Abstract: According to analysis of the properties of spindle-power of CNC machine tool, a kind of automatic identification and machining progress monitoring method for machined workpiece, based on dynamic time warping technology and spindle-power information, is proposed. Its purpose is to solve some problems in job shop, such as machining progress information mainly relying on manual statistics incurring frequent error in case of mixed type machining, some workpieces are scrapped because of unauthorized change of machining parameters. At first, through processing processes for the possible spindle-power information of various workpieces, such as selection of characteristic vector, forming of vector sequence, time warping, computation of matching distance, as well as clustering of K-means algorithm, characteristic template library is formed. During machining of workpiece, spindle power information is collected in real time and then processed to form characteristic vector sequence; this sequence is processed by time warping and computation of matching distance, etc., together with template in the characteristic template library. According to the size of matching distance and the set threshold value, the type of machined workpiece or the correctness of machining parameter are judged. The case study in equipment workshop shows that the method is practical.

Key words: Identification, Job shop, Monitoring, Workpiece