The Study Of Control Strategy For PHEV Based On Traveling Mileage Prediction

With the increasing market demand for new energy vehicles,the major car manufacturers are putting the target into new energy cars.As the transition models between traditional cars and pure electric vehicles,Plug-in hybrid electric vehicles(PHEV)have the both advantages of pure electric vehicles and traditional cars,which make themselves more favored by consumers.And for PHEV,energy management strategy is the core technology.At present,the energy management strategy commonly used on PHEV is the Charge Depleting-Charge Sustaining(CD-CS)strategy,which uses electric power to drive the vehicle and drives the vehicle through the engine after the battery runs out.This strategy can ensure the priority of the electricity's using,but fails to take the energy distribution of the overall driving conditions into account.once the mileage exceeds the pure electric mileage,it can only use the engine to drive in the excess part.So in this case it can not guarantee that the engine works in high efficient work area.Therefore,CD-CS energy management strategy can not play PHEV's energy-saving potential.A reasonable energy management strategy should consider the whole day's driving conditions so that not only the full use of electricity can be ensured,but also the working conditions of the engine are taken into account,and the fuel consumption will be minimized further.Therefore,this paper takes the actual travel data as the research basis,processes the collected data and obtains the travel route.The on-line identification of the travel route is realized by the classification algorithm.The analysis on the energy distribution method is based on the simulation results of the rule control strategy,and the research on the energy management strategy of PHEV is based on this energy distribution method.So,the energy management strategy developed in this paper can be more adapted to the actual driving conditions and can improve fuel efficiency better.For the collection of travel data,first,the car owners being tested are expected to use handheld GPS devices to collect their daily travel data,then the daily data need to be preprocessed,including data filtering and transforming and the travel route classifying.For the recognition of the route,the naive Bayesian algorithm is applied in this paper to realize the current route on-line identifying with the information of the number of the current travel section and the current time period,and to make the judgment for the subsequent control strategy selection.For the analysis of energy distribution,this paper analyzes the use of electricity under different energy management strategies for fixed working conditions,and establishes the basic principles of power consumption.Then,the relationship between the power consumption and the characteristic parameters of the working condition is obtained by simulation.The power consumption and the characteristic parameters of the different operating mode are analyzed with regression analysis.Combined with the basic principle of power consumption,the theoretical SOC curve under arbitrary condition is obtained for the subsequent.Finally,for the study of energy management strategies,this paper develops a fuzzy control strategy under the condition of determining the working conditions,allocates the torque of the engine and the motor while following the theoretical SOC curve,and simulates its fuel saving effect.For uncertain conditions of frequent travel routes,this paper uses the neural network to predict the mileage in the time series,and the fuzzy control algorithm is used to adjust the control parameters by using the prediction results.The whole power distribution is adjusted according to the basic principle of electricity consumption.The effect of the algorithm is verified by simulation.