Study On Energy Management Optimization Strategy Of Hybrid Heavy-Duty Truck Based On Driving Cycle Prediction

Due to the existence of dual power sources,hybrid vehicles can flexibly switch their working modes under different working requirements.Compared with the working characteristics of traditional engines,hybrid vehicles have unique advantages in improving the economy of the whole vehicle.Traditional heavy-duty trucks consume more than 60% of road energies.With the further popularization of hybrid vehicles,the research on the application of hybrid power to improve the economy of heavy-duty trucks is also of great significance.This thesis studies the energy management strategy of P2 single axle parallel heavy truck under complex working conditions.According to the vehicle configuration and transmission system characteristics of the research object,this thesis establishes the energy consumption simulation model of hybrid heavy truck.Firstly,different working modes of P2 hybrid system are introduced to prepare for the subsequent establishment of rule-based energy management strategy.Secondly,according to the basic principles of different key components of the vehicle,the vehicle model based on Simulink is built,which contains of engine model,motor model,gearbox model and so on.According to the dynamic programming algorithm,the engine mode switching rules are optimized,and on this basis,the control strategy optimized by the dynamic programming algorithm is formulated.The vehicle dynamic performance and speed following test are verified under specific drving cycle.The off-line simulation results confirm the reliability of the controlled object model and provide the fuel consumption reference for the following tests.In order to improve the adaptability of energy management strategy under working conditions,the grey prediction model based on grey system theory is used to predict the vehicle speed under short-term working conditions in the future.The vehicle speed information of 5 seconds in the future is predicted by collecting the vehicle’s 10 s historical speed data,and the ergodic iterative method is used to optimize the adjacent weight coefficient in the prediction model.In this thesis,the standard working conditions suitable for heavy vehicles are selected to build a typical working condition database for pattern recognition,and 15 characteristic parameters representing the working conditions are preliminarily selected.The dimensionality of the parameters is reduced by principal component analysis and correlation coefficient method,and the classification results of hierarchical clustering method are used to compare and analyze the characteristics of the two dimensionality reduction methods.Therefore,seven key parameters are selected as the basis for working cycle classification and identification,and the typical working cycles are divided into six categories.The recognition model based on BP neural network is established and compared with the support vector machine model.The two recognition models are trained with a large number of driving cycle segment samples respectively.It is concluded that the average accuracy of the recognition model based on BP neural network is 88.04%,which is higher than the average accuracy of 61.37% of the support vector machine model.Therefore,the former is selected for subsequent research.According to the principle of Equivalent Consumption Minimization Strategy,a real-time energy management strategy model is established.For each typical driving cycle,the adjacent weight coefficient,equivalent factor and penalty factor in the control strategy model are optimized by ergodic iteration and particle swarm optimization algorithm,and six sets of control parameters suitable for different driving cycles are obtained.The built energy management strategy is tested by d SPACE real-time online test platform.The results show that compared with the traditional constant-parameter ECMS strategy,the adaptive energy management strategy based on driving cycle prediction proposed in this thesis improves the fuel economy by 0.27%,2.39% and1.22% respectively under China World Transient Vehicle Cycle(C-WTVC),China Heavy-duty Commercial Vehicle Test Cycle Trunk(CHTC-HT)and composite driving cycle,which verifies the effectiveness and real-time performance of the strategy.