Research On Gearshift Of An Electric Vehicle With Dual Clutch Transmission Based On Multi-objective Optimization

With the continuous development of electrification in automobile industry,the requirement for performance of electric vehicles(EVs)was getting higher and more diverse.To satisfy with the demand of economy and power performance in EVs,many researchers have focus on the application researches of multi-speed transmission(MST)in EVs.Compared with other MST,dual-clutch transmission(DCT)with the advantages of lower cost,higher transmission efficiency and more compact structure,has become the objective in many application researches of MST in EVs.Currently,the researches mainly concentrated on the shift schedule,torque estimation problems,shift control strategy and so on.This paper selected a passenger car equipped with two-speed DCT as the research object.Based on the modelling of transmission system and the establishment of shift schedule,the researches on the multi-objective optimization of shift trajectories and the application researches of Pareto solution set in shift control strategy were conducted.And the hardware-in-the-loop experiments were conducted under the NEDC cycle condition to verify the effectiveness of the control strategy.Firstly,in this paper,aimed at a two-speed DCT in EVs,the shift process and working principle was analyzed and explained.The shift process was divided into three phase,which were torque phase,power reflux phase and inertial phase.The analysis of the powerflow in each phase and dynamic characteristic of each component was conducted.With the comprehensive consideration of torsional dynamic characteristics of the components in the powertrain,a 11-DOF dynamic model was build.On the basic of the above researches,the dynamic shift schedule was formulated according to the max acceleration principle,while the economic shift schedule was formulated according to the max motor efficiency principle.Secondly,based on the researches of the shift process of the two-speed DCT in EVs,this paper emphatically concentrated on the researches of the multi-objective optimization in shift trajectory planning problem.The friction work,jerk and duration,which were associated with the evaluating indicators of shift quality,were selected as the optimization objectives.The rotational speeds of the motor and tire,and the torque of the motor,Clutch 1 and Clutch 2 were selected as the state variables.The derivatives of the torque of motor,Clutch 1 and Clutch 2 were selected as the control variables.With the comprehensive consideration of the path constraints,point constraints and link constraints between phases,the shift trajectory planning problem was transferred into a multi-objective optimization problem.With the iterations of Legendre pseudospectral method(LPM),the optimization results were presented in the form of Pareto solution set.Under the discussion of Pareto solution set and sample analysis,the relationships of the evaluating indicators and their formation mechanism were explained.Then,the research on the application of Pareto solution set in the shift control strategy was conducted.Aimed at the application of multi-objective optimization on shift trajectories in the real vehicle,a novel control strategy architecture,which applied the Pareto solution set in the real shift control according to the judgement on driver’s intention,was proposed to reflect the requirements of the driver.This control architecture was composed of shift trajectory planning module,feedforward and feedback control module,state estimation module and actuator module.The fuzzy logic controller is applied to identify the driver’s intention,so as to judge the driver’s demand for shift quality.According to the different requirement of shift quality,the corresponding optimal trajectories of motor torque,Clutch 1 torque and Clutch 2 torque were choose as the input of the feedforward controller,while the corresponding optimal trajectory of rotational speed difference was selected as the target trajectory in the feedback controller to realize the shift control with the feedforward and feedback module.Finally,to verify the feasibility and effectiveness of the control architecture,the hardware-in-the-loop(HIL)experiment was carried out on the Simulink real-time simulation platform.The aforementioned shift control architecture was transferred into an executable shift logic controller,while the 11-DOF dynamic model was transferred into a real-time simulation model.The HIL experiments was conducted on the sample points obtained from the above multi-objective optimization.With the comparison between optimal results and HIL experiment results,the feasibility of the HIL experiment was verified.On this basis,the shift control architecture was applied in a NEDC cycle condition based HIL experiment to verified its effectiveness.