Study On Damage Identiifcation Of Structure Based On Genetic Algorithm

Bridge as a major component of the transport system, its safety, durability andnormal performance of the use more and more arrest people’s attention with the con-tinuous development of China’s transport.Now economic quickly develop, plus corro-sion of the natural environment, leads to more and more traffic and other diseases oflower bearing capacity and structural damage. Promptly discover the bridge structuredamage can make sure the traffic safety and provide scientific decisionmaking for thebridge repair and reinforcement. Therefore, the bridge stru-cture damage identificationstudy has significant economic and social benefits.This dissertation investigates model updating based genetic algorithm and thenidentify the damage. The objective function is built through frequencies and modeshapes to modifiy model. Structural damage can be identified by contrast the model pa-rameters before and after. In optimization problems, improvement strategies are pro-posed based on standard genetic algorithm. Individual fitness value is scaled by estab-lished fitness function to enlarge it for improving the selectivity and preventing excep-tional individual rule the entire group and changing selection pressure. The crossoverrate and mutation rate of the genetic algorithm are changed with the number of itera-tions to improve the searching ability. And pattern search algorithm is mixed in theiterative process to further improve the search capabilities.Freely supported beam is simulated. The results show that recognition works wellin the simple structure and accurate modal analysis. The noise immunity of algorithm issubsequently proved well.Reinforced concrete T-beam is poured in laboratory as research object, let itcracks by loading to analog damage. The identification results show that the proposedmethod can correctly identify the damage location and degree of reinforced concreteT-beam.