| Diesel partially premixed compression ignition(PPCI)can reduce nitrogen oxides(NOx)and particulate matter(PM)while ensuring economy.PPCI engine has many control parameters.Aiming at the problem that the combustion process is difficult to accurately control,this paper develops combustion and emission modeling based on two-stage injection parameters,and applies the predictive model to the model predictive controller(MPC)to achieve two-stage injection control that optimizes performance and emissions.PPCI combustion control parameters sensitivity experimental research was carried out.The influence of control parameters such as exhaust gas recirculation(EGR)rate,two-stage injection timing,injection interval on output parameters such as start of combustion(SOC),crank angle of 50 %heat release(CA50),pressure rise rate(PIR),NOx emission,PM emission,etc.is analyzed,which provide support for control-oriented modeling.A combustion prediction model and an emission prediction model based on injection parameters have been established to improve the prediction accuracy.The measured heat release rate(HRR)of PPCI is linearized by Wiebe function,and the HRR segmentation phenomenon is obvious.Based on this phenomenon,a linearized combustion prediction modeling method is proposed.The maximum error of CA50 and indicated mean effective pressure(IMEP)predicted by the established combustion model is 0.88 °CA and 0.033 MPa.The correlation analysis of emissions using fuel injection parameters and combustion parameters shows that the correlation between emissions and combustion parameters is obvious.Based on the neural network,a NOx and PM prediction model with combustion parameters as input was established,and the prediction accuracy reached 95 %.Based on combustion and emission prediction models and different optimization goals,a model prediction controller for multiple input and multiple output(MIMO)PPCI engine is designed.In target parameters tracking control,the steady state error of CA50 and IMEP is less than 2 %,and the system adjustment time is within 8 cycles.In the cooperative optimization control of combustion parameters and emissions,the IMEP tracking and the cooperative control of PIR,NOx and PM are realized.After applying MPC,NOx emissions were reduced by 36.8 %,and PM emissions were reduced by 50.2 %. |
