基于信息熵的混凝土结构损伤动力识别与反演研究

Internal damage which endangers the safety of concrete structures is often difficult to detect, even by means of modern technology is not yet accurate or effective to identify and evaluate the structures' health status. This project uses the research foundation and application methods of information entropy theory in the disciplines of mechanical structure for reference, and introduces information entropy theory into concrete structures' damage identification and inversion study. Entropy would be applied to describe the damage evolution direction of concrete structures. By theoretical analysis, numerical simulation and model experimental methods (by means of high-precision experiments and refined numerical simulations), crack propagation and evolution would be studied from mesoscopic level, and the change law of entropy would be discussed and extracted. Then, the intrinsic link between entropy theory and fracture of concrete structures would be built up; a set of robust and general entropy function model would be constructed and screened, which is sensitive to structural damage severity, damage location and damage factors. Damage measurement methods, standards and criterion of typical components and damage modes would be established for each optimized entropy index. The basic framework of the concrete structure damage identification system based on information entropy would be established. The validity, economy, sensitivity and robustness of information entropy would be studied. This project is expected to expand the existing concrete damage recognition technology, and have a competitive advantage in the main performance together with the practical application prospects.

危及混凝土结构安全的内部损伤往往难以察觉，即使通过现代技术手段也尚不能得到准确有效的识别和评判。本项目借鉴信息熵理论在机械结构学科的研究基础和应用方法，将其引入到混凝土结构损伤识别与反演之中，使用熵来描述混凝土结构损伤演化的方向，通过理论分析、数值模拟和模型实验等方法，借助于高精度实验表征和精细化数值模拟的技术手段，从细观层面去研究裂纹的开展和演化，探讨并提取结构损伤的熵变规律，建立熵理论与混凝土结构损伤断裂之间的内在联系，构建并筛选出一套对混凝土结构损伤程度、损伤位置和致损因素的敏感性强，又能兼具鲁棒性和通用性的熵函数模型。针对典型构件、典型损伤模式下各优选的熵指标技术，拟定损伤度量的方法、标准和判据。建立基于信息熵的混凝土结构损伤识别基本框架体系，研究其损伤识别的有效性、经济性、敏感性和鲁棒性。通过本项目研究，有望拓展已有的混凝土损伤识别技术，并在主要性能上具有竞争优势和实际应用前景。

危及结构安全的内部损伤往往难以察觉，即使通过现代技术手段也不能得到准确有效的筛查、识别和评判。本项目利用信息熵理论，开展了结构损伤识别与反演研究，取得的主要成果有：使用传递熵来表征结构损伤后结构响应信息传递的线性和非线性程度，通过理论分析、数值模拟和模型实验等方法，探讨了结构损伤的熵变规律，建立了兼具鲁棒性和通用性的传递熵指标，采用传递熵结合替代数据的方法，可用于识别结构损伤（非线性）程度和损伤位置；针对典型构件（风电塔筒/混凝土梁）、典型损伤模式（塔筒法兰盘存在间隙和扭转错位损伤/梁截面裂缝损伤）下各优选的熵指标技术，拟定基于熵指标的损伤度量方法、标准和判据；建立了基于传递熵结合替代数据的结构损伤识别基本框架体系，揭示了其损伤识别的有效性、经济性、敏感性和鲁棒性。成果已应用于大型风电塔筒法兰盘损伤的筛查和定位。本项目研究拓展了现有的结构损伤识别技术，具有重要的理论价值和实际应用前景。