摘要
优化分离膜制备条件可以提高膜的性能，扩大其应用范围，具有十分重要的意义。目前，膜制备条件的优化方法以正交试验和单因素实验法为主，即在多种条件下制备膜，根据膜的性能及其变化趋势优化制备条件。这种方法实验工作量大，只能得到较优的制备条件。因此，建立普适的膜制备条件与性能之间关系的数学模型，具有十分重要的意义。本研究基于Flory-Huggins理论分析得到了多种添加剂的最大加入量，制备了多种条件下的聚醚酰亚胺（PEI）超滤膜，并使用纯水通量截留率测定实验与图像分析两种方法对膜进行了表征。利用这些基础数据，建立了优化PEI超滤膜制备条件的反向传播神经网络（BPNN）和遗传算法（GA）混合模型。该模型预测了不同制备条件下PEI超滤膜和聚二甲基硅氧烷（PDMS）/陶瓷渗透汽化复合膜的性能，与实验数据吻合良好，具有较好的普适性。还预测出了高纯水通量与期望截留率的PEI超滤膜和高渗透通量与期望选择性的PDMS/陶瓷渗透汽化复合膜的制备条件。该模型减少了实验量，实现了根据需要进行膜的设计制备。
利用Flory-Huggins理论研究了PEI为聚合物，N, N-二甲基乙酰胺（DMAc）为溶剂，水（H2O）为非溶剂，乙醚（DE）、聚乙二醇（PEG400）、正丁醇（BuOH）和丁内酯（GBL）为添加剂的铸膜液体系的成膜热力学过程，计算得到了H2O/DMAc/PEI、DE/DMAc/PEI、PEG400/DMAc/PEI和BuOH/DMAc/PEI体系的双节线和旋节线，分析得到了H2O、DE、PEG400和BuOH的最大加入量，并根据铸膜液的热力学性能进行了膜的制备实验。使用MATLAB软件开发了膜孔结构参数统计程序，对膜的表面扫描电镜照片进行分析，得到了膜的平均孔径、最大孔径和离散孔径等参数。提出利用离散孔径与筛分方程计算膜的截留曲线、截留分子量和截留分子尺寸比利用平均孔径准确，并通过牛血清白蛋白（BSA）和葡聚糖分子的截留实验进行了验证。
在上述实验数据的基础上，分别建立了BPNN和径向基函数网络（RBFNN）模型回归相转化法五个关键超滤膜制备条件（PEI浓度、添加剂种类与浓度、停留蒸发时间、凝胶浴温度）与纯水通量、BSA截留率之间的关系。BPNN易陷于局部极值，模型多次收敛结果差别较大；而RBFNN的容错性不理想，本研究结合BPNN很强的局部收敛能力与遗传算法优异的全局寻优能力建立了BPNN-GA混合模型，并用“试差法”得到了较优的模型结构参数。尽管多种添加剂的成膜机理各异，膜性能变化趋势多样，但是BPNN-GA混合模型都能准确预测膜的纯水通量与BSA截留率，大部分测试数据的偏差都小于10%，证明该模型的预测精度良好。该模型还预测出GBL是本研究BuOH、DE、PEG400、聚乙烯吡咯烷酮（PVP）和GBL中制备PEI超滤膜最优的添加剂，预测出截留率为80-90%，纯水通量高达1.15-0.95m~3m~(-2) h~(-1)的超滤膜的制备条件，与实验结果吻合良好。
为了进一步考察BPNN-GA混合模型的普适性，本研究利用文献中的实验数据预测了不同制备条件下PDMS/陶瓷渗透汽化膜的性能，并与文献中的响应曲面法进行了比较。混合模型训练数据和测试数据的最大偏差分别为11.01%和7.17%，均小于响应曲面法对应的最大偏差（17.79%和15.56%）。混合模型预测值与实验数据的一致性，说明该模型具有较好的精度。BPNN-GA混合模型预测出PDMS浓度、交联剂浓度与涂膜时间对渗透汽化复合膜性能的影响都很大，还预测出选择性为6和7，渗透通量高达10.5和9kg·m~(-2)·h~(-1)的膜的制备条件。
Optimization of membrane preparation condition which can enhance membraneperformances and expand application range of membranes is of great significance. Nowadays,it mainly relies on single factor experiment and orthogonal test. That is, membranes undervarious preparation conditions are fabricated and characterized. According to the membraneperformances, preparation conditions are optimized. Obviously, this method has disadvantagesof large experimental data scattering, and the optimal preparation condition is not guaranteed.Therefore, there is an urgent need to develop a universal mathematical model to capture therelationship between preparation conditions and membrane performances. In this study,Flory-Huggins theory was employed to analyse the maximal addition amount of additive, andseveral membranes were fabricated, then image analysis and determination experiment of purewater flux and rejection ratio were used to characterize the membranes. Based on theexperimental data, hybrid models based on backpropagation neural network (BPNN) andgenetic algorithm (GA) were established to optimize the preparation conditions ofpolyetherimide (PEI) ultrafiltration membrane via dry/wet phase inversion. The hybrid modelspredict performances of PEI ultrafiltration membranes and polydimethylsiloxane(PDMS)/ceramic pervaporation composite membranes under various preparation conditions.The hybrid models can contribute to designing the preparation conditions to obtain desiredmembrane performances and avoiding large experimental data scattering in the fabrication ofmembranes.
Flory-Huggins theory was employed to investigate the thermodynamic mechanism of thecasting solutions whose polymer is PEI, solvent is N, N-dimethylacetamide (DMAc),nonsolvent is water (H2O), and additives are diethyl ether (DE), polyethylene glycol(PEG400),1-Butanol (BuOH) and1,4-butyrolactone (GBL). The binodal curves and thespinodal lines for H2O/DMAc/PEI, DE/DMAc/PEI, PEG400/DMAc/PEI andBuOH/DMAc/PEI systems were calculated, which reveal the maximal addition amount ofadditive H2O, DE, PEG400and BuOH. According to the thermodynamic property of thecasting solution, preparation conditions of membranes were primarily explored. A statisticalprocedure was developed to measure microstructure parameters of membranes. It disposed thescanning electronic microscope (SEM) images, including gray translation, binarization andnoise reduction of the images, and then maximal pore size, discrete pore size and average poresize were obtained. It was proposed that using discrete pore size and sieving equation tocalculate rejection curve, molecular weight cutoff and molecular size cutoff of the membrane was more accurate that using average pore size, which was proved by the rejection experimentsof bovine serum albumin (BSA) and dextran.
Based on the experimental data, BPNN and radial basis function neural network (RBFNN)models were constructed to capture the relationship of five key prepatation conditions (PEIconcentration, additive type and concentration, evaporation time in air and temperature ofcoagulation bath) to the performance of membranes, i.e., pure water flux and BSA rejectionratio. BPNN is easy to be convergent at suboptimal solutions, and there are great deviationsbetween several convergences. RBFNN has dissatisfactory fault tolerance. Therefore, hybridmodels which united perfect local convergent ability of BPNN and ideal global searchcapability of GA were proposed, whose model arctectures were optimized by trial-and-errormethod. Membrane formation mechanisms of various additives are numerous, and theperformance trends were different, but the predictions of the hybrid models were accurate, withmost of deviation in testing data less than10%. The hybrid models can predict membraneperformances under different preparation conditions and hereby indicate H2O/DMAc/PEI/GBLis the best of six casting systems in the study. In addition, the hybrid models can contribute todesigning preparation conditions to obtain higher performances of ultrafiltration membranes(BSA rejection is80-90%and pure water flux is up to1.15-0.95m3m-2h-1) and avoiding largeexperimental data scattering in the fabrication of phase inversion membranes.
The hybrid models were used to optimize the preparation conditions of PDMS/ceramiccomposite membrane with the experimental data from the literature to discuss the modeluniversality. The maximal deviations of the training and testing data between the experimentsand the hybrid model predictions were11.01%and7.17%, smaller than those between theexperiments and response surface methodology (RSM) model predictions in the literature. Theaccordances between the the experiments and the hybrid model predictions show that thehybrid models have sufficient accuracy. Connection weight analyses show that PDMSconcentration, crosslinking agent concentration and dip-coating time have great influences onthe performances of pervaporation membranes. In addition, the models predict the preparationconditions to fabricate pervaporation membranes whose permeation fluxes reached10.5and9kg m-2h-1and selectivities were6and7.
引文
[1] VAN DE WITTE P, DIJKSTRA P J, VAN DEN BERG J W A, et al. Phase separation processes inpolymer solutions in relation to membrane formation [J]. J. Membr. Sci.,1996,117(1-2):1-31.
[2]俞三传,高从堦.浸入沉淀相转化法制膜[J].膜科学与技术,2000(05):36-41.
[3]陆茵, PVDF相转化成膜机理及制膜规律研究[D].杭州:浙江大学,2003.
[4]王建琴,聚砜膜的制备及其性能研究[D].杭州:浙江大学,2006.
[5]朱利平,聚醚砜、聚醚砜酮多孔膜的结构可控制备及其表面改性[D].杭州:浙江大学,2007.
[6]高嵩,低截留分子质量PPES超滤膜的制备与性能研究[D].大连:大连理工大学,2008.
[7]刘小芬,相转化法制备改性聚偏氟乙烯（PVDF）多孔膜的研究[D].杭州:浙江大学,2009.
[8]左丹英,溶液相转化法制备PVDF微孔膜过程中的结构控制及其性能研究[D].杭州:浙江大学,2005.
[9]胡亮平,聚醚酰亚胺平板超滤膜的制备研究[D].大连:大连理工大学,2007.
[10] HERN NDEZ A, CALVO J I, PR DANOS P, et al. Surface structure of microporous membranes bycomputerized SEM image analysis applied to Anopore filters [J]. J. Membr. Sci.,1997,137(1-2):89-97.
[11] JEON J-D, KIM S J, KWAK S-Y.1H nuclear magnetic resonance (NMR) cryoporometry as a tool todetermine the pore size distribution of ultrafiltration membranes [J]. J. Membr. Sci.,2008,309(1–2):233-238.
[12] BEATON N C. Polymer Science and Technology: Ultrafiltration Membrnaes and Applications [J].National Meeting of the American Chemical Society1980.
[13]陈燚,聚偏氟乙烯中空纤维多孔膜孔径及孔径分布测试方法的研究[D].天津:天津工业大学,2003.
[14]高以烜,叶凌碧.膜分离技术基础[M].北京:科学出版社,1989:167-175.
[15] MARCEL MULDER著,李琳译.膜技术基本原理[M].北京:清华大学出版社,1999:107-122.
[16]时钧,袁权,高从堦.膜技术手册[M].北京:化学工业出版社,2001:371-376.
[17]田敬霞,基于人工神经网络的胎儿体重及孕妇分娩方式预测[D].济南:山东大学,2008.
[18] WIKIPEDIA, Frank Rosenblatt,2013, http://en.wikipedia.org/wiki/Frank_Rosenblatt.
[19] MOODY J, DARKEN C J. Fast learning in networks of locally-tuned processing units [J]. NeuralComput.,1989,1(2):281-294.
[20] MADAENI S S, GHESHLAGHI A, REKABDAR F. Membrane treatment of oily wastewater fromrefinery processes [J]. Asia-Pac. J. Chem. Eng.,2013,8(1):45-53.
[21] HAGAN M T, DEMUTH H B, BEALE M. Neural Network Design [M]. Boston, USA: PWSpublishing company, a division of Thomson learning,1996:327-331.
[22] DEMUTH H, BEALE M. Neural Network Toolbox For Use with MATLAB [M]. Natick, Massachusetts,USA: MathWorks, Inc.,1997:137.
[23]张立明.人工神经网络的模型及其应用[M].上海:复旦大学出版社,1992:43-47.
[24]林开平,人工神经网络的泛化性能与降水预报的应用研究[D].南京:南京信息工程大学,2007.
[25] MOLGA E J. Neural network approach to support modelling of chemical reactors: problems,resolutions, criteria of application [J]. Chem. Eng. Process.,2003,42(8-9):675-695.
[26] AL-ABRI M, HILAL N. Artificial neural network simulation of combined humic substance coagulationand membrane filtration [J]. Chem. Eng. J.,2008,141(1-3):27-34.
[27]彭黔荣,烟叶的化学成分与烟叶质量的人工神经网络预测[D].成都:四川大学,2004.
[28]周开利,康耀红.神经网络模型及其MATLAB仿真程序设计[M].北京:清华大学出版社,2005:70-82.
[29]曹刚,基于神经网络的厌氧反应器模拟预测及其运行状态估计[D].杭州:浙江大学,2003.
[30] OU S, ACHENIE L E K. A hybrid neural network model for PEM fuel cells [J]. J. Power Sources,2005,140(2):319-330.
[31] NIEMI H, BULSARI A, PALOSAARI S. Simulation of membrane separation by neural networks [J]. J.Membr. Sci.,1995,102:185-191.
[32] BOWEN W R, JONES M G, YOUSEF H N S. Prediction of the rate of crossflow membraneultrafiltration of colloids: A neural network approach [J]. Chem. Eng. Sci.,1998,53(22):3793-3802.
[33] BOWEN W R, JONES M G, YOUSEF H N S. Dynamic ultrafiltration of proteins-A neural networkapproach [J]. J. Membr. Sci.,1998,146(2):225-235.
[34] DELGRANGE N, CABASSUD C, CABASSUD M, et al. Neural networks for prediction ofultrafiltration transmembrane pressure-application to drinking water production [J]. J. Membr. Sci.,1998,150(1):111-123.
[35] RICHARD BOWEN W, JONES M G, WELFOOT J S, et al. Predicting salt rejections at nanofiltrationmembranes using artificial neural networks [J]. Desalination,2000,129(2):147-162.
[36] CABASSUD M, DELGRANGE-VINCENT N, CABASSUD C, et al. Neural networks: a tool toimprove UF plant productivity [J]. Desalination,2002,145(1-3):223-231.
[37] RAZAVI S M A, MOUSAVI S M, MORTAZAVI S A. Dynamic prediction of milk ultrafiltrationperformance: A neural network approach [J]. Chem. Eng. Sci.,2003,58(18):4185-4195.
[38] SHETTY G R, MALKI H, CHELLAM S. Predicting contaminant removal during municipal drinkingwater nanofiltration using artificial neural networks [J]. J. Membr. Sci.,2003,212(1-2):99-112.
[39]熊日华,王世昌.神经网络在膜技术中的应用[J].膜科学与技术,2003(06):44-48.
[40] OH H K, YU M J, GWON E M, et al. KNT-artificial neural network model for flux prediction ofultrafiltration membrane producing drinking water [J]. Water Sci Technol,2004,50(8):103-110.
[41] ABBAS A, AL-BASTAKI N. Modeling of an RO water desalination unit using neural networks [J].Chem. Eng. J.,2005,114(1-3):139-143.
[42] AYDINER C, DEMIR I, YILDIZ E. Modeling of flux decline in crossflow microfiltration using neuralnetworks: the case of phosphate removal [J]. J. Membr. Sci.,2005,248(1-2):53-62.
[43] BOWDEN G J, MAIER H R, DANDY G C. Input determination for neural network models in waterresources applications. Part2. Case study: forecasting salinity in a river [J]. Journal of Hydrology,2005,301(1-4):93-107.
[44] CHELLAM S. Artificial neural network model for transient crossflow microfiltration of polydispersedsuspensions [J]. J. Membr. Sci.,2005,258(1-2):35-42.
[45] CURCIO S, SCILINGO G, CALABR V, et al. Ultrafiltration of BSA in pulsating conditions: anartificial neural networks approach [J]. J. Membr. Sci.,2005,246(2):235-247.
[46] FU R-Q, XU T-W, PAN Z-X. Modelling of the adsorption of bovine serum albumin on porouspolyethylene membrane by back-propagation artificial neural network [J]. J. Membr. Sci.,2005,251(1-2):137-144.
[47] ZHAO Y, TAYLOR J S, CHELLAM S. Predicting RO/NF water quality by modified solution diffusionmodel and artificial neural networks [J]. J. Membr. Sci.,2005,263(1-2):38-46.
[48] CHEN H, KIM A S. Prediction of permeate flux decline in crossflow membrane filtration of colloidalsuspension: a radial basis function neural network approach [J]. Desalination,2006,192(1-3):415-428.
[49] CINAR O, HASAR H, KINACI C. Modeling of submerged membrane bioreactor treating cheese wheywastewater by artificial neural network [J]. J. Biotechnol.,2006,123(2):204-209.
[50] DARWISH N A, HILAL N, AL-ZOUBI H, et al. Neural Networks Simulation of the Filtration ofSodium Chloride and Magnesium Chloride Solutions Using Nanofiltration Membranes [J]. Chem. Eng.Res. Des.,2007,85(4):417-430.
[51] AIDAN A, ABDEL-JABBAR N, IBRAHIM T H, et al. Neural network modeling and optimization ofscheduling backwash for membrane bioreactor [J]. Clean Technol. Envir.,2008,10(4):389-395.
[52] HILAL N, OGUNBIYI O O, AL-ABRI M. Neural network modeling for separation of bentonite intubular ceramic membranes [J]. Desalination,2008,228(1-3):175-182.
[53] CURCIO S, CALABR V, IORIO G. Reduction and control of flux decline in cross-flow membraneprocesses modeled by artificial neural networks and hybrid systems [J]. Desalination,2009,236(1-3):234-243.
[54] YANAGAWA F, ONUKI Y, MORISHITA M, et al. Prediction of permeability of FD-4through porouspoly (2-hydroxyethyl methacrylate) membrane by multiple linear regression and artificial neuralnetwork [J]. Pharmazie,2009,64(5):311-315.
[55] SARGOLZAEI J, HAGHIGHI ASL M, HEDAYATI MOGHADDAM A. Membrane permeate flux andrejection factor prediction using intelligent systems [J]. Desalination,2012,284(0):92-99.
[56] ROSTAMIZADEH M, RIZI S M H. Predicting gas flux in silicalite-1zeolite membrane using artificialneural networks [J]. J. Membr. Sci.,2012,403–404(0):146-151.
[57] RAHMANIAN B, PAKIZEH M, MANSOORI S A A, et al. Prediction of MEUF process performanceusing artificial neural networks and ANFIS approaches [J]. Journal of the Taiwan Institute of ChemicalEngineers,2012,43(4):558-565.
[58] JAAFARZADEH N, AHMADI M, AMIRI H, et al. Predicting Fenton modification of solid wastevegetable oil industry for arsenic removal using artificial neural networks [J]. Journal of the TaiwanInstitute of Chemical Engineers,2012,43(6):873-878.
[59] RAHMANIAN B, PAKIZEH M, MANSOORI S A A, et al. Application of experimental designapproach and artificial neural network (ANN) for the determination of potential micellar-enhancedultrafiltration process [J]. J. Hazard. Mater.,2011,187(1–3):67-74.
[60] PIRON E, LATRILLE E, REN F. Application of artificial neural networks for crossflow microfiltrationmodelling:"black-box" and semi-physical approaches [J]. Comput. Chem. Eng.,1997,21(9):1021-1030.
[61] DELGRANGE N, CABASSUD C, CABASSUD M, et al. Modelling of ultrafiltration fouling by neuralnetwork [J]. Desalination,1998,118(1-3):213-227.
[62] TEODOSIU C, PASTRAVANU O, MACOVEANU M. Neural network models for ultrafiltration andbackwashing [J]. Water Res.,2000,34(18):4371-4380.
[63] SHETTY G R, CHELLAM S. Predicting membrane fouling during municipal drinking waternanofiltration using artificial neural networks [J]. J. Membr. Sci.,2003,217(1-2):69-86.
[64] LIU Q-F, KIM S-H, LEE S. Prediction of microfiltration membrane fouling using artificial neuralnetwork models [J]. Sep. Purif. Technol.,2009,70(1):96-102.
[65] CHOI Y-J, OH H, LEE S, et al. Investigation of the filtration characteristics of pilot-scale hollow fibersubmerged MF system using cake formation model and artificial neural networks model [J].Desalination,2012,297(0):20-29.
[66] GHANDEHARI S, MONTAZER-RAHMATI M M, ASGHARI M. A comparison betweensemi-theoretical and empirical modeling of cross-flow microfiltration using ANN [J]. Desalination,2011,277(1–3):348-355.
[67] TAVAKOLMOGHADAM M, SAFAVI M. An Optimized Neural Network Model of Desalination byVacuum Membrane Distillation Using Genetic Algorithm [J]. Procedia Engineering,2012,42(0):106-112.
[68] KHAYET M, COJOCARU C. Artificial neural network model for desalination by sweeping gasmembrane distillation [J]. Desalination,2013,308(0):102-110.
[69] JING G, DU W, GUO Y. Studies on prediction of separation percent in electrodialysis process via BPneural networks and improved BP algorithms [J]. Desalination,2012,291(0):78-93.
[70] GODINI H R, GHADRDAN M, OMIDKHAH M R, et al. Part II: Prediction of the dialysis processperformance using Artificial Neural Network (ANN)[J]. Desalination,2011,265(1–3):11-21.
[71] CHAKRABORTY M, BHATTACHARYA C, DUTTA S. Studies on the applicability of artificial neuralnetwork (ANN) in emulsion liquid membranes [J]. J. Membr. Sci.,2003,220(1-2):155-164.
[72] EL-SHARKH M Y, RAHMAN A, ALAM M S. Neural networks-based control of active and reactivepower of a stand-alone PEM fuel cell power plant [J]. J. Power Sources,2004,135(1-2):88-94.
[73] LEE W-Y, PARK G-G, YANG T-H, et al. Empirical modeling of polymer electrolyte membrane fuel cellperformance using artificial neural networks [J]. Int. J. Hydrogen Energy,2004,29(9):961-966.
[74] CHEN Y H, CAO G Y, ZHU X J. Neural network modeling and control of proton exchange membranefuel cell [J]. Journal of Central South University of Technology,2007,14(1):84-87.
[75] SAENGRUNG A, ABTAHI A, ZILOUCHIAN A. Neural network model for a commercial PEM fuelcell system [J]. J. Power Sources,2007,172(2):749-759.
[76] LOBATO J, CANIZARES P, RODRIGO M A, et al. The neural networks based modeling of apolybenzimidazole-based polymer electrolyte membrane fuel cell: Effect of temperature [J]. J. PowerSources,2009,192(1):190-194.
[77] WU S J, SHIAH S W, YU W L. Parametric analysis of proton exchange membrane fuel cellperformance by using the Taguchi method and a neural network [J]. Renewable Energy,2009,34(1):135-144.
[78] KHAJEH-HOSSEINI-DALASM N, AHADIAN S, FUSHINOBU K, et al. Prediction and analysis ofthe cathode catalyst layer performance of proton exchange membrane fuel cells using artificial neuralnetwork and statistical methods [J]. J. Power Sources,2011,196(8):3750-3756.
[79] MADAENI S S, ZAHEDI G, AMINNEJAD M. Artificial neural network modeling of O-2separationfrom air in a hollow fiber membrane module [J]. Asia-Pac. J. Chem. Eng.,2008,3(4):357-363.
[80] SHOKRIAN M, SADRZADEH M, MOHAMMADI T. C3H8separation from CH4and H2using asynthesized PDMS membrane: Experimental and neural network modeling [J]. J. Membr. Sci.,2010,346(1):59-70.
[81] MADAENI S S, GHESHLAGHI A, REKABDAR F. Membrane treatment of oily wastewater fromrefinery processes [J]. Asia-Pac. J. Chem. Eng.,2012: n/a-n/a.
[82]周激流,遗传算法理论及其在水问题中应用的研究[D].成都:四川大学,2000.
[83]金芬,遗传算法在函数优化中的应用研究[D].苏州:苏州大学,2008.
[84] JIAN X, DAI Y, HE G, et al. Preparation of UF and NF poly (phthalazine ether sulfone ketone)membranes for high temperature application [J]. J. Membr. Sci.,1999,161(1-2):185-191.
[85] GE J, CUI Y, YAN Y, et al. The effect of structure on pervaporation of chitosan membrane [J]. J. Membr.Sci.,2000,165(1):75-81.
[86] DAI Y, JIAN X, ZHANG S, et al. Thermostable ultrafiltration and nanofiltration membranes fromsulfonated poly(phthalazinone ether sulfone ketone)[J]. J. Membr. Sci.,2001,188(2):195-203.
[87]张欣宇,李荣先,王臻,等.氧化铝/二氧化钛复合膜的制备及其工艺优化[J].新技术新工艺,2006(08):24-28.
[88] ZHOU C, WANG Z, LIANG Y, et al. Study on the control of pore sizes of membranes using chemicalmethods Part II. Optimization factors for preparation of membranes [J]. Desalination,2008,225(1-3):123-138.
[89] SALJOUGHI E, SADRZADEH M, MOHAMMADI T. Effect of preparation variables on morphologyand pure water permeation flux through asymmetric cellulose acetate membranes [J]. J. Membr. Sci.,2009,326(2):627-634.
[90] BARMALA M, MOHEB A, EMADI R. Applying Taguchi method for optimization of the synthesiscondition of nano-porous alumina membrane by slip casting method [J]. J. Alloys Compd.,2009,485(1-2):778-782.
[91] KOOLS W F C, Membrane formation by phase inversion in multicomponent polymer systems.Mechanisms and morphologies [D]. Enschede: Chemical Engineering University of Twente,1998.
[92] RAHIMPOUR A, MADAENI S S, SHOCKRAVI A, et al. Preparation and characterization ofhydrophile nano-porous polyethersulfone membranes using synthesized poly(sulfoxide-amide) asadditive in the casting solution [J]. J. Membr. Sci.,2009,334(1-2):64-73.
[93] BAKERI G, ISMAIL A F, SHARIATY-NIASSAR M, et al. Effect of polymer concentration on thestructure and performance of polyetherimide hollow fiber membranes [J]. J. Membr. Sci.,2010,363(1-2):103-111.
[94] WANG Z, YU H, XIA J, et al. Novel GO-blended PVDF ultrafiltration membranes [J]. Desalination,2012,299:50-54.
[95] ALBRECHT W, WEIGEL T, SCHOSSIG-TIEDEMANN M, et al. Formation of hollow fibermembranes from poly(ether imide) at wet phase inversion using binary mixtures of solvents for thepreparation of the dope [J]. J. Membr. Sci.,2001,192(1-2):217-230.
[96] SHEN L-Q, XU Z-K, XU Y-Y. Phase separation behavior of poly(ether imide)/N,N-dimethylacetamide/nonsolvent systems [J]. J. Appl. Polym. Sci.,2003,89(4):875-881.
[97]富海涛,杨大令,张守海,等. PPESK纺丝液相分离行为与气体分离膜结构性能的关系[J].高分子学报,2007(07):615-620.
[98] SHI L, WANG R, CAO Y, et al. Effect of additives on the fabrication of poly(vinylidenefluoride-co-hexafluropropylene)(PVDF-HFP) asymmetric microporous hollow fiber membranes [J]. J.Membr. Sci.,2008,315(1–2):195-204.
[99] HAN J, YANG D, ZHANG S, et al. Effects of compatibility difference in the mixed solvent system onthe performance of PPES hollow fiber UF membrane [J]. J. Membr. Sci.,2010,365(1–2):311-318.
[100] REN J, ZHOU J, DENG M. Morphology transition of asymmetric polyetherimide flat sheet membraneswith different thickness by wet phase-inversion process [J]. Sep. Purif. Technol.,2010,74(1):119-129.
[101] ZHANG J, ZHANG Y, ZHAO J. Thermodynamic study of non-solvent/dimethylsulfoxide/polyacrylonitrile ternary systems: effects of the non-solvent species [J]. Polym. Bull.,2011,67(6):1073-1089.
[102] ALTENA F W, SMOLDERS C A. Calculation of liquid-liquid phase separation in a ternary system of apolymer in a mixture of a solvent and a nonsolvent [J]. Macromolecules,1982,15(6):1491-1497.
[103] YILMAZ L, MCHUGH A J. Analysis of nonsolvent–solvent–polymer phase diagrams and theirrelevance to membrane formation modeling [J]. J. Appl. Polym. Sci.,1986,31(4):997-1018.
[104] KIM J Y, LEE H K, BAIK K J, et al. Liquid-liquid phase separation in polysulfone/solvent/watersystems [J]. J. Appl. Polym. Sci.,1997,65(13):2643-2653.
[105] MATSUYAMA H, TERAMOTO M, NAKATANI R, et al. Membrane formation via phase separationinduced by penetration of nonsolvent from vapor phase. I. Phase diagram and mass transfer process [J].J. Appl. Polym. Sci.,1999,74(1):159-170.
[106]林阳政,薛亚卓,李继定,等.聚合物PPES/PPEK/PPESK-NMP-H_2O三元体系的相行为研究(Ⅱ)相图的计算[J].膜科学与技术,2008(03):33-38+44.
[107] VIVIER H, PONS M-N, PORTALA J-F. Study of microporous membrane structure by image analysis[J]. J. Membr. Sci.,1989,46(1):81-91.
[108] ZEMAN L. Characterization of microfiltration membranes by image analysis of electron micrographs.:Part II. Functional and morphological parameters [J]. J. Membr. Sci.,1992,71(3):233-246.
[109] ZEMAN L, DENAULT L. Characterization of microfiltration membranes by image analysis of electronmicrographs.: Part I. Method development [J]. J. Membr. Sci.,1992,71(3):221-231.
[110] CALVO J I, HERN NDEZ A, CARUANA G, et al. Pore Size Distributions in Microporous Membranes:I. Surface Study of Track-Etched Filters by Image Analysis [J]. J. Colloid Interface Sci.,1995,175(1):138-150.
[111] CALVO J I, HERN NDEZ A, PR DANOS P, et al. Pore Size Distributions in Microporous MembranesII. Bulk Characterization of Track-Etched Filters by Air Porometry and Mercury Porosimetry [J]. J.Colloid Interface Sci.,1995,176(2):467-478.
[112] KIM K J, STEVENS P V. Hydraulic and surface characteristics of membranes with parallel cylindricalpores [J]. J. Membr. Sci.,1997,123(2):303-314.
[113] MASSELIN I, DURAND-BOURLIER L, LAINE J-M, et al. Membrane characterization usingmicroscopic image analysis [J]. J. Membr. Sci.,2001,186(1):85-96.
[114] WANG L, WANG X. Study of membrane morphology by microscopic image analysis and membranestructure parameter model [J]. J. Membr. Sci.,2006,283(1-2):109-115.
[115] SUN W, CHEN T, CHEN C, et al. A study on membrane morphology by digital image processing [J]. J.Membr. Sci.,2007,305(1-2):93-102.
[116] WU Q, WU B. Study of membrane morphology by image analysis of electron micrographs [J]. J.Membr. Sci.,1995,105(1–2):113-120.
[117] SHE F H, TUNG K L, KONG L X. Calculation of effective pore diameters in porous filtrationmembranes with image analysis [J]. Robotics and Computer-Integrated Manufacturing,2008,24(3):427-434.
[118] ISMAIL A F, LAI P Y. Development of defect-free asymmetric polysulfone membranes for gasseparation using response surface methodology [J]. Sep. Purif. Technol.,2004,40(2):191-207.
[119]韩少卿,陶汝胜,彭奇均.聚砜超滤膜制备工艺的优化[J].江南大学学报(自然科学版),2004(04).
[120] IDRIS A, KORMIN F, NOORDIN M Y. Application of response surface methodology in describing theperformance of thin film composite membrane [J]. Sep. Purif. Technol.,2006,49(3):271-280.
[121] XIANGLI F, WEI W, CHEN Y, et al. Optimization of preparation conditions for polydimethylsiloxane(PDMS)/ceramic composite pervaporation membranes using response surface methodology [J]. J.Membr. Sci.,2008,311:23-33.
[122]谭翎燕,张浩勤,刘金盾.用均匀实验设计法优化磺化聚醚砜超滤膜制膜工艺[J].膜科学与技术,2008,28(03):54-58.
[123] YI S, SU Y, QI B, et al. Application of response surface methodology and central composite rotatabledesign in optimizing the preparation conditions of vinyltriethoxysilane modifiedsilicalite/polydimethylsiloxane hybrid pervaporation membranes [J]. Sep. Purif. Technol.,2010,71(2):252-262.
[124] MADAENI S S, ARAST N, RAHIMPOUR F, et al. Fabrication optimization of acrylonitrile butadienestyrene (ABS)/polyvinylpyrrolidone (PVP) nanofiltration membrane using response surfacemethodology [J]. Desalination,2011,280(1–3):305-312.
[125]宋来洲,白明华,盛义平,等.聚丙烯酸／聚偏氟乙烯中空纤维微滤复合膜的制备工艺[J].2006.
[126]刘臻,袁其朋,杜雪岭.神经网络优化复合纳滤膜制备工艺的研究[J].计算机与应用化学,2007,24(08):1055-1058.
[127] MADAENI S S, HASANKIADEH N T, KURDIAN A R, et al. Modeling and optimization ofmembrane fabrication using artificial neural network and genetic algorithm [J]. Sep. Purif. Technol.,2010,76(1):33-43.
[128]陈桂娥,许振良,施亚钧.相转化法制备聚醚酰亚胺超滤膜[J].华东理工大学学报,1998(05):16-20.
[129]许振良, T.S.CHUNG.聚醚酰亚胺中空纤维超滤膜的研究(Ⅰ)膜结构、热性能和机械性能[J].膜科学与技术,2001,21(2):33-36.
[130] KHAYET M, FENG C Y, MATSUURA T. Morphological study of fluorinated asymmetricpolyetherimide ultrafiltration membranes by surface modifying macromolecules [J]. J. Membr. Sci.,2003,213(1-2):159-180.
[131]陈桂娥,许振良.聚醚酰亚胺中空纤维气体分离膜及结构[J].膜科学与技术,2003,23(05):1-4.
[132] KIM I-C, LEE K-H. Effect of poly(ethylene glycol)200on the formation of a polyetherimideasymmetric membrane and its performance in aqueous solvent mixture permeation [J]. J. Membr. Sci.,2004,230(1-2):183-188.
[133]胡亮平,贺高红,张玲玲,等.聚醚酰亚胺平板超滤膜的制备[J].高分子材料科学与工程,2008(11):169-171+175.
[134]沈立强,聚醚酰亚胺中空纤维超滤膜的研究[D].杭州:浙江大学,2002.
[135] WANG D, LI K, TEO W K. Phase Separation in Polyetherimide/Solvent/Nonsolvent [J]. J. Appl. Polym.Sci.,1998,71:1789-1796.
[136]张玮,金属腐蚀形貌特征提取用于腐蚀诊断的研究[D].大连:大连理工大学,2005.
[137]吴冰,秦志远.自动确定图像二值化最佳阈值的新方法[J].测绘学院学报,2001(04):283-286.
[138]徐晓明,指纹图像的预处理及特征提取[D].大连:大连理工大学,2005.
[139]齐丽娜,张博,王战凯.最大类间方差法在图像处理中的应用[J].无线电工程,2006(07):25-26+44.
[140] OTSU N. A Threshold Selection Method from Gray-Level Histograms [J]. IEEE Trans. Sys., Man.,Cyber,1979,9(1):62-66.
[141]邢藏菊,王守觉,邓浩江,等.一种基于极值中值的新型滤波算法[J].中国图象图形学报,2001(06):25-28.
[142]高浩军,杜宇人.中值滤波在图像处理中的应用[J].电子工程师,2004(08):35-36.
[143] WIKIPEDIA,对数正态分布,2013,http://zh.wikipedia.org/zh/%E5%AF%B9%E6%95%B0%E6%AD%A3%E6%80%81%E5%88%86%E5%B8%83.
[144] FERRY J D. Ultrafilter Membranes and Ultrafiltration [J]. Chem. Rev.,1936,18(3):373-455.
[145] FERRY J D. Statistical evaluation of sieve constants in ultrafiltration [J]. The Journal of GeneralPhysiology,1936,20(1):95-104.
[146] RENKIN E M. Filtration, diffusion, and molecular sieving through porous cellulose membranes [J].The Journal of General Physiology,1954,39(5):225-243.
[147] LADENBURG R. über die innere Reibung z her Flüssigkeiten und ihre Abh ngigkeit vom Druck [J].Annalen der Physik,1907,327(2):287-309.
[148] BACON L R. Measurement of absolute viscosity by the falling sphere method [J]. Journal of theFranklin Institute,1936,221(2):251-273.
[149] BELL D W. Ceramic ultrafiltration membranes with photocatalytic properties [M]. Wisconsin, USA:University of Wisconsin--Madison,1999:9.
[150] SINGH S, KHULBE K C, MATSUURA T, et al. Membrane characterization by solute transport andatomic force microscopy [J]. J. Membr. Sci.,1998,142(1):111-127.
[151] TYN M T, GUSEK T W. Prediction of diffusion coefficients of proteins [J]. Biotechnol. Bioeng.,1990,35(4):327-338.
[152]司徒勇.溶解度参数的基本原理和应用[M].上海:上海大学,2007:15-16.
[153] BOOM R M, VAN DEN BOOMGAARD T, SMOLDERS C A. Equilibrium Thermodynamics of aQuaternary Membrane-Forming System with Two Polymers.1. Calculations [J]. Macromolecules,1994,27(8):2034-2040.
[154]赵长生,钟银屏,欧阳庆,等.聚醚砜制膜液的相分离2.四元体系[J].水处理技术,1997(01):23-26.
[155]钟银屏,赵长生,欧阳庆,等.聚醚砜四元制膜液体系的相图计算[J].膜科学与技术,1997(01):63-68.
[156]魏永明,许振良,杨晓天.水-N-甲基吡咯烷酮-添加剂(小分子醇类)-聚醚砜四元体系相图计算[J].华东理工大学学报(自然科学版),2006(03):245-248+258.
[157] LINDVIG T, MICHELSEN M L, KONTOGEORGIS G M. A Flory–Huggins model based on theHansen solubility parameters [J]. Fluid Phase Equilib.,2002,203(1-2):247-260.
[158] WEI Y-M, XU Z-L, YANG X-T, et al. Mathematical calculation of binodal curves of apolymer/solvent/nonsolvent system in the phase inversion process [J]. Desalination,2006,192(1–3):91-104.
[159] MIKOS A G, PEPPAS N A. Flory interaction parameter χ for hydrophilic copolymers with water [J].Biomaterials,1988,9(5):419-423.
[160] PEPPAS N A, LISA B-P. Controlled release of fragrances from polymers I. Thermodynamic analysis [J].J. Controlled Release,1996,40(3):245-250.
[161]许元泽,赵得禄,吴大诚,等.聚合物的性质性质的估算及其与化学结构的关系[M].北京:科学出版社,1981:114-115.
[162]孙树东, PSF-DMAc-IBA体系制得的高度非对称聚砜膜[D].成都:四川大学,2003.
[163] HANSEN C M. The Universality of the Solubility Parameter [J]. Product R&D,1969,8(1):2-11.
[164] LIU Y, SHI B. Determination of Flory interaction parameters between polyimide and organic solventsby HSP theory and IGC [J]. Polym. Bull.,2008,61(4):501-509.
[165] CHUNG T-S, XU Z-L. Asymmetric hollow fiber membranes prepared from misciblepolybenzimidazole and polyetherimide blends [J]. J. Membr. Sci.,1998,147(1):35-47.
[166] FREDENSLUND A, JONES R L, PRAUSNITZ J M. Group-contribution estimation of activitycoefficients in nonideal liquid mixtures [J]. AICHE J.,1975,21(6):1086-1099.
[167]朱自强.流体相平衡原理及其应用[M].杭州:浙江大学出版社,1990:322-349.
[168] LI S, JIANG C, ZHANG Y. The investigation of solution thermodynamics for thepolysufone-DMAC-water system [J]. Desalination,1987,62(0):79-88.
[169] LAI J-Y, LIN F-C, WANG C-C, et al. Effect of nonsolvent additives on the porosity and morphology ofasymmetric TPX membranes [J]. J. Membr. Sci.,1996,118(1):49-61.
[170] LIU L, CHAKMA A, FENG X. A novel method of preparing ultrathin poly(ether block amide)membranes [J]. J. Membr. Sci.,2004,235(1–2):43-52.
[171] CHEN S-H, LIOU R-M, LAI J-Y, et al. Effect of the polarity of additional solvent on membraneformation in polysulfone/N-methyl-2-pyrrolidone/water ternary system [J]. Eur. Polym. J.,2007,43(9):3997-4007.
[172] ROSENBLATT F, The perception: a probabilistic model for information storage and organization in thebrain, in Neurocomputing: foundations of research, James A A, Edward R, Editors.1988, MIT Press:Cambridge, Massachusetts, USA. p.89-114.
[173] FENG X, HUANG R Y M. Preparation and performance of asymmetric polyetherimide membranes forisopropanol dehydration by pervaporation [J]. J. Membr. Sci.,1996,109(2):165-172.
[174] BINDAL R C, HANRA M S, MISRA B M. Novel solvent exchange cum immersion precipitationtechnique for the preparation of asymmetric polymeric membrane [J]. J. Membr. Sci.,1996,118(1):23-29.
[175] KIM I-C, YOON H-G, LEE K-H. Formation of integrally skinned asymmetric polyetherimidenanofiltration membranes by phase inversion process [J]. J. Appl. Polym. Sci.,2002,84(6):1300-1307.
[176]王国庆,张守海,杨大令,等.新型聚芳醚腈酮超滤膜制备:非溶剂添加剂的影响[J].功能高分子学报,2005(01):105-110.
[177]孙本惠.热力学及动力学因素对L-S相转化法制备非对称膜结构与性能的影响[J].膜科学与技术,2011, v.31;No.146(01):1-11.
[178] WU L, SUN J, WANG Q. Poly(vinylidene fluoride)/polyethersulfone blend membranes: Effects ofsolvent sort, polyethersulfone and polyvinylpyrrolidone concentration on their properties andmorphology [J]. J. Membr. Sci.,2006,285(1–2):290-298.
[179] CABASSO I, KLEIN E, SMITH J K. Polysulfone hollow fibers. I. Spinning and properties [J]. J. Appl.Polym. Sci.,1976,20(9):2377-2394.
[180] WIENK I M, OLDE SCHOLTENHUIS F H A, VAN DEN BOOMGAARD T, et al. Spinning of hollowfiber ultrafiltration membranes from a polymer blend [J]. J. Membr. Sci.,1995,106(3):233-243.
[181] CRANFORD R J, DARMSTADT H, YANG J, et al. Polyetherimide/polyvinylpyrrolidone vaporpermeation membranes. Physical and chemical characterization [J]. J. Membr. Sci.,1999,155(2):231-240.
[182] XU J, XU Z-L. Poly(vinyl chloride)(PVC) hollow fiber ultrafiltration membranes prepared fromPVC/additives/solvent [J]. J. Membr. Sci.,2002,208(1–2):203-212.
[183] ARTHANAREESWARAN G, SRIYAMUNA DEVI T K, MOHAN D. Development, characterizationand separation performance of organic–inorganic membranes: Part II. Effect of additives [J]. Sep. Purif.Technol.,2009,67(3):271-281.
[184] KNEIFEL K, PEINEMANN K V. Preparation of hollow fiber membranes from polyetherimide for gasseparation [J]. J. Membr. Sci.,1992,65(3):295-307.
[185] KIM I-C, LEE K-H. Effect of various additives on pore size of polysulfone membrane byphase-inversion process [J]. J. Appl. Polym. Sci.,2003,89(9):2562-2566.
[186] YUN Y, TIAN Y, SHI G, et al. Preparation, morphologies and properties for flat sheet PPESKultrafiltration membranes [J]. J. Membr. Sci.,2006,270(1-2):146-153.
[187] XIANGLI F, CHEN Y, JIN W, et al. Polydimethylsiloxane (PDMS)/Ceramic Composite Membranewith High Flux for Pervaporation of Ethanol/water Mixtures [J]. Ind. Eng. Chem. Res.,2007,46(7):2224-2230.
[188] CARLEY K M, KAMNEVA N Y, REMINGA J. CASOS Technical report: Response surfacemethodology [M]. Pittsburgh, USA: Carnegie Mellon University,2004:2.
[189] JENNY N M, FOLEY G. Dead-end filtration of yeast suspensions: Correlating specific resistance andflux data using artificial neural networks [J]. J. Membr. Sci.,2006,281(1-2):325-333.
[190] KIM H, KIM H-G, KIM S, et al. PDMS–silica composite membranes with silane coupling forpropylene separation [J]. J. Membr. Sci.,2009,344(1-2):211-218.
[191] GUO J X, ZHANG G J, WU W, et al. Dynamically formed inner skin hollow fiberpolydimethylsiloxane/polysulfone composite membrane for alcohol permselective pervaporation [J].Chem. Eng. J.,2010,158(3):558-565.
[192] WANG X-P. Modified alginate composite membranes for the dehydration of acetic acid [J]. J. Membr.Sci.,2000,170(1):71-79.
[193] HYDER M, HUANG R, CHEN P. Composite poly(vinyl alcohol)–poly(sulfone) membranescrosslinked by trimesoyl chloride: Characterization and dehydration of ethylene glycol–water mixtures[J]. J. Membr. Sci.,2009,326(2):363-371.
[194] CHEN J, LI J, CHEN J, et al. Pervaporation separation of ethyl thioether/heptane mixtures bypolyethylene glycol membranes [J]. Sep. Purif. Technol.,2009,66(3):606-612.
[195] KHAYET M, COJOCARU C. Artificial neural network modeling and optimization of desalination byair gap membrane distillation [J]. Sep. Purif. Technol.,2012,86(0):171-182.