Research On Mapping Model Between Drilling Process And Monitoring Signals

Research on mapping model between drilling process and monitoring signals is suitable to study drilling mechanism in a meticulous way, evaluate the quality of work- pieces accurately and realize the online monitoring for drilling process, so how to establish the mapping model becomes the purpose of this thesis. In this dissertation, surrounding the problems that must be studied in establishing the mapping model, a detailed study on the aspects of designing the drilling monitor system, dividing drilling process stages, filtering power frequency and reducing noise from monitoring signals, and extracting the instantaneous frequency of monitoring signals has been done.(1) According to the characteristics of drilling process, the vibration of the machine is monitored by three-axis acceleration sensor, the wear condition of the tool is detected by AE sensor, and the change of the spindle motor current is monitored by Hall sensor. Monitoring signals data are acquired by writing LabVIEW software, and then the experiment is successfully finished. Based on the contacting state of the work piece and the drill edge, the whole drilling process is divided into three stages which are named the stage of drilling in, the stage of drilling and the stage of drilling out. Four instantaneous points are also defined in accordance with the change of contact position between the drill blade and the work piece.(2) There is obvious power interference for axial vibration (Az) signal and noise for spindle power (Sp) signal after frequency domain analysis of drilling monitoring signals. Independent component analysis (ICA) method and wavelet threshold de-noising method are respectively applied in filtering the power interference from Az signal and removing the noise from Sp signal after completely comparative analysis of methods of power filtering and noise reduction. In order to obtain a better result, the algorithms of power filtering and noise reduction are improved in this paper. The results indicate that the improved methods can filter power interference and de-noise efficiently.(3) From the definition of instantaneous frequency, the necessary conditions of instantaneous frequency when having its definite physical meaning are firstly analyzed; Intrinsic mode functions (IMFs) which proposed by Huang are secondly discovered to satisfy the conditions; The method of empirical mode decomposition (EMD) and its end effect restrain, and the algorithm of IMF feature selection are thirdly investigated; A conclusion that the orthogonal maximum envelope method is more suited to extract the instantaneous frequency of drilling monitoring signals in comparison with the analytic signal method is fourthly brought to the upshot; Finally the mapping model between drilling process and monitoring signals is established within 1% based on the instantaneous points of mutation instantaneous frequency features of monitoring signals.Sub-problems which are the software and hardware design for monitoring system, power frequency filtration and de-noising for monitoring signals, EMD and IMF feature selection, and instantaneous frequency extraction are involved in the process of establishing mapping model. The methods applied in sub-problems can be used as reference for other monitoring systems. The above achievements of this dissertation play a positive role for deepening the study for monitoring sensor information of drilling process. In sum, the prominent contribution of this paper is the establishment of a favorable base for further study on working information for different drilling stages and quality testing of work- pieces.