Abstract: In order to realize accurate identification of wear degree of roadheader cutting picks, an analysis and recognition method of wear degree of roadheader cutting picks was proposed. Three-direction vibration signal and current signal in cutting process of picks with different wear degree were tested and extracted. Sample evidence set for each cutting characteristic signal was established, and on-line recognition of wear degree of picks was realized by using D-S combination recognition model based on modified model. The experimental results show that the proposed method can accurately identify wear degree of picks according to multi cutting characteristic signals, and has high recognition precision and reliability, which provides important basis for determining the optimal maintenance and replacement period of picks.