Research Of Statistical Model Of Automobile Transportation Corporation Energy Consumption

Countrywide energy saving is more and more significant along withincreasing severity of energy crisis, and energy used by the industry of automobiletransportation is especially great, so energy saving in the industry of automobiletransportation is crucial to the healthy and persistent development of whole society.Statistical work of energy consumption of automobile transportation is in troubleon account of system innovation. In many enterprises, statistical data are feignedmostly because of no such departments, thus there are non-trustful statistical resultsof energy consumption in the whole industry, which is so troublesome forestablishment of policies of primary departments that it can not lead the industryaccurately.Aim at the facts, this paper statistic energy consumption with energyconsumption statistics models. This method fits the actuality of statistic inautomobile transportation industry by argumentation and study in this paper. Thepaper makes fuel wastage of passenger car as dependent variable y, as well asturnover volume, transport volume, vehicle's age, driving age, average velocity,actual passenger transport rate as independent variables x to establish energyconsumption statistics models.y = β₀ +β₁x₁+β₂x₂+β₃x₃+β₄x₄+β₅x₅+β₆x₆+β₇x₇+ε, where y isdependent variable, x₁ is turnover volume, x₂ is transportation volume, x₃vehicle age, x₄ is driving age, x₅ is average velocity, x₆ is actual passengertransport rate, x₇ is the distance covered, β₀, β₁…,β₇ are regression parameters,ε is random variable, and ε N (0,σ²).At first, sample size are confirmed by the formula 22222σσNtnNt= ?x +, where n issample size, N is total unit, ? xis sampling limit error of sample average, σ 2istotal variance, t is probability. Then, using the formulanNnn1 0/0= +,2n 0 = ( μr YaS) based on total quantities of different vehicle models, where N iscalendar days of a sort of vehicle in a investigating month, Y is mean of fuelwastage of sample vehicle each month, S 2is variance of sample vehicle eachmonth, μ a is quantile with two sides of standard normal school, its value is 3.17with confidence measure 99% in this paper, r is allowable maximal relative error,its value chooses 0.03, selected sample data are put into the formula to gettheoretical sampling days.Secondly, data of sample vehicle run in period of sampling days are collectedto put into the energy consumption statistics model to get β 0, β 1,β7 whichare coefficients of the model, namely confirming the model. Statistic index inenergy consumption statistics model can be reduced by optimizing the model withmethod of gradual regression, and the final statistics model optimizedis y? = ?64.12+0.001x1 ?6.310x3?18.017x4+0.235x7. The result calculatedthrough the optimized model with collected data of all vehicles run in the month is1292381.74 L, and the error compared with actual fuel wastage 1215749L in themonth is 6.3%, however, the result calculated by former sampling method, namelymultiplying average fuel wastage of sample vehicle in period of investigation, daysof a month and quantities of the vehicles is 1035156.86L, the error compared withactual fuel wastage 1215749L is 14.9%, thus it can be seen statistic precision onfuel consumption of vehicle transportation enterprise with energy consumptionstatistics model excels traditional methods.It only needs a part of vehicles in the course of statistics of energyconsumption with energy consumption statistics models as well as several targetsdemanded in the model operation, then eliciting the model and putting collecteddata into it, total energy wastage can be calculated finally. It can make statisticalresults more exact, the work more effectual, and the workload reduced, the datamemorized, picked up and exported automatically, thus statistical tables designedbeforehand can be made, if the FCMS03 fuel management system mentioned inthis paper can be used.Software for statistic is exploited combined with the method in the paper,which can pick up data the model operation needed to obtain total energy wastageultimately with energy consumption statistics models. The software can stillcalculate detailed statistical data of vehicle and show detailedly run circumstanceof different vehicles.The study of this paper would make the energy sources statistic working of thewhole industry of vehicle transportation easy and high efficient.If do statisticworking of the whole industry by this method, we only need to select severaltypical transportation corporation in our country(if fuel management systemsuch as FCMS03 can be installed, it would greatly save Stat. Cost and improvestatistic precision),and then collect the vehicle running data of each corporation(theindex of enregy consumption statistics model), then make out the statistic modelof every province, collect data of every province transportation vehicle, in the end,to calculate the total fuel wastage of every province transportation vehicle. Basedon turnover volume, we can calculate the single wastage of varies of vehicle type inevery province.