Yarn Diameter Measurements Using Coherent Optical Signal Processing

A method to measure variations in yarn diameter using coherent optical signal processing based on a single photodiode plus additional electronics is described. The approach enables us to quantify yarn irregularities associated with diameter variations which are linearly correlated with yarn mass variations. A robust method of system auto-calibration, eliminating the need for a temperature and humidity controlled environment, is also demonstrated. Two yarns that span the diameter ranges commonly used in the textile industry were used to verify the system linearity and ascertain its resolution. The results obtained have been verified using image analysis. Moreover, a diameter characterization was performed under real-world conditions for three types of yarns and a correlation with capacitive measurements is also presented. The system with sensitivity of 0.034 ${rm V}/{rm mm}^{2}$ is able to detect minute variations of yarn diameter and characterize irregularities starting at very low thresholds (commercial systems generally characterize variations of 30% or greater relative to the average value).      

Optical Processing of Lattice Images

An optical image processing facility designed specifically for the treatment of lattice images is described. A range of optical processing techniques for improving the interpretation of lattice images are outlined and illustrated with examples. These operations include methods of signal-to-noise enhancement and pseudocolour encoding of spatial frequencies.     

Non-uniform Sampling in Optical Processing

The use of non-uniform sampling in optical processing is considered. The filtering operations needed to recover the spectrum of a non-uniformly sampled band-limited function are examined. Serious limitations are shown to exist for markedly non-uniform sampling.     

PSD 信号处理电路的研究

目的简化PSD信号处理电路,降低噪声信号和温漂的技术.方法用软件代替部分硬件电路的方法.结果简化了电路,提高了信噪比.结论这种方法适于多PSD计算机信号处理的智能测试系统.     

信号多频率成份参数识别

利用最小二乘原理,提出了一种新的对信号进行多频率成份识别的算法.振动信号中往往含有多种频率成份,如果事先知道这些主要的频率,就可用该算法得到信号的解析表达式,从而能识别出各种频率成份的幅值和相位.通过数值模拟,多频率成份参数识别方法表现出了FFT所不具有的一些优点.本文以轴承油膜动特性系数测试中所用的激振力为例,先借助FFT确定其中的主要频率成份,再用一种优化方法来精确辨识其主激振频率,从而实现了对激振力信号的多频率成份识别.