基于遗传算法的冲击信号拟合  

Research of shock signal curve fit based on genetic algorithm

作　　者：曹源[1] 汪凤泉[1] 桂益俊[2] 

机构地区：[1]东南大学土木工程学院,南京210096 [2]东南大学信息科学与工程学院,南京210096

出　　处：《东南大学学报（自然科学版）》2007年第2期320-323,共4页Journal of Southeast University：Natural Science Edition

摘　　要：为便于冲击响应谱的工程应用,以正弦衰减基波组合的形式对冲击信号的拟合进行了研究.通过该拟合形式与遗传算法的结合,将拟合问题转化为参数的搜索与优化问题.考虑到冲击信号在时域内表现出的复杂特征,对其在频域内进行拟合优化,并最终得到优化的信号拟合形式.在观察与分析冲击信号的频域特征以后,采取信号尺度调整、优化参数归一化以及插值运算等措施提高算法的拟合精度.仿真试验表明,冲击信号在频域内的拟合具有较强的适用性,遗传算法的全局优化能力适合于该组合形式下的冲击信号拟合.算法中所采取的改进措施有利于运算效率和拟合精度的提高.Shock response signal curve fitting in the form of assembling the damped sin wavelets is researched to make the shock response spectrum＇s （SRS） engineering application more concenient. Combining the fitting form and the genetic algorithm （GA）, the curve fitting is translated into the parameters＇ searching and optimizing. Considering the complex characters of shock response signal in time domain, the signal is fitted and optimized in frequency domain, and the fit form after optimizing is acquired in the end. After the view and analysis of shock response signal character in frequency domain, the steps such as signal scale adjustment, normalization of optimized parameter and the interpolating operation are taken to improve the fit precision. Simulation test shows that better applicability is presented by fitting the shock response signal curve in frequency domain, and the allround optimization ability of GA is adaptable in fitting the signal under this form. The improved measurements in algorithm are favourable to the increasing of operation efficiency and curve fit precision.

关 键 词：冲击 曲线拟合 遗传算法 插值 

分 类 号：TB123[理学—工程力学]