IC Engine Dynamic oil Life Prediction Using Machine Learning Approach

<div class="section abstract"><div class="htmlview paragraph">Conventional resource conservation for automotive traction and its utilization to maximize efficiency is not only important but also essential for climate change. Diesel engine lubricating oil is one such component, which has enough potential considering the current methods of servicing. Present work is focused on developing/validating a process, which can predict critical properties of oil at any given instance. This prediction is independent of vehicle operating conditions and vehicle mileage. As the process is independent of above two parameters, even the oil quality after a top up can be identified. Experimental setup consisted of single cylinder stationary engine (Kirloskar - TV1) delivering 5.2kW power @1500 RPM is tested for detailed analysis of the oil characteristics such as Kinematic viscosity, Acid number, Base number and Carbon residue with respect to different engine parameters.Test run was conducted for 150 hrs and the said characteristics for samples were critically analysed at fixed intervals. A ML model was built using data obtained from the test to classify oil quality and predict the remaining mileage. This model was validated with different oil samples (fresh, intermediate and deteriorated). While this model was deployed over raspberry pi for live computations, the results (classification - ok/not ok and mileage) were relayed to a mobile application to simulate end user real time monitoring. The future work will focus on real time testing of a heavy diesel engine integrating both experimental and ML approaches to predict the life of lubricating oil coupled with dynamic scenarios and potential oil failure modes.</div></div>