Research On Model Predictive Control Of Diesel Engine Air System Based On Mean Value Model

With the promulgation and implementation of CHINA ? and higher emission regulations,diesel exhaust emission regulations become stricter and stricter.For the CHINA VI emission regulations,it gives priority to the use of internal cleaning technology to control NOx and PM emissions in a certain range for bare engine.It is difficult to meet the stringent emission requirements of NOx and PM by adjusting exhaust gas recirculation(EGR)and variable geometry turbocharger(VGT)in CHINA ?/? diesel engine air systems.Therefore,the intake throttle must be added so as to achieve higher EGR rate to meet the air-fuel ratio control requirements,and the emission of NOx and PM can be controlled in a certain range.With the aftertreatment technology can meet the requirements of emission regulations.The following studies are carried out in this paper:1)A four cylinder diesel engine model with EGR,VGT and throttle is established in GT-POWER and validating the model in steady and transient state with full intake throttle.The results show that the GT-POWER model can accurately simulate the state of the real engine.2)The intake manifold pressure,intake air intercooler pressure,exhaust manifold pressure and the rotational speed of the turbine are selected to establish a mean-value model of a four cylinder diesel engine with EGR,VGT and throttle.The mean-value model is verified by the GT-POWER model.The results show that the average accuracy of the proposed model is 96%and 93%in steady and transient state respectively,the calculation of the model is also greatly reduced.3)Model predictive control algorithm for diesel engine air system are designed based on mean value model.The control objective is to make the intake air flow rate(MAF)and the intake pressure(MAP)to track the target value to reduce PM and NOx emissions.At the same time,in order to compare with the model predictive control,an observer based state feedback controller is designed,in which the state observer is used to estimate the state of the inlet intercooler pressure.The results show that the model predictive control algorithm has higher control precision and robustness.