Matching and optimization to minimize fuel consumption and NOx emission for a marine diesel engine with turbo-assisted exhaust gas recirculation

Exhaust gas recirculation (EGR) and two-stage turbocharging technology are widely used in marine diesel engines due to their advantages of reducing nitrogen oxides (NOx) emission and improving fuel economy respectively. The matching of EGR, two-stage turbocharging and diesel engine is significant to realize both advantages simultaneously. Turbo-assisted EGR (TA EGR) can achieve high EGR rate and lower fuel consumption due to a turbocharger driving EGR. In this study, a 1D model of the diesel engine with TA EGR coupling two-stage turbocharging system was established by GT-Power. A theoretical matching method was proposed based on the requirements of excess air coefficient and EGR rate. The matching laws of the coupling system were studied, and the preliminary matching results achieved good effects. A fully connected neural network was used as the optimization agent model of the 1D model, and the NSGA II algorithm was used to carry out the multi-parameter and multi-objective optimizing matching. Considering constrained and unconstrained conditions, independent optimizing matching was carried out under each single load, and the weighted brake specific fuel consumption (BSFC) and weighted NOx under all loads were taken as the target to carry out comprehensive optimizing matching. The results showed that the optimization performance is worse under constrained conditions, while TA EGR coupling adjustable two-stage turbocharging system can achieve comprehensive optimal BSFC and NOx, with weighted BSFC decreasing by 3.57% and weighted NOx decreasing by 75.28% compared with experimental results of the original engine.