Optimisation of Small Diesel Engine Performance Using Response Surface Methodology and Genetic Algorithm

ABSTRACT Even though significant studies on diesel engine optimisation have been made, most existing approaches lack comprehensive integration of methods for effective multi‐objective optimisation (MOOP) frameworks that effectively balance engine performance and emission trade‐offs under diverse engine operating conditions. This highlights the need for more integrated, data‐driven MOOP frameworks. So, in the present work, an effective MOOP of diesel engine performance and emission carried out via adaptation of integrated methods involving Box–Behnken design of experiment (DOE), response surface methodology (RSM), Box‐Cox plot, and genetic algorithm (GA). A comprehensive experimental investigation based on Box–Behnken DOE performed to evaluate the combined effects of key engine operating parameters—fuel injection pressure (FIP), load torque, and engine speed on performance metrics including brake specific fuel consumption (BSFC) and brake thermal efficiency (BTE), as well as emission characteristics, such as unburned hydrocarbons (HC), carbon monoxide (CO), nitrogen oxides (NO), and smoke opacity. RSM is used to construct predictive models and derive objective functions for each response with Box‐Cox transformation ensuring model reliability. MOOP using GA is performed to simultaneously achieve enhanced performance and reduced emissions. The integrated method‐based optimisation result is validated with optimised engine input‐based experimental results with an engine settings of 2270 rpm, 12 Nm and 250.4 bar FIP, and found to be in good agreement with modelled‐optimisation results with less error < 3% for all except BTE (6.735%) and CO (11.11%) which are well acceptable in practical scenario. Consequently, the same integrated method‐based MOOP can be applied to other fuels based engine running condition effectively.