Artificial Neural Networks Based Algorithm for Identifying Engine Oil Parameters

Engine oil is an essential type of lubricant normally used by internal combustion engines. The main job is to lubricate moving part of the engine. Engine oil is also performed to clean, inhibits corrosion and cools the engine by sinking the heat away from the main structure of the engine. However, the viscosity of the engine oil varies with respect to changes in mileage and the change of lubricant on schedule is always inaccurate. As a result, engine oil becomes out of condition for after a period of usage, resulting poor performance and increase cost and maintenance. Viscosity is a parameter to measure resistance of a fluid to flow. Through the theoretical and experimental method, this paper verifies the measurement of transmittance percentages of viscosity in terms of infrared wavelength with respect to several classes´ mileage distances. The most robust networks are also identified to be applied in this study. Finally, this information can be used in designing sensor to detect the degree of viscosity and development of the decay of lubricant evaluation index.