Integral Barrier Function-Based Adaptive Output Feedback Robust Force Control for Pneumatic Compliant End-Effectors With Complex Hysteresis

Haoqi Tang, Qingxiang Wu, Zhuoqing Liu, Yongchun Fang, Ning Sun

Abstract: Passive compliant grinding systems actuated by double-acting cylinders (DACs) have output forces in both the thrust and tension directions, offering significant advantages such as high response speed and convenient application. However, the complex hysteresis nonlinearities, model uncertainties, and external disturbances of DACs pose serious challenges for high-precision grinding. Additionally, the grinding force between grinding tools and workpiece surfaces is difficult to measure in real time by force sensors. To address these challenges, we establish the equivalent model of serial elastic actuators (SEAs) to transform grinding force into measured displacements. Subsequently, we propose an inner–outer control framework for "bipolar" compliant pneumatic end-effectors. Specifically, in the inner loop, the inversion of established generalized Prandtl–Ishlinskii (GPI) models is constructed. Based on this, we propose an integral barrier function-based adaptive output feedback robust force control method, which includes a disturbance estimation item. Also, the closed-loop stability is rigorously proven by using Lyapunov methods. Further, to estimate large grinding-force tracking errors, we propose an outer loop controller consisting of an SEA and a hysteresis compensator. Finally, experiments are conducted on our self-built experimental platform. Experimental results demonstrate that our proposed methods exhibit satisfactory tracking performance and strong robustness.