Research On The Construction Of Truck Driving Cycles Based On Driving Style And Geographical Features Of Routes

According to the 2018 National Economic and Social Development Statistical Bulletin^[1],13.6862 million trucks accounted for 76.8%of China’s total freight,and truck fuel consumption accounted for 49.2%of the total oil consumption of automobiles.In the actual operation of the truck,there are two practical problems,one is the difficulty of matching the vehicle-road,and the other is the difficulty of matching the person-vehicle.Therefore,it is necessary to develop a more representative driving condition curve for drivers and driving routes as a support for the coordinated matching of human-vehicle-road.The traditional driving cycles construction method mainly considers the influence of traffic characteristics on vehicle speed without considering the differences in vehicle,driver and line characteristics,but the actual driving cycles will be affected by traffic flow,road structure and driver,and the uncertainty is large.In order to make the fuel consumption of the constructed driving cycles closer to the actual situation,this paper proposes a construction method of truck driving cycles based on driving style and geographical features of the route.Firstly,aiming at the problem that the driving cycles of trucks are greatly affected by the driver’s operation,a method of identifying the driving style of the driver for different driving cycles is proposed.According to the analysis of the driving route data,it is determined that the main driving areas of the trucks are in the high-speed and suburban areas,and the characteristic parameters of the heavy-duty trucks in the high-speed and suburban conditions are extracted respectively.The feature parameters are reduced by factor analysis,and the driving style is divided into three categories by the k-means clustering method,and a route driving database reflecting the driving style is established.Then,in view of the problem that the driving cycles of trucks are greatly affected by road characteristics,a method is proposed to divide the three-parameter state division of vehicle speed,acceleration and slope for the driving data of high-speed and suburban areas,and the Markov property of driving cycles is verified.The transition probabilities between states under different driving styles and working conditions were established,and a state transition probability matrix reflecting driving styles and geographical features was established.Aiming at the problem of low synthesis efficiency of Markov stochastic process,a method of designing multi-swarm genetic algorithm based on Markov chain to optimize the construction process of working conditions is proposed.The fitness function is used to redesign each operator according to the state transition probability matrix.According to the working condition type sequence of the line,the working condition curve is synthesized in sections,and finally the driving condition curve of the entire line is formed.The construction process based on the multi-swarm genetic algorithm and the working condition based on the genetic algorithm is compared,and the conclusion is obtained based on the multi-swarm genetic algorithm.It can effectively avoid the premature problem;compared the operating efficiency of using the multi-group genetic algorithm and the Markov random method,the running time of the two methods is 2.69h and 21.3h respectively,and the operating efficiency of the multi-group genetic algorithm combined driving cycles is improved.7.9 times.Finally,the rationality of the representative driving cycles of the line is verified from the characteristic parameter deviation.The maximum speed,average speed,standard deviation of the vehicle speed and the average relative deviation of the acceleration and deceleration of the synthetic driving cycles of different driving styles are all less than 5%.The absolute deviation of the addition and subtraction idling ratio of the aggressive synthetic driving cycle is the maximum 1.06%,the absolute deviation of the addition and subtraction idling ratio of the general synthetic driving cycle is the maximum 1.18%,and the addition and subtraction idling ratio of the mild synthetic driving cycle is at most 1.18%.The absolute deviation is a maximum of 0.99%.The relative deviation of road gradient-related characteristic parameters is less than 10%.The joint distribution errors of speed-acceleration under the synthetic conditions of different driving styles are:1.99%,1.06%,and 2.97%,respectively;the fuel consumption deviations are:0.13%,0.46%,and 2.21%,respectively.The validity of the driving cycles is verified from the matching optimization of driver-vehicle-road.The driving cycles under the three driving styles are input into the twin models of the five models,and the fuel consumption deviation between the driving cycles and the actual driving data is less than 2%.According to the fuel economy index,it is concluded that the mild driver driving model B is the best combination.The effectiveness of this method is verified by experimental data.