test paradigm. Then, through feature extraction and statistical analysis of objective physiological data, physiological index variables with strong correlation with KSS sleepiness rating were found. Finally, it is corresponding with the subjective KSS rating scale to obtain the popular objective physiological data of different KSS rating. Through machine learning, an effective KSS rating prediction model is established, and the prediction accuracy of the model is 76.8%, which has a good prediction performance.