Smooth and safe stop: Fixed-time fault tolerant control for heavy legged robot with active identification on tolerance capability

Abstract

Heavy-legged robots (HLRs), integral to optimizing efficiency in manufacturing and transportation, rely on advanced active servo fault diagnosis and fault-tolerant control (FTC) mechanisms. This study presents an FTC framework with active fault status identification, fault tolerance capability assessment, and model uncertainty handling. A key contribution is the introduction of an active servo fault state estimator (ASFSE), which enables real-time monitoring of servo status by comparing residual differences between servo and controller outputs. The system’s tolerance capability interval (TCI) is tied to the servo state, with the dual-line particle filters (DPF) algorithm predicting when the HLR exceeds the TCI under faults. Subsequently, a target trajectory modifier (TTM) and fixed-time backstepping controller (FTBC) are proposed. The TTM promptly adjusts the trajectory when the HLR surpasses the TCI, while the FTBC ensures fixed-time convergence based on the predicted failure time for precise trajectory tracking. As the HLR approaches its fault tolerance limits, the TTM and FTBC ensure a smooth stop, thus mitigating equipment damage caused by servo faults. Mathematical stability proof and simulation validations confirm the effectiveness of the FTC framework.