Research On Response Reconstruction Method Based On State-Space Model

Structural health monitoring system is widely used in the safety assessment of large engineering structures,with analysis data originating from a network of sensors installed in various parts of the structures.However,due to economic conditions and the inherent installation restrictions of the structure itself and other practical factors,many structural parts and even critical locations cannot be installed with sensors,which results in incomplete monitoring data.The emergence of structural response reconstruction technology provides a solution to such problems by utilizing the response information at the limited measurement locations of the structure to reconstruct the required information at the unmeasured locations.Taking the state-space model as the entry point,this thesis carries out the work related to the full-text structural response reconstruction.Details of the research are as follows:(1)Aiming at the uncertain structural response reconstruction problem under unknown excitation,the state-space model is established,and the simultaneous input and state estimation algorithm is used to reconstruct the unknown structural response.Through the numerical simulation of the simply supported two-dimensional truss model,it is verified that the proposed method can reconstruct the unmeasured velocity and acceleration responses well under two different types of unknown excitation.(2)A structural response reconstruction method based on particle swarm Kalman filter and maximum correlation entropy criterion is proposed to address the problem that the traditional Kalman filter algorithm is not effective in response reconstruction under non-Gaussian noise environment.Firstly,the measurement noise variance is determined using the particle swarm optimization algorithm.Then,the cost function of traditional Kalman filter is modified by introducing correlation entropy considering the general non-Gaussian noise environment interference.Finally,numerical example and experimental analysis of simply supported beam are given to verify the proposed method.The results show that the proposed method still has high accuracy and robustness under the conditions of different types of non-Gaussian noise interference and excitation transformation,and the reconstructed time history curves match well with the real measurement curves,which has stronger applicability in the practical response reconstruction problem solving.(3)Aiming at the problems of particle degradation and reduction of particle diversity in the process of structural response reconstruction by the particle filter algorithm,a response reconstruction method based on improved particle filter algorithm is proposed.Firstly,the particle swarm optimization algorithm is introduced into the importance sampling phase of the particle filter algorithm to update the state of the particles using the latest observation information.Then,its weights are optimally combined to slow down particle degradation and enhance the diversity of particles.Finally,the numerical simulation and experimental verification of space truss and simply supported beam are carried out respectively.The results show that the proposed improved algorithm has higher accuracy in response reconstruction compared with the response reconstruction methods of the standard particle filter algorithm and the particle swarm optimization particle filter algorithm.