Multi-objective optimization of combustion and emission characteristics of ammonia-diesel dual-fuel engines under HPDI strategy based on ELM-MOPSO

• A coupling chemical reaction kinetics model for ammonia-diesel engines is established. • The use of ammonia fuel can significantly reduce carbon emissions. • The novel injection strategy has the capability to strengthen combustion processes. • Model based ELM-MOPSO algorithm was used to obtain optimal CO 2 and NO x emissions. Due to its carbon-free nature, ammonia is increasingly being explored as an alternative fuel for heavy-duty diesel engines. Ammonia's slow flame propagation and high ignition temperature present challenges for efficient combustion. Therefore, this topic carries out research on ammonia diesel dual fuel (ADDF) engine. This study employed a three-dimensional numerical simulation approach to develop and validate a direct-injection ADDF engine model. On this basis, characteristics of engine combustion and emissions under varying ammonia energy fraction (AEF), ammonia nozzle diameters and intake temperature were numerically analyzed. The results showed that reducing the diameter of ammonia spray holes was conducive to significantly improving the fuel-air mixing efficiency and leading to more complete combustion. In addition, appropriately increasing the intake temperature was conducive to promoting the mixing of fuel and air and reducing the emission of unburned ammonia. Finally, the Extreme Learning Machine (ELM) model was employed to train and test the CFD simulation data, and the results were systematically analyzed and validated. The optimized Pareto frontier diagram was obtained through the multi-objective particle swarm optimization (MOPSO) algorithm. The analysis results showed that the minimum emissions of CO 2 and NO x after optimization had decreased by 55.42 % and 48.57 % respectively compared with the initial results.