This work is concerned with the design and implementation of an automatic system for detection and classification of yarn faults. Opto -electronic data processing makes it possible to detect and classify faults and identify their source. The method is based on Image processing, neural networks, and expert systems. We present the principles of yarn fault detection modeling. Image processing is a new technique for the continuous measurement and supervision of yarn faults. The new opto- electronic technique can measure yarn thickness and yarn faults. The automated vision system captures data from the running lengths of yarn to be stored and subsequently examined in detail. Examples of thickness histogram, the yarn fault profiles and statistical features such as, the mean thickness, standard deviation, average deviation, coefficient of variation, variance, and maximum and minimum yarn diameters. Neural networks are designed and applied to the classification of yarn faults. We have studied the ability of networks to correctly classify both training and testing examples. Multi- layered neural networks were trained to classify the faults using the error back propagation learning algorithm. Noisy faults could be efficiently recognized. We evaluate the performance of the back propagation technique on recognition of yarn faults. Results indicate a high percentage of correct recognition and fault tolerance capability. The most effective number of neural network layers and the number of units in the hidden layers are conducted through extensive experimental work on yarn faults. Expert system is introduced to identify the source of a recognized yarn fault. An expert system is a program that aids in solving complex real world problems that usually require a human expert. The basic structure of an expert system is organized around two fundamental modules, which are: .The knowledge base .The inference engine. The knowledge base may be divided into two parts involving a set of facts
By: A.E.Amin*, R.El-Bealy **and A.Tolba*** -Quality Control Eng. Dakahlia Spinning and Weaving Company.
Submit Date: 6/7/2010 18:00