Search

Your search keyword '"Fu, Guiyuan"' showing total 24 results
Start Over Author "Fu, Guiyuan"
24 results on '"Fu, Guiyuan"'

2. Signal Processing Meets SGD: From Momentum to Filter

Author

Yao, Zhipeng, Fu, Guiyuan, Li, Ying, Zhang, Yu, Li, Dazhou, and Yu, Rui

Subjects
Computer Science - Machine Learning, Electrical Engineering and Systems Science - Signal Processing
Abstract

In deep learning, stochastic gradient descent (SGD) and its momentum-based variants are widely used for optimization, but they typically suffer from slow convergence. Conversely, existing adaptive learning rate optimizers speed up convergence but often compromise generalization. To resolve this issue, we propose a novel optimization method designed to accelerate SGD's convergence without sacrificing generalization. Our approach reduces the variance of the historical gradient, improves first-order moment estimation of SGD by applying Wiener filter theory, and introduces a time-varying adaptive gain. Empirical results demonstrate that SGDF (SGD with Filter) effectively balances convergence and generalization compared to state-of-the-art optimizers.

Published
2023

21. Robust adaptive tracking control of MIMO nonlinear systems in the presence of actuator hysteresis.

Author

Fu, Guiyuan, Ou, Linlin, and Zhang, Weidong

Subjects
*ROBUST control, *TRACKING control systems, *MIMO systems, *NONLINEAR systems, *ACTUATORS, *HYSTERESIS, *MATHEMATICAL bounds, *LOWPASS electric filters
Abstract

Adaptive tracking control of a class of MIMO nonlinear system preceded by unknown hysteresis is investigated. Based on dynamic surface control, an adaptive robust control law is developed and compensators are designed to mitigate the influences of both the unknown bounded external uncertainties and the unknown Prandtl–Islinskii hysteresis. By adopting the low-pass filters, the explosion of complexity caused by tedious computation of the time derivatives of the virtual control laws is overcome. With the proposed control scheme, the closed-loop system is proved to be semi-globally ultimately bounded by the Lyapunov stability theory, and the output of the controlled system can track the desired trajectories with an arbitrarily small error. Finally, numerical simulations are given to verify the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Published
2016
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results