On the Estimation of the Useful Lifespan of Lubrication Oil under Constrained Functioning Conditions an ANN / Fuzzy Logic Approach

Abstract

Lubrication oil in automotives is a multi-billion-pound business but the non-optimization of its lifespan entails colossal harm to societies, global resources and the environment. Losses are caused by premature oil change or by machinery wearing due to deteriorated oil. The actual practice in the automotive field follows a predetermined routine-replacement policy that does not consider the wide spectrum of operating conditions. In this paper a decision support model is developed for the determination of the optimum life span of oil under specific working conditions. A data gathering scheme is set to capture the most relevant oils' characteristics from real samples over specific ranges of operation. The relationship between the causal factors and the resulting condition of oil is programmed in an ANN which is complemented with a fuzzy-logic approach in order to predict the optimum lifespan of oil under any set of causal factors. The approach is applied on a case study in Egypt; the model is tested, validated and is believed to fulfil its objectives satisfactorily.