In order to monitor structural safety in-time, a model updating strategy with high calcula-tion speed and easy-to-convergence capacity is proposed.First, the relationship between the element stiffness and the structural responses is established by calculating the instantaneous energy.Second-ly, the instantaneous energy is substituted into the Kalman filter.The structural element real stiffness is obtained by updating according to the difference of the prediction energy and measured energy in every time step.Finally, simulation tests are carried out on a highway sign support truss in network for engineering earthquake simulation ( NEES) in the United States.The results show that without noise, the initial stiffness of the elements without damage and with 20%, 40%, 60%, 80%damage can be updated to real stiffness in 0.4 s.Even with 5% environmental noise, the updated stiffness errors obtained by this strategy are less than 12%.The time cost of CPU ( central processing unit) in every time step is far less than the sampling time.Therefore, the energy theory and Kalman filter can update the structure with unknown stiffness promptly and effectively.%为了实时监测结构的安全状态，提出了一种计算速度快、易收敛的模型修正策略．首先通过计算瞬时能量来建立结构单元刚度和结构响应之间的关系；然后，将结构瞬时能量代入Kal-man滤波器中，根据每一时间步能量预测值和实际测量值的差异进行修正，得到结构的真实刚度；最后，以美国地震工程模拟中心数据库（ NEES）中的美国某州际公路指示牌支撑桁架为例进行数值验证，结果表明：无噪声干扰情况下，刚度发生20％，40％，60％，80％损伤的杆件和未发生损伤的杆件均能在0.4s内从初始刚度收敛到各自的真实刚度；在5％随机噪声干扰下，利用该策略修正得到的刚度误差均小于12％；每一时间步所消耗的CPU时间远小于采样周期．因此，利用能量原理和Kalman滤波器能够快速有效地对未知刚度的结构进行实时模型修正．
展开▼
