Construction The Vehicle Self-driving Cycle Based On The Restored Road Conditions

With the development of economy,roads have been reconstructed all over our country,car ownership has also increased significantly.Resulting in traffic conditions are very different compared with before,which makes our original use of European driving cycle can not well reflect the kinematic characteristics of China’s car driving.Therefore,combined with the actual situation of China’s roads to establish the corresponding vehicle driving cycle is also on the agenda.In addition,China has 9.6 million square kilometers of land area.For the cities that are far apart,their economic development level,climate and other natural conditions as well as traffic,road conditions are also very different,thus making the influence of these factors between the automobile driving conditions of each city also exist obvious differences.Therefore,it is important to build the vehicle driving conditions based on the city’s own road conditions.This paper constructs the working conditions curve from the city road conditions and the driving conditions of the test vehicle itself.First,the bad data and abnormal data collected in the process of data collection are pre-processed.Then,a suitable interpolation function is selected to restore the vehicle driving trajectory,invert the road parameters,and calculate the energy consumption conversion rate and fuel consumption conversion rate as two measurement indexes.Then,for the respective component factors before and after the introduction of the above measurement indexes,the principal components related to the kinematic features are extracted by principal component analysis,which is used to optimize the classification criteria.Then the optimized classification criteria are used to classify the kinematic segments using the K-means clustering method,and they are divided into two types: smooth road and non-smooth road.Then,the proportion of uphill,downhill and flat roads are used as classification criteria to classify smooth and non-smooth roads into uphill,downhill and flat roads respectively by K-means clustering method.The characteristic parameters of the six types of kinematic fragment classified after optimizing the classification criteria were counted and evaluated,and it was found that the characteristic parameters obtained after optimizing the classification criteria were more consistent with the collected data.Finally,the driving cycle curves were fitted to the characteristic parameters of the six types of kinematic fragment obtained after optimizing the classification criteria,and then the conditions were summarized according to the percentage of movement time of each kinematic fragment type.