A Yarn Detection Method Which Based On Tension

Quality control of yarn is an important approach for textile technology and the key step that will define the quality of fabric.Most of traditional ways for detection of the yarn quality such as resistive based on metal strain,capacitive based on yarn density,opto-electronics based on yarn appearance and machine vision method based on image sensor,are low-accuracy,susceptible to external environmental which has humidity and dust,unstable,expensive and difficult to popularize.This paper presents a yarn detection method online which based on piezoelectric.This sensor made of a piezoelectric ceramic sensor for detecting the yarn tension signal,using a adaptive way,according to the density of the yarn and tension are a fixed proportional on the suffered detection fragment,converted the tension signal to a voltage signal.This method is environmental adaptability,low cost,detection status reliable,can be used to achieve real-time online information collection yarn,avoiding the waste-sampling detection.Because the original signal from the piezoelectric sensor carried a lot of higher harmonics which generated by mechanical vibrations,the collected signals goes through a low-pass filter and a Butterworth low-pass filter that made of two active second-order low-pass filter in order to achieve the signal from analog noise signal in the hardware design.Amplifier used an operational amplifier to consist of the integration circuit in order to accumulate the yarn signal in the time domain to achieve a yarn defect classification,and then use of programmable amplifiers amplifying threshold and gain linear,so that the system can adapt to a wide type range of yarn.Spectrogram diagram analysis which come from detecting yarn on line could give a theoretical basis for improving the process of textile and detection the mechanical fault diagnosis of the production process.For yarn signal is non-stationary signals,the wavelet transform method may suitable to calculate spectrogram that break the traditional way though the Fourier transform method calculation spectrogram.This paper analysis the wavelet transformation time and frequency domain decomposition by processing the yarn signal using the wavelet transform.According to their characteristics,the design uses 5 layer decomposition 85 channels decimated and sampled discrete yarn signal after filtering,which are 85 band-pass filter,analyze the reasons of aliasing by using complex wavelet computing spectrum,choose wavelet function from the minimum variance approach and the maximum squared signal principle,and then use the self-constructed wavelet analysis function to achieve spectrogram of yarn signal.Finally,the same kind yarn detection results obtained with the use of this method were compared with DQSS-60 electronic yarn clearer and YG135 evenness tester(all equipment based on capacitance detection),obtain satisfactory results.