Adaptive control systems for structures subjected to earthquake and wind loadings

Current vibration dissipation systems are inadequate for effectively managing varying load patterns and structural responses to excitations. A robust vibration control system capable of handling multidirectional structural response interactions is crucial, particularly in controlling both story drift and acceleration, as improving one can often exacerbate the other. Traditional model-based control algorithms have limited adaptivity due to their reliance on explicitly defined structural properties and the need for extensive training data, making them less adaptable to unpredictable excitations. This research addresses these challenges in two parts. In Part One, a Fuzzy Logic Multi-Verse Optimal Control (FLMVOC) system is developed as a new adaptive real-time vibration control mechanism for structures subjected to seismic and wind loads. This system integrates a stochastic optimization method with fuzzy logic techniques, employing a magnetorheological damper (MR) as a controllable vibration damping system. The fuzzy logic controller (FLC) utilizes story drift and absolute acceleration as inputs and outputs the MR commanding voltage. The FLC is trained using the multi-objective multi-verse algorithm (MOMVO) within a hybrid MATLAB-Simulink and finite element simulation framework to optimize the controller parameters. The FLMVOC system is tested on both seismic-excited structures and high-rise buildings subjected to wind loads, demonstrating its effectiveness in reducing dynamic structural responses. In Part Two, an Online Fuzzy Logic Multi-Verse Optimal Vibration Control system (Online-FLMVOC) is developed to address the problem of system dependency on training data. This novel system combines fuzzy logic inference with the multi-verse optimization algorithm and introduces a decentralized approach with memory-saving mechanisms. The Online-FLMVOC is evaluated on a three-story shear building subjected to seismic loads and a 76-story benchmark building facing multidirectional wind loads, showcasing its real-time adaptability and effectiveness in minimizing structural responses. Results show that the FLMVOC system reduced structural drifts by 60%, 53%, and 41% during the El Centro, Kobe, and Northridge earthquakes, respectively, with corresponding reductions in floor absolute acceleration of 38%, 17%, and 10%. In wind-induced structures, the system kept floor acceleration within comfort criteria while reducing story drift. The Online-FLMVOC system further improved performance, reducing structural drifts by 75% and 48% during the El Centro and Northridge earthquakes, with reductions in floor absolute acceleration of 60% and 26%, respectively. For wind-induced responses, the system achieved reductions in story displacement by 37.5%, 14.5%, and 12% for across wind, along wind, and rotational responses, respectively. The Online-FLMVOC system achieved similar control effectiveness to reference algorithms but with only 45% to 66% of the required control force. In conclusion, the developed FLMVOC and Online-FLMVOC systems successfully address the challenges outlined, offering competitive performance against other vibration control strategies under varying load patterns. However, further investigation is needed to assess the reliability of these adaptive vibration control systems on real-prototype structures.

Text FK 2023 19 - Declaration Form.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (513kB) Text FK 2023 19 - Full Text.pdf Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (11MB) Text FK 2023 19.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (11MB) Official URL or Download Paper: http://ethesis.upm.edu.my/id/eprint/18936

Actions (login required)

View Item