Structural Damage Identification Method Based On Chaotic Particle Swarm Random Subspace

The existing large-scale engineering structures in China are affected by complex loads and adverse environment,the structural safety problems are becoming increasingly prominent,and the structural collapse and failure continue to occur.Modal parameters are an important index to measure the overall performance of the structure.The structural state can be judged by the change of modal parameters,but the traditional modal identification method has a large number of false modes,so it is necessary to propose an on-line identification method suitable for largescale structural damage.In this paper,the structural damage identification method based on Chaotic Particle Swarm Optimization(CPSO)and covariance random subspace(COV-SSI)is developed,the numerical simulation model of high-rise building structure is established,the scale model test of long-span cable-stayed bridge is carried out,and the research on multi-point damage location and quantitative identification of complex structure is realized.The main contents include the following aspects:(1)Based on chaotic particle swarm optimization algorithm,the covariance random subspace method is improved,and a framework for structural modal parameter identification in unknown excitation environment is proposed.The particle swarm optimization algorithm is optimized by chaotic local search,multiple windows are set,the window size and window position are set randomly,the acceleration signal is windowed and truncated,the modal identification of each sub signal is carried out by covariance random subspace method,the modal parameters with the most similar identification results are taken as the optimal result,and the window size is adjusted by chaotic particle swarm optimization method.The results show that the chaotic particle swarm random subspace method(CPSO-SSI)has good modal identification accuracy.(2)A damage discrimination model based on the square difference of vibration mode curvature is established to realize the multi-point damage location and quantitative evaluation of large structures in noise environment.The modal parameters before and after structural damage are obtained by CPSO-SSI,and then the square difference of modal curvature is compared as an index to realize structural damage location and identification.Taking the numerical simulation model of service high-rise building structure as an example,12 groups of damage conditions with different degrees are set for low,medium and high-rise buildings.The response signals under white noise excitation are collected,and the response noise signal ratio is set to 10%.The results show that this method has strong robustness and can realize the quantitative identification of single point and multi-point damage of high-rise building structure.(3)The typical damage identification scale model test of long-span cable-stayed bridge is carried out,and the damage identification model based on proportional flexibility matrix and uniformly distributed load curvature is constructed.The natural frequency and vibration mode of the structure before and after the added mass of the structure are obtained by CPSO-SSI method,and then the proportional flexibility matrix of the structure is calculated to obtain the deflection curvature of the structure under uniformly distributed load,and the location and degree of structural damage are judged by the relative change of deflection curvature.Based on the principle of stiffness similarity,the damage identification test of scale model of cable-stayed bridge is carried out,and 8 groups of cable breaking conditions and 15 groups of main beam damage conditions are set.The results show that CPSO-SSI method can identify the typical damage of cable and main beam quantitatively and accurately.