Starting Evaluation Of Pure Electric Bus Based On Driving Cycle

For a long time,the bus plays an extremely important role in the city’s public travel.Due to the task demand,the bus needs to start and stop frequently in the cycle,so the reasonable starting mode of the driver can not only effectively reduce the energy consumption of the vehicle,but also significantly improve the comfort and public safety level of the passengers.On the other hand,with the rapid popularization of pure electric bus in public transportation,it is possible to obtain the dynamic driving data of bus such as instantaneous energy consumption and acceleration and deceleration.Therefore,based on the actual driving data mining of pure electric bus,this paper constructs a highly interpretable driver starting evaluation model(1)The data preprocessing and starting process reconstruction of pure electric bus are carried out.Taking a certain type of pure electric bus as the research object,after preprocessing the daily operation data and Feature Engineering,the starting process and cycle conditions of pure electric bus are reconstructed,and the data subset representing the starting behavior is divided from all the driving data,which reduces the complexity of data mining analysis.(2)Starting style recognition and starting process motion characteristics analysis of pure electric bus driver.Firstly,the operation characteristics of pure electric bus in urban road network are analyzed from a macro perspective,and the energy consumption per unit mileage and the correlation analysis of energy consumption are carried out to select the driving characteristics that affect the operation economy.Then,by using the characteristic parameters,principal component analysis,K-means clustering and other methods,the driver’s starting style is divided into mild type,moderate type and radical type.Taking vehicle speed,acceleration,impact degree,specific power of motor vehicle and energy consumption per unit mileage during acceleration as key evaluation indexes,the comprehensive evaluation of riding safety and operation economy of three starting types is carried out.Finally,combined with the evaluation index,the optimization and acceleration suggestions of pure electric bus are given.(3)The start evaluation system of pure electric bus driver is constructed.Aiming at the key indicators of driver’s starting evaluation,this paper constructs a score card model by using logistic regression algorithm,realizes the integration of starting style classification and scoring,and solves the problem that the current driving style classification methods are difficult to describe its "degree" in the corresponding categories.Compared with the previous classification models,the process and results of model construction have higher interpretability.The accuracy of the score card model in the test set is 94.34%,and the AUC value is 0.99,which shows that the model has a good classification effect.Finally,the driver comprehensive starting evaluation system is constructed by using the model.The results of statistical analysis show that the evaluation system can effectively evaluate the rationality of drivers’ starting operation from the aspects of safety and economy,which has important guiding significance for strengthening the supervision of drivers’ behavior and improving the level of road traffic safety.