Research On On-Line Tension Detection Of Polyester Filament Based On Wavelet Denoising

Tension is an important parameter in the spinning process,which directly affects the quality of fiber products and the subsequent process.The on-line testing of polyester filament tension is the data base of product quality control in the later period.At present,contact sampling inspection method is widely used in the industry.Under the condition of high speed spinning,contact sampling inspection is easy to cause the damage of fiber bundles and seriously affect the quality of fiber products.The existing non-contact tension detection methods based on image processing have difficulty in parameter identification,large amount of data processing and real-time acquisition.The technology of laser Doppler vibration measurement provides a way to realize the non-contact online tension detection of polyester filament,but there are a lot of noise interference in the actual sampled signal,so it is necessary to process the vibration signal of polyester filament with proper denoising method.Based on the existing laser Doppler vibration measurement technology,a non-contact polyester filament tension detection method based on wavelet denoising is studied in this paper.Firstly,an optical path system is built to collect vibration signals of polyester filament.The optical path system is designed based on the laser Doppler effect and interference effect,and the optical fiber path system is chosen to solve the problem that the traditional prism path system is difficult to realize the interference of measurement light and reference light.The vibration signal of polyester filament is collected by the optical path system.Secondly,the mathematical model of vibration fundamental frequency and tension of polyester filament is constructed.Kelvin model was used to characterize the viscoelastic model of the polyester filament.In order to obtain the transverse vibration equation,a segment of polyester filament before winding was selected for stress analysing.The differential vibration model was obtained by Galerkin truncation method,and the general mathematical model of tension and vibration fundamental frequency was established.The vibration signal of polyester filament was collected by the system and the fundamental frequency was obtained by analyzing the signal.The tension was calculated based on the constructed tension detection mathematical model,and verified by the tension detected by contact tensiometer.Then,a vibration signal processing system based on wavelet threshold denoising is built.The improved wavelet threshold denoising algorithm is used to process vibration signals.The variational modal decomposition algorithm further separates the signal and noise.Multiwindow spectrum analysis and energy centroid method were used to extract the fundamental frequency from the decomposed modal components.Finally,the non-contact tension detection of polyester filament was verified by the simulation winding test bench and laser vibration optical path system.Vibration signals were processed to extract the fundamental frequency of vibration of polyester filament,and the non-contact online tension detection of polyester filament was simulated based on the constructed tension detection mathematical model.The experimental results show that the tension value obtained by the non-contact tension detection method fluctuates around the preset tension,and there is little difference between the value and that obtained by the contact tensiometer,and the maximum error is within 13%.