基于神经网络的石墨烯弹性参量识别方法研究
|
摘要
石墨烯的弹性参量是准确研究其力学性能的前提和基础.将神经网络的BP算法应用于石墨烯弹性模量和剪切弹性模量的预测,考虑石墨烯薄膜的长度、宽度、长宽比、手性、层数和温度6个影响因素,通过选取84组训练和检验样本,建立了石墨烯弹性参量的BP神经网络预测模型.将预测结果进行误差分析,其平均相对误差均小于3%,从而验证了该方法的适用性和可行性.将训练好的网络模型进行扩展计算,基于L_(25)(5~6)正交表试验理论分析了石墨烯弹性参量对各影响因素的敏感性.为同类材料性能的预测提供了参考.
Elastic parameters of graphene is the premise and foundation for the research of its material mechanics performances. The BP neural network is used to predict the elastic modulus and shear modulus of graphene. Considering the length, width, aspect ratio, chiral, layers and temperature of graphene as the main influence factors and choosing 84 groups of data as training and forecasting sample, BP neural network model is established. The errors of forecasting results are analyzed, and the average relative errors are less than 3 %, which proves the applicability and feasibility of this method. Based on the calculation results, the sensitivity of influence factor to the graphene elastic parameter is analyzed by using L_(25)(5~6) orthogonal table, which may provide a reference to the performance prediction of similar material.
Elastic parameters of graphene is the premise and foundation for the research of its material mechanics performances. The BP neural network is used to predict the elastic modulus and shear modulus of graphene. Considering the length, width, aspect ratio, chiral, layers and temperature of graphene as the main influence factors and choosing 84 groups of data as training and forecasting sample, BP neural network model is established. The errors of forecasting results are analyzed, and the average relative errors are less than 3 %, which proves the applicability and feasibility of this method. Based on the calculation results, the sensitivity of influence factor to the graphene elastic parameter is analyzed by using L_(25)(5~6) orthogonal table, which may provide a reference to the performance prediction of similar material.
引文
[1]NOVSELOW K.S.,GEIM A.K.,MOROZOW S.V.,et al.Electric field effect in atomically thin carbon films[J].Science,2004,(306):666-669.
[2]CORDELIA Sealy.Graphene goes from strength to strength[J].Materials Today,2008,11(9):12-18.
[3]LI Peiyuan,XIE Zhijiang,LI Xinxia.Research into fault diagnosis of large rotating machinery on BP network and the data source of network[J].Journal of Southwest University for Nationalities Natural Science Edition,2004,30(3):386-390.
[4]SHOKRI S H M,SHOKRI E H,ROHAM Raffiee.Prediction of Young’s modulus of graphene sheets and carbon nanotubes using nanoscale continuum mechanics approach[J].Materials and Design,2010,31(2):790-795.
[5]余晓红.BP神经网络的MATLAB编程实现及讨论[J].浙江交通职业技术学院学报,2007,8(4):45-48.YU Xiaohong.Matlab implementation and discussion of BP neural network[J].Journal of Zhejiang Institute of Communications,2007,8(4):45-48.
[6]Venkatesh.Predicting the mechanical characteristics of hydrogen functionalied graphene sheets using artificial netural network approach[J].Journal Of Nanostructure in Chemistry,2013,(3):83-87.
[7]尹海莲,胡自力.基于BP神经网络的复合材料性能预测[J].南京航空航天大学学报,2006,38(2):234-238.YI Hailian,HU Zili.Prediction of composite material properties basedon bp algorithm of artificial neutral etwork[J].Journal of Nanjing University of Aeronautics&Astronautics,2006,38(2):234-238.
[8]白光辉,孟鹤松,杜善文,等.基于神经网络炭/炭复合材料烧蚀性能预测[J].复合材料学报,2007,26(4):83-88.BAI Guanghui,MENG He Song,DU Shanwen,et al.Prediction on the ablative performance of carbon/carbon composites based on artificial neutral network[J].Acta Materie Compositae Sinica,2007,26(4):83-88.
[9]李东波.基于ANN的碳纤维楠竹锚杆锚固力预测研究[J].力学与实践,2013,35(2):40-45.LI Dongbo.Anchorage force prediction for the cfrp-bamboo bolt based on artificial neural network[J].Mechanics in Engineering,2013,35(2):40-45.
[10]王伟.人工神经网络入门与应用[M].北京:北京航空航天大学,1995.WANG wei.The introduction and application of artificial neural network[M].Beijing:Beijing University of Aeronautics and Astronautics Press,1995.
[11]Tho K.K.,SWADDIWUDHIPONG S,LIU Z S,et al.Artificial neural network model for material characterization by indentation[J].Modelling and Simul.Mater.Sci.Eng,2004,12(5):1055-1062.
[12]朱熹育,王社良,朱军强.基于Sugeno型模糊神经网络的空间杆系结构的压电驱动器主动控制[J].工程力学,2013,30(8):272-277.ZHU Xiyu,WANG Sheliang,ZHU Junqiang.Sugeno type fuzzy neural network active cortrol of space frame structure based on piezoelectric actuator[J].Engineering Mechanics,2013,30(8):272-277.
[13]沈乐.石墨烯薄膜的等效弹性参数和力学行为研究[D].上海:上海交通大学大学,2010.SHEN Le.Effective elastic properties and mechanical behavior of single layer graphene sheets[D].Shanghai:Shanghai Jiao Tong University,2010.
[14]XU Yumou,SHEN Huishen,ZHANG Chenli.Nonlocal plate model for nonlinear bending of bilayer graphene sheets subjected to transverse loads in thermal environments[J].Composite Structures,2013,98(9):294-302.
[15]韩同伟,贺同飞,王健,等.石墨烯拉伸力学性能温度相关性的数值模拟[J].同济大学学报,2009,37(12):1638-1641.HAN Tongwei,HE Pengfei,WANG Jian,et al.Numerical simulation of temperature dependence of tensile mechanical properties for single graphene sheet[J].Journal of Tongji University,2009,37(12):1638-1641.
[16]韩同伟,贺鹏飞,王健,等.单层石墨烯薄膜拉伸变形的分子动力学模拟[J].新型炭材料,2010,25(4):261-266.HAN Tongwei,HE Pengfei,WANG Jian,et al.Molecular dynamics simulation of a single graphene sheet under tension[J].New Carbon Materials,2010,25(4):261-266.
[2]CORDELIA Sealy.Graphene goes from strength to strength[J].Materials Today,2008,11(9):12-18.
[3]LI Peiyuan,XIE Zhijiang,LI Xinxia.Research into fault diagnosis of large rotating machinery on BP network and the data source of network[J].Journal of Southwest University for Nationalities Natural Science Edition,2004,30(3):386-390.
[4]SHOKRI S H M,SHOKRI E H,ROHAM Raffiee.Prediction of Young’s modulus of graphene sheets and carbon nanotubes using nanoscale continuum mechanics approach[J].Materials and Design,2010,31(2):790-795.
[5]余晓红.BP神经网络的MATLAB编程实现及讨论[J].浙江交通职业技术学院学报,2007,8(4):45-48.YU Xiaohong.Matlab implementation and discussion of BP neural network[J].Journal of Zhejiang Institute of Communications,2007,8(4):45-48.
[6]Venkatesh.Predicting the mechanical characteristics of hydrogen functionalied graphene sheets using artificial netural network approach[J].Journal Of Nanostructure in Chemistry,2013,(3):83-87.
[7]尹海莲,胡自力.基于BP神经网络的复合材料性能预测[J].南京航空航天大学学报,2006,38(2):234-238.YI Hailian,HU Zili.Prediction of composite material properties basedon bp algorithm of artificial neutral etwork[J].Journal of Nanjing University of Aeronautics&Astronautics,2006,38(2):234-238.
[8]白光辉,孟鹤松,杜善文,等.基于神经网络炭/炭复合材料烧蚀性能预测[J].复合材料学报,2007,26(4):83-88.BAI Guanghui,MENG He Song,DU Shanwen,et al.Prediction on the ablative performance of carbon/carbon composites based on artificial neutral network[J].Acta Materie Compositae Sinica,2007,26(4):83-88.
[9]李东波.基于ANN的碳纤维楠竹锚杆锚固力预测研究[J].力学与实践,2013,35(2):40-45.LI Dongbo.Anchorage force prediction for the cfrp-bamboo bolt based on artificial neural network[J].Mechanics in Engineering,2013,35(2):40-45.
[10]王伟.人工神经网络入门与应用[M].北京:北京航空航天大学,1995.WANG wei.The introduction and application of artificial neural network[M].Beijing:Beijing University of Aeronautics and Astronautics Press,1995.
[11]Tho K.K.,SWADDIWUDHIPONG S,LIU Z S,et al.Artificial neural network model for material characterization by indentation[J].Modelling and Simul.Mater.Sci.Eng,2004,12(5):1055-1062.
[12]朱熹育,王社良,朱军强.基于Sugeno型模糊神经网络的空间杆系结构的压电驱动器主动控制[J].工程力学,2013,30(8):272-277.ZHU Xiyu,WANG Sheliang,ZHU Junqiang.Sugeno type fuzzy neural network active cortrol of space frame structure based on piezoelectric actuator[J].Engineering Mechanics,2013,30(8):272-277.
[13]沈乐.石墨烯薄膜的等效弹性参数和力学行为研究[D].上海:上海交通大学大学,2010.SHEN Le.Effective elastic properties and mechanical behavior of single layer graphene sheets[D].Shanghai:Shanghai Jiao Tong University,2010.
[14]XU Yumou,SHEN Huishen,ZHANG Chenli.Nonlocal plate model for nonlinear bending of bilayer graphene sheets subjected to transverse loads in thermal environments[J].Composite Structures,2013,98(9):294-302.
[15]韩同伟,贺同飞,王健,等.石墨烯拉伸力学性能温度相关性的数值模拟[J].同济大学学报,2009,37(12):1638-1641.HAN Tongwei,HE Pengfei,WANG Jian,et al.Numerical simulation of temperature dependence of tensile mechanical properties for single graphene sheet[J].Journal of Tongji University,2009,37(12):1638-1641.
[16]韩同伟,贺鹏飞,王健,等.单层石墨烯薄膜拉伸变形的分子动力学模拟[J].新型炭材料,2010,25(4):261-266.HAN Tongwei,HE Pengfei,WANG Jian,et al.Molecular dynamics simulation of a single graphene sheet under tension[J].New Carbon Materials,2010,25(4):261-266.