Research data for “Reinforcement learning-based control to suppress the transient vibration of semi-active structures subjected to unknown harmonic excitation” (DOI: 10.1111/mice.12920)

This project looks at using reinforcement learning (RL) algorithms to control and reduce temporary vibrations in semi‑active structures that are hit by unknown harmonic forces. Reducing vibration is important to keep civil and mechanical structures safe when they experience dynamic loads. Conventional control methods often have trouble when the external forces are unknown or change. With reinforcement learning, we can train an adaptive control policy that changes the damping settings of semi‑active devices based on real‑time feedback from the structure’s state. The method uses state‑space models of the structure’s nodes, treated as a multi‑degree‑of‑freedom system, to define the RL agent’s state, action, and reward. Numerical simulations run on a high‑performance

computer to check how well the trained agent converges, stays stable, and suppresses vibrations. The project gives detailed instructions on setting up state‑space variables, creating reward functions that penalize both displacement and velocity, and testing the controller’s robustness against excitation frequencies that were not modeled. This work provides a base for building intelligent structural control systems that can adapt in real time.

Message: Hi FE, I need help with “Research data for "Reinforcement learning-based control to suppress the transient vibration of semi-active structures subjected to unknown harmonic excitation" (DOI: 10.1111/mice.12920)” in “Computer Science & Engineering”