Structural Parameter Identification Based On Atom Search Optimization

The civil engineering structures have been facing the coupling effects of environmental erosion,aging of building materials,long-term loading and various sudden external factors during the service process,and these problems have resulted in structural resistance deteriorating and damages accumulation,which makes bad influences on structural safety and durability.With the rapid development of scientific researches and technological innovation,structural health monitoring(SHM)has become a research hotspot in the field of civil engineering.The structural parameters are inevitably changed after the structures are damaged,therefore,structural parameter identification is the basic component of structural damage diagnosis and safety assessment.With the tremendous advances of computer technologies,a large number of novel intelligent optimization algorithms have been proposed by scholars and applied in various fields,the performance of various optimization varies for different application scenarios.Based on the above background,this paper studies the feasibility and effectiveness of the newly proposed Atom Search Optimization algorithm(ASO)in structural parameter identification.The main contributions and research results are given as follows:(1)The previous methods of structural parameter identification and its current research status are briefly summarized at the beginning of this paper.Due to the requirements of structural prior information and complete response information,the traditional methods of structural parameter identification have many limitations in practical applications under complex environmental conditions.In order to address the challenges caused by the traditional methods,a structural parameter identification method based on ASO algorithm is proposed in this study.(2)After the necessary illustration of basic principle and theoretical background of the ASO algorithm,the concrete procedure of structural parameter identification based on ASO algorithm is proposed in this study.Simulation of physical parameter identification on a fivestory frame structure is conducted,and the recognition results of the proposed method are compared with DE,PSO and GA methods.It shows that the physical parameters can be accurately identified when the structural masses information are completely unknown and structural responses information with noises are incomplete,and the recognition results are better than its homogeneous competitive algorithms.Moreover,the robustness test of the structural parameter identification method based on ASO algorithm is also carried out in study,and the test results show that the proposed method is robust even under the strong noise disturbance condition.(3)Considering that the ASO algorithm is easy to fall into the local optimal value and the flat area of the search space when applying in the structure with a large number of degrees of freedom(DOFs),a roulette selection strategy,a random walk strategy and an elite retention strategy are combined with the original ASO algorithm to improve the identification accuracy of structural parameters.The numerical simulation results show that the improved algorithm(RREASO)has better recognition results than the original ASO algorithm when identifying structures with a large number of DOFs.(4)The number of structural parameters to be identified will increase proportionally to the number of DOFs of the structure.In order to improve the accuracy and efficiency of parameter identification results for large complex structures,RREASO method is enhanced with a substructure division technique to estimate the structural parameters of a 27-DOF shear building structure,and the simulation verification results are compared with the results using RREASO method without substructure division.It shows that the RREASO algorithm enhanced with substructure approach can independently estimate,the structural parameters of the substructures one by one,which is more accurate and efficient than the method using the RREASO algorithm to directly identify the structural parameters of the entire structure.(5)A shaking table test of a 5-story metal frame structure model was designed.The accelerometers mounted on the test structure were used to obtain the structural excitation of base and response signals,and the structural parameters were identified by the RREASO algorithm.The experimental results show that the RREASO algorithm can accurately identify the physical parameters of the structure under different damage scenarios,and can further determine the damage location and damage severity of the structure.