Multi-objective optimization of CRDI engine parameters fueled with blends of diesel and sterculia foetida biodiesel: A comparative study of RSM composite desirability and meta-heuristic algorithm

This study presents a comprehensive multi objective optimization of common rail direct injection (CRDI) engine
parameters using response surface methodology (RSM) with composite desirability approach and three metaheuristic algorithms: genetic algorithm with desirability function (GADF), differential evolution with desirability (DE-DF), and particle swarm optimization with desirability (PSO-DF). A central composite design (CCD) with 50 experimental runs was conducted to evaluate the effects of fuel blend percentage, engine load, injection pressure, injection timing, and EGR rate on engine performance and emissions. The study aimed to maximise torque, brake power, BMEP, brake thermal efficiency, mechanical efficiency, and volumetric efficiency while minimising specific fuel consumption, CO, HC, and NOx emissions. RSM composite desirability optimisation yielded optimal conditions at 53% fuel blend, 98% engine load, 999.97 bar injection pressure, 6.0o BTDC timing, and 0% EGR. Mata-heuristic validation showed DE-DF achieving 95.3%similarity to RSM results, PSO-DF
demonstrating 97.9% similarity, and GADF providing 88.9% agreement. The study validates the effectiveness
of meta-heuristic algorithms as robust alternatives to traditional RSM approaches for complex engine optimization problems.