Research On Eco-driving Evaluation Method Of Bus Drivers Based On Big Data

Under the background of the current "dual carbon" strategy,the transportation industry,as a major energy consumer and polluter,must make contributions to achieving this goal.Vehicle energy consumption is the result of the combined action of drivers,vehicles,and scenarios.Therefore,reducing energy consumption can be solved from the perspectives of vehicles,scenarios,and drivers.But it is limited by the fact that traditional fuel-saving technologies have almost reached the limit of fuel-saving potential,urban congestion that is difficult to change in the short term,and slower market penetration speed in alpine regions,improving driving behavior is a feasible way to achieve energy saving and emission reduction in the short term.Public transportation is an important part of the transportation industry.The promotion of ecodriving technology in public transportation enterprises will not only achieve good social benefits but also reduce energy costs,resulting in significant economic benefits.Based on the actual operation data of pure electric buses,and considering the influence of traffic conditions,ambient temperature,and passenger load on energy consumption,this paper proposes an economic evaluation method of bus driver driving behavior with energy consumption as a single evaluation index for quantitative evaluation the driver’s eco-driving level.And by analyzing the driving behavior parameters,it can identify the driver’s bad driving behavior and guide to improve the driving behavior.On the one hand,the evaluation and analysis results can be fed back to the driver to help the development of eco-driving behavior and the long-term maintenance of the energy-saving effect,and on the other hand,it can be fed back to the operating company as an evaluation basis for the incentive mechanism.Specifically,under the general framework of the economic evaluation method of driving behavior proposed in this paper,the influence of traffic conditions,ambient temperature,and passenger load on energy consumption is analyzed separately,and based on the analysis results,the influence of these three factors is excluded.The economic evaluation method of driving behavior is established.The main work and innovations of this paper include:First,a traffic conditions recognition method based on low-frequency data is proposed.Given the low frequency of the data,spline interpolation is proposed to interpolate the vehicle speed data to restore some vehicle speed and acceleration information;Considering the particularity of bus operation,it is proposed to divide short trips based on stations to avoid misjudgment of traffic conditions due to the uncertainty of passengers getting on and off the bus;By comparing the consistency of the characteristic parameters of the interpolated data and the characteristic parameters of the high-frequency data,the average speed,the deceleration ratio,and the idle speed ratio are selected as the traffic conditions evaluation indicators,and the traffic condition is divided based on K-means clustering.The final result shows that the above three characteristic parameters can distinguish the difference in energy consumption due to different traffic conditions while distinguishing the traffic state well.Secondly,the influence of ambient temperature on vehicle energy consumption is analyzed.From the two aspects of the impact of ambient temperature on battery performance and air conditioning accessories,the temperature is divided into four temperature ranges(-∞,-10℃]、(-10℃,0℃]、(0℃,26℃]、(26℃,+∞).And based on the Mann-Whitney rank-sum test,it is proved that the difference in energy consumption in different temperature ranges is significant.The proposed method of dividing the temperature range can not only avoid the problem of the insufficient number of samples at the same temperature value caused by the temperature value as the sample dividing standard but also ensure that the temperature has a similar degree of influence on energy consumption in the same temperature range.Third,the time-varying characteristics of driving energy consumption due to changes in passenger load are explored.During the operation of buses,the change of passenger load has obvious regularity.Therefore,to analyze the influence of passenger load on energy consumption,the concept of time-varying characteristics of bus energy consumption is proposed for the first time.Based on historical data,this paper analyzes the difference in energy consumption caused by changes in passenger load during different periods of the day on weekdays and holidays and depicts the time-varying characteristics of driving energy consumption caused by passenger load based on Gaussian function and polynomial function respectively.Finally,based on the above research work,an eco-driving scoring method is proposed.The one-year travel energy consumption of 19 4-way buses in Changchun is converted into eco-driving scores.The results show that the driving behavior economic evaluation method proposed in this paper can distinguish drivers with different driving levels very well,and the score depends only on the driver’s economic driving level and is not affected by traffic conditions,ambient temperature,and passenger load.Meanwhile,the differences in driving behavior parameters of drivers of different levels are analyzed,and the analysis results can be used as the basis for personalized guidance for different drivers,to help drivers improve bad driving behavior and improve ecodriving ability.Based on vehicle big data,this paper takes the three main factors that have a significant impact on bus energy consumption: traffic conditions,ambient temperature,and passenger load as the research objects,and quantitatively analyzes the impact of traffic conditions,ambient temperature,and passenger load on bus energy consumption,realizes the decoupling analysis of the influence of the above factors on energy consumption,and establishes a reasonable eco-driving scoring method.The research work of this paper lays the foundation for the promotion and application of eco-driving in the context of big data,and also provides ideas for exploring the new application value of vehicle big data and new modes of application development.