Structural Damage Identification Based On The Artificial Immune Algorithm

Because of natural and man-made reasons, damage always occurs in the structuresof civil engineering. When damage is accumulated to a certain extent, performance ofthe structures will decline, even lose the ability to work, thus it will pose a significantthreat on the country, society and people’s life, property, and the lose of the country,society and people caused is inestimable. Therefore, it is an important significance fordamage detection and positioning of the existing structure. The structural dynamiccharacteristics are determined by the physical parameters of the structure (including themass, stiffness, damping) decision. Structures occur after damage, the various physicalparameters will change (especially the stiffness), thereby will lead the dynamismcorresponding with the structure to changing. Therefore, it can be realized for structuraldamage identification to establish the mapping relationship between the structuraldynamic characteristics and structural damage. In the practical engineering, obtainingthe time domain response is relatively simple and fast. Structural damage also willchange statistical characteristics of the time domain response of structural. Time seriesmodel can condense a few large number of time-domain response data into a few of themodel parameters (regression coefficient, sliding coefficient of mean value and modelresidual). Since the regression coefficient contains the inherent characteristics of thesystem, according to which we can diagnosis the state of the system. The rise ofartificial intelligence theory including neural network, fuzzy mathematics, geneticalgorithm, artificial immune algorithm etc, providing new ideas and methods fordevelopment of structural damage identification technology. In this paper, utilizingresearch results in the field of artificial immune algorithm, an artificial immunealgorithm is applied to the structure damage identification, mainly to complete the workof the following respects:①this paper studies several major artificial immune algorithm, elaborates theprinciple, main structure of the human immune system and the important mechanism,typical characteristics of artificial immune system. Comparison of several algorithms,this paper uses the general immune algorithm as the damage identification algorithm,and makes some improvement.②The truss structure is preliminary positioning structure analysis used modalstrain energy dissipation rate theory. Then the structure was further damage location and quantitative analysis by using the artificial immune algorithm, in order to verify thefeasibility of artificial immune algorithm.③It conducts dynamic time history analysis for the six interlayer shear structuralby using the MATLAB language programming, then conducts data process for timeseries of structural dynamic time history response by using AR model of time seriesanalysis method, and detect the structural injury comparing with the artificial immunealgorithm.