The Research Of Operating State Recognition For Machine Tools Based On The Analysis Of Load Current

Machine tool as a important equipment of modern production, the efficiencydirectly affect the enterprise’s production cost. Acquire the information of theproperties, including working load, processing progress etc, can provide the basis forproduction manager to make reasonable production plan and task allocation. Thesignificance of running state monitoring not only exploit the potential of theequipments through optimizing production processes, but also improve the productionefficiency of management, and reduce the production cycle.At present the majority research of machine tools（No CNC）, are focused on thefault detection and diagnosis of the overall unit, scarcely pay attention to the furtherclassification of normal operation state of the machine tools, whether the operatingmachine tools produce profits is not clear, they may doing something useless. Thisthesis submit a new method for running state recognition based on load currentanalysis, by the analysis of load current we acquire the information when does themachine tool produce benefit, and when doing useless thing, through the statistic analysis, we getthe utilization information, and this could promote the quality of produce plan and control.In this thesis, the same type of ordinary lathes are taken for analysis purposes, collect thesignal in different state such, single, multiple, processing, idling, and many kinds of cuttingparameters, using the single variable analysis method to analyses the signal with the help of fouriertransform, wavelet packet transform, bispectrum analysis and the combination of above method.Search the characteristics that can distinguish the different state, and make a preliminaryassessment about the cause of those differences. Through a large number of experimental analysisand statistics, obtained the results as follows:(1) By analysing the starting signal accomplish the goal that monitoring which machine toolis starting.(2) By the means of the CZT spectrum-zoom analysis to find the characteristics of frequency, by these features identify which machine tools are running accurately.(3) Using the wavelet packet energy spectrum analysis recognize several kinds of typicalprocessing method and cutting depth.(4) Using wavelet packet energy spectrum combine bispectrum to analyses which iscuttinging or idling of three machine tools.