计算机辅助棉纺质量控制系统开发

【作者】 王恩清；

【导师】 赵书林；

【作者基本信息】 天津工业大学 ， 纺织工程， 2008， 硕士

【摘要】 棉纺质量控制是一个系统工程,在温湿度一定时,需要对原料、设备、工艺进行综合考虑,顾此失彼则很难达到良好控制质量的目的,因此定位于棉纺质量控制的应用软件在开发时,应从上述三个角度出发为实施质量控制提供相应的分析工具,以确保良好的成纱质量。针对上述特点,本课题以纺纱工艺配置优化为中心,从纺纱原料、纺纱工艺、纺纱设备及成纱质量离线检测等角度出发,对实施成纱质量控制进行了探讨并开发了相应的应用程序。从原料角度出发利用棉纤维质量指标对成纱条干、强度实施预测并进行质量预测子系统开发,经验证该子系统预测精度可满足纺纱厂要求,预测结果可用于指导纺纱实践活动。从工艺角度出发学习与分析了常见试验设计方法及数据分析方法,以其原理为依据实施工艺优化子系统开发,此部分针对单因素及多因素分别开发了不同的数据分析功能,借助于此实现纺纱工艺参数的优化配置。从成纱质量角度出发开发了成纱质量综合评定功能、布面规律性机械波波长计算功能、规律性机械波跨工序诊断功能等,借助于系统的该部分功能不但可以确定产生纱疵的工序,而且还可在生产前根据纱线用途进行防疵重点查询,使防疵工作做到重点突出。在该系统的开发过程中,从模型构建到编程方法都进行了一些探索。在进行成纱质量预测模型的构建时引入了人工神经网络技术,利用BP人工神经网络构建成纱质量预测模型对成纱质量进行预测,经验证效果较为理想;采用了模糊分析方法进行成纱综合质量评定模型的构建,并通过引入加权系数使评定结果更具有实际意义;在进行质量预测子系统的开发时,采用了MATLAB与VB混合编程技术,从而将二者优势相结合,缩短了程序开发周期,提高了用户界面的友好性。更多还原

【Abstract】 Cotton-spinning quality control is a systematic project, when the humidity is set, it is necessary to think about fiber material, spinning machine and process altogether, or else it is hard to ensure good yarn quality, so the software designed for Cotton-spinning quality control should offer analyzing methods in those three aspects; According to the characteristic of Cotton-spinning quality control, this essay discusses the means to control yarn quality and develops programs based on the Cotton-spinning quality control means.Yarn evenness and tenacity are predicted through cotton property; meanwhile the predicting subsystem is developed. It is proved that this system can predict yarn qualities with the accuracy satisfying cotton spinning mill requirement and the predicted result can be used to guide spinning practice.The common experiment design methods and data analysis principles, based on which the subsystem for optimizing process is programmed, are studied. This subsystem can be used to analyze data for single factor or multiple factors in order to optimize process parameters.This system also includes some other functions to evaluate yarn qualities, compute mechanical wave length, and diagnose mechanical wave to locate the process which has incurred that yarn fault. With the help of those functions, it is quick to find out the malfunction parts, at the same time, the central work for preventing yarn impurities can be decided according to ultimate use of the yarn to be produced.In the process of developing this system, some new methods are tried both in model building and programming method. The artificial neural network techniques are used to build predicting model, which is proved much accurate; the model used to assess yarn qualities is built with fuzzy decision and coefficients are added to the model to make the assessment more practical; the subsystem for predicting yarn qualities is developed with the technique of VB and MATLAB merging programming so that the advantages of one are incorporated to that of the other .The developing period is shortened, but the UI is much friendly.更多还原