Research On Vibration Measuring Tool Holder System And Signals' Singularity Analysis For Online Tool Wear Condition Monitoring In Milling

Intelligent manufacturing is the main direction of the strategic development plan of"Made in China 2025",and the development of intelligent manufacturing informatizationized equipment based on technological and theoretical innovation is the priority development direction of the new generation of information technology industry.In the field of machining,the online identification of tool wear condition and tool replacement decision during manufacturing still relies on workers' experience,and lack of scientific and effective support of theory,technology and equipment,which limits the improvement of production efficiency and product quality.The online tool wear condition monitoring system based on sensor technology can effectively identify the tool wear condition to make full use of the machining potential of tools,which has gradually gained widespread attention and application in the manufacturing industry.And it is considered as an important supporting technology to realize automated manufacturing and unmanned production.Aiming to establish an integrated milling process monitoring system with the ability of "sense-analysis-decision",relevant theories,methods,and technologies are researched around the scientific issues involved in intelligent sensing equipment integration and time-frequency processing theory and algorithms on sensory signals.Aiming at the problems of vibration attenuation,inconvenient installation,and poor transplantability of traditional wired vibration acquisition systems in cutting process,a rotary vibration measuring tool holder system for online monitoring of the milling process has been developed.Through the integration of an industrial-grade triaxial accelerometer,the development of signal acquisition and wireless communication module and pre-conditioning module,the real-time measurement of triaxial vibration signals during the milling process is realized.The dynamic characteristics of the tool holder system are evaluated by means of the single-point excitation modal experiment and finite element modal analysis technology.A comparative experiment of vibration signals during milling is carried out to test the signal acquisition performance of the rotary vibration measuring tool holder system.The experimental results show that the developed rotary vibration measuring tool holder system can meet the application requirements.In view of the close relationship between the signal waveform changes of various sensors and the tool wear condition during the milling process,a singularity analysis theory based on wavelet transform is established,and a de-noising algorithm based on the evaluation of wavelet transform modulus maxima(WTMM)is formed,then the quantitative characterization of small changes in the sensory signal waveform is realized.The most suitable wavelet basis for the singularity evaluation of cutting force signals is determined by establishing a three-dimensional cutting force model with considering tool wear.A wavelet base selection method based on the estimation of the WTMM de-noising algorithm is established to provide a qualitative basis for the evaluation of unknown singularity types of cutting vibration and sound signals.Based on the developed rotary vibration measuring tool holder system,an experimental platform for milling processing including three kinds of sensors such as cutting force,vibration,and sound is set up.Then the full life cycle cutting wear experiments of the solid cemented carbides are designed and carried out scientifically.Based on the established singularity analysis algorithm,Holder Exponent(HE)indexes of various sensory signals are calculated.By analyzing the probability density distribution function and typical statistical features of the HE indexes,a correlation mechanism with the tool wear condition is established.Based on the developed rotary vibration measuring tool holder system,an online tool condition monitoring system is established.The feature selection algorithm based on mutual information is used to screen the statistical features of HE indexes estimated from various sensory signals.Based on the support vector machine algorithm,the tool condition recognition models of cutting force,vibration and sound signals are established respectively.The sensitivity to the change of different tool wear states of statistical features of HE indexes calculated from vibration signals,tool condition monitoring model trained by the vibration signal is optimized.Finally,the host software for online monitoring of the milling process is built,which has the ability of real-time acquisition and display of vibration signals from the rotary vibration measuring tool holder system,on-line tool wear condition monitoring,and time-frequency analysis and singularity evaluation in offline state.