A composite adaptive robust control approach for ball screw feed system with mismatched uncertainty

Abstract

In practical operation, time-varying bounded uncertainties such as dynamic parameter variations and external disturbances significantly affect the performance of the ball screw feed system. An innovative composite adaptive robust control approach is proposed to handle these uncertainties and improve the system’s trajectory tracking performance. Firstly, the system dynamics are modeled and decomposed into nominal and uncertain parts for separate control design. Then, a composite adaptive law incorporating the dead-zone and an exponential leakage term is proposed to estimate the unknown uncertainty boundaries. This method only requires the knowledge that a boundary exists, without needing any additional information about the boundary itself. Furthermore, a controller is designed that can uniformly handle both matched and mismatched uncertainties. Crucially, to address the challenge of determining the dead-zone size, a dual-layer adaptive hierarchy is introduced: the superior-layer adaptive law determines the value of the dead-zone in real-time, providing parameters for the foundational-layer, which in turn provides parameters for the controller. This strategy achieves parameter adaptation, replacing the traditional approach of engineers determining values based on personal experience. Finally, simulations and experiments are conducted, and the results verified that the proposed control method significantly improves the system’s tracking performance, proving its effectiveness.