Article
Chemistry, Multidisciplinary
Shun-Yuan Wang, Chuan-Min Lin, Chen-Hao Li
Summary: The study proposes an adaptive Takagi-Sugeno-Kang fuzzy self-organizing recurrent cerebellar model articulation controller (ATFSORC) for synchronization and control of chaos. By utilizing different types of recurrent CMAC, self-organizing CMAC, and TSK fuzzy rules, it aims to reduce structure complexity, improve control performance, and increase learning speed. Experimental results show that the proposed method has favorable control performance in chaotic systems.
APPLIED SCIENCES-BASEL
(2021)
Article
Environmental Sciences
Vahid Moosavi, Ayoob Karami, Ramyar Aliramaee
Summary: This study conducted a comparative analysis of PSO-based optimized CMAC and PSO-based optimized GMDH in generating high-resolution surface soil moisture (SM) estimates over a Mediterranean agro-ecosystem using sentinel-2 imagery. The results showed that the PSO-CMAC method outperformed the PSO-GMDH method in terms of accuracy for both global and local approaches.
SCIENCE OF THE TOTAL ENVIRONMENT
(2023)
Article
Automation & Control Systems
Thanh-Quyen Ngo, Dinh-Khoi Hoang, Trong-Toan Tran, Thanh-Thuan Nguyen, Van-Tho Nguyen, Long-Ho Le
Summary: This paper presents an intelligent controller for robotic systems with uncertainties. The controller combines a cerebellar model articulation controller and sliding mode control, and adjusts its structure automatically using a self-organizing approach. Experimental results demonstrate the effectiveness of the proposed controller.
INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL
(2022)
Article
Automation & Control Systems
Ning Xu, Yun Chen, Anke Xue, Huanqing Wang, Xudong Zhao
Summary: This article presents an effective method to address the tracking control problem in uncertain switched high-order nonlinear system. By adopting the adding a power integrator approach in the framework of backstepping, a novel tracking controller is developed to guarantee an asymptotic tracking performance under the approximation error caused by neural networks.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2022)
Article
Automation & Control Systems
Ashia C. Wilson, Ben Recht, Michael Jordan
Summary: Accelerated optimization methods, like Nesterov's accelerated gradient method, play a significant role in optimization and some methods are provably optimal under standard oracle models. The difficult technique of estimate sequences led researchers to develop more intuitive methods. Research has shown equivalence between estimate sequences and a family of Lyapunov functions.
JOURNAL OF MACHINE LEARNING RESEARCH
(2021)
Article
Mathematics
Borja Sanchez-Lopez, Jesus Cerquides
Summary: This work aims to provide a geometric explanation to convergence results and to state and identify conditions for the convergence of not exclusively optimization methods but any stochastic process. The expected directions of a stochastic process are related to conservative vector fields, with the possibility of ensuring convergence under certain conditions. The study translates existing convergence results into processes resembling particular conservative vector fields, offering a new perspective on identifying Lyapunov functions for proving convergence.
Article
Computer Science, Information Systems
Mohammed Yousri Silaa, Oscar Barambones, Aissa Bencherif
Summary: This paper presents an adaptive PID control algorithm using stochastic gradient descent with momentum (SGDM) for a proton exchange membrane fuel cell (PEMFC) power system. By minimizing the cost function and adapting the PID parameters according to perturbation changes, this method achieves fast convergence and high robustness in control performance.
Article
Automation & Control Systems
Huifang Min, Shengyuan Xu, Baoyong Zhang, Qian Ma, Deming Yuan
Summary: This paper investigates the fixed-time stability theorem and state-feedback controller design for stochastic nonlinear systems. An improved fixed-time Lyapunov theorem is proposed with a more rigorous and reasonable proof procedure. A state-feedback controller is skillfully designed based on the backstepping technique and the addition of a power integrator method. It is proved that the proposed controller can render the closed-loop system fixed-time stable in probability.
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2022)
Article
Mathematics, Interdisciplinary Applications
Jia Tang
Summary: This paper proposes two fractional gradient descent algorithms for switching models. Each submodel is assigned a weight which can determine the identity of the submodel in each sampling instant. By using the fractional gradient descent algorithms, the parameters of each submodel can be obtained, and then the weights of all the submodels can be estimated based on the self-organizing maps method. These two algorithms can deal with different kinds of switching models on a case by case basis. In addition, compared with the traditional identification algorithms, the proposed methods have two advantages: (1) has faster convergence rates; (2) has less computational efforts. Simulation example demonstrates the effectiveness of the proposed methods.
CHAOS SOLITONS & FRACTALS
(2023)
Article
Automation & Control Systems
Van-Phuong Ta, Dinh-Nhon Truong, Nguyen-Thanh Nhan
Summary: Water Distribution System (WDS) is a crucial phase of the Water Treatment Plant (WTP) that plays a vital role in plant, animal, and human life. To maintain a continuous, stable water supply, water pressure in the pipe network must be maintained at desired set points despite uncertainties, disturbances, and noises. In order to save energy, a Variable Frequency Driver (VFD) and Recurrent Cerebellar Model Articulation Control System (RCMACS) were used to control the speed of the Water Boost Pump (WBP) and maintain water pressure at the desired level.
INTELLIGENT AUTOMATION AND SOFT COMPUTING
(2022)
Article
Automation & Control Systems
Tien-Loc Le
Summary: In this paper, the intelligent control using interval type-2 fuzzy systems and cerebellar model articulation networks (CMAN) has been studied. To address the issues of learning rate choice and adaptive laws for updating network parameters, an improved gray wolf optimizer (IGWO) and an adaptation law were proposed. Additionally, a three-dimensional Gaussian membership function (3DGMF) was developed to handle external disturbances and system uncertainties. The numerical simulation results on MEMS motion control validated the effectiveness of the proposed control method.
INTERNATIONAL JOURNAL OF FUZZY SYSTEMS
(2023)
Article
Engineering, Mechanical
Amardeep Mishra, Satadal Ghosh
Summary: This paper proposes a novel tuning law based on variable gain gradient descent for the critic neural network, aiming to solve the optimal tracking problem of continuous time nonlinear systems. The proposed update law adjusts the learning rate based on HJB approximation error, improving the convergence time of critic neural network weights and resulting in smaller oscillations in state error.
NONLINEAR DYNAMICS
(2022)
Article
Engineering, Mechanical
Robab Ebrahimi Bavili, Ahmad Akbari, Reza Mahboobi Esfanjani
Summary: This paper utilizes the IDA-PBC notion to design a novel controller for teleoperation systems, improving position and force tracking performance by shaping kinetic and potential energies, and deriving synthesis conditions in terms of LMIs to ensure stable tracking. Comparative simulation results demonstrate the superiority of the proposed method over recent rival methods, and laboratory experiments verify its real-world applicability.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2021)
Article
Automation & Control Systems
Rujun Fan, Yunhua Li
Summary: An adaptive control system utilizing a terminal sliding mode controller and a new neural network controller has been proposed for trajectory tracking of an electro-hydraulic shovel (EHS), with only the system position signal being utilized in the control law. The newly designed hybrid control structure comprises an RBFNN preprocessor and a CMAC controller, resulting in faster learning and reduced computational cost. The control law is derived based on Lyapunov stability theory, enabling finite-time convergence and the introduction of an adaptive compensation term to address combination errors. Experimental results demonstrate that the proposed method can effectively handle lumped uncertainties with high computational efficiency.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2021)
Article
Automation & Control Systems
Mingjun Dai, Zelong Zhang, Xiong Lai, Xiaohui Lin, Hui Wang
Summary: Due to the issues with SGD optimization algorithms, the Adam algorithm, although it can adaptively adjust the update direction and learning rate, still suffers from overshoot phenomenon and slow convergence. In this study, the concept of PID controller is borrowed to propose a complete adaptive PID optimizer, which successfully alleviates the overshoot phenomenon and speeds up the convergence.
IET CONTROL THEORY AND APPLICATIONS
(2023)