Work

A Hybrid MLS-HEPSO Framework for Fuzzy Fractional-Order Adaptive Robust Control of Under-Actuated Systems

Year: 2026
Type: article
Abstract: Tracking control of nonlinear under-actuated systems is difficult due to coupled dynamics, model uncertainties, and external disturbances. To handle these issues, we propose a control framework and apply it to a ball–wheel system. Our framework combines a moving least squares (MLS) approximation layer with a high-exploration particle swarm optimization (HEPSO) algorithm. These tools optimize a fuzzy fractional order adaptive robust feedback linearization controller. The controller itself integrates approximate feedback linearization, decoupled sliding surfaces, Grünwald–Letnikov fractional derivatives, adaptive laws, and fuzzy inference. Within this structure, the MLS layer provides a continuous approximation of the control law. We treat the fictitious nodal values as design parameters and optimize them offline using HEPSO. This optimization minimizes a cost function based on the integral absolute errors of the ball and wheel angles. After this offline tuning, the optimized MLS layer handles the online control. We tested the system under nominal and disturbed conditions. The simulations show better tracking accuracy and disturbance rejection than baseline methods. Sensitivity tests also confirm that tracking errors and control effort stay bounded without causing actuator saturation.
Language: English
Open Access status: gold
Estimated APC paid: $3,258
OpenAlex ID: W7235447141
FWCI:
Cites: 28
Cited by:
Related to: 10