Artificial neural network to identify RCCI combustion mathematical model for a heavy-duty diesel engine fueled with natural gas and diesel oil

The aim of this study is to apply the artificial neural network to identify the mathematical model of the combustion of a reactivity-controlled compression ignition (RCCI) engine fueled with natural gas and diesel oil. In this study, a single-cylinder heavy-duty diesel engine is set on operation at 9.4 bar gross indicated mean effective pressure. To implement the identification process, the effects of three control factors, namely the intake temperature, the intake pressure (both at the intake valve closing time), and the single-stage diesel fuel injection timing on the engine performance are assessed. Based on the design of experiments–fractional factorial concept, the randomized treatment combinations of chosen levels from the three selected control factors are employed to simulate the RCCI combustion. According to the engine’s responses derived from the RCCI combustion’s simulation, an artificial neural network is trained directly to identify the mathematical model of the RCCI combustion. The results show that the fractional factorial method is capable to produce an appropriate database for the artificial neural network’s training. Moreover, this method can predict more efficient scenarios in which the engine operation under the RCCI combustion has a desirable behavior. Also, artificial neural network as a powerful tool is effectively capable to predict the range of the excessive combustion noise occurrence, misfire occurrence, and the desirable engine load.