Study On Application Of Fiber Bragg Grating Senor In Structural Damage Recognition

Fiber Bragg grating strain sensor is a new sensor that is developing rapidly in recent years. It possesses many virtues that include distributing sensing, anti-electromagnetic interference, high precision and good stability, and so on. The extensive interest has been produced because of attractive application of the new sensor in health inspection of large complex structures, such as aviatic space implements, oceanic flat, many-floor builds, and large-span bridges. However, there is no application research about structural damage recognition of BP neural networks basing on Fiber Bragg Grating strain sensing array. This paper has developed primary research about integrating Fiber Bragg grating strain sensing measurement using artificial neural networks. The main work is as follows:(1) The paper firstly presents the research progress about the application of Fiber Bragg Grating strain sensor and the basic methods of structural damage recognition, and then introduces basic methods of employing BP neural networks based on Fiber Bragg grating strain-sensing arrays in the structural damage recognition.(2) Basic theory of Fiber Bragg Grating strain sensing is briefly analyzed. The excellent linearity, sensitivity and repetition of Fiber Bragg grating strain sensor were confirmed by comparing the strains of Fiber Bragg grating strain sensor and resistance strain gauge from the equivalent-strength standard beam under the same static state force.(3) The dynamic response characteristic of Fiber Bragg strain sensor correspond to the flexible beam being vibrated indicates the high sensitivity and precision to the strain response of Fiber Bragg grating strain sensor, which especially suits to monitor structure healthy during dynamic strain state.(4) Base on the basic theory of finite-element analysis, hypostatic modeling for four-side simple supporting plank was upbuilt with ANSYS software. Three-dimensions hypostatic model and plane model were present by simulating the strain distribution of four-side simple supporting plank with damage using finite-element analysis software. However, the strain measurement test indicates that the plane model excels to three-dimension hypostatic model. The influence of the sizes of damage aperture on the four-side simple supporting plank was also investigated using finite-element analysissoftware.(5) The stickup location of Fiber Bragg grating strain sensing array on the four-side simple supporting plank was ascertained on the base of the result of finite-element analysis. The comparison analysis of the strain distribution of the simple supporting plank has carried out employing Fiber Bragg grating strain sensing array and resistance strain gauge. To obtain the study sample and the checkout sample that will be used to recognize the damage location of four-side simple supporting plank with BP neural networks based on Fiber Bragg Grating strain sensing array, the strain distribution of the simple supporting plank with damage was emphatically investigated when it was loaded at the differ location with the same loads.(6) The topology structure of BP neural network based on Fiber Bragg Grating strain sensing array was built. At the same time, the program charts of C language of BP arithmetic were present in the paper. BP neural network based on Fiber Bragg grating strain sensing array proposed in this paper not only can recognize whether there is a damage in the four-side simple supporting plank, but also can reason out which orientation the damage site is corresponding to load spot.