Model Based Control Of Camless Electro-hydraulic Variable Valve Actuator For Internal Combustion Engines

The increasingly stringent regulation for fuel economy and emission bring the world a new revolution for the automotive industry.Electrification is one of the most important trend.The camless variable valve actuation(VVA)is a milestone for the internal combustion engine(ICE)electrification,since it replaces the camshaft and realize fully flexible and fast engine breathing.This will help further improve the ICE performance.Without the constrains of a camshaft,the camless VVA relies on the electronic control of the valve timing,lift,profile area and seating velocity to guarantee the reliability and accuracy of air charging process,making the VVA control one of the most important task during engine operation.Therefore,In this dissertation,model based control of a novel electric-hydraulic variable valve actuator(EHVVA)for the ICE is studied.The target is to overcome the challenging of temperature and pressure variation of the hydraulic system and guarantee the accurate VVA motion.First,detailed operation characteristics analysis of the EHVVA is conducted.The system structure,functional principle is introduced and the test bench is developed in this section.The system dynamics in different time scope(in-cycle and cycle-by-cycle)and the parameter affecting pattern are analyzed using the test data,revealing the system nature of time-delay,time-varying,nonlinearity and disturbance.Two control problems of the system are identified,the valve timing control under the valve delay variation due to the temperature and the disturbance of pressure variation,and the nonlinear control of the valve profile area(response).Second,detailed dynamic modeling is conducted to provide a platform for both simulation study and control-oriented modeling during control design.The system components are modeled using the mechanical,electro-magnetic,and flow dynamics and a high-order time-delay,time-varying,and nonlinear model is achieved.To reduce computation effort in online implementation,the high-order model is reduced using simplification and decoupling,and also linearized using the linear-parameter-varying(LPV)method.The model parameters are calibrated and identified using the test data and finally the model is verified.The challenge of the EHVVA control is the temperature sensitivity of the oil viscosity and variation of the oil pressure,which bring the valve timing and response variation.Taking the challenge,the key parameter affecting valve timing and response,the unmeasurable supply pressure point before valve open,is adaptively estimated online.To increase the transient convergence rate and steady-state accuracy of estimation,and to provide one-step/cycle prediction for the valve delay to provide feedforward valve timing control,a model-guided adaptive estimation is proposed.The prosed algorithm is validated through both simulation and bench test.To control the valve timing,with the measured temperature,the estimated and predicted supply pressure are used to look-up the valve delay and thus a feedforward can be realized for the valve timing control.Since the conventional open-loop feedforward control has steadystate error and the PID control has one-step delay,an adaptive disturbance compensation based optimal tracking control algorithm is proposed.The receding horizon method is used to realize moving optimization during the tracking control.The effectiveness of the proposed algorithm is validated both in simulation and bench tests.The profile area control problem is equivalently transformed into a response tracking problem and the active regulation of the oil supply system(DC motor)is used for control input.Since the valve response/rising time varies nonlinearly with the supply pressure,and the engine speed variation bring further disturbance,the conventional open-loop and PID control can hardly achieve good performance.Therefore,the trajectory linearization and optimal tracking control are combined.The trajectory linearization is used to decouple the nonlinearity and disturbance from the linear portion of the system,and an equilibrium point feedforward optimal tracking controller is designed at each operation point with Kalman optimal state estimation.Comparison validation is conducted both in simulation and bench tests.As a summary,the accurate EHVVA motion control is realized by a systematic methodology,during which theoretic methods are used for the system modeling and control design.