Improved Methods Of Structural Modal Identification Under Ambient Exitation And Software System Development

Modal parameter identification and damage identification technology ofengineering structures as a new subject to adapt to the actual needs of the project havedeveloped for almost20years, they have strong engineering background and importantpractical value. In recent years, the health monitoring and damage assessment ofengineering structure have been paid more and more attention, modal parameteridentification and damage identification as its core technology and theoretical basishave become research hot spot in civil engineering field. It is too difficult to exertartificial excitation and efficiently measure excitation signal on engineering structuresfor their huge volume, complex bind condition, mixed material, etc. Thereforetraditional modal parameter identification theories and methods with testing input andoutput signal are difficult to apply on engineering structures. Modal parameteridentification under ambient excitation has been widely used in the engineering fieldwithout applying artificial incentives, without affecting the normal working, withoutmeasuring the excitation signal, with low-cost and more in line with the actual situation,etc. The existing modal parameter identification methods have many deficiencies inaccuracy, robustness, efficiency and economic performance indicators, so they are stillin the developed stage, so further study still be needed to perfect these methods.This article research on engineering structural modal parameter identificationunder ambient excitation, owing to the shorcomings of these methods, improvedmethods have been proposed. To sum up, the main contents and conclusions are asfollows:①The importance of engineering structures was briefly introduced and the significanceof structural health monitoring and the damage identification were also demonstrated.Then the research background and the significance of modal parameter identificationunder ambient excitation were discussed detailedly. A literature review on the mainmodal parameter identification method under ambient excitation had been given. Finallythe main work was described.②The theory and identification process of empirical mode decomposition method weresystematically studied. The problems of end effects and the mode mixing of EMD wereillustrated by some numerical example. To solve the prior dramatic problem, the methodof depending on support vector regression was used to inhibit the problem of end effects. To solve the latter dramatic problem, the method of depending on Hilbert character wasused. The experimental data indicated that the improved methods were effect to inhibitthe problem of end effects and to eliminate the mode mixing phenomenon. Futhermorethe improved methods were applied to the experimental data of four-story steel framestructure, the results were correct that certify the correctness and feasibility of the newimproved methods.③Depth studies focused on the method ofNatural Excitation Technique method andfive modal parameter identification method (STD method, ITD method, the complexexponential method, ARMA method, ERA method,). Advanced method was proposedaming at NExT method and the new method componding with the above five modalparameter identification methods were used to identify the modal parameter of structure.These methods were applied to a12-story reinforced concrete model shaking table testof Tongji University. The modal parameters were identified under small earthquake andmajor earthquake. The results proved that the improved NExT method had highaccuracy.④The research was made on developing the visual operating system of structuralmodal parameter identification, operaters could use this system to identify the modalparameter. The visual operating system included file-reading module, geometric datamodule, vibration data module(includes data processing, such as filter technique, theincrement of singular entropy de-noising, ect), handling analysis module(includesvarious modal parameter identifying method, such as stochastic subspace method, HHTmethod, STD method), result processing module. A steel truss suspension bridge wassimulated by MIDAS, the identified results under vehicle load were shown on thedeveloped visual operating system. The identified prosess was clear and accurate.