一种非线性梁系统动载荷识别算法

A dynamic load identification method for nonlinear beam system

载荷识别是结构健康监测的重要组成部分,而非线性梁系统在工程应用中扮演重要角色,为了在结构健康监测的同时方便利用最优化算法对非线性梁系统进行有效控制,提出了基于容积卡尔曼滤波器(Cubature Kalman filter,简记为CKF)的载荷识别算法。此算法在卡尔曼滤波器的体系下,通过CKF产生的增益矩阵、新息序列、一步估计值和协方差矩阵,利用最小二乘算法在线估计载荷的大小和位置,附录给出了算法的详细推导过程。为了验证算法的有效性,采用受高斯白噪声影响的大变形梁系统作为仿真对象,根据系统响应依次识别不同噪声影响下的正弦、方波和锯齿波载荷。实验方面,采用受非线性弹簧约束的Bernoulli-Euler梁作为对象,利用光纤光栅传感器测得的应变值识别载荷。结果表明提出的动载荷识别算法能够很好地抑制噪声,并且具有良好的稳定性。

Load identification plays an important role in structural health monitoring,and nonlinear system is getting more and more attention.After evaluating the structural condition,the new method which is based on Cubature Kalman filter(CKF)and a recursive least-squares algorithm are proposed to identify load for nonlinear systems.For the proposed method,the gain matrix?residual innovation sequences?priori state estimate?and innovation covariance generated by CKF are employed to iden-tify load by using a least-squares algorithm.The detailed derivation of the nonlinear estimator can be found in appendix.To verify the effectiveness of this identification method,numerical simulations subjected to white Gaussian noise and three types of load are employed.For the experiment,Bernoulli-Euler beam which is constrained by a nonlinear spring is employed,and strain values getting from Fiber Bragg Grating(FBG)sensors are used as system response.The simulation results demonstrate that the proposed method can suppress noise well and has a good stability.Experimental results show the method can identify load accurately in engineering application.