Modeling And Regeneration Control Of Gasoline Particle Filter

With the widespread application of gasoline direct injection technology in gasoline engines,while improving fuel economy,it also brings more soot emissions than traditional port injection engines.In order to meet strict emission standards,most Automobile manufacturers choose the important technology of gasoline particle filter.In the normal driving,it is usually impossible to completely oxidize all the soot particles in the particle filter,resulting in a decline in economy and power.In order to improve this situation,it is necessary to control the active regeneration of the particle filter from three main directions: regeneration timing,regeneration means and regeneration interruption,so as to restore the filtration efficiency and reduce the exhaust back pressure.Therefore,constructing a reasonable regeneration control strategy is the most effective way to ensure the working efficiency of the gasoline particle filter.This paper builds a visual general regeneration control framework based on Matlab/Simulink,divides the control strategy into seven sub-modules,realizes the determination of active regeneration demand based on the soot load and other regeneration conditions,determines whether activing regeneration according to the engine operating conditions,and coordinates the use of different engines operating conditions to active regeneration,and interruption of active regeneration for safety and economic reasons.According to the control strategy requirements,the temperature changes of the exhaust pipe,three-way catalytic converter,particle filter and pressure drop of the particle filter are predicted based on the model.And on the 1.5 GDI engine bench,the control strategy and object model are designed and tested,and the obtained dynamic data is used to test the strategy and model.The main findings and conclusions of this article are as follows:(1)The soot load estimation in this paper is mainly analyzed from the perspective of soot production and soot oxidation.The soot production is mainly affected by engine operating conditions,coolant water temperature and other factors.Soot oxidation is based on model prediction.Then,the measurement value of the pressure difference sensor is used as an auxiliary to realize the determination of the soot load.The demand for active regeneration is based on a series of judgments based on soot loading,regeneration usually by increasing the exhaust temperature or increasing the oxygen content in the exhaust,such as delaying ignition or increasing the excess air coefficient.When regeneration is in progress,it is necessary to monitor the temperature changes inside the particle filter at all times.When the temperature is too high or continues to be high,it is necessary to reduce the oxidative combustion of soot particles by reducing the excess air coefficient.(2)Using a non-parametric model such as a MAP in a control strategy makes it easier to obtain parameters that cannot be expressed in formulas or that from very complex formulas,and can also be applied to the judgment when multiple input conditions.While ensuring accuracy,the response speed of the control strategy is greatly improved.(3)The zero-dimensional numerical model is used in the object model,the exhaust gas is regarded as an ideal gas,and the exhaust gas temperature and concentration in the threeway catalytic converter are evenly distributed,the exhaust gas in the gasoline particle filter is evenly distributed,and the soot particles are even settlement can not only make rapid predictions,but also meet the needs of control strategies after calibrating the model with corresponding data.(4)Test verification results show that the accuracy of all object models is above 98%,and the deviation of the soot load predicted by the control strategy is less than 10%.It shows that the control strategy and object model built in this paper have high precision and can be applied to the actual control strategy development.