DIAGNOSTICS OF COMPRESSED AIR CAR ENGINES USING AI

Abstract

Effective diagnostics of the technical condition of a vehicle ensures its reliability and productivity during the performance of transport processes. Modern information technologies and data processing methods make it possible to detect and predict potential malfunctions at the early stages of their occurrence, which contributes to increasing road safety and reducing maintenance costs.
The paper analyzes the methods and models used in information systems for diagnosing engines operating on compressed air. The possibilities of using machine learning algorithms for processing sensor data and detecting engine malfunctions are considered. The structure of the information support of the diagnostic system is proposed, which includes a relational database and machine learning models implemented on the TensorFlow platform.
The possibility of using classification, regression and anomaly analysis methods to determine the technical condition of engine elements and predict their resource is shown. The proposed approach allows you to automate the diagnostic process, increase the accuracy of fault detection and ensure system integration with telecommunication and cloud services.