Vehicle driving condition construction method based on improved short-stroke method

The invention relates to the technical field of automobile driving performance testing, and discloses an automobile driving condition construction method based on an improved short-stroke method, which mainly comprises the following steps of: acquiring automobile speed signal data, dividing the data into a plurality of motion segments, and carrying out data optimization processing on idle speed part data for multiple times to obtain an optimal driving condition; then, after optimization is completed, feature extraction is conducted on the motion segments, feature dimension reduction is conducted on the features of the motion segments through a principal component analysis method, unsupervised clustering analysis is conducted on the motion segments subjected to dimension reduction through a K-Means + + clustering method, a plurality of candidate working conditions are obtained finally, feature parameters which do not contain extreme values in the candidate working conditions are calculated, and finally the mot