<p>Due to nature’s complexity, buildings, vehicles, aircraft, and machines are subject to disturbances of unknown magnitude, frequencies, and duration, causing unwanted mechanical vibrations. These vibrations need to be controlled, and the vibration control can be passive, active, or hybrid. Smart materials, such as shape memory alloys (SMAs), have been widely studied for passive vibration control. Coupled with some active control techniques, the resulting hybrid control produces redundancy for the system. However, the energy dissipation mechanism of SMA produces hysteresis, a recognized nonlinear phenomenon within the field of control systems. In addition, other nonlinearities can arise, such as saturation and dead zone in the control signal. It’s also common to not fully understand the dynamic model in complex systems. An intelligent control (IC) technique was proposed to cope with these challenges. This novel approach combines the traditional sliding mode control (SMC) methodology with a radial-basis function (RBF) network, and its convergence and stability were analytically proved using the Lyapunov stability theory. Within the proposed control, online training updates the neural network’s weight vectors, improving the ability to predict plant dynamics during task execution. Numerical simulations were performed, and their results were compared. The results revealed that the IC exhibited a notably smaller boundary layer in the error state space than the SMC. The proposed IC obtained a steady state error 77 % lower than the SMC. Several performance indices were computed, and the IC achieved a superior performance in all compared to the classic SMC.</p>