Research On Intelligent Control Of Launch And Gearshift For Dual Clutch Transmissions

The control of launch and gearshift are important factors affecting the smoothness and the powerfulness of the vehicle with dual clutch transmissions(DCT).The clutch control is the core of the launch control and gearshift control of DCT.However,open loop control strategy is usually used to control the clutch pressure in engineering,which cannot fully consider the influence of the driving intention on the launch and gearshift.The closed loop PID control is generally used to compensatory control the clutch torque,which is difficult to adapt to the complex transient changes and aging characteristics of clutch friction coefficients.Therefore,in the paper,the research on intelligent control of launch and gearshift for DCT is carried out.The main research contents are as follows:(1)The modeling and optimization of the launch and gearshift of DCT.The dynamic model of powertrain and transmission system is established,and the optimal control problem of launch and gearshift are designed.The jerk,friction work,and launch or gearshift duration are taken as sub-objective functions,which are weighted and summed according to driving intention.Accordingly,the multi-objective optimization problem is transformed into a single-objective optimization problem.The constraints are designed to ensure that the launch and gearshift can be completed successfully while the optimization results can be more in line with the driving intention.The optimal control problem is transformed into a nonlinear programming problem by pseudo-spectral method,which is solved by sequence quadratic programming.The engine and clutch desired speed,which satisfy multi-objective comprehensive optimality,are obtained to lay the foundation for subsequent closed loop controller design.(2)The adaptive controller design of the launch and gearshift of DCT.A speed closed loop control strategy of DCT is proposed.The optimal desired speed obtained by pseudo-spectral optimization is called online by neural network and rule-based table.The synchronization of clutch acceleration is ensured by micro-slip control strategy at the moment of clutch speed synchronization.A clutch model,which can adapt to the variable friction coefficients,is established based on radial basis function neural network.An adaptive sliding mode controller is designed based on model-data hybrid driven,which can control the clutch pressure to track the clutch desired speed during launch and gearshift process.The engine controller is designed based on sliding mode control and PID control,which can adjust the throttle opening to track the engine desired speed during launch and gearshift process.The simulation verification of launch control and gearshift control of DCT is carried out.The results show that the proposed control method maintain good launch and gearshift quality at the time-variant acceleration pedal opening.The controller has high adaptive ability and strong robustness.The self-adaptation of the variable friction coefficients of clutch is achieved.(3)The optimization of the control parameters of DCT on cloud services.A strategy,which the control parameters of DCT is optimized and updated on cloud services,is proposed.The clutch torque change rate is estimated from the driving data based on the multi-sensor and multi-model fusion filter and moving average filter.Taking the clutch sliding speed and the estimated value of clutch torque change rate as the training data,the radial basis function neural network is trained offline based on improved genetic algorithm.The training results is updated to the adaptive clutch controller as the initial value of neural network.The simulation verification of the offline update of DCT control parameters is carried out.The results show that the update of DCT control parameters by cloud services can effectively improve the control performance of DCT when the friction coefficients are changed due to the aging of clutch.(4)The experiment of the launch and gearshift of DCT on the vehicle.The experiment scheme of the launch and gearshift of DCT on the vehicle is designed.The designed control programs are run in host computer,which communicates with the transmission control unit by CAN analyzer.The parameters of the proposed control strategy and control algorithm are calibrated,and the experiment of launch and gearshift is carried out in the vehicle.The results show that the proposed control method maintain the good launch and gearshift quality.The control strategy can obtain the desired speed that meets the driving intention.The control algorithm based on model-data hybrid driven can accurately track the desired speed.The self-adaptation of the variable clutch friction coefficients is realized,and the intelligent level of transmission control with DCT is improved.