The paper presents usage of a Particle Swarm Optimization [1] based algorithm for spectral-based modal parameters identification. The main algorithm consists of two groups of swarms, namely, scouts and helpers. For the first group additional penalizing process is provided to force separation of scouting swarms in frequency space. The swarms have an ability to communicate with each other. At first stage, each swarm focuses on a different natural frequencies dominating in the signal. Next, helper swarms are associated, one by one, with scouting swarms focus on frequency spaces provided by scouting swarms in order to achieve the best representation of a given signal in the frequency domain. At this stage amplitudes and damping coefficients are identified as well. The results from all helpers are then combined to create final identification results.
Authors
Additional information
- Category
- Publikacja monograficzna
- Type
- rozdział, artykuł w książce - dziele zbiorowym /podręczniku o zasięgu krajowym
- Language
- angielski
- Publication year
- 2017
