Vehicle Engine Fault Diagnosis Approach Based on a Decision Tree and Knowledge Base

This paper proposes an approach to vehicle engine fault diagnosis utilizing a decision tree, knowledge bases, databases, and fuzzy logic. It focuses on identifying patterns that allow us to determine the cause of irregular behaviors in engines, implementing a practical system in mechanic workshops. The study is based on the use of machine learning techniques, especially decision trees, which allow representing fault diagnosis in a branching manner, depending on the presented characteristics or driving conditions. These characteristics are stored in a knowledge base that establishes conditions for making the appropriate and timely diagnosis, initially fed by a database that contains OBDII codes from the vehicle's computer with the cause and the elements embedded. In order to improve the precision and accuracy of the proposal, the decision tree is pruned. Finally, the proposal is validated to probe its feasibility with various evaluation metrics to obtain an accurate diagnosis in the automotive field.