Research On Energy Management Strategy Of Extended-range Electric Vehicle Based On Trip Characteristics

With the coming of industry 4.0 era,people pay more and more attention to resource shortage and environmental pollution.The traditional internal combustion engine vehicle consumes a lot of oil,and the exhaust emission of the vehicle has become an important source of air pollution.Many countries are vigorously advocating the development of new energy vehicles and related infrastructure construction,which can not only reduce the dependence on oil,but also greatly reduce urban pollution.It is the general trend of automobile development.Although pure electric vehicle has the advantages of no emission and no pollution,the development of pure electric vehicle is limited due to the energy density of power battery.Extended range electric vehicle is a type of new energy vehicle,which can solve the problem of relatively short running mileage of pure electric vehicle.It has been widely studied due to simple structure and high fuel saving.This paper build the complete vehicle model and control strategy based on Cruise.The simulation results show that the strategy is reasonable.Based on the trip data collected by the driver,the characteristic analysis is carried out to extract the typical driving cycles,and the control parameters of the control strategy are optimized based on the driving cycles,which improves the fuel economy and reduces emissions of the vehicle.The main research contents are as follows:Get the tester's travel data using GPS devices,preprocess the unreasonable segments in the driving data,and use wavelet de-noising to denoise the preprocessed trip data.The characteristic parameters are extracted from the kinematic segment of the driving data,and comparedwith the foreign standard driving cycles and the Chinese driving cycle to verify the rationality of the collected driving data.The correlation analysis of the extracted feature parameters is carried out to remove some feature parameters with high correlation.Principal component analysis is used to reduce the dimension of characteristic parameters,and then ISODATA algorithm is used to cluster the working condition segment samples,and several typical driving cycles are obtained.According to the results of principal component analysis,the characteristic parameters are selected as the parameters of cycle recognition,and the extreme learning machine is used to identify the condition,and the recognition effect is good.The working mode of the vehicle is analyzed,and the simulation model is built in Cruise software,the simulation results meet the technical requirements.Three vehicle control strategies of "thermostat" control strategy,power following control strategy and multi-points control strategy are developed and their characteristics are analyzed to verify the rationality of the strategy.The simulation results of the three strategies are compared and analyzed in terms of the the engine and power battery's working state,power,economy and emission.Finally,a multi-points control strategy based on trip characteristics is built.For each typical driving cycle,the equivalent fuel consumption and emissions are taken as the objective functions,and the dynamic performance is taken as the constraints.The adaptive simulated annealing algorithm is used to optimize the control parameters and engine working points.Using these optimized control parameters,a control strategy based on cycle recognition is developed.Compared with the multi-points control strategy without cycle recognition,the economy and emission of the vehicle are greatly improved.