Damage Recognition Based On Probability Statistics Under The Action Of Temperature

The civil engineering structure inevitably exist certain degree of damage in the process of actual use. Structural health monitoring help people in time find the degree of structural damage accumulation to guarantee the safety and durability of the structure. Structural damage identification is the key issue of structural health monitoring. Temperature has seriously affected the reliability and stability of identification during damage identification. In the paper, the combination based on vibration identification method and probability statistics method is used to determine the damage threshold under different temperature to identify damage.Considering the influence of temperature on damage identification, it studies the temperature distribution model in this paper to determine the variation regularity of temperature for a long time. Atmospheric temperature data in Dalian area for40years are used to set up temperature probability distribution model, and the reliability of the model is verified through the experiments. The generalized probability density evolution equation with finite difference method and the structural response analysis of the structure analyze the model to obtain the response probability density. It determines the validity and reliability of this method through the numerical simulation of five-bar truss structure.To determine the effect of temperature and identify damage, Response probability distribution model under different temperature is established, combining with the probability statistics method to establish the damage index. Damage identification thresholds are determined through the hypothesis test for damage identification. The feasibility and validity of this method is verified through the steel beam experiments and numerical simulation. Damage identification results under different temperature and different degree of damage identification is analyzed.Research results are mainly listed as follow:The statistical analysis is used to obtain the return period of50years and the extreme forecast of100years. Temperature significant impact on structural damage identification is determined. Probability density curve of structural response at any time and any temperature are received through the probability density evolution method, which can not only identify the damage especially to identify the degree of minor damage, but also the probability of damage is determined through the probability test.