摘要
由于间歇染色过程难以在线直接测量出染机内织物色泽,而只能采用人工离线测色法进行测量染色终点的色泽,导致产品质量不稳定,生产的一次合格率低,成本、能源消耗和废水排放量增加,因此,色泽在线测量技术是染整生产提高产品质量,降低生产成本,实现节能减排的关键技术。本文针对间歇染色生产色泽在线测量这一染整行业关键技术难题,提出了通过测定间歇染机内染液的染料浓度来获得织物色泽的软测量方法,并设计了实际染机色泽在线软测量装置。
本文主要的创新性研究工作包括:
1.基于调整系数的染料单位浓度K/S值算法研究
为了掌握各类染料上染不同织物的染色机理,研究织物K/S值与染料浓度关系,提出基于调整系数的单位浓度K/S值算法。该算法通过引入调整系数,可根据不同浓度对光谱数据进行合理地调整,使织物K/S值与染料浓度具有更好的线性关系,采用最优化方法确定调整系数,以获得准确的染料单位浓度K/S值,使计算机测色模型和自动校色算法更为准确、可靠,为在线测色以及计算补色配方提供了保障。
2.染色动力学模型及仿真研究
从染色热力学和动力学角度出发,研究染料的上染机理,分析染料浓度及助剂、染色温度、染液流速等因素对上染率的影响。在菲克(Fick)定律的基础上,导出染色过程染料上染率动力学模型,实时预测染料的上染状态,再根据计算机测色模型来估计织物色泽的变化。同时,基于染色动力学模型,设计了染色过程计算机仿真系统,模拟实际染色过程染料的上染情况以及织物颜色变化,有助于掌握染料的染色机制,为间歇染色过程的工艺优化和控制奠定基础。
3.低浓度染液多组分浓度同时测定的色泽软测量方法研究
在分光光度法同时测定混合染液各染料浓度的基础上,本文提出了一种基于卡尔曼滤波(Kalman Filter,KF)的染液多组分染料浓度的软测量方法。该方法通过测定间歇染机内混合染液在其各染料最大吸收波长处的吸光度数值,基于吸光度模型,采用卡尔曼滤波算法来估计染液中各组分染料浓度,滤除噪声。最后根据计算机测色模型计算出织物色泽三刺激值,有效地解决低浓度混合染液间歇染色过程织物色泽在线测量问题。
4.软测量系统非线性滤波算法研究
非线性滤波是软测量技术中减少测量噪声,提高测量精度和稳定性的主要方法。扩展卡尔曼滤波(Extended Kalman Filter,EKF)和粒子滤波(Particle Filter,PF)是最常见的非线性滤波方法,其理论和应用研究对提高软测量技术及非线性滤波理论和应用水平具有重要作用。本文在同时测定染液组分浓度非线性模型的基础上,分别设计了EKF和PF算法,对染液各组分浓度进行最优估计,并在实际间歇染色过程色泽在线软测量中验证了其有效性和可行性。
As it is difficult to measure the color of fabric online in the batch dyeing process, there is the only way to measure color offline by the worker when the dyeing process is finished. However, this way causes many problems such as bad quality of products, low pass-rate, high cost, much energy consumption, serious pollution and so on. So on-line color measurement technology is very helpful for stabilizing quality of products and saving energy as well as reducing pollution in the dyeing industry. For solving this technical problem, a soft-sensing method of color measurement is presented, with which the fabric color can be measured by testing the absorbance of the dye solution in the dyeing machine, and an actual soft-sensing device of fabric color is also designed in this paper.
The main contributions of this thesis are as follows:
1. Study of the K/S value of dyes based on adjustment factors
For mastering the mechanism of different kinds of dye dyeing different fabric, the relationship between the fabric K/S value and dye concentration is studied and a calculation method of K/S value of dyes based on adjustment factors is presented. By introducing adjustment factors to correct the source data reasonably according to the different concentration of dyes, the relationship between the fabric K/S value and dye concentration will turn linear. With the optimization methods, adjustment factors could be worked out and K/S value of dyes could be obtained according. In this case, both the computer color measuring model and the automatic color correction algorithm are more accurate and reliable, which could be helpful for the online color measurement and the calculation formula of complementary color.
2. Study on kinetics model of dyeing and simulation
From the dyeing thermodynamics and kinetics theories, it is known that there are many factors effecting the uptake rate of dye including dye concentration and dyeing temperature as well as flow rate. Considering these factors, kinetics model of uptake rate is built based on the Fick’s law, with which the uptake rate can be predicted at any time, and the color of fabric can be also estimated in the dyeing process. Foremore, a simulation system of dyeing process is designed, which helps to master the dynamic characteristics of the dye and lays the foundation for optimization and control of the process.
3. The soft-sensing technology of fabric color in batch dyeing process with low concentration of multi-dyes
Based on the determination technology of concentrations of multi-component, a soft-sensing method of dye concentration measurement using Kalman filter is presented. The absorbance of the dye liquor in the dyeing machine is tested by the spectrophotometer in the visible wavelength range. Then the concentrations of dyes are calculated by the Kalman filter algorithm, thus the soft-sensing measurement of the dye concentrations in the batch dyeing machine is achieved, and the color of fabric can be also estimated baesd on the model of color measurement.
4. Study on nonlinear filter algorithm applied in soft-sensing system
Nonlinear filter algorithm is the main way for reducing measurement noise and improving accuracy in the soft-sensing system. The common nonlinear filtering method such as extended Kalman filter (EKF) and particle filter (PF), on which the theoretical and applied research is very important for improving the level of nonlinear filter. Based on the nonlinear absorbance model, EKF and PF are applied to determine the concentrations of dyes respectively. The feasibility and validity of these methods have been verified by experiments.
引文
[1]中国纺织工程学会编著.纺织科学技术学科发展报告[R]北京:中国科学技术出版社,2007:1-129.
[2]王岩.中国纺织工业发展战略研究[D]吉林大学,2003.
[3]俞亦政.21世纪纺织行业所面临的水环境问题及对策[D]东华大学,2005.
[4]石碧,王双飞,郑庆康等.轻化工程导论[M]北京:化学工业出版社,2010:151-202.
[5]Syed Z.Kazmi, Perry L. Grady, Gary N. Mock et al. On-line color monitoring in continuous textile dyeing[J].ISA Transactions,1996 (35):33-43.
[6]陈兴梧,田学飞.纺织品颜色的在线测量[J]计量学报,1997,18(2):150-155.
[7]姚祖亮.针织物的连续化加工设备[J]针织工业,2004,10(25):72-74.
[8]牟晶晶,徐海松,徐东晖.在线色差检测的误差修正方案设计[J]光电工程,2007,34(2):37-40.
[9]顾秋基.色差的在线测控[J].印染, 1998,24(12):24-27.
[10]陈立秋.浸轧机色差控制的技术进步[J].染整技术, 1995,7(1):17-18.
[11]Martin Ferus-Comelo. Analysis of the factors influencing dye uptake on jet dyeing equipment [J].Coloration Technology, 2006(122):289-297.
[12]叶早萍(译),何叶丽(校).染色[J]印染,2008(9):47-49.
[13]邱红娟(译),唐人成(校).用FIA-HPLC分析技术在线监控活性染料的竭染和水解[J]国外丝绸,2000(6):18-22.
[14]Santos.Jorqe G,Soares.Graca M.B,Hrdina.Radim et al. A study on the spectral changes of reactive textile dyes and their implications for online control of dyeing processes [J].Coloration Technology, 2009, 125(1):8-13.
[15]Reddy. Madhukar, Lee. Gordon, McGregor. Ralph etal. Modeling of the batch dyeing process [J] Proceedings of the American Control Conference, 1995(3):2180-2184.
[16]Jasper, Warren J., Reddy, Madhukar Y. Real-time system for data acquisition and control of batch dyeing [J] IEEE Annual Textile, Fiber and Film Industry Technical Conference, 1994.
[17]田世昌,张谦,王东云.间歇式染色机的计算机控制[J].郑州纺织工学院学报,1995,6(2):22-27.
[18]汪岚,金福江.染色过程色差与染液吸光度模型[J].纺织学报,2008,29(4):83-85.
[19]汪岚,黄彩虹.基于MATLAB色差预测多元回归模型的研究[J].计算机与应用化学,2008,25(8):1015-1018.
[20]荆志巍.间歇生产过程的控制系统设计及控制算法研究[D]华北电力大学,2002.
[21]单洪亮.间歇蒸煮过程的模型辨识与监控[D]浙江大学,2004.
[22]侯迪波.间歇生产过程中的知识发现方法研究[D]浙江大学,2004.
[23](英国)T.维克斯太夫著.董亨荣,水佑人(译).染色物理化学[M].北京:纺织工业出版社,1981:58-59.
[24]Lakkumi, N., Sundaram, A., Study on dyeing of silk using reactive dyes[J].Textile Dyer & Printer, 1991,24(4):32-36.
[25]傅忠君,于鲁汕.染色动力学数学模型的研究[J]染料与染色. 2003.40(1):23-29.
[26]Peters, R. H., Textile Chemistry (Vol.3) [M].New York: Amsterdam, Physical Chemistry of Dyeing, Elsevier Scientific Publishing Company, Netherland. 1975:158-161. [27(]日)黑木宣彦著.陈水林(译).染色理论化学(上)[M].北京:纺织工业出版社, 1981:1-93.
[28](苏)格.叶.克里切夫斯基著.高敬琮(译).染色和印花过程的吸附与扩散[M]北京:纺织工业出版社,1985:178-192.
[29]汪辉雄.纤维染色学[M]台中:大学图书供应社,1982:1657-1661.
[30]屠天民,骆钦,施睿等.染色过程中染料浓度的在线监测[J].印染,2009(9):19-22.
[31]殷先雄,王栋,于永海等.测定双组分染料浓度的两种方法比较[J].辽宁化工,2008,37(2):131-133.
[32]杨玉杰,孙剑辉,赵颖等.卡尔曼滤波分光光度法同时测定活性艳红活性紫混合体系[J]河南师范大学学报1996,24(1):89-91.
[33]王强,马沛生,汤红梅等.扩展Kalman滤波器同时测定苯酚和邻氯苯酚[J].光谱学与光谱分析,2006,26(5):899-903.
[34]宋志勇.非线性随机离散系统推广卡尔曼滤波方法收敛性分析[J]控制理论与应用,2002 17(2):264-269.
[35] Song T.L., The modified gain extended Kalman filter and parameter identification in linear system [J]. Automatic, 1986, 22(1):59-75.
[36]Fisher L. Factorized extended Kalman filter [J]. Proceedings of SPIE. 1985:119-129.
[37]Juiler S. J, Uhlmann J.K,Durrant-whyteh.F.A new method for the nonlinear transformation of means and covariances in filters and estimators [J] .IEEE Transactions on Automatic Control, 2000, 45(3):477-482.
[38] Juile S.J., J.K.Uhlmann. Unscented filtering and nonlinear estimation [J]. Proceedings of the IEEE2004,92(3):401-421.
[39]Gordon N.Salmond D. Novel approach to nonlinear and non-Gaussian Bayesian state estimation[J].Proc of Institute Electric Engineering,1993,140(2):107-113.
[40]Musso C,Oudjanen,and Legland F.Improving regularized particle filters [M]. New York:Springer-Verlag.200l:247-272.
[41]朱志臣.基于改进边缘化粒子滤波器的机动目标跟踪[J].武汉理工大学学报,2008,30(6):118-121.
[42]朱志臣.基于高斯粒子滤波器和TVAR模型的语音识别增强技术[J].仪器仪表学报,2008,29(9):1903-1907.
[43]俞金寿,刘爱伦,张克进.软测量技术及其在石油化工中的应用[M]北京:化学工业出版社,2000:1-120.
[44]李海青,黄志尧.软测量技术原理及应用[M]北京:化学工业出版社,2000:1-134.
[45]杨敏,胡斌,费正顺等.基于DPCA-RBF网络的工业流化床乙烯气相聚合过程的软测量研究[J].仪器仪表学报,2010,3(31):481-487.
[46]金福江,王慧,陆浩等.蒸煮过程有效碱浓度在线软测量技术研究[J].仪器仪表学报,2003,24(4):380-383.
[47]孙玉坤,王博,黄永红等.基于聚类动态LS-SVM的L-赖氨酸发酵过程软测量方法[J].仪器仪表学报,2010,2(31):404-409.
[48]金福江,周丽春.化工软测量技术研究进展[J].化工进展,2005,24(12):1379-1382.
[49]Qi Li, Cheng Shao. Soft Sensing Modelling Based on Optimal Selection of Secondary Variables and Its Application [J].International Journal of Automation and Computing, 2009, 6(4):379-384.
[50]Shrikant A.Bhat, Deoki N.Saraf,Sanjay Gupta et al. Use of Agitator Power as a Soft Sensor for Bulk Free-Radical Polymerization of Methyl Methacrylate in Batch Reactors [J]. Ind.Eng.Res. 2006, 45(12):4243-4255.
[51]Tham, M.T., G. A. Montague. Soft-sensors for process estimation and inferential control [J]. J. Process Control, 1991, 1(1): 3-14.
[52]张颖.改进的T-S模糊神经网络在化工软测量中的应用[J].电子测量与仪器学报,2010,6(24):585-589.
[53]张光新,周泽魁.烧碱蒸发过程中出料浓度软测量技术研究[J].仪器仪表学报,2004,25(4):675-677.
[54]杨茹,邱法林.分光光度学[M].北京:机械工业出版社,1998:1-48.
[55]雷波.分光光度法对混合染料浓度的同时测定[J].染整技术,2003,25(3):35-37.
[56]李昌厚,孙呤秋.杂散光与吸光度误差和吸光度真值关系的研究[J].仪器仪表学报,2001,22(1):54-58.
[57]李志良,刘一鸣,石乐明等.卡尔曼滤波分光光度法用于多组分分析[J].分析测试学报,1989,8(6):38-43.
[58]王雁鹏,董旭辉,陈岩等.卡尔曼滤波分光光度法同时测定混合氨基酸[J].光谱实验室,2006,23(6):1166-1169.
[59]蔡卓,赵静,江彩英等.卡尔曼滤波紫外分光光度法同时测定盐酸异丙嗪和盐酸氯丙嗪[J].广西大学学报(自然科学版),2009,3(34):340-346.
[60]余煜棉,张音波,刘春英等.卡尔曼滤波分光光度法同时测定扑热息痛四组分的研究[J].光谱学与光谱分析,2003,23(5):1005-1007.
[61]朱志臣,汤红梅,倪丽琴等.Kalman滤波同时测定苯酚和2,4-二氯苯酚[J].天津城市建设学院学报,2008,14(3):219-222.
[62]王海峰,陈海相,邵建中.HPLC法研究C.I Reactive Red24水解反应动力学[J]染料与染色,2004(06):331-333.
[63]王正佳,邵敏,邵建中.乙烯砜型活性染料水解动力学的HPLC研究[J]纺织学报,2006(09):9-13.
[64]M.Klancnik.温度对一氯均三嗪活性染料的水解和醇解反应的动力学的影响[J],国外纺织技术,2001(07):26-29.
[65]范炜,李勇.近似二阶扩展卡尔曼滤波方法研究[J]空间控制技术与应用,2009,35 (1):30-35.
[66]K.K. Ah, D.Q.Truong.Online tuning fuzzy PID controller using robust extended Kalman filter[J].Journal of process control,2009 (19):1011-1023.
[67]范文兵,陈达.卡尔曼滤波器在状态和参数估计中的应用[J]郑州大学学报(理学版),2002,34 (4):44-47.
[68]仲卫进.基于扩展卡尔曼滤波的动态负荷建模与参数辨识[D]上海交通大学,2007.
[69]李书进,虞晖.基于扩展卡尔曼滤波的非线性带滑移滞变系统的实时估计[J].武汉大学学报(信息科学版),2004,29(1):89-92.
[70]Bolic Miodrag. New resampling algorithms for particle filter [J] IEEE ICASSP, 2003, 2: 589-592.
[71]谷雨,李平,韩波.基于分层粒子滤波的地标检测与跟踪[J].浙江大学学报(工学版),2010,44(4):687-691.
[72]宋钦涛,汪立新,潘慈颜.基于粒子滤波的全息检测技术的研究[J].杭州电子科技大学学报,2009,29(6):23-25.
[73]李子昱,秦红磊.改进重采样粒子滤波算法在GPS中的应用[J].计算机工程与设计,2010,31(11):2523-2526.
[74]王法胜,赵清杰.一种用于解决非线性滤波问题的新型粒子滤波算法[J].计算机学报,2008,31(2):346-352.
[75]张瑞华,雷敏.粒子滤波技术的发展现状综述[J].噪声与振动控制,2010,(2):1-4.
[76]张琪,胡昌华,乔玉坤.基于权值选择的粒子滤波算法研究[J].控制与决策,2008,23(1):117-120.
[77]郭文艳,韩崇昭.基于统计线性回归的粒子滤波方法[J].电子与信息学报,2008,30(8):1905-1908.
[78]赵丰,汤磊,张武等.一种高实时粒子滤波重采样算法[J]系统仿真学报,2009,21(18):5789-5793.
[79]Li Yuanqiu, Shen Yi, Liu Zhiyan.A new particle filter for nonlinear tracking problems[J].Chinese Journal of Aeronautics,2004,17 (3):170-175.
[80]王菊生.染整工艺原理(第三册)[M]北京:中国纺织出版社,2000.
[81]郑光洪,蒋学军,杜宗良.印染概论[M]北京:中国纺织出版社,2000.
[82]刘辅庭(译),金鲜英(校).染色[J]印染,2005(10):47-50.
[83]张建新,魏俊民.基于嵌入式Linux技术的间歇式染色机控制器的研究[J]天津工业大学学报,2005,24(2):40-42.
[84]白凤翔,曾华,伊继东等.基于Kubelka-Munk理论的线性规划配色法[J].涂料工业,1994(4):21-23.
[85]金福江,汤仪平.计算机配色中活性染料单位浓度K/S值研究[J].武汉理工大学学报,2008,30(1):83-85.
[86]郝文静,赵秀萍. Kubelka-Munk单常数配色理论与实践[J].中国印刷与包装研究,2009,3(1):43-47.
[87]薛朝华.颜色科学与计算机测色配色实用技术[M]北京:化学工业出版社,2003.
[88]徐海松.计算机测色与配色新技术[M]北京:中国纺织出版社,1999.
[89]D.H.Brainard, B. A. Wandell. Asymmetric color matching: how color appearance depends on the illuminant [J] J. Opt. Soc. Am., 1992:1433-1439.
[90]荆其诚,焦书兰等.色度学[M]北京:科学出版社,1979.
[91]董振礼,郑宝梅.测色及电子计算机配色[M]北京:中国纺织出版社,1996:168-197.
[92]ROY S. BERNS编著.颜色技术原理[M]北京:化学工业出版社, 2002:96-146.
[93]Improvement to Industrial Color-Difference Evaluation[R].Vienna: Gentral Bureau of the CIE.2001.
[94]Luo M R, Cui G, Rigg B. The Development of the CIE 2000 Color-Difference Formula: CIEDE2000 [J]. Color Res. 2001.26(5): 340-350.
[95]宋钧才.漫谈色差评定[J].中国纤检2006(4):29-31.
[96]郑元林,杨淑蕙.CIE1976LAB色差公式的均匀性研究[J].包装工程, 2005, 26(2):48-65.
[97]陈东辉,马仁汀.色差判定的研究[J].纺织学报, 1998, 19(3):142-145.
[98]徐海松,叶关荣.计算机自动配色预测算法研究[J]光学学报, 1996,16 (11).
[99]陈宝林.最优化理论与算法[M]北京:清华大学出版社,2005: 229-286
[100]陈开周.最优化计算方法[M]陕西:西北电讯工程学院出版社,1985:1-102.
[101]Grank, J., The Mathematics of Diffusion [M].Oxford: Clarendon Press, 1956:104-352.
[102]Jost, W., Diffusion in Solids, Liquids and Gases [M].New York: Academic press inc., Publishers, 1960(23):102-105.
[103]金咸穰.染整工艺实验[M]北京:中国纺织出版社,1982:74-93.
[104]赵静,但琦.数学建模与数学实验[M]北京:高等教育出版社,2000:199-222.
[105]刘承平,数学建模方法[M]北京:高等教育出版社,2002:24-33.
[106]黄华江.实用化工计算机模拟-MATLAB在化学工程中的应用[M]北京:化学工业出版社,2004:: 41 -83.
[107]武彬,王禄山,张为灿等.应用美兰吸附法测定纤维素表面积的讨论[J]纤维素科学与技术,2004(4):23-30.
[108]朱海东,刘俊康,凌伟等.改进BP神经网络-残液法同时测定多组分拼混染料上染率[J]计算机与应用化学,2009,26(1):117-120.
[109]纵瑞龙,王建民,郝新敏. K/S值与Integ值差异的探讨[J]印染,2006(24):30-33.
[110]孔祥玉,韩崇昭,魏瑞轩等.一种全最小二乘算法及其在非线性滤波器中的应用[J]信号处理,2005,21(2):140-143.
[111]曲从善,许华龙,谭营.非线性贝叶斯滤波算法综述[J]电光与控制,2008,15(8):64-69.
[112]徐钟济.蒙特卡罗方法[M].上海:上海科学技术出版社,1985.
[113]Haug A J. A tutorial on Bayesian estimation and tracking techniques applicable to nonlinear and non-Gaussian processes [J]. Mclean: MITREN Corporation, 2005.