基于耦合振动响应的轨道车辆载荷识别与损伤识别新技术研究

The load identification and damage identification problem of the rail vehicle greatly affect the safety of the structure. In This project, A method is proposed to identify the inverse problem of key components of the track vehicle structure based on the coupled vibration response. The loads and damage parameters of the structural components are extremely important in the dynamics and fatigue strength of the rail vehicle. And some key parameters are difficult to measure directly. Based on the research of domestic and foreign literatures, and the theory of vehicle dynamics and inverse problem analysis, the basic research of identification technology based on coupled vibration response will be carried out. The main contents are listed follows: (1) Based on Multi-body Dynamics (MBD) theory and finite element method (FEM), the basic theory and simulation algorithm of hybrid identification technology using TIKHONOV regularization method and artificial intelligence algorithm are combined with vehicle coupling vibration response and resonance fatigue analysis technique. (2) The experimental study on load identification and damage identification of the Beam-Plate structure will be carried out. Based on the sensors optimization technology and the coupling vibration measured data, the engineering application technology research is carried out from the view point of experimental. The study of dynamic load identification technology for fatigue damage of vehicle structure can provide theoretical support for the safety design of actual vehicle structure.

轨道车辆的载荷识别和损伤识别难题极大影响着结构的安全性，本项目拟提出一种基于耦合振动响应的轨道车辆结构关键部件的逆问题识别计算方法。结构部件的载荷与损伤参数在轨道车辆动力学和疲劳强度中极为重要，而且一些关键参数难以直接测量。本项目拟在国内外相文献的研究基础上，基于车辆动力学和逆问题分析理论，开展基于耦合振动响应的识别技术的基础研究。具体内容包括：（1）在多体动力学和有限元法的基础上，结合车辆耦合振动响应和共振疲劳分析技术，使用TIKHONOV正则化方法和人工智能算法开展混合识别技术的基础理论和仿真算法研究；（2）开展板梁结构与比例车体结构的载荷识别与损伤识别的实验技术研究，根据传感器布置优化技术及耦合振动测试数据，从实验的角度开展工程应用技术研究。通过针对车辆结构疲劳损伤的动载荷识别技术研究可以为实际车辆结构的安全性设计提供理论支持。

轨道车辆关键部件的载荷识别和损伤识别技术问题存在着识别理论不完善、预测精度不高、计算效率低、可靠性差等难题。本项目主要以板梁结构、轮轨、（比例）车体结构等为具体研究对象，建立了反映实际结构服役的损伤特征参数的新型数据驱动的车辆耦合动力学模型，提出一种自适应和机器学习的车辆结构混合识别新方法，建立了车下设备的局部共振效应下振动响应特征和混合识别模型，分析典型疲劳裂纹影响下的结构刚度、模态振型相关的计算方法及系统响应下异常动载荷的影响，对动力学参数劣化特征下的关键部件的裂纹识别和载荷识别算法开展了跟踪技术研究。重点解决车体结构及车下悬挂设备在共振效应下的耦合振动、裂纹识别及异常动载荷识别的复杂问题。利用精准识别算法、建模和仿真、试验台设计和试验验证等方法初步解决了车辆结构裂纹跟踪和溯源技术难题。同时，本项目利用多目标优化算法和逆问题分析等技术针对轮轨作用力、受电弓-接触网的接触力、车体及车下设备共振载荷等进行了混合识别算法优化。通过数值仿真与试验验证等技术验证了提出方法的有效性。本项目的研究可以改善高速列车智能诊断过程中混合识别的精度与抗噪性问题，对于完善轨道车辆结构智能运维的理论体系有着积极的推进作用，具有重要的学术研究意义和广阔的工程应用前景。