部分观测下结构质量及非线性恢复力免模型识别

摘要

强动力荷载作用下结构构件的恢复力是其非线性行为的最直观描述,不同结构的非线性特性复杂且往往难以事先用准确的参数化形式描述,而且活荷载的存在也导致结构质量也需要识别,此外测量完整结构所有自由度的动力响应较为困难.为此,提出一种结合等效线性理论与无迹卡尔曼滤波(Unscented Kalman Filter,UKF)的迭代算法,仅利用结构部分自由度上的动力响应,实现结构质量与非线性恢复力的免参数化模型的同时识别.在一个线性多自由度系统中引入磁流变阻尼器模拟非线性元件,在不同质量初始值情况下,当一处或多处存在不同类型的非线性构件时,可实现结构恢复力及质量识别,通过将识别结果与理论值的比较验证了该方法的有效性.%Nonlinear restoring forces (NRF) of structures excited by strong dynamic loadings provide a direct description of the initiation and development of structural damage. However, due to the complexity of different structural nonlinearities, it is difficult to describe the nonlinear behavior of various structures by using a parametric model in advance. Moreover, it is always difficult to estimate structural mass due to the existence of live load and to measure dynamic responses at all degree-of-freedoms (DOFs) accurately for structure identification. In this study, an iterative mass and nonlinear restoring force simultaneous identification approach using partial structural dynamic responses by combining equivalent linear theory and Unscented Kalman Filter (UKF) is proposed without using any parametric model of the NRF. By introducing different magnetorheological (MR) dampers and considering different initial mass values and different nonlinear components at different locations, a multi-DOF nonlinear structure is simulated and the mass and restoring force identification is realized. And the effectiveness of the proposed approach for both NRF and mass identification is validated.