The cycle recognition algorithm based on Daubechies wavelet and fuzzy C-means clustering

A traditional recognition method for driving cycles is based on single or double parameters. Aiming at the double power sources characteristics of hybrid electric vehicles, a driving cycle recognition method based on multi-source information fusion is proposed in this paper for improving the identification accuracy of driving cycles. The features of the time series collected from multi-sensor are extracted by means of Daubechies wavelet. Accordingly the data feature information of different frequency bands for each sensor is obtained. Then for these data feature information, clustering recognition is carried out based on a fuzzy C-means clustering method with variable weights. Eventually, the membership values belonging to a certain category under different frequency bands are weighted to realize the final driving cycle recognition. The experiment results verify the effectiveness of the proposed method in this paper.