The target of this paper is to study two combustion control strategies, partial premixed compression ignition (PPCI) and dual fuel combustion (DDF), to solve the problems of low efficiency and high emissions of micro-ignition dual fuel engines. Firstly, the effects of different control parameters on engine performance and emissions in two combustion modes were studied. Then, the encoder-decoder convolutional neural network-gated recurrent unit (ED CNN-GRU) is used for the first time to establish a predictive regression model between the operating parameters and performance of the diesel micro-ignition dual-fuel engine. Finally, NSGA III is used to drive ED CNN-GRU to perform multi-objective optimization of engine performance. The results show that under 75 % load, the predicted results of the model and test results show that the corresponding THC emissions in the PPCI combustion mode are 50.65 % and 53.18 % lower than those in the DDF combustion mode, the corresponding CO emissions in the PPCI combustion mode are 69.05 % and 70.26 % lower than those in the DDF combustion mode respectively. The prediction results of the model and the test results meet the Tier III emission regulations, and the emission is minimized while taking into account the economy. Under the propulsion characteristics, PPCI combustion mode is selected for 0–75 % load, and DDF combustion mode is selected for more than 75 % load. The control strategy of micro ignition dual fuel engine can effectively promote the development of micro ignition dual fuel engine in the field of marine engine.
Read full abstract