Self-updating inverse model for magnetorheological dampers

Inverse models for magnetorheological (MR) devices can be used to calculate the command current required for the MR damper to generate the desired forces. Such models are of great importance in the development of semi-active vibration control. Nevertheless, it is difficult to generate inverse dynamic models because of the inherent nonlinearity of MR dampers. In this paper, we present an adaptive method for establishing the inverse model and describe the corresponding optimization algorithm. Unlike most inverse models, our self-updating model can be adjusted to describe the dynamic characteristics of MR dampers under changes in the external incentives. Experiments under different conditions verify the effectiveness of the proposed inverse model. A comparison between the proposed model and a traditional model with fixed parameters shows that the error can be reduced by approximately 20%. This demonstration of a self-updating inverse model expands the description of MR dampers and provides a new idea for vibration control under varying conditions.