基于BP神经网络的聚偏氟乙烯/聚丙烯梯度复合滤料工艺优化

康乐; 王立志; 高晓平

doi:10.13801/j.cnki.fhclxb.20210913.005

基于BP神经网络的聚偏氟乙烯/聚丙烯梯度复合滤料工艺优化

doi: 10.13801/j.cnki.fhclxb.20210913.005

1.
内蒙古工业大学　轻工与纺织学院，呼和浩特 010010
2.
内蒙古自治区纤维质量监测中心，呼和浩特 010010

基金项目: 内蒙古科技计划项目(2020GG0282)

详细信息

通讯作者:
高晓平，博士，教授，博士生导师，研究方向为静电纺纳米纤维及碳玻复合材料　 E-mail: gaoxp@imut.edu.cn

中图分类号: TS174.8
计量
- 文章访问数: 785
- HTML全文浏览量: 266
- PDF下载量: 46
- 被引次数: 0
出版历程
- 收稿日期: 2021-08-03
- 修回日期: 2021-08-19
- 录用日期: 2021-08-27
- 网络出版日期: 2021-09-14
- 刊出日期: 2022-08-31

Process optimization of polyvinylidene fluoride/polypropylene gradient composite filter media based on BP neural network

1.
College of Light Industry and Textile, Inner Mongolia University of Technology, Hohhot 010010, China
2.
Inner Mongolia Fiber Quality Monitoring Center, Hohhot 010010, China

摘要

摘要: 口罩是防止病毒通过呼吸系统和黏膜进入人体的重要防疫屏障。一次性口罩存在过滤效率随静电衰减下降快、呼吸阻力大、使用寿命短等问题。将静电纺纳米纤维膜与熔喷布复合，减少颗粒物过滤性能对静电作用的依赖，实现长效过滤。以N, N-二甲基甲酰胺(DMF)为溶剂，基于静电纺丝技术制备聚偏氟乙烯(PVDF)纳米纤维膜，与聚丙烯(PP)熔喷基布覆合，制备PVDF/PP纳/微米复合纤维膜。实验研究静电纺丝工艺参数对复合结构纤维膜气溶胶过滤性能的影响规律。建立三元二次多项式模型优化纺丝工艺，同时构建反向传播(BP)神经网络模型，预测不同工艺下的纤维膜过滤阻力。结果表明，电压、接收距离、注射速度、纺丝液浓度和纤维膜面密度对过滤效率和过滤阻力有着一致的影响规律。纺丝液浓度为15wt%、面密度为3 g/m²时，优化纺丝工艺参数为：电压30 kV，接收距离16.8 cm，注射速度1.6 mL/h。应用多项式模型预测的过滤阻力值为76.79 Pa，相对误差为9.23%，误差变异系数(CV)值为59%。BP神经网络预测的过滤阻力值为81.25 Pa，相对误差为1.99%，误差CV值为48%。实验证明，三元二次模型和BP神经网络具有较高的预测准确度。
- 纳/微复合纤维膜 /
- 静电纺丝工艺优化 /
- 空气过滤 /
- 非线性拟合 /
- 三元二次多项式 /
- BP神经网络
Abstract: Mask was an important epidemic prevention barrier to prevent virus from entering human body through respiratory system and mucous membrane. The disposable mask had some problems, such as rapid decline of filtration efficiency with electrostatic attenuation, large respiratory resistance, short service life and so on. Electrospun nanofiber membrane was compounded with melt blown cloth to reduce the dependence of particle filtration on static electricity and realize long-term filtration. Polyvinylidene fluoride (PVDF) nanofiber membrane was prepared by electrospinning with N, N-dimethylformamide (DMF) as solvent. Then it was coated with polypropylene (PP) melt blown base cloth to prepare PVDF/PP nano/micron structure composite fiber membrane. The effect of electrospinning process parameters on the aerosol filtration performance of composite fiber membrane was experimentally studied. The ternary quadratic polynomial model was established to optimize the spinning process and predict the fiber membrane resistance. At the same time, the back propagation (BP) neural network model was constructed to predict the fiber membrane resistance. The results show that the effects of voltage, receiving distance, injection speed, spinning solution concentration and fiber membrane surface density on the filtration efficiency and filtration resistance are consistent. When the concentration of spinning solution is 15wt% and the area density is 3 g/m², the optimized spinning process parameters are voltage of 30 kV, receiving distance of 16.8 cm and injection speed of 1.6 mL/h. The filtration resistance predicted by polynomial model is 76.79 Pa, the relative error is 9.23%, and the error coefficient of variation (CV) value is 59%. The filtration resistance predicted by BP neural network is 81.25 Pa, the relative error is 1.99%, and the error CV value is 48%. The experiments show that the ternary quadratic model and BP neural network have high prediction accuracy.
- nano/micro composite fiber membrane /
- optimization of electrospinning parameters /
- air filtration /
- nonlinear fitting /
- ternary quadratic polynomial model /
- BP neural network

HTML全文

图 1 人工神经网络结构

Figure 1. Structure of back propagation neural network

x—Input layer; m—Hidden layer; y—Output layer

下载: 全尺寸图片幻灯片

图 2 不同纺丝液浓度下聚偏氟乙烯(PVDF)纤维形貌

Figure 2. Morphologies of polyvinylidene fluoride (PVDF) fiber under different concentration of spinning solution

下载: 全尺寸图片幻灯片

图 3 电压对聚偏氟乙烯/聚丙烯纳/微米复合纤维膜(PVDF/PP纳/微米复合纤维膜(NMCM))过滤效率及阻力影响

Figure 3. Effect of voltage on filtration efficiency and resistance of polyvinylidene fluoride/polypropylene nano/micro composite membrane (PVDF/PP nano/micro composite membrane (NMCM))

下载: 全尺寸图片幻灯片

图 4 接收距离对NMCM过滤效率及阻力影响

Figure 4. Effect of receiving distance on filtration efficiency and resistance of NMCM

下载: 全尺寸图片幻灯片

图 5 注射速度对NMCM过滤效率及阻力影响

Figure 5. Effect of injection speed on filtration efficiency and resistance of NMCM

下载: 全尺寸图片幻灯片

图 6 纺丝液浓度对NMCM过滤效率及阻力影响

Figure 6. Effect of spinning solution concentration on filtration efficiency and resistance of NMCM

下载: 全尺寸图片幻灯片

图 7 面密度对NMCM过滤效率及过滤阻力影响

Figure 7. Effect of area density on filtration efficiency and resistance of NMCM

下载: 全尺寸图片幻灯片

图 8 NMCM纺丝工艺的反向传播 (BP) 神经网络拟合

Figure 8. Fitting of back propagation (BP) neural network for NMCM spinning process

R—Fit coefficient

下载: 全尺寸图片幻灯片

表 1 因子水平

Table 1. Factor level

Level	Voltage /kV	Receiving distance/cm	Injection speed /(mL·h⁻¹)	Concentration/wt%	Area density/(g·m⁻²)
1	25	10	0.5	11	1
2	26	13	1.0	12	2
3	27	16	1.5	13	3
4	28	19	2.0	14	4
5	29	22	2.5	15	5

下载: 导出CSV

表 2 实验方案

Table 2. Experimental scheme

Group	Voltage/kV	Receiving distance/cm	Injection speed /(mL·h⁻¹)	Concentration/wt%	Area density/(g·m⁻²)
1	25	16	1.5	12	3
2	26	16	1.5	12	3
3	27	16	1.5	12	3
4	28	16	1.5	12	3
5	29	16	1.5	12	3
6	27	10	1.5	12	3
7	27	13	1.5	12	3
8	27	16	1.5	12	3
9	27	19	1.5	12	3
10	27	22	1.5	12	3
11	27	16	0.5	12	3
12	27	16	1.0	12	3
13	27	16	1.5	12	3
14	27	16	2.0	12	3
15	27	16	2.5	12	3
16	27	16	1.5	11	3
17	27	16	1.5	12	3
18	27	16	1.5	13	3
19	27	16	1.5	14	3
20	27	16	1.5	15	3
21	27	16	1.5	12	1
22	27	16	1.5	12	2
23	27	16	1.5	12	3
24	27	16	1.5	12	4
25	27	16	1.5	12	5

下载: 导出CSV

表 3 NMCM纺丝工艺优化模型验证

Table 3. Verification of NMCM spinning process optimization model

Group	Actual value	Predicted value of filtration resistance /Pa		Absolute value of relative error /%
Group	Actual value	Ternary quadratic polynomial	BP neural network	Ternary quadratic polynomial	BP neural network
1	214.97	210.23	219.00	2.21	1.88
2	143.43	160.32	143.50	11.77	0.05
3	145.77	128.06	143.75	12.15	1.38
4	100.73	113.46	102.50	12.63	1.75
5	119.17	116.51	122.00	2.23	2.38
6	131.13	136.37	133.50	4.00	1.80
7	143.70	134.60	147.00	6.34	2.30
8	145.77	128.06	143.75	12.15	1.38
9	102.17	116.77	104.50	14.29	2.28
10	107.33	100.72	110.00	6.16	2.48
11	125.93	119.31	128.50	5.26	2.04
12	108.50	130.69	110.50	20.45	1.84
13	145.77	128.06	143.75	12.15	1.38
14	97.90	111.43	102.50	13.82	4.70
15	83.10	80.80	85.00	2.77	2.29

下载: 导出CSV

参考文献(21)

[1]	张煌忠. 聚丙烯熔喷非织造材料的空气过滤效果研究[J]. 合成纤维工业, 2015, 38(4):28-30. ZHANG Huangzhong. Air filtration efficiency of melt-blown polypropylene nonwoven[J]. China Synthetic Fiber Industry,2015,38(4):28-30(in Chinese).
[2]	侯冠一, 武文杰, 万海肖, 等. 口罩聚丙烯熔喷布的静电机理及其影响因素的研究进展[J]. 高分子通报, 2020(8): 1-22. HOU Guanyi, WU Wenjie, WAN Haixiao, et al. Research progress of static-electricity mechanism and influencing factors of polypropylene melt-blown nonwovens in mask[J]. Polymer Bulletin, 2020(8): 1-22(in Chinese).
[3]	刘学洋. PVDF/PSU复合抗菌纳米纤维空气过滤材料的制备及其在口罩中的应用研究[D]. 上海: 东华大学, 2016. LIU Xueyang. Fabrication of antibacterial polyvinylidenefluoride/polysulfone nanofibrous membranes for air filtration and application in masks[D]. Shanghai: Donghua University, 2016(in Chinese).
[4]	王珊. 聚偏氟乙烯/聚四氟乙烯复合驻极纳米纤维膜的制备及其空气过滤性能研究[D]. 上海: 东华大学, 2017. WANG Shan. Electret polyvinylidene fluoride nanofibers/polytetrafluoroethylene composite fibrous membrane for air filtration[D]. Shanghai: Donghua University, 2017(in Chinese).
[5]	刘延波, 陈志军, 郝铭, 等. 电纺膜替代荷电熔喷布解决口罩短缺问题的可能性分析[J]. 纺织导报, 2020, 3(7):64-70. LIU Yanbo, CHEN Zhijun, HAO Ming, et al. Possibility analysis on replacement of electrostatic melt-blown nonwoven fabric to electrospun membraneto solve mask shortage problem[J]. China Textile Leader,2020,3(7):64-70(in Chinese).
[6]	丁彬, 斯阳, 俞建勇. 静电纺纳米纤维材料在环境领域中的应用研究进展[J]. 中国材料进展, 2013, 32(8):492-502. DING Bin, SI Yang, YU Jianyong. Progress in the research of electrospun nanofibers for environmental applications[J]. Materials China,2013,32(8):492-502(in Chinese).
[7]	付文丽, 康为民, 程博闻, 等. 静电纺纳米纤维的工艺原理、应用及发展前景[J]. 现代纺织技术, 2009, 17(1):51-54. FU Wenli, KANG Weimin, CHENG Bowen, et al. Process principle, application and development prospect of electrospun nanofibers[J]. Advanced Textile Technology,2009,17(1):51-54(in Chinese).
[8]	康卫民, 程博闻, 庄旭品, 等. 静电纺纳米级纤维复合膜及其过滤性能[J]. 纺织学报, 2006, 27(10):10-12. KANG Weimin, CHENG Bowen, ZHUANG Xupin, et al. Electrospun nano-fiber composite membrane and its filtration properties[J]. Journal of Textile Research,2006,27(10):10-12(in Chinese).
[9]	FACCINI M, AMANTIA D, AUBOUY L, 等. 基于静电纺纳米纤维的纺织防护材料[J]. 国际纺织导报, 2011, 39(9):68. doi: 10.3969/j.issn.1007-6867.2011.09.019 FACCINI M, AMANTIA D, AUBOUY L, et al. Textile protec-tive materials based on electrospun nanofibers[J]. Melliand China,2011,39(9):68(in Chinese). doi: 10.3969/j.issn.1007-6867.2011.09.019
[10]	王利娜, 娄辉清, 辛长征, 等. 空气过滤用电纺聚偏氟乙烯-聚丙烯腈/熔喷聚丙烯无纺布复合材料的制备及过滤性能[J]. 复合材料学报, 2019, 36(2):277-282. WANG Lina, LOU Huiqing, XIN Changzheng, et al. Preparation and filtration properties of electrospun poly (vinylidene fluoride)-polyacrylonitrile/melt-blow polypropylene nonwoven composite filtration materials[J]. Acta Materiae Compositae Sinica,2019,36(2):277-282(in Chinese).
[11]	李娟, 章明清, 章赞德, 等. 三元非结构肥效模型提高水稻施肥推荐的可靠性[J]. 植物营养与肥料学报, 2019, 25(2):149-158. LI Juan, ZHANG Mingqing, ZHANG Zande, et al. Increasing precision of fertilizer recommendation using ternary non-structural fertilizer response model[J]. Journal of Plant Nutrition and Fertilizers,2019,25(2):149-158(in Chinese).
[12]	王蕊, 董祥旻, 何卫苹. 一种多元非线性回归模型的建立方法及其应用[J]. 中国考试, 2010(11):17-22. WANG Rui, DONG Xiangming, HE Weiping. A kind of multiple nonlinear regression model and its application[J]. China Examinations,2010(11):17-22(in Chinese).
[13]	刘天舒. BP神经网络的改进研究及应用[D]. 哈尔滨: 东北农业大学, 2011. LIU Tianshu. The research and application on BP nerual network improvement[D]. Harbin: Northeast Agricultural University, 2011(in Chinese).
[14]	中国国家标准化管理委员会. 医用防护口罩技术要求: GB 19083—2010[S]. 北京: 中国标准出版社, 2010. Standardization Administration of the People’s Republic of China. Technical requirements for protective face mask for medical use: GB 19083—2010[S]. Beijing: China Standards Press, 2010(in Chinese).
[15]	李萍, 曾令可, 税安泽, 等. 基于MATLAB的BP神经网络预测系统的设计[J]. 计算机应用与软件, 2008, 25(4):149-150. LI Ping, ZENG Lingke, SHUI Anze, et al. Design of forecast system of back propagation neural network based on matlab[J]. Computer Applications and Software,2008,25(4):149-150(in Chinese).
[16]	张媛. 静电纺载药串珠纤维形态和尺寸的可控性研究[D]. 上海: 东华大学, 2015. ZHANG Yuan. Research on the morphology and size controlling of electrospun beaded nanofibers for drug encapsulation[D]. Shanghai: Donghua University, 2015(in Chinese).
[17]	梁幸幸, 杨颖, 许德平. 静电纺丝法制备空气过滤膜及应用[J]. 化学工业与工程, 2015, 32(3):59-67. LIANG Xingxing, YANG Ying, XU Deping. Preparation and application of air filtration membrane by electrospinning[J]. Chemical Industry and Engineering,2015,32(3):59-67(in Chinese).
[18]	覃小红, 李妮, 杨恩龙, 等. 静电纺纳米纤维毡在新型高效过滤器上使用的优势与前景[J]. 产业用纺织品, 2007, 4(199):33-37. QIN Xiaohong, LI Ni, YANG Enlong, et al. The advantage and foreground of electrospinning nanofiber mats applied on the high-effect filters[J]. Technical Textiles,2007,4(199):33-37(in Chinese).
[19]	尹桂波, 臧传峰. 聚丙烯腈静电纺/聚丙烯熔喷复合材料的制备及过滤性能研究[J]. 纺织科技进展, 2017(11):17-21. YIN Guibo, ZANG Chuanfeng. Preparation and filtering performance of PAN electro-spun/PP melt-blown compo-sites[J]. Progress in Textile Science & Technology,2017(11):17-21(in Chinese).
[20]	王磊, 王汝凉, 曲洪峰, 等. BP神经网络算法改进及应用[J]. 软件导刊, 2016, 15(5):38-40. WANG Lei, WANG Ruliang, QU Hongfeng, et al. Improved BP neural network algorithm and its application[J]. Software Guide,2016,15(5):38-40(in Chinese).
[21]	潘文婵, 刘尚东. BP神经网络的优化研究与应用[J]. 计算机技术与发展, 2019, 29(5):74-76. PAN Wenchan, LIU Shangdong. Optimization research and application of BP neural network[J]. Computer Technology and Development,2019,29(5):74-76(in Chinese).