Comparative Evaluation of Semiactive Control Strategies for the Milad Telecommunication Tower Using MR Dampers Under Spectrally Matched Near‐ and Far‐Field Earthquakes

ABSTRACT Tall telecommunication towers are highly vulnerable to seismic and wind‐induced vibrations due to their slender geometry, low damping, and long fundamental periods. Existing control systems often struggle to balance adaptability, energy efficiency, and real‐time feasibility. This study introduces a fully data‐driven semiactive control framework for the Milad Tower, integrating magnetorheological (MR) dampers modeled via artificial neural networks (ANNs) with four control strategies: linear quadratic regulator (LQR), fuzzy logic controller (FLC), model predictive controller (MPC), and adaptive neuro‐fuzzy inference system (ANFIS). The tower was subjected to 21 spectrally matched ground motions, categorized into near‐field with pulse, near‐field without pulse, and far‐field earthquakes, tailored to its long‐period dynamics using wavelet‐based spectral matching. Among all strategies, MPC achieved the greatest peak and RMS response reductions, lowering RMS displacement and acceleration by 73.3% and 40.3%, respectively, but required the highest control effort (60.9%). ANFIS matched or slightly exceeded MPC's performance while consuming less energy (54.6%) and demonstrated superior adaptability, especially under high‐pulse near‐field events by preventing control saturation. FLC consistently outperformed LQR but lagged behind ANFIS in adaptability. These results underscore the effectiveness of combining ANN‐based damper modeling with intelligent, data‐driven controllers, offering a high‐performance and energy‐efficient solution for real‐time seismic mitigation in tall telecommunication structures.