A wavelet-based strategy for structural identification and damage detection using modified Gas

سال انتشار: 1400
نوع سند: مقاله کنفرانسی
زبان: انگلیسی
مشاهده: 237

فایل این مقاله در 16 صفحه با فرمت PDF قابل دریافت می باشد

استخراج به نرم افزارهای پژوهشی:

لینک ثابت به این مقاله:

شناسه ملی سند علمی:

ICSAU07_1209

تاریخ نمایه سازی: 21 دی 1400

چکیده مقاله:

This paper presents a robust wavelet-based approach for structural identification and damage detection using modified genetic algorithms (MGAs). The parallel decoding capability of genetic algorithm, in addition with the possibility of applying the search space reduction technique and its reliability make it still popular in optimization community. In this study, in order to enhance the practicality of this powerful tool in structural identification problems, the previously established method using the family of Chebyshev wavelet, Legendre and Haar wavelets is employed to accurately simulate the structural responses in achieving a computationally effective approach. An efficient operator of derivative is used for sensitivity analysis of accelerations capable of using various wavelets. This operator is inherently very sensitive to small changes of accelerations considering small sampling rates especially in dealing with online damages imposed to the structures. It is concluded that, the GA-based fitness evaluation is significantly enhanced towards the most reliable strategy compared with state-of-the-art methods. This allows an evolutionary multi-species GA to be used through an adaptive search space reduction method, thus the GA-based exploration and exploitation phases ensued very fast and resulting in a cost-effective procedure in dealing with large structural systems.

نویسندگان

Seyed Hossein Mahdavi

Dr., Higher Education Complex of Bam

Kimiya Azimbeik

MSc., Sharif University of Technology