Development Of Driving Cycle For Representative Roads In Fuzhou

With the continuous development of Chinese economy,the national vehicle sales continued to grow,air pollution and energy shortages become urgent problems to be solved in China,especially in recent years there is a wide range of foggy weather,giving the Chinese atmospheric environment sounded the alarm.Driving cycle,also known as automotive test cycle,is to describe the vehicle speed-time curve,used to determine vehicle emissions,fuel consumption,new vehicle development and evaluation,is the core technology of the automotive industry.At present,Chinese automobile driving conditions in the study far behind Europe,the United States,Japan and other auto industry developed countries.In the automobile test cycle,the new European Driving Cycle adopted in China as a standard,the study shows that the driving cycle do not meet the driving characteristics of China's actual road,so the development of China's urban road driving conditions have a certain significance.This paper takes the light vehicle of Fuzhou as the research object,established the data acquisition route by referring to the frequency of the bus route of Fuzhou city,and sets up the corresponding data acquisition scheme according to the daily travel characteristics of Fuzhou residents,through the filter and the corresponding screening criteria of the test data for pretreatment.In driving data analysis,firstly,the micro-trip segmentation method is used to divide the experimental data and the 15 characteristic parameters of each micro-trip fragment are calculated.Secondly,the principal component analysis method is used to reduce the characteristic parameters of the micro-trip segments,and get the four principal components.Thirdly,the principal component score is used for K-means clustering analysis and the Silhouette function is divided into two categories as the final classification results.According to the classification results,the reasonableness of the classification results is verified by the speed-acceleration frequency distribution and the time distribution ratio of four driving states in the classification result.In the construction of automobile driving cycle,draw lessons from domestic and foreign scholars research results,this paper combines Markov with K-means clustering analysis results.Firstly,the data of the classification result is divided by four driving states.Secondly,the transition probability matrix of each state is calculated.Thirdly,the driving cycle is randomly synthesized according to the transition probability matrix.Finally,the candidate driving cycle is obtained by the six parameters.At the same time,this paper also introduces the traditional K-means clustering analysis method to build the driving cycle.By comparing the six special parameters,the speed-acceleration frequency distribution,the difference between the fuel consumption simulation results and the actual fuel consumption,to find the differences between the construction methods and the reasonable results.The Fuzhou driving cycle and foreign standard cycle were compared to analyze the difference between the two.According to the research results,using MATLAB GUI programming,try to develop the driving cycle build software.