Research On Longitudinal Control Of Unmanned Dual-motor Electric Vehicle Based On Traffic Sign Recognition

With the impact of the new scientific and technological revolution,Chinese automobile industry has ushered in great changes.Because dual-motor electric vehicle has good performance,the related configuration and control strategies have been studied.Vehicle intelligence also promotes the development of autonomous driving.The information fusion obtained by sensors and the internal information of the vehicle is the basis of autonomous driving.If the traffic signs information and road slope information from environment can be applied to the vehicle control strategy,which can ensure the safe operation and improve the economy of the vehicle better.Firstly,based on the analysis of the common configuration of pure electric vehicles,this paper selects the front and rear axle dual-motor configuration as the research object.According to the cost characteristics,efficiency characteristics and loss characteristics of the motor,the motor type and the parameter matching of the relevant components of the power system are completed.The motor model,battery model,driver model and vehicle longitudinal dynamics model are developed in the MATLAB/Simulink platform,and AVL＿Cruise software and MATLAB software are used to carry out parameter matching verification.Secondly,to ensure coordinated control and reduce energy consumption of the dual-motor electric vehicle,the driving and braking control strategy of the front and rear axle dual-motor pure electric vehicle is developed.Based on the on-board camera,the motor’s optimal working efficiency is taken as the object,and the dual-motor electric vehicle is divided into three working modes in drive level.Especially in the dual-motor drive,the vehicle’s driving efficiency is taken as the objective function,and the adaptive particle swarm optimization algorithm is used to obtain the optimal front and rear wheel split torque to achieve the optimal energy consumption under autonomous driving.On the basis of ensuring braking stability,an adaptive regenerative braking energy recovery strategy based on multi-information fusion under on-board camera,driving speed,road slope and battery SOC is established.The strategy has automatic adjustment ability under different braking modes,aiming to make the motor work in the high efficiency range as far as possible,which provides the possibility for realizing the high efficiency recovery of braking energy.Thirdly,in order to further integrate traffic information into the longitudinal control of dual-motor electric vehicles,Multi-feature fusion algorithm and Fast Normalized Cross Correlation(FNCC)algorithm are used to recognize and match numbers in the speed limit traffic signs.The elevation information of a road near Shandong University of Technology is obtained by GPS,and the slope prediction is carried out by the nonlinear autoregressive neural network with external input(NARX).The real-time adjustment of vehicle power under information fusion is realized by combing the speed limit traffic signs,the feedback vehicle speed,and the road slope information.Then,the development of speed control strategy under autonomous driving is completed.Finally,the co-simulation platform is built on MATLAB platform and Prescan platform.Finally,in order to verify the effectiveness of the proposed driving and braking strategies and achieve the dynamic goal of unmanned drive speed control and the economic goal of reducing energy consumption of pure electric vehicles,three conditions of acceleration,deceleration and acceleration-deceleration on the road are designed.Then the variable controlled method is adopted to set the comparison strategy for simulation and verification.According to the analysis of the results obtained from the MATLAB/Prescan co-simulation platform based on the multi-information fusion,it can be seen that the unmanned speed control strategy proposed in this paper has good effect.The control strategy can significantly improve the overall working efficiency of the motor and reduce the energy consumption.Besides,the energy recovery of pure electric vehicles is greatly improved.The results provide a reference for control optimization of pure electric vehicle.