Clustered Firefly Algorithm for Global Parameter Estimation in Cyber-Physical Systems: Applications in Wastewater Treatment and Residential Heating
Author
Term
4. term
Education
Publication year
2025
Submitted on
2025-06-11
Pages
83
Abstract
This thesis presents a novel variant of the Firefly Algorithm—called the Clustered Firefly Algorithm (cFA)—for efficient parameter estimation in dynamic models used within the context of Model Predictive Control (MPC). By integrating k-Means clustering with the original Firefly Algorithm, the proposed method reduces time complexity while maintaining strong global search capabilities. Benchmark experiments across several standard optimization functions demonstrate that the cFA converges faster and more reliably to high-quality solutions compared to the original Firefly Algorithm and other baseline methods. The algorithm's effectiveness is further validated through two real-world applications: parameter estimation for a theoretical wastewater treatment plant (ASM1 model) and a thermal building model used for heat pump control. Results indicate that the algorithm handles complex, multimodal search landscapes effectively. While the results are promising, further experimental validation is recommended to fully assess the cFA's applicability across diverse MPC scenarios.
Keywords
Documents