The Study Of Batch Drilling Process Quality Inspection Based On Multi-sensor Instantaneous Feature

Quality detection is a key step to guarantee the drilling quality. And the present detection methods have many problems such as incompleteness of detection, high cost, great difficulty for on-line detection, missing detection, thus there are showing high technical requirements. Taking the implement of batch drilling quality detection as paramount, in this dissertation we presents a detailed study on the aspects of building up the drilling monitor system, the extraction of the instantaneous feature points, the extraction of the statiscal characteristics and the principal component analysis which is used to decline the dimensions of feature matrix:1. According to the properties of drilling process, the change of the spindle power is monitored by HALL power sensor, the wear condition of the drill cutter is detected by AE sensor, and the vibration of the workpieces is monitored by three-axis acceleration sensor. LabVIEW software is also applied to construct the platform of drilling monitoring system.2. Based on the changes of contact environment between the drill edge and the workpiece, the drilling process is divided into three stages which are the stage of drilling guide, the stage of drilling and the stage of drilling out. After a series of processing such as filtering, RMS analysis and derivation for spindle power (SP) signals and acoustic emission (AE) signals, the instantaneous feature points of each stage can be extracted. The feature points can also be extracted by use of the orthogonal envelop method.3. In accordance with the extracted instantaneous feature points, firstly the drilling process is accurately divided into three stages and the machining time can correspondingly be calculated. Secondly the mean, the standard deviation, the range interquartile and the entropy value of drilling signals are served as process features. Thirdly these characteristics are be treated as elements to build quality matrix, Lastly the obtained data are be standardized.4. Principle component analysis is applied to reduce the dimensions of statistic feature matrix which is constructed by instantaneous points. After drafting the quality distribution map of the features under the first and the second principle component coordinate, the efficiency assessment of batch drilling quality could be implemented.24 drilling quality are classified and the batch drilling quality inspection is implemented by use of the method mentioned in this dissertation. This new method can solve the problems which mentioned in the first graph of this abstract and offer a good reference for analysis of batch drilling quality.