基于多尺度动力学分离的结构非线性连接状态诊断

项目名称： 基于多尺度动力学分离的结构非线性连接状态诊断

项目编号： No.U1530139

项目类型： 联合基金项目

立项/批准年度： 2016

项目学科： 数理科学和化学

项目作者： 徐超

作者单位： 西北工业大学

项目金额： 63万元

中文摘要： 针对复杂结构系统中局部非线性连接状态诊断的工程需求和因连接界面力学行为具有非线性、多尺度和不确定性的特点而难以进行辨识的科学挑战，本项目提出了将信号的多尺度分解与非线性快慢流动力学建模技术相结合的学术思想，建立了一种结构非线性连接状态诊断新方法。该方法首先对结构宏观动力学响应信号进行多尺度分解，建立不同尺度的波动信号与非线性系统不同快变频率下的慢流动力学模型之间的关系；然后，构造表征连接界面非线性状态的力学模型，并将复杂连接结构的高维非线性动力学模型进行降维；再利用分离出的多尺度信号辨识降阶模型，最终提取出定量表征不同非线性状态的特征参数，实现对复杂结构中非线性连接状态的诊断。项目研究的主要创新点是揭示了信号的多尺度特征与非线性快慢变过程的关联关系，构造了一种利用非线性降阶模型表征接触状态的定量化诊断方法。

中文关键词： 复杂结构系统；非线性识别；接触建模；慢流动力学；信号分解

英文摘要： Facing the urgent needs to diagnose local nonlinear joint condition for complicated engineering structural systems and the great challenge of joint identification due to nonlinear, multi-scale and uncertain mechanics behaviors existing on the interface, the proposal presents a new idea combing signal multi-scale decomposition with fast-slow flow modeling of nonlinear dynamic systems. A novel nonlinear joint condition diagnosis method is proposed. Multi-scale signal decomposition is used first to decompose system response into many single-frequency intrinsic mode functions. The physical relation between multi-scale response and system slow-flow dynamics is built. Then, a nonlinear constitute relationship that can describe interface contact condition is developed and used to build system full order model. Model reduction is thus performed to yield a nonlinear reduced model, and model parameters are identified from decomposed multi-scale response signal. Finally, feature parameters are constructed to identify nonlinear joint condition. The two innovation points are building the relationship between multi-scale response decomposition and system fast-slow dynamic processes and presenting a novel joint condition diagnosis method based on identification of the nonlinear reduced-order model.

英文关键词： complicated structural systems;nonlinear identification;contact modelling;slow-flow dynamics;signal decompostion