A robust adaptive pressure control strategy is proposed for a novel indexing variable-displacement pump. In the proposed approach, parametric uncertainties and unmodeled dynamics are identified to the extent possible using a model free learning network and used to de-couple the dynamics using physical insights derived from careful reduced order modeling. The swash plate motion control is then carefully designed to provide the desired pressure response characteristics showing improved performance with learning. The proposed control framework and designs are validated using a detailed nonlinear simulation model.

This content is only available via PDF.
You do not currently have access to this content.