【作者】 王利英；

【导师】 杨绍普；

【作者基本信息】 北京交通大学 ， 车辆工程， 2014， 博士

【摘要】 摘要：架桥机是一种将预制钢筋混凝土梁段吊装在桥梁支座上的专用施工机械。架桥机一旦发生故障或事故,将会造成人员伤亡和巨大的经济损失,甚至是灾难性的后果。主梁是架桥机最主要的承载部件之一,其结构的安全性尤其重要,因此,对架桥机的主梁进行结构损伤识别具有重要的现实意义。分析了结构损伤的判断过程,构建了安全因子集,从定义、公式和判断依据三个方面对安全因子进行了详细的说明,并用模型梁作为算例,对安全因子的适用性进行了分析,找到了比较适合梁类结构的安全因子。介绍了布谷鸟搜索算法,从动态发现概率、步长和莱维飞行三个方面对其进行了改进,改进后的布谷鸟搜索算法收敛速度非常快,精度较高。并在此基础上提出了适用于高维函数的协同布谷鸟搜索算法。然后把损伤识别问题看作约束优化的数学问题,以固有频率和模态应变能建立目标函数,利用改进后的布谷鸟搜索算法进行损伤识别,并于其它算法进行对比,分析结果表明,改进后的布谷鸟搜索算法能较好的对架桥机主梁进行损伤识别。建立了基于核模糊C-聚类算法和多分辨小波核相关向量机的结构损伤模型,并且以频率变化率、模态应变能和加加速度等安全因子作为输入向量,分别对单损伤和多损伤样本进行无噪声和有噪声两种情况下的损伤识别,仿真结果表明该模型识别精度高、速度快。其次,利用提出的方法对损伤进行了预测分析,分析结果表明该方法有良好的预测能力。利用大型分析软件ANSYS建立架桥机主梁的有限元模型,对其进行网格划分,利用前面提到的安全因子集模拟不同的损伤工况并对他们进行分析,找到适合于架桥机主梁的安全因子,为大型施工机械的安全监测提供理论依据。对架桥机的主梁进行了损伤试验研究,采用矩形梁来模拟架桥机的主梁,分别进行了刚度试验、模态试验和动载试验,试验结果和仿真结果基本吻合。更多还原

【Abstract】 ABSTRACT:The Bridge Erection Machine is a special construction machinery which can hoist prefabricated reinforced concrete beam on the bridge bearing. The failure or accident of bridge erection machine will cause casualties and huge economic loss, even catastrophic consequences. The girder is one of the most important bearing parts and then its safety is particularly vital; therefore, it has a great practical significance to identify the structural damage of the girder.The judging process of structural damage is analyzed. The security factor subset is built; the definition, formula and judgment are described in detail and the beam model as an example, the security factors are carried on applicability analysis, and then the security factors more suitable for beam structure are found.Cuckoo search algorithm is introduced and improved from the dynamic discovery probability, the step length and the levy flight, and the improved cuckoo search algorithm has a faster convergence speed and a higher precision. And on this basis the cooperative cuckoo search algorithm suitable for high dimensional functions is developed. Then the damage identification is formulated as a constrained optimization problem, and the combination of natural frequency and modal strain energy are used to establish the objective function, which can be identified by the improved cuckoo search algorithm. The simulation results show that the improved cuckoo search algorithm can better identify the damage of structure.The kernel fuzzy means clustering algorithm and the relevance vector machine based on multi-wavelet kernel are presented respectively to establish a damage identification model for the girder of bridge erection machine. The safety factors such as modal strain energy, frequency variation, and the derivative of acceleration are the input vector, the single damage and damages samples were used to identify the damage under noise and without noise, thus illustrating its high accuracy and rapid speed through the simulation results.The large analysis software ANSYS is used to establish the finite element model of girder and its meshing is performed, the aforementioned security factors are used to simulate different damage conditions and analyze them, thus the security factors suitable for the girder are acquired, which can provide a theoretical basis for safety monitoring of large construction machinery. The damage experiments are conducted for the main girder, the rectangular beam is used to simulate the main girder, and then the stiffness tests; the modal tests and the dynamic load tests are finished respectively, the test results and the simulation ones are almost unanimous.更多还原