Title :
The Crack Injury Detection Based CMR and Natural Frequency with BP Neutral Network
Author :
Liu, Mingjun ; Li, Li ; Zhao, Xiuyang ; Liu, Hua
Author_Institution :
Sch. of Electr. Eng. & Autom., Harbin Inst. of Technol. Harbin Heilongjiang, Harbin, China
Abstract :
In this paper a new approach based curvature mode ratio (CMR) and natural frequency with BP neutral network for crack injury detection is proposed. Taking a simple cantilever beam with multiple damage and perfect structure as study object, injury indices are gained by way of analysis the merit and drawback of natural frequency and the curvature mode, then use BP neutral network training tests that regards the natural frequency, the amount of the frequency relative decline and the first-order maximum CMR as the input parameter, crack location and injury degree as the output parameter. The simulated results confirm that the method can be very well applied to the crack damage identification.
Keywords :
backpropagation; beams (structures); cantilevers; crack detection; neural nets; structural engineering computing; BP neutral network; CMR; crack injury detection; crack location; curvature mode ratio; injury degree; multiple damage cantilever beam; natural frequency; Artificial neural networks; Automation; Feedforward neural networks; Frequency; Function approximation; Injuries; Iterative algorithms; Neural networks; Structural beams; Testing; BP Neutral Network; CMR; Natural Frequency;
Conference_Titel :
Future Networks, 2010. ICFN '10. Second International Conference on
Conference_Location :
Sanya, Hainan
Print_ISBN :
978-0-7695-3940-9
Electronic_ISBN :
978-1-4244-5667-3
DOI :
10.1109/ICFN.2010.38
