基于遗传算法的高速铁路桥梁损伤识别

Damage identification of high-speed railway bridge based on genetic algorithm

针对高速铁路桥梁结构服役局部损伤问题,文章提出一种改进损伤识别因子并结合遗传算法的损伤识别方法。推导了任意单元长度组合的车辆-轨道-桥梁耦合单元及其动力方程,遗传算法采用锦标赛策略及最优保存策略2种选择算子,引入灾变及邻域搜索机制。通过对不同损伤工况进行多次独立重复计算,对算法的稳定性与准确性进行统计分析。结果表明:在信噪比(signal-to-noise ratio,SNR)大于25 dB时,识别效果好,极少数误判的结果大概率存在于真实损伤位置的附近或对称处;与单个单元损伤情况相比,多个单元损伤工况平均第一次出现最优解的迭代次数略高,但整体都偏低;SNR与识别率(identification rate,IR)呈“扁S”型Logistic曲线关系。

Aiming at the problem of local damage in service of high-speed railway bridge structure,this paper proposes a damage identification method based on improved damage identification factor and genetic algorithm.The train-rail-bridge coupling element and its dynamic equation of arbitrary element length combination are derived.The genetic algorithm adopts two selection operators,tournament strategy and optimal preservation strategy,and introduces disaster and neighborhood search mechanism.The stability and accuracy of the algorithm are analyzed statistically by repeated calculation of different damage conditions.The results show that when the signal-to-noise ratio(SNR)is greater than 25 dB,the identification effect is good,and a small number of misjudgments have a high probability of being near or symmetric to the real damage location.Compared with the damage condition of a single unit,the average number of iterations of the first optimal solution in multi-unit damage conditions is slightly higher,but the overall number is lower.SNR and identification rate(IR)form a flat S-shaped Logistic curve relationship.