Multi Factor Analysis Of Heavy Truck Fuel Consumption Based On Driving Data Before And After Covid-19

At present,heavy-duty diesel trucks have become the main means of transportation for medium and long-distance cargo transportation in China’s logistics enterprises.According to statistics,the fuel consumption cost of heavy diesel trucks accounts for about 30% of the total transportation cost.In the actual driving process,how to adopt economic and green driving mode is the key factor to determine the transportation cost.Scientific optimization of the fuel consumption of heavy trucks can not only improve the service life of vehicles,promote the cost reduction and efficiency increase of small and medium-sized logistics enterprises,but also help to alleviate the current environmental and energy problems.In order to explore the various factors affecting the fuel consumption of heavy trucks before and after the occurrence of COVID-19,the paper takes Dingchen International Logistics Co.,Ltd.as an example,and obtains the natural driving data of 34 heavy trucks from September2019 to April 2020 through the company’s internal intelligent vehicle supervision system,including the fuel consumption of 100 kilometers,the running status of the whole vehicle,the micro driving behavior characteristics and the external environment.Considering that the time span of the data covers the two stages before the occurrence of the COVID-19 and the initial outbreak of the COVID-19,based on the analysis of the influencing factors of heavy truck fuel consumption,this study further discusses the change rules of driving behavior characteristics and heavy truck fuel consumption characteristics under similar public health emergencies.Finally,according to the research results,practical and feasible differentiated fuel consumption control strategies are proposed for different social subjects.The details are as follows:Firstly,focusing on the normalization period before the outbreak and the initial stage of the outbreak,the paper uses the methods of general linear regression,Pearson chi square test and support vector regression to test the influence of single factors on the fuel consumption of heavy trucks one by one.It is found that there are differences between the normalization period and the initial stage of the outbreak,the quantitative influence law is obtained.Secondly,this paper discusses the changes of various factors on the efficiency of fuel consumption in the multi factor coupling situation.Using disordered multi classification logistic,CART decision tree and random forest algorithm for analysis,it is found that the factors with significant changes under the impact of the epidemic are idle speed,average speed and altitude difference,while whether it is a holiday,economic speed,parking,neutral taxiing and mileage are less significant before and after the COVID-19.Thirdly,based on the above analysis results,the importance of the influencing factors is calibrated based on the extreme gradient lifting tree XGBoost,the action mechanism of the important factors is visualized with the help of Shapley value method,and the factor analysis method is used to reduce the dimension of the extracted factors.On the basis of verifying the effectiveness of the factors,a composite index that can comprehensively reflect the fuel consumption level and driving behavior characteristics is constructed to evaluate the fuel consumption.Finally,based on the conclusions of the above objective analysis,combined with the subjective questionnaire survey,this paper analyzes the differences of fuel-saving driving behavior of heavy truck drivers before and after the COVID-19;According to the results of subjective and objective comparative analysis,this paper puts forward differentiated management and control strategies to reduce costs and increase efficiency for different subjects such as government departments,third-party logistics enterprises and drivers.