Research On The Characteristics Of The Narrow Dimension Of Features For Damage Detection And Its Application In Bridges

The operating of actual bridges is always affected by various environmental factors.It is a great challenge to accurately identify the damage state of the bridges under time-varying environment.Aiming at this issue,a concept for the characteristics of the narrow dimension of the damage features of bridges is proposed,that is,correlation of damage features of structuras under environmental effects.For the research on the characteristics of the narrow dimension of the damage feaures of bridges,this study focuses on the damage identification of bridges under the environmental effects from three factors,i.e.,the characteristics of the linear narrow dimension and the characteristics of the nonlinear narrow dimension of the damage features of bridges,the damage identification of bridges based on the characteristics of the narrow dimension of the damage features,reconstruction methods of the damage features of bridges for the characteristics of the nonlinear narrow dimension.The main contents of this study are described as following.Aiming at the issue of damage identification of bridges under time-varying environment,the principle of the damage identification algorithm based on latent environmental variable-based methods is studied.Then,the influencing factors and applicable conditions of the latent environmental variable-based methods are systematically discussed,and the influence of the correlations and variances of the damage features of bridges for the latent environmental variable-based methods is summarized.The concept for the characteristics of the narrow dimension of the damage features of bridges is proposed,and a discriminant factor is established to evaluate the characteristics of the narrow dimension of the damage features.Regarding the issue of how to use the characteristics of the narrow dimension to improve the accuracy of damage identification of bridges,the influencing factors of the damage diagnosis factors based on the latent environmental variable-based methods are analyzed.The noise reduction algorithm of the damage diagnosis factor of bridges based on the variational modal decomposition is established.The influence of the characteristics of the linear narrow dimension on the noise reduction algorithm is discussed.A damage identification method of bridges based on the characteristics of the linear narrow dimension of damage features is proposed.Numerical simulation and actual bridge examples are used to verify that the proposed method is effective in improving the probability of damage identification of bridges.Facing the problem of damage identification of nonlinear damage features of bridges,the limitations of damage identification methods based on linear assumptions and kernel methods are discussed,and the functional dependence representation method of nonlinear damage features under environmental effects is explored.The characteristics of the nonlinear narrow dimension of damage features is proposed.A piecewise linearization method of the damage features for the characteristics of the nonlinear narrow dimension is proposed,and a damage identification method of bridges based on the characteristics of the nonlinear narrow dimension is established;numerical simulation and an actual bridge examples are used respectively to verify the reasonable linearization of the damage features of the proposed method and the effectiveness of improving the efficiency of damage identification of bridges.Regarding the reconstruction of damage features that do not possess the characteristics of the nonlinear narrow dimension,the effect of the nonlinear mapping of the polynomial kernel function on the correlation of the damage features is studied,and the change rule of the correlation of the linear damage features under the nonlinear mapping is deduced.Then,a reconstruction method of linear damage features based on polynomial kernel function mapping is proposed.Features extraction ability of low-dimensional manifolds for nonlinear damage features is explored.A low-dimensional embedded manifold extraction method for damage features under environmental influences is constructed.The influence of the inverse operation of manifold learning on the spatial distribution of nonlinear damage features is analyzed,and a reconstruction method of nonlinear damage features based on local tangent space arrangement is proposed.Numerical examples are used to verify the effectiveness of the proposed method on damage features reconstruction.The above research results have certain guiding significance for the further development of the characteristics of the linear narrow dimension of damage features on practical applications of bridges.These methods may improve the accuracy of damage identification of bridges under time-varying environments,and expand the applicability of the characteristics of the linear narrow dimension to different types of damage features of bridges.