TOWARDS ARTIFICIAL INTELLIGENCE BASED DIESEL ENGINE PERFORMANCE CONTROL UNDER VARYING OPERATING CONDITIONS USING SUPPORT VECTOR REGRESSION

Diesel engine designers are constantly on the look-out for performance enhance-ment through efficient control of operating parameters. In this paper, the concept of an intelligent engine control system is proposed that seeks to ensure optimised per-formance under varying operating conditions. The concept is based on arriving at the optimum engine operating parameters to ensure the desired output in terms of efficiency. In addition, a support vector machines based prediction model has been developed to predict the engine performance under varying operating conditions. Experiments were carried out at varying loads, compression ratios and amounts of exhaust gas re-circulation using a variable compression ratio diesel engine for data acquisition. It was observed that the support vector machines model was able to predict the engine performance accurately. Key words: variable compression ratio, diesel engine, support vector machines, performance, exhaust gas re-circulatio