| Traditional structural health monitoring methods have many problems in practical bridge health monitoring,such as low efficiency,low accuracy,limited quantity,difficult installation,poor durability,and difficult damage identification.This paper explores a method for structural dynamic response and damage identification based on image sequences and wavelet analysis,starting from machine vision and combining the principles of traditional modal identification and wavelet transform damage identification methods.The main research work of this paper is as follows:(1)A target tracking algorithm based on template matching and optical flow was used to realize the measurement of structural dynamic displacement tracking based on image sequences.Through vibration experiments of a suspension bridge model,the dynamic displacement time history signal based on image sequences was compared and verified with the traditional laser ranging displacement sensor,verifying the accuracy and reliability of the two image-based dynamic displacement tracking algorithms for displacement measurement.(2)A method for identifying modal parameters based on image sequences was proposed to extract the structural modal parameters.By using optical flow tracking algorithm to monitor the multi-point structural vibration displacement of a cantilever beam model in the laboratory and combining with traditional modal identification theory,the structural modal parameters were extracted and compared with the modal parameters results obtained from traditional laser ranging sensors and numerical simulations,verifying the feasibility and reliability of the proposed method for structural vibration monitoring and modal parameter identification.(3)A damage identification method based on image sequences and wavelet transform was proposed.Firstly,the modal parameters of a simply supported beam in four different states were calculated by analyzing and calculating the numerical model of the structure through finite element simulation.Then,the displacement mode was taken as the input signal,and discrete wavelet transform was performed on it.A new damage index,the square difference of wavelet coefficients,was established to determine the location of structural damage.The proposed method was applied to dense point layout structures in the laboratory during a simple-supported beam damage test,obtaining the structural modal parameters(frequency,mode shape).Finally,the proposed wavelet coefficient square difference was used to identify and locate structural damage.The experimental results demonstrate that this damage identification method based on image sequences and wavelet transform is feasible and accurate. |