Research On Energy Management Strategy Of Four-wheel Drive PHEV Based On Multi-source Information Fusion

In recent decades,energy shortage and environmental pollution problems have plagued the development of China’s automobile industry from the beginning to the end,and energy saving and environmental protection have become important themes in the development of the current automobile industry.Plug-in Hybrid Electric Vehicles(PHEV)has become the most suitable new energy model for development under the current technology level because of its advantages of both pure electric and hybrid electric vehicles(HEV).Therefore,the development of a reasonable and efficient energy management control strategy has become a hot topic in the current research of plug-in hybrid electric vehicles.This paper takes Four-wheel Drive Plug-in Hybrid Electric Vehicle(4WD PHEV)as the research object and focuses on its whole-vehicle matching modeling and energy management strategy,specifically from the following aspects.First,the plug-in two-axis drive hybrid system configuration is determined,its power system characteristics are analyzed and eight operating modes are reasonably classified;based on the vehicle design requirements,the core power components are matched;and the4WD PHEV model is built based on the MATLAB virtual simulation platform,which provides the basis for the subsequent research on energy management control strategies.Secondly,the rule-based energy management strategy is developed based on MATLAB/Simulink/Stateflow,and the correctness of the proposed 4WD PHEV model and the effectiveness of the rule strategy are verified;to further improve the working condition adaptability of the rule strategy under different driving conditions,adaptive weights are used based on a variety of international typical working condition combinations,with the overall vehicle driving cost as the objective function The rule threshold parameters are optimized by particle swarm algorithm.After the optimization,the fuel economy and service condition adaptability of the vehicle are further improved,and the energy consumption of the vehicle is reduced by 7.21%under the combination of service conditions.Then,a Genetic Algorithm(GA)optimized BP neural network for vehicle speed prediction considering multi-factor feature information fusion of longitudinal driving intention D_T,historical vehicle speed V_T,and relative motion characteristics M_T of the workshop is proposed.Based on the fuzzy control principle,the driver’s driving intention D_T is accurately and effectively identified;the relative vehicle speed V_M and relative vehicle distance D_M with the vehicle in front are obtained using the millimeter wave radar sensor.A back-propagation neural network speed prediction model optimized by genetic algorithm is established,and the above-mentioned multi-source information is fused as the input of the prediction model to obtain the speed prediction results.In order to further verify the effectiveness and accuracy of the proposed speed prediction method,different prediction time domains are selected for comparison with the single feature speed prediction method which only considers the historical speed.The prediction results show that the prediction accuracy is improved by 12.95%,22.20%and 8.29%in different prediction time domains of 5s,10s and 20s,respectively,which provides an effective vehicle speed prediction model for the following research on energy management strategy.Finally,based on the speed prediction results,the vehicle demand torque in the future finite time domain is obtained indirectly.Combined with Model predictive control(MPC)framework,Dynamic programming(DP)algorithm is chosen as the solver for rolling optimization in the prediction time domain.And develop a SOC reference trajectory method based on mileage traveled,establish an energy management strategy based on Multi-source information fusion(MIF).The simulation results show that the optimization effect of local optimum and global approximate optimum in the predicted time domain is achieved under this strategy.Compared with the rule strategy,vehicle economy improved by 25.11%,which effectively reduces the fuel consumption of the whole vehicle and verifies the effectiveness of the MIF strategy.