Uniformity in Full-State Error Prescribed Performance Control via Error-Driven Flexibility for Input-Saturated Systems With External Disturbances

This paper proposes a fuzzy control scheme that enforces a unified prescribed performance for full-state errors in input-saturated systems with external disturbances. The proposed scheme is characterized by three key innovations: First, a series of functional transformations are designed to guarantee multiple performance behaviors within a unified control framework, enabling desired behaviors through parameter selection without controller redesign. Second, a novel performance function for virtual errors completely eliminates strict initial value constraints, thereby removing offline verification computations and streamlining the design/implementation process. Third, an error-driven mechanism is developed to prevent singularities induced by input saturation and disturbances. Unlike existing flexible prescribed performance control methods that rely on a control input-driven mechanism, this mechanism offers two distinct advantages: it directly adjusts only a single boundary for a more straightforward adjustment, and operates without dependency on auxiliary system integration. This design eliminates adjustment delays while minimizing performance degradation caused by boundary relaxation. Simulations confirm the scheme’s efficacy and superiority. Note to Practitioners—This work addresses a critical challenge in industrial control systems: maintaining prescribed performance under actuator physical limits (input saturation) and external disturbances (e.g., load variations, noise). Traditional prescribed performance methods often degrade or fail under such combined conditions, leading to inefficient operation or safety risks. Our proposed error-driven flexible prescribed performance control offers a practical solution by dynamically adjusting performance boundaries only during saturation or disturbance events (activated when errors exceed predefined thresholds). The unified design framework supports four configurable performance boundary types (unconstrained, upper/lower-bounded, or fully constrained initial errors) through a single control architecture, eliminating the need for case-specific controller redesign. The marriage of unified constraints and error-driven mechanism provides saturation- and disturbance-aware performance adaptation: preserving strict boundaries normally while allowing temporary, adaptive relaxation during saturation or disturbance-induced deviations (e.g., robotic arm overload or sudden external torque), all within a unified control architecture.