Torque-related Control Strategies For Mode Transition And Gear Shifting In HEVs

In comparison with the conventional powertrain,the hybrid electric powertrain has advantages in terms of reducing energy consumption and cutting emissions.The P2 hybrid electric powertrain equipped with an electric automated manual transmission has advantages in terms of cost and maintenance,but is subject to drivability issues.The clutch engagement resulting from the mode transition and the gear shifting affects the responsiveness and smoothness of the propulsion torque.During the gear shifting,the duration and jerk should be reduced simultaneously.In the P2 hybrid electric powertrain,the control problems of the clutch engagement and the active synchronization should be addressed.In compliance with the torque-based control architecture,the important component models of the P2 hybrid electric powertrain are proposed.While modeling the engine,clutch and motor,by analyzing the mean value model and conducting experiments,time constants are used to describe the torque response characteristics.While modeling the clutch,the LuGre friction model is used to simplify the model architecture,and to ensure the computational speed.The coefficient of the LuGre friction model is tuned to reduce the difference between the model results and the experiment results.The powertrain model is implemented by using S-Functions,and comparison with the SimDriveline model shows the proposed model can obtain identical results and has advantages in terms of the architecture.By analyzing changing speeds of the state variables,the important state variables are chosen,and the dynamic continuous-time state-space model of the P2 hybrid electric powertrain is obtained.The DMPC(discrete-time model predictive control)based torque control algorithm for the clutch engagement is designed.By analyzing the clutch engagement control problem,the cost functions related to the torque responsiveness during the clutch engagement,the duration and smoothness of the clutch engagement,the engine and motor torque,and the torque smoothness after the clutch engagement are proposed.The discrete-time model predictive control is chosen to address the multi-objective optimization problem with constraints.The equilibrium state of the clutch engagement is proposed to describe the ideal state of the powertrain when the clutch is engaged,and the cost functions are transferred to a target state.By discretizing the P2 hybrid electric powertrain dynamic continuous-time model,the dynamic discrete-time state-space model is obtained.The discrete-time Laguerre functions are used to formulate the clutch engagement optimization problem,and then the number of the parameters to be solved is reduced from about 200 to 10,so the computational speed is ensured.The explicit solution of the control algorithm is obtained,and comparison with the standard solvers shows the proposed method has advantages in terms of the computational speed.Comparison with the simulation results of the open loop control of the clutch torque shows the DMPC-based control algorithm can ensure the smooth engagement of the clutch,the responsiveness and smoothness of the propulsion torque,the smoothness of the propulsion torque after the clutch engagement within the identical duration.During the mode transition,after the clutch engagement,the extra torque manipulation is not needed.Furthermore,the simulation results show the DMPC-based control algorithm is robust to the variation of the driving conditions and uncertainty of the model.The DLQR(discrete-time linear quadratic regulator)based torque control algorithm for the active synchronization is designed.By analyzing the active synchronization control problem,the cost function related to the duration is proposed.The discrete-time linear quadratic regulator is chosen to address the active synchronization control problem.The equilibrium state of the active synchronization is proposed to describe the ideal state of the gears engagement,and in the ideal state,the angular speed difference,angular acceleration difference,changing speed of the angular acceleration difference are equal to zero.Then the cost function is transferred to a target state.By discretizing the primary shaft dynamic continuous-time model,the dynamic discrete-time state-space model is obtained.The model is identified by designing and conducting experiments.The explicit solution of the control algorithm is obtained,and the experiment show the computational time of the program is about 3ms.The program architecture and procedure of the electric automated manual transmission are designed.The test bench is designed and constructed,and the experiment procedure is designed.The experiment results show the DLQR-based control algorithm can ensure the motor torque command and motor torque converge to the primary shaft resistance torque,the motor speed converges to the target speed,and the convergence speeds are almost identical.So the experiment results demonstrate that the proposed control algorithm can reduce the jerk.During the experiments,the active synchronization duration can be 89.7ms,and the total gear shifting duration can be 299.1ms.Furthermore,the experiment results validate the robustness of the proposed algorithm.