Automation of Experimental Modal Analysis Using Bayesian Optimization

The dynamic characterization of structures by means modal parameters offers many valuable insights into the vibrational behavior these structures. However, parameter estimation has traditionally required expert knowledge and cumbersome manual effort such as, for example, selection poles from a stabilization diagram. Automated approaches which replace user inputs with set rules depending on input data have been developed to address this shortcoming. This paper presents an alternative approach based Bayesian optimization. way, possible solution space is kept as widely open while ensuring high accuracy final model. proposed was validated both synthetic test experimental analysis machine tool. Furthermore, it benchmarked against similar tool well-known numerical computation software application.