.png){kind=icon}
.png){kind=icon}
Return
Signed Learning Intensity in Fuzzy Neural Self-Learning Control for Model-Free Attitude Tracking
.png){kind=icon}
PRE
AI
X
Q
C
Z
DOI：10.1109/taes.2026.3714377.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This article presents a signed self-learning-based fuzzy neural network control scheme for spacecraft attitude tracking. The proposed model-free controller integrates self-learning control (SLC) with a fuzzy neural network (FNN) using signed learning intensity. The combined architecture leverages SLC to achieve rapid transient convergence, while the FNN compensates for disturbances in the steady state. A Gaussian-type neural network learning rate, aligned with the FNN's radial basis structure, simplifies stability analysis. Crucially, the signed learning intensity can take negative values near equilibrium, actively counteracting neural estimation errors and mitigating chattering in fuzzy neural adaptive systems. Simulations show a 77% improvement in steady-state accuracy and effective chattering reduction compared to tanh-type methods under disturbances.
Keywords:
Fuzzy neural network (FNN)
self-learning control (SLC)
signed learning intensity
Journal
IF:
5.7
Papers:
651
Citations:
2.4W
Organization
Cited Papers