itsayushthada / frequency-modelling

This is the project work related to the modelling of Frequency and associated parameters using Machine Learning.

Frequency Modelling of Fized Ended Beam

Paper Link: https://www.sciencedirect.com/science/article/abs/pii/S0888327021003101

If you use the code, please cite

@article{THADA2021107915,
title = {Machine learning based frequency modelling},
journal = {Mechanical Systems and Signal Processing},
volume = {160},
pages = {107915},
year = {2021},
issn = {0888-3270},
doi = {https://doi.org/10.1016/j.ymssp.2021.107915},
url = {https://www.sciencedirect.com/science/article/pii/S0888327021003101},
author = {Ayush Thada and Shreyash Panchal and Ashutosh Dubey and Lokavarapu {Bhaskara Rao}},
keywords = {Machine learning, Non-parametric statistics, Frequency, Cracked beam, Design of experiment, Experimental bias},
abstract = {Detection of cracks in structures has always been an important research topic in the industrial domain closely associated with aerospace, mechanical, marine and civil engineering. The presence of the cracks alters the dynamic response properties. Hence, it becomes crucial to locate these cracks in the structures to avoid any catastrophic failures and maintain structural integrity and performance. The study's objective is to propose two distinct statistical procedures for conducting the machine learning experiment for modelling the frequency and show the effect of experiment design on the results. In the study, the predictive performance of machine learning models and their ensembles is compared within each experiment design and between two experimental designs for the task of prediction of first six natural frequencies of a fixed ended cracked beam. The study highlights the significance of more than one experimental design to reduce the confirmation bias in the research and discusses the proposed methods' generalizability over the different modelling constraints and modelling parameters. The study also discusses a real-world implementation of the learned machine learning models from the perspective of Bayesian optimization.}
}