A system-level optimisation framework for marine diesel engine calibration and performance mapping

This study presents a sequential optimisation framework for marine diesel engine calibration and performance mapping. The framework couples a one-dimensional WAVE engine simulation model with a MATLAB-based nonlinear constrained optimiser, enabling systematic tuning of key control variables across the engine load diagram under realistic thermodynamic, operational, emissions, and combustion-stability constraints. The methodology is structured into three connected modules. The first module calibrates mechanical losses using an optimised friction model based on the Chen–Flynn correlation, ensuring accurate representation of friction mean effective pressure (FMEP). The second module calibrates nitrogen oxides (NOx) emissions by tuning the Arrhenius coefficients of the extended Zeldovich mechanism. The third module embeds the calibrated friction and NOx submodels as fixed components within a system-level load-diagram optimisation framework. The optimisation adjusts turbocharger speed, fuel rate, start of injection, and injection duration while satisfying constraints on brake power, maximum pressure, maximum temperature, exhaust temperature, air–fuel ratio (AFR), NOx emissions, and combustion stability. AFR is used as a constrained air-management objective rather than as a direct surrogate for minimum brake-specific fuel consumption (BSFC). A BSFC-based constrained optimisation is also performed as a benchmark. The results demonstrate smooth, constraint-compliant, and physically consistent marine engine maps.