基于遗传小波神经网络的疲劳短裂纹演变规律研究
|
摘要
疲劳短裂纹的演变历程是一种非线性动力学变化过程,为对其进行深入研究,本文采用遗传小波神经网络对其进行数值分析。结合神经网络的自学能力和小波分析的快速衰减、非线性逼近特性,以及遗传算法宏观搜索和全局优化的特点,遗传小波神经网络可在综合考虑多个影响因素的情况下,反映各因素相互之间隐含的非线性特性。基于2种加载频率下光滑漏斗形圆棒试样的疲劳短裂纹复型试验及其在"有效短裂纹准则"体系下的复型膜观察结果,计算试样有效短裂纹密度和主导有效短裂纹扩展率,并运用遗传小波神经网络对其分别进行仿真模拟比较。仿真结果表明,遗传小波神经网络应用于疲劳短裂纹演化行为研究具有合理性和有效性。
To study the complicated nonlinear dynamics process of the fatigue short crack evolution behavior,a method using the genetic wavelet neural network was adopted to carry out numerical analysis on the evolution laws.Combined with the fast decay and the nonlinear approximation ability of wavelet analysis,the self-learning ability of neural network,and the macroscopic searching and global optimization of genetic algorithm,the genetic wavelet neural network can reflect the implicit complex nonlinearity between the various factors when considering multi-influencing factors synthetically.Besides,the effective short crack density and the dominant effective short fatigue crack propagation rate of the specimen were calculated based on the fatigue short crack test of a smooth funnel shaped bar specimen under two loading frequencies and the observed results of the replica film under the effective short fatigue crack principle.The effective short fatigue crack density and the dominant effective short fatigue cracks were simulated and compared respectively by the genetic wavelet neural network.The simulation results show the rationality and validity of the application of genetic wavelet neural network to the study of the evolution behavior of fatigue short cracks.
To study the complicated nonlinear dynamics process of the fatigue short crack evolution behavior,a method using the genetic wavelet neural network was adopted to carry out numerical analysis on the evolution laws.Combined with the fast decay and the nonlinear approximation ability of wavelet analysis,the self-learning ability of neural network,and the macroscopic searching and global optimization of genetic algorithm,the genetic wavelet neural network can reflect the implicit complex nonlinearity between the various factors when considering multi-influencing factors synthetically.Besides,the effective short crack density and the dominant effective short fatigue crack propagation rate of the specimen were calculated based on the fatigue short crack test of a smooth funnel shaped bar specimen under two loading frequencies and the observed results of the replica film under the effective short fatigue crack principle.The effective short fatigue crack density and the dominant effective short fatigue cracks were simulated and compared respectively by the genetic wavelet neural network.The simulation results show the rationality and validity of the application of genetic wavelet neural network to the study of the evolution behavior of fatigue short cracks.
引文
[1]邓增杰,周敬恩.工程材料的断裂与疲劳[M].北京:机械工业出版社,1995:108.
[2]徐强,张幸红,韩杰才,等.人工神经网络在材料科学中的应用与展望[J].材料科学与工艺,2005,13(4):352-356.XU Qiang,ZHANG Xinghong,HAN Jiecai,et al.Application and Prospect of Artificial Neural Networks in Materials Science[J].Materials Science&Technology,2005,13(4):352-356.
[3]ZHANG Q H,BENVENISTE A.Wavelet Networks[J].IEEE Transactions on Neural Networks,1992,3(6):889-898.
[4]林雪原.基于遗传小波神经网络的CPS/SINS组合导航系统算法[J].兵工自动化,2011,30(4):42-45.LIN Xueyuan.CPS/SINS Integrated Navigation Algorithm Based on Genetic Wavelet Neural Network[J].Ordnance Industry Automation,2011,30(4):42-45.
[5]张秀玲,高美静,谷芳春.新型神经网络模型参考自适应控制系统设计[J].系统工程学报,2003,18(2):148-152.ZHANG Xiuling,GAO Meijing,GU Fangchun.Novel Design of Neural Network Model Reference Adaptive ControlSystems[J].Journal of System Engineering,2003,18(2):148-152.
[6]JIANG Y G,LI X.A Method for Face Recognition Based on Wavelet Neural Network[C]//Proceedings 2010Second WRI Global Congress on Intelligent Systems.New York:IEEE,2010:133-136.
[7]韩志艳,王健,伦淑娴.基于遗传小波神经网络的语音识别分类器设计[J].计算机科学,2010,37(11):243-246.HAN Zhiyan,WANG Jian,LUN Shuxian.Design of Speech Recognition Based on Genetic Wavelet Neural Network[J].Computer Science,2010,37(11):243-246.
[8]刘美容.基于遗传算法、小波与神经网络的模拟电路故障诊断方法[D].长沙:湖南大学,2009:25-30.
[9]高美静,赵勇,谈爱玲.基于遗传小波神经网络的多传感器信息融合技术的研究[J].仪器仪表学报,2007,28(11):2103-2107.GAO Meijing,ZHAO Yong,TAN Ailing.Study on Genetic Wavelet Neural Based Multi-Sensor Information Fusion Technique[J].Chinese Journal of Scientific Instrument,2007,28(11):2103-2107.
[10]李建平.小波分析与信号处理——理论、应用及软件实现[M].重庆:重庆出版社,1997:259-281.
[11]胡新辉,陈莉.基于遗传算法的小波神经网络研究及应用[J].西北大学学报(自然科学版),2009,39(2):203-207.HU Xinhui,CHEN Li.Research and Application of Wavelet Neural Network Based on Genetic Algorithm[J].Journal of Northwest University(Natural Science Edition),2009,39(2):203-207.
[12]罗玉春,都洪基,崔芳芳.基于MATLAB的BP神经网络结构与函数毕竟能力的关系分析[J].现代电子技术,2007,30(24):88-90.LUO Yuchun,DU Hongji,CUI Fangfang.Analysis of Relation between the Structure of BP Feed-Forward Neural Network and Precision Proximate Based on MATLAB[J].Modern Electronics Technique,2007,30(24):88-90.
[13]SUH C M,LEE J J,KANG Y G,et al.A Simulation of the Fatigue Crack Process in Type 304Stainless Steel at538℃[J].Fatigue and Fracture of Engineering Masterials and Structure,1992,15(4):671-684.
[14]SHE S,LANDES J D.Statistical Analysis of Fracture in Graphite[J].International Journal of Fracture,1993,3(2):89-200.
[15]BAI Y L,KE F J,XIA M F.Formulation of Statistical Evolution of Microcracks in Solids[J].Acta Mechanica Sinica,1991,7(1):59-66.
[16]MILLER K J,MOHAMMED H J,et al.The Behavior of Short Fatigue Cracks[M].London:Mechanical Engineering Publications,1986:491-511.
[17]HOBSON P D,BROWN M W,RIOS E R D L.Two Phases of Short Crack Growth in a Medium Carbon Steel[M]//Miller K J.The Behavior of Short Fatigue Cracks.London:Mechanical Engineering Publications,1986:441-459.
[18]OBRTLIK K,POLAK J,HAJEK M,et al.Short Fatigue Crack Behavior in 316LStainless Steel[J].International Journal of Fatigue,1997,19(6):471-475.
[19]刘长虹,陈虬,丁宇,等.疲劳短裂纹扩展中的混沌现象[J].大型铸锻件,2001(3):10-14.LIU Changhong,CHEN Qiu,DING Yu,et al.The Chaotic in the Development of Fatigue Short Crack[J].Heavy Casting and Forging,2001(3):10-14.
[20]赵永翔,高庆,王金诺.不锈钢管道焊缝金属疲劳短裂纹行为的试验研究Ⅰ:材料微观结构和研究方法[J].金属学报,2000,36(9):931-936.ZHAO Yongxiang,GAO Qing,WANG Jinnuo.Experimental Observations on the Short Fatigue Crack Behaviour of a Stainless Steel Pipe-Weld MetalⅠ:Material Microstructures and Research Approach[J].Acta Metallurgica Sinica,2000,36(9):931-936.
[21]赵永翔,高庆,王金诺.不锈钢管道焊缝金属疲劳短裂纹行为的试验研究Ⅱ:裂纹萌生、扩展与交互作用[J].金属学报,2000,36(9):937-943.ZHAO Yongxiang,GAO Qing,WANG Jinnuo.Experimental Observations on the Short Fatigue Crack Behaviour of a Stainless Steel Pipe-Weld MetalⅡ:Crack Initiation,Growth and Interaction[J].Acta Metallurgica Sinica,2000,36(9):937-943.
[22]赵永翔,杨冰,高庆.1Cr18Ni9Ti管道焊缝金属疲劳短裂纹萌生与早期扩展[J].核动力工程,2003,24(22):127-132.ZHAO Yongxiang,YANG Bing,GAO Qing.Equation of Short Fatigue Crack Growth Law of 1Cr18Ni9Ti Weld Metal[J].Nuclear Power Engineering,2003,24(2):127-132.
[23]赵永翔,杨冰,高庆.1Cr18Ni9Ti管道焊缝金属的物理疲劳短裂纹扩展试验研究[J].核动力工程,2005,26(6):584-589.ZHAO Yongxiang,YANG Bing,GAO Qing.Experiment of Physical Short Fatigue Crack Propagation of1Cr18Ni9Ti Weld Metal[J].Nuclear Power Engineering,2005,26(6):584-589.
[2]徐强,张幸红,韩杰才,等.人工神经网络在材料科学中的应用与展望[J].材料科学与工艺,2005,13(4):352-356.XU Qiang,ZHANG Xinghong,HAN Jiecai,et al.Application and Prospect of Artificial Neural Networks in Materials Science[J].Materials Science&Technology,2005,13(4):352-356.
[3]ZHANG Q H,BENVENISTE A.Wavelet Networks[J].IEEE Transactions on Neural Networks,1992,3(6):889-898.
[4]林雪原.基于遗传小波神经网络的CPS/SINS组合导航系统算法[J].兵工自动化,2011,30(4):42-45.LIN Xueyuan.CPS/SINS Integrated Navigation Algorithm Based on Genetic Wavelet Neural Network[J].Ordnance Industry Automation,2011,30(4):42-45.
[5]张秀玲,高美静,谷芳春.新型神经网络模型参考自适应控制系统设计[J].系统工程学报,2003,18(2):148-152.ZHANG Xiuling,GAO Meijing,GU Fangchun.Novel Design of Neural Network Model Reference Adaptive ControlSystems[J].Journal of System Engineering,2003,18(2):148-152.
[6]JIANG Y G,LI X.A Method for Face Recognition Based on Wavelet Neural Network[C]//Proceedings 2010Second WRI Global Congress on Intelligent Systems.New York:IEEE,2010:133-136.
[7]韩志艳,王健,伦淑娴.基于遗传小波神经网络的语音识别分类器设计[J].计算机科学,2010,37(11):243-246.HAN Zhiyan,WANG Jian,LUN Shuxian.Design of Speech Recognition Based on Genetic Wavelet Neural Network[J].Computer Science,2010,37(11):243-246.
[8]刘美容.基于遗传算法、小波与神经网络的模拟电路故障诊断方法[D].长沙:湖南大学,2009:25-30.
[9]高美静,赵勇,谈爱玲.基于遗传小波神经网络的多传感器信息融合技术的研究[J].仪器仪表学报,2007,28(11):2103-2107.GAO Meijing,ZHAO Yong,TAN Ailing.Study on Genetic Wavelet Neural Based Multi-Sensor Information Fusion Technique[J].Chinese Journal of Scientific Instrument,2007,28(11):2103-2107.
[10]李建平.小波分析与信号处理——理论、应用及软件实现[M].重庆:重庆出版社,1997:259-281.
[11]胡新辉,陈莉.基于遗传算法的小波神经网络研究及应用[J].西北大学学报(自然科学版),2009,39(2):203-207.HU Xinhui,CHEN Li.Research and Application of Wavelet Neural Network Based on Genetic Algorithm[J].Journal of Northwest University(Natural Science Edition),2009,39(2):203-207.
[12]罗玉春,都洪基,崔芳芳.基于MATLAB的BP神经网络结构与函数毕竟能力的关系分析[J].现代电子技术,2007,30(24):88-90.LUO Yuchun,DU Hongji,CUI Fangfang.Analysis of Relation between the Structure of BP Feed-Forward Neural Network and Precision Proximate Based on MATLAB[J].Modern Electronics Technique,2007,30(24):88-90.
[13]SUH C M,LEE J J,KANG Y G,et al.A Simulation of the Fatigue Crack Process in Type 304Stainless Steel at538℃[J].Fatigue and Fracture of Engineering Masterials and Structure,1992,15(4):671-684.
[14]SHE S,LANDES J D.Statistical Analysis of Fracture in Graphite[J].International Journal of Fracture,1993,3(2):89-200.
[15]BAI Y L,KE F J,XIA M F.Formulation of Statistical Evolution of Microcracks in Solids[J].Acta Mechanica Sinica,1991,7(1):59-66.
[16]MILLER K J,MOHAMMED H J,et al.The Behavior of Short Fatigue Cracks[M].London:Mechanical Engineering Publications,1986:491-511.
[17]HOBSON P D,BROWN M W,RIOS E R D L.Two Phases of Short Crack Growth in a Medium Carbon Steel[M]//Miller K J.The Behavior of Short Fatigue Cracks.London:Mechanical Engineering Publications,1986:441-459.
[18]OBRTLIK K,POLAK J,HAJEK M,et al.Short Fatigue Crack Behavior in 316LStainless Steel[J].International Journal of Fatigue,1997,19(6):471-475.
[19]刘长虹,陈虬,丁宇,等.疲劳短裂纹扩展中的混沌现象[J].大型铸锻件,2001(3):10-14.LIU Changhong,CHEN Qiu,DING Yu,et al.The Chaotic in the Development of Fatigue Short Crack[J].Heavy Casting and Forging,2001(3):10-14.
[20]赵永翔,高庆,王金诺.不锈钢管道焊缝金属疲劳短裂纹行为的试验研究Ⅰ:材料微观结构和研究方法[J].金属学报,2000,36(9):931-936.ZHAO Yongxiang,GAO Qing,WANG Jinnuo.Experimental Observations on the Short Fatigue Crack Behaviour of a Stainless Steel Pipe-Weld MetalⅠ:Material Microstructures and Research Approach[J].Acta Metallurgica Sinica,2000,36(9):931-936.
[21]赵永翔,高庆,王金诺.不锈钢管道焊缝金属疲劳短裂纹行为的试验研究Ⅱ:裂纹萌生、扩展与交互作用[J].金属学报,2000,36(9):937-943.ZHAO Yongxiang,GAO Qing,WANG Jinnuo.Experimental Observations on the Short Fatigue Crack Behaviour of a Stainless Steel Pipe-Weld MetalⅡ:Crack Initiation,Growth and Interaction[J].Acta Metallurgica Sinica,2000,36(9):937-943.
[22]赵永翔,杨冰,高庆.1Cr18Ni9Ti管道焊缝金属疲劳短裂纹萌生与早期扩展[J].核动力工程,2003,24(22):127-132.ZHAO Yongxiang,YANG Bing,GAO Qing.Equation of Short Fatigue Crack Growth Law of 1Cr18Ni9Ti Weld Metal[J].Nuclear Power Engineering,2003,24(2):127-132.
[23]赵永翔,杨冰,高庆.1Cr18Ni9Ti管道焊缝金属的物理疲劳短裂纹扩展试验研究[J].核动力工程,2005,26(6):584-589.ZHAO Yongxiang,YANG Bing,GAO Qing.Experiment of Physical Short Fatigue Crack Propagation of1Cr18Ni9Ti Weld Metal[J].Nuclear Power Engineering,2005,26(6):584-589.