차량 DTC 고장 예측을 위한 딥러닝 적용 사례 연구

Abstract

In recent times, there has been a growing focus on integrating information technologies and arti- ficial intelligence into vehicle condition diagnosis and prediction technologies, enabling proac- tive identification of potential vehicle malfunctions. The provision of automated vehicle condition assessments and timely predictive maintenance services to drivers holds substantial significance. This research deals with a case study dedicated to the differentiation between faults and normal states in commercial vehicles using an unsupervised deep learning approach, based on DTC (Diagnostic Trouble Code) data. We construct and evaluate three distinct deep learning models to forecast fault occurrences. The outcomes of this case study are deliberated upon in conjunction with its limitations and prospects for future research directions.