Research On Predictive Gear Decision-making Of Electric Vehicle

With the continuous expansion of the electric vehicle market,intelligent and energy-saving electric vehicles have gradually become the mainstream direction of research and development.Similarly,with the development of electric vehicles,multigear research of electric vehicles also rises.As the control performance of the drive motor is far better than that of the engine,light-duty electric vehicles with centralized drive usually adopt two driving modes: reducer direct drive or two-gear automatic transmission.Compared to direct-drive electric cars,electric cars equipped with automatic transmissions can improve the efficiency of the driving motor and the utilization of battery energy.Gear decision is the core problem of EV automatic transmission affecting vehicle performance.It is very important to make a reasonable gear decision method for the improvement of vehicle performance.In this paper,a pure electric passenger car equipped with a two-gear parallel shaft automatic transmission is taken as the research object.In view of the poor applicability of static shift strategy and the shifting strategy solved by dynamic programming cannot be solved in real time,this paper aims at improving vehicle economy and uses Cruise-Matlab co-simulation platform to systematically study the shifting strategy of two-gear electric vehicles.Finally,an online optimization strategy of vehicle running gear was developed,which completed the real-time solution of vehicle running gear and the optimization of vehicle performance.The main research contents are as follows:(1)Build an electric vehicle simulation model and select the optimal transmission ratio of the transmission system.In this paper,an electric passenger car equipped with a two-gear parallel shaft automatic transmission is selected as the research object.The longitudinal dynamics simulation model of the vehicle is established by AVL-Cruise,the shift control model is established by Matlab-Simulink,and the Cruise-Matlab cosimulation platform is built.At the same time,aiming at the minimum energy consumption under the fixed cycle condition,the greedy algorithm is used to solve the optimal transmission ratio of different gear positions of the vehicle.(2)Solve the static shift point and dynamic optimal shift point of the vehicle.According to the external characteristic curve and efficiency MAP curve of the vehicle drive motor and other component characteristics,the optimal power shift MAP and the optimal economic shift MAP of the vehicle were solved.Based on the multi-objective optimization problem,the static shift MAP which could maximize the vehicle power performance under the condition of low velocity and maximize the vehicle economy under the condition of medium and large velocity was formulated.When the global driving conditions are known,the dynamic programming algorithm is used to discretize the vehicle cycle conditions,making the solution of the vehicle shift point become a typical multi-stage decision-making problem,and finally the dynamic optimal solution of the vehicle shift point is obtained.(3)Build a vehicle driving state prediction model based on Markov principle.Related parameters of vehicle running state,such as vehicle speed,acceleration and jerk were collected,the transfer matrix of Markov prediction model was obtained by using the training set data,and the real-time prediction of vehicle speed and acceleration was finally completed,and the accuracy of the prediction model was verified on the simulation platform.At the same time,considering the real-time performance of the prediction model,the Markov transfer matrix is optimized,and the prediction time domain is selected by analyzing the prediction results.(4)The model predictive control algorithm is used to solve the online shifting points of vehicles.Considering the influence of vehicle speed and accelerator pedal opening on the operating point of the driving motor,based on the prediction of vehicle driving parameters,the nonlinear equation of state of shifting strategy was established,and the Levenberg-Marquardt method was used to solve the nonlinear equation,and the real-time optimization of shifting point was completed.In addition,genetic algorithm is used to optimize the coefficients of each cost function in the objective function,so as to improve the optimization effect of the online gear solving algorithm.(5)Simulation verification and comparison.The simulation platform was used to compare the energy consumption of vehicles and the reduction of battery SOC in cycle conditions with different shifting strategies.Meanwhile,the influence of each cost function in the objective function on the vehicle performance was analyzed.The effectiveness of the online solution method of shifting point based on the model predictive control algorithm was verified.