The Study On Intake Flow Estimation Method Of Range-extender Engine And Experimental Verification

Range extender is one of the important components in series hybrid power system,it is air-fuel ratio control(AFC)has direct effect to energy exchange efficiency for the system.The air mass into the cylinder is the feedforward information of the AFC and its estimation accuracy has an important influence on the accuracy of AFC,which indirectly affects the engine fuel economy and exhaust pollutant emissions.During the engine working cycle,the air mass into the cylinder is affected by volumetric efficiency which has time-variant characteristic with respect to the engine working conditions for the air intake system.Therefore,it is an effective method to improve the effect of air mass flow estimation by improving the real-time identification accuracy of the volumetric efficiency under various working conditions.Focus on the structural characteristics of the air intake system of range-extender engine,and the corresponding air intake system model can be established by the analysis of the air dynamic in the air intake process.In every working cycle,the variation of engine speed will lead to the unevenness of fixed-step sampling.Therefore,it is necessary to discretize the mean-value model and adjust the sampling step in real time according to engine speed,so as to achieve uniform sampling based on crank angle.In consideration of the influence of intake fluctuations on signal acquisition,the moving average filtering algorithm is used to filter the signals of inlet flow at the throttle and intake pressure,the periodic fluctuations of the signals can be eliminated by the real-time filtering.Aiming at the problem of air mass flow estimation,an adaptive observer is used to identify the volumetric efficiency in real time to obtain the relatively accurate inlet flow estimation results.In order to further improve the estimation effect under transient conditions,the self-learning algorithm is combined with the original estimation algorithm to make the engine gradually obtain the MAPof volumetric efficiency through self-learning and correct itself in the process of running.The identification speed and accuracy of volumetric efficiency under transient conditions can be improved effectively and corresponding transient inlet flow will be estimated more accurately meanwhile.The engine test bench is designed and established to verify the estimation effect of the proposed algorithm.dSPACE is also used to realize the real-time calibration of the algorithm,and the actuators are drived by self-designed engine drive circuit.The two methods proposed in this paper are verified respectively by experiments.During the experiment,the first method is compared with other methods,and the experimental results of the second method are analyzed according to the shortcomings of the first method.In addition,the estimation effects of the two methods are verified by comparing the estimation results of inlet flow with the measured values at throttle.