噪声背景下大型井架钢结构微损伤识别及其在风载荷下的安全评估

Large derrick steel structure is widely used in various fields such as machinery, oil, mine, construction and so on. Due to the poor working conditions and varied load case, derrick will inevitably emerge early and local micro damage of derrick. These micro damages are very dangerous, which may cause disastrous incidents especially under wind loads. Therefore, the study of damage identification and safety assessment of the large bearing steel structure is very important. This project proposes novel method of micro damage signal feature identification under noise background and new safety assessment method of the large derrick steel structure under wind loads. The specific content includes: The damage dynamic model of large derrick steel structure is established, and the characteristics of the micro damage signal of the structure are analyzed, and the mechanism of the micro damage is deduced. A new early micro damage identification method of large derrick steel structures under noise background is proposed based on the combination of new multi stable stochastic resonance and AMD-HVD. The large derrick steel structural damage pattern function is constructed; the quantitative relationship between the damage characteristics and performance reaction of large derrick steel structure is determined, and micro damage identification method of steel structure is studied. Utilizing transient excitation and wind load excitation method, the safety evaluation method of derrick steel structures with micro damage structures under wind loads is studied based on multiple criteria optimization safety criterion. Finally, the experimental researches are carried out based on experimental platform of large derrick steel structure.

大型井架钢结构在机械、石油、矿山、建筑等行业应用广泛。由于工作环境恶劣、载荷多变，不可避免发生早期和局部微损伤。这些微损伤极其危险，尤其在风载荷作用下，可导致结构失稳倒塌等灾难性事故。因此，井架早期微损伤识别和风载荷下的安全评估研究十分重要。本项目提出大型井架钢结构在噪声背景下微损伤信号特征识别新方法及其在风载荷下安全评估新方法。具体内容包括：建立大型井架钢结构微损伤动力学模型，分析微损伤信号特征规律，揭示井架钢结构微损伤机理；探索基于新型多稳随机共振和改进AMD-HVD相结合的噪声背景下大型井架钢结构微弱损伤信号提取新方法；构建井架钢结构损伤特征模式函数，确定损伤特征与大型井架钢结构性能反应的定量关系，研究井架钢结构微损伤的识别方法；采用瞬态激励和风载荷模拟激励等方法，基于多指标优选的安全准则，研究含微损伤的大型钢结构在风载荷下的安全评估方法；最后在大型井架钢结构实验平台上进行实验研究。

本项目主要通过以下五个方面来进行噪声背景下大型井架钢结构微损伤识别及其在风载荷下的安全评估的研究:1.建立含微损伤大型井架钢结构动力学模型。通过梁-壳子结构动力学模型的建立模拟微损伤并对井架钢结构的动力学特性进行分析，其中模态振型、模态曲率、模态应变能等特征相结合可以作为损伤识别指标对损伤进行判断；2.研究了在噪声背景下井架钢结构的微弱信号特征提取方法。主要集中于随机共振模型理论对噪声信号利用的进一步加强。得到了双稳态及多稳态随机共振信号提取方法，本部分研究属交叉学科，在旋转机械故障诊断领域也得到了应用；3.开发了井架钢结构微损伤特征信号识别算法。包括基于小波包分析、HVD分解与不同神经网络的结合算法、基于多传感信息融合、以及基于深度学习的智能诊断算法，通过仿真分析得出，这些方法均能有效的识别井架钢结构的微损伤；4.井架钢结构风载荷模拟及风载荷激励下的安全评估系统研究。采用线性滤波法模拟风载荷作用，对风载激励下微损伤井架钢结构进行了分析，基于位移、应力、可靠度的安全评估准则对井架钢结构模型进行了安全评估；5.实验研究和风载作用下井架钢结构损伤诊断及安全评估系统设计。包括实验台搭建、损伤识别算法的实验验证、风载激励下井架微损伤识别的实验验证，最后开发了基于算法的风载安全评估软件。. 通过上述研究，建立了一个完善的井架钢结构微损伤故障诊断的流程，对于长期暴露于恶劣环境（海洋平台、戈壁荒漠等）下工作的复杂结构的安全性能评估提供了参考，对于减少事故发生概率，做到防患于未然，对于保护人民生命财产安全具有重要意义。