Barrier function-based dynamic surface control with global stability guarantees: Theory and implementation

Classical dynamic surface control (CDSC) is widely adopted to alleviate the complexity explosion inherent in integrator backstepping control (IBC), where each virtual control input is processed through a linear low-pass filter. Nevertheless, the introduction of such filters inevitably induces additional errors, which may compromise the global boundedness of the closed-loop system. This paper develops an improved dynamic surface control (DSC) framework for strict-feedback nonlinear systems (SFNSs) that rigorously guarantees global stability. In contrast to CDSC approaches, the proposed method replaces linear filters with barrier function-based nonlinear filters, which effectively mitigate the complexity issue while constraining the filtered errors. A salient feature of the proposed framework is its ability to ensure global uniform boundedness (GUB) of all closed-loop signals, whereas existing CDSC methods typically provide only semiglobal uniform boundedness (SGUB). Extensive simulation and experimental results demonstrate that the proposed approach achieves an average reduction of 57.15% in tracking error compared with CDSC and preserves accurate tracking performance even under a large filtering time constant (ψ=1), thereby highlighting its improved robustness and superior control performance.