Neural network based semi-active control strategy for structural vibration mitigation with magnetorheological damper

This paper presents a neural network based semi-active control method for a rotary type magnetorheological (MR) damper. The characteristics of the MR damper are described by the classic Bouc-Wen model, and the performance of the proposed control method is evaluated in terms of a base exited shear frame structure.

As demonstrated in the literature effective damping of flexible structures is obtained by a suitable combination of pure friction and negative damper stiffness. This damper model is rate-independent and fully described by the desired shape of the hysteresis loops or force-displacement trajectories. The proposed neural network controller is therefore trained based on data derived from these desired forcedisplacement curves, where the optimal relation between friction force level and response amplitude is determined explicitly by simply maximizing the damping ratio of the targeted vibration mode of the structure.

The neural network control is then developed to reproduce the desired force based on damper displacement and velocity as network input, and it is therefore referred to as an amplitude dependent model reference control method. An inverse model of the MR damper is needed to determine the damper current based on the derived optimal damper force.

For that reason an inverse MR damper model is also designed based on the neural network identification of the particular rotary MR damper. The performance of the proposed controller is compared to that of an optimal pure viscous damper. The top floor displacement and acceleration of the base excited shear frame structure are selected as the performance parameters of this comparison.

It is found by the simulations that the proposed control design yields a reduction in the structural response compared to the viscous case.