基于统计模式识别理论, 采用AMRA 时序分析方法, 通过长自回归模型计算残差法和最小二乘法的结合来 估计模型参数, 从而建立了系统模型。运用均值控制图的方法对压力管道的振动信息特征进行提取、选择, 从而有 效地判别压力管道的无损与损伤问题。数值模拟结果表明, 基于统计模式识别的均值控制图的结构异常检验方法, 能够准确诊断检测结构的损伤状态, 而且对于损伤程度和损伤位置都有很强的敏感性。

Based o n the statistical patt ern recog nition theor y, t he AMRA tim ing analysis method combined w ith the long auto re2 g ressive mo del residual method and the least squares method w ere used to estimate the model parameter s and then to develo p the sy stem mo del. The mean co nt rol chart method w as used to ex tract and select the v ibratio n information and feature of the pr essure pipe, therefo re the damag e of the pressur e pipe can be ident ified effectively. T he simulation r esult s show ed that struc2 tur al abno rmality test metho d of the mean contro l char t, w hich is based on the statist ical patter n recog nitio n theo ry , can diag2 nose the structur al damage accurately, and is sensitive to the damage deg r ee and location.

甘肃省科技支撑计划研究项目( 1304FKA055) ; 甘肃省高等学校科研项目(2013B2 124) ; 国家电网甘肃电力科技项目( 5227011600D9)