Research On A-ECMS Energy Management Control For Hybrid Electric Vehicle Based On Road Condition Prediction

The growing energy crisis and environmental crisis in the 21 st century forced people to shift their development direction of vehicle to pure electric vehicles.However,due to the fact that the technology maturity is not good enough,the hybrid electric vehicle is the main object of the current research.The core of the hybrid car's ability to improve the high fuel consumption and high emission of traditional vehicles lies in the design and control of the system,and the control strategy is the key to the realization of multi-power source energy management for hybrid electric vehicles.In this article,parallel hybrid electric vehicle is the research objectives,and realizes the adaptive energy management control of short-term road conditions in the future through algorithms.The main research results are as follows:First,Matlab/Simulink simulation model of a hybrid electric vehicle was built in this paper,analysis and resolution the shortcomings of the gray prediction GM(1,1)algorithm.Three different types of standard conditions are used at the same time to verifies that the short-term vehicle speed prediction in the future can be accurately achieved with the improved grey prediction GM(1,1)model and can be adapted to different driving conditions.Then,Kalman filter algorithm and the least square method with forgetting factor was discussed based on the longitudinal dynamics of the vehicle,and then set up road conditions and parameters of the main components in Carsim,validity and accuracy of both road gradient estimate algorithm is verified co-simulation with Matlab / Simulink.After a comprehensive analysis,it was decided to use Kalman filter algorithm based on the longitudinal dynamics of the vehicle as the main algorithm for the road gradient estimation.Finally,after analysis of the geometric characteristics of the road surface,proposed to fit the slope estimated by Kalman filter with history data,and then predict short-term road grade in the future by the size of radius of curvature through road fitting,simulation analysis also demonstrated the feasibility of prediction methods.Second,Pontryagin's minimum principle was introduced leads to the principle of equivalent consumption minimization strategy(ECMS),and simulation results shows that main problem of ECMS is that the fixed equivalent factor in the whole condition,so that trajectory of SOC cannot be optimized and beneficial control effect cannot be obtained too.Three kinds of adaptive equivalent consumption minimization strategy(A-ECMS)algorithms were introduced to solve problems of ECMS.Taking the most commonly used SOC-based feedback A-ECMS as an example,the simulation results of SOC-based feedback A-ECMS are compared to ECMS under the three conditions of WVUCITY?WVUSUB and WVUINTER.It is found that the SOC-based feedback A-ECMS significantly improves the control trajectory of SOC,improves the fuel economy at the same time,and also has certain optimization in the demand torque distribution.Results show that trajectory of SOC improves a lot as well as fuel economy and distribution of demanded torque based on SOC-based feedback A-ECMS.Finally,VAIL2 NREL standard condition with road grade was chosen to verify the effectiveness of SOC-based feedback A-ECMS.But analysis results show that the control strategy could not reach control effect that we expected.Then,a road condition prediction-based A-ECMS was proposed aimed at real-time equivalent factor adjustment by short-term road conditions.Then,the simulation results under the same condition with SOC-based feedback A-ECMS are compared.The results show that the proposed algorithm not only achieves better results in optimizing trajectory of SOC and distribution of demanded torque,but also improves fuel economy by 6.54%.