Article
Computer Science, Artificial Intelligence
Anuli Dass, Smriti Srivastava, Monika Gupta, Manju Khari, Javier Parra Fuente, Elena Verdu
Summary: Identification and control of non-linear and complex dynamical systems are important and challenging topics in adaptive control systems. This paper introduces an optimization algorithm called intelligent water drop (IWD) algorithm and demonstrates its application in modelling and controlling non-linear dynamic systems.
Article
Computer Science, Information Systems
Shihan Huang, Hua Dang, Rongkun Jiang, Yue Hao, Chengbo Xue, Wei Gu
Summary: This study introduces a multi-layer hybrid fuzzy support vector machine model which effectively addresses some issues in human emotion recognition. By combining a feature extraction layer and multi-layer classifiers, it has successfully improved the stability and accuracy of emotion recognition systems.
Article
Computer Science, Information Systems
Dan Bao, Xiaoling Liang, Shuzhi Sam Ge, Zhiwei Hao, Baolin Hou
Summary: This paper introduces a framework for adaptive fuzzy control and optimization of nonlinear systems under uncertainties and disturbances. The barrier Lyapunov function (BLF) technique is used to determine output constraints. Fuzzy logic systems (FLSs) are applied to approximate nonlinear terms and improve tracking performance. Bayesian optimization and particle swarm optimization (BO-PSO) are combined for gains optimization to enhance control performance. Multilayer neural networks (MNNs) are employed as surrogate models of nonlinear systems with interval parameters to improve the computational efficiency of the optimization process. Two simulations are conducted to demonstrate the effectiveness of the proposed framework.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Aijun Kou, Xiaojun Li
Summary: In order to reduce errors in intelligent control systems while ensuring system performance, this study proposes a new Particle Swarm Optimization (PSO) optimization scheme and innovatively improves neural network intelligent control based on the MPSO algorithm. The experimental results show that the improved controller controls the error within 0.01 within 0.02 seconds, and the control error of MPSO has decreased by 98.95% and 93.06% compared to PSO and WOA algorithms, respectively. It not only has the best control effect but also has the shortest system response time.
Article
Automation & Control Systems
Islam Helmy, Wooyeol Choi
Summary: Precise focus is crucial for quality astronomical observations and scientific research. This study proposes a fuzzy logic-based focus measure and uses particle swarm optimization to optimize its parameters. The results show that the proposed method outperforms traditional focus operators, providing improved accuracy for high magnification astronomical images.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Claudiu Pozna, Radu-Emil Precup, Erno Horvath, Emil M. Petriu
Summary: This article presents a hybrid metaheuristic optimization algorithm that combines particle filter (PF) and particle swarm optimization (PSO) algorithms. The algorithm is applied to the optimal tuning of proportional-integral-fuzzy controllers for position control of integral-type servo systems, resulting in reduced energy consumption. A comparison with other metaheuristic algorithms is provided at the end of the article.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2022)
Article
Nuclear Science & Technology
Mengwei Zhao, Zhi Chen, Longtao Liao, Kai Xiao, Qingyu Huang
Summary: An intelligent multi-step predictive scheme is proposed in this paper to support the development of autonomous control for small modular reactors (SMRs). This scheme constructs an adaptive prediction model using case-generated training data of reactor system model, and employs a real-time adaptive algorithm for updating model parameters to improve prediction accuracy. Particle swarm optimization (PSO) is also used to search for the optimal decision. The simulation results demonstrate the effectiveness and good performance of the proposed approach.
ANNALS OF NUCLEAR ENERGY
(2022)
Article
Computer Science, Information Systems
Qiang Zhang, Ping Liu, Quan Deng, Angxin Tong, Juergen Pannek
Summary: This paper proposes a novel intelligent operator-based sliding mode control scheme to address the problem of trajectory tracking control in the presence of bounded model uncertainty and external disturbance. By using the operator-based robust coprime factorization method, robust stability is guaranteed. A finite-time integral sliding mode control law is also designed to achieve asymptotic tracking and enhance responsiveness to disturbance. The controller's parameters are automatically adjusted using stabilizing particle swarm optimization with linear time-varying inertia weight.
Article
Multidisciplinary Sciences
Majad Mansoor, Adeel Feroz Mirza, Fei Long, Qiang Ling
Summary: The proposed MPPT strategy based on the tunicate swarm algorithm shows promising results in improving power tracking efficiency and robustness under partial shading conditions in photovoltaic systems.
ADVANCED THEORY AND SIMULATIONS
(2021)
Article
Automation & Control Systems
Song Liu, Shumin Zhou, Xiujuan Lu, Fang Gao, Feng Shuang, Sen Kuang
Summary: This paper presents a Lyapunov control scheme to drive finite-dimensional closed and Markovian open quantum systems into any target pure state with high fidelity and short time. The control law is established using a Lyapunov function and the optimal eigenvalues are searched using the quantum-behaved particle swarm optimization algorithm. A improved constrained QPSO algorithm is proposed for open systems with small denominator in the control law. Numerical simulations on different quantum systems demonstrate the effectiveness of the proposed control scheme.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2023)
Article
Computer Science, Information Systems
Nur Syazreen Ahmad
Summary: This paper studies the control optimization strategies for the ball and beam system and introduces a novel intelligent control approach that combines Particle Swarm Optimization (PSO) and Whale Optimization Algorithm (WOA) for optimal parameter search. The proposed hybrid approach demonstrates superior performance compared to traditional methods and individual optimization algorithms in simulations.
Article
Energy & Fuels
Jue Hou, Zhou Liu, Shaorong Wang, Zhe Chen, Ji Han, Wei Xie, Chen Fang, Jinsong Liu
Summary: This paper proposes an intelligent coordinated damping control strategy to improve the dynamic characteristics of ADN with multiple devices. The strategy, composed of supplementary control and coordinated strategy, effectively enhances the dynamic stability of the system and provides a powerful solution to address the dynamic instability problem of ADN.
Article
Engineering, Electrical & Electronic
Arvin Ghasemi, Mostafa Sedighizadeh, Ahmad Fakharian, Mohammad Reza Nasiri
Summary: Control of critical micro-grid factors such as voltage/frequency is complex and challenging. This study proposes an intelligent control scheme for primary and secondary microgrids, which optimizes the voltage/frequency restoration and compensates for deviations. The proposed algorithm performs optimally and efficiently under various scenarios.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2022)
Article
Automation & Control Systems
Mustafa Yavuz Coskun, Mehmet Itik
Summary: In this research, an intelligent PID (i-PID) controller is developed for position control of a nonlinear electrohydraulic system with uncertain valve characteristics and supply pressure variations. The controller utilizes an estimation of the ultra-local model of the system. The controller parameters are optimized through the particle swarm optimization method, based on a nominal nonlinear model, and its performance is evaluated under uncertainties caused by valve characteristics, supply pressure variations, and friction between the piston and the hydraulic cylinder.
Article
Chemistry, Multidisciplinary
Samia Charfeddine, Attia Boudjemline, Sondess Ben Aoun, Houssem Jerbi, Mourad Kchaou, Obaid Alshammari, Zied Elleuch, Rabeh Abbassi
Summary: This paper addresses the control problem of nonlinear disturbed polynomial systems using output feedback linearization and sliding mode control, aiming to ensure the asymptotic stability of unstable equilibrium points. By incorporating meta-heuristic techniques, a robust control strategy was designed to decouple system output from disturbances and achieve desired trajectory tracking. The effectiveness and efficiency of the technique were evaluated through numerical simulation analysis on a benchmark model of a continuous stirred tank reactor (CSTR).
APPLIED SCIENCES-BASEL
(2021)
Article
Operations Research & Management Science
Zai-Yun Peng, Jing-Jing Wang, Ka Fai Cedric Yiu, Yun-Bin Zhao
Summary: This paper establishes existence results for the solution of the generalized symmetric set-valued quasi-equilibrium problem (GSSQEP) and introduces new forms of the problem. Sufficient conditions for the existence of solutions to GSSQEP are developed using fixed point method, maximal element principle, and nonlinear scalarization technique. The paper also provides applications to related problems and improves existing results.
Article
Computer Science, Artificial Intelligence
Jonah Ong, Ba-Tuong Vo, Ba-Ngu Vo, Du Yong Kim, Sven Nordholm
Summary: This paper presents an online multi-camera multi-object tracker that can be trained with a monocular detector and is independent of the multi-camera configurations. It operates in the 3D world frame and provides 3D trajectory estimates of objects. The proposed algorithm integrates track management, state estimation, clutter rejection, and occlusion/misdetection handling into a single Bayesian recursion using a high fidelity yet tractable 3D occlusion model.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Computer Science, Information Systems
Mingjie Gao, Ka-Fai Cedric Yiu
Summary: In this paper, we investigate the asymptotic behaviors of an estimator for the number of operating sensors in a sensor network using the Good-Turing estimator. We obtain the asymptotic normality, moderate deviations, and deviation inequalities of the estimator. Our approach is based on tail probability estimates and moderate deviations for occupancy problems. By applying these asymptotic behaviors, we provide a performance analysis for the estimator of the number N of operating nodes when the deviations of the estimator are within (root N, o(N)). These estimates also offer a method for constructing confidence intervals for N.
IEEE TRANSACTIONS ON INFORMATION THEORY
(2023)
Article
Computer Science, Artificial Intelligence
N. Man, S. Guo, K. F. C. Yiu, C. K. S. Leung
Summary: Optical Coherence Tomography (OCT) is a non-invasive method for early diagnosis of ocular diseases. This research focuses on retinal layer segmentation in OCT images, exploring algorithms and network structures, and proposing a method to reduce complexity when training a large volume of data on a cloud platform.
Article
Mathematics
Zeshan Aslam Khan, Naveed Ishtiaq Chaudhary, Waqar Ali Abbasi, Sai Ho Ling, Muhammad Asif Zahoor Raja
Summary: A recommender system aims to gain users' confidence and reduce their time and effort. In this study, an improved, confidence-integrated denoising auto-encoder (DAE) is proposed to enhance the performance of recommender systems. The proposed model achieves improved scores in various evaluation metrics and proves to be efficient and accurate in generating recommendations.
Article
Mathematics, Applied
Mingjie Gao, Ka-Fai Cedric Yiu
Summary: This paper addresses the importance of beamforming in signal enhancement and proposes a method that considers both filters and microphone positions as design variables. The Gauss-Newton algorithm is employed to simultaneously update these two variables during iterations. The effectiveness of the proposed method is demonstrated through several design examples.
NUMERICAL ALGEBRA CONTROL AND OPTIMIZATION
(2023)
Article
Computer Science, Artificial Intelligence
W. M. Tang, K. F. C. Yiu, K. Y. Chan, K. Zhang
Summary: Accurate traffic forecasting is crucial for regional traffic management. The proposed NSDNN approach combines DNN and subset selection method to extract useful inputs from nearby roads. By selecting appropriate input subsets based on congestion cycle patterns, the method reduces input data dimensions and avoids artificial high correlations. Experimental results show that NSDNN achieves higher accuracy compared to other conventional methods. It is also comparable to NSLSTM when the same selected input subset is used. The forecasting system can benefit logistic companies in route planning and fleet management.
APPLIED SOFT COMPUTING
(2023)
Article
Automation & Control Systems
Maolin Wang, Xinsong Yang, Shuoyu Mao, Ka Fai Cedric Yiu, Ju H. Park
Summary: This article studies the leader-following consensus problem in multi-agent systems (MASs) with time-varying switching subject to deception attacks. The one-sided Lipschitz (OSL) condition is used for the nonlinear functions, resulting in more general and relaxed results than those obtained using Lipschitz condition. Nonidentical double event-triggering mechanisms (ETMs) are adopted for only a fraction of agents, and each agent transmits data according to its own necessity. The switching topology is modeled using semi-Markov process with time-varying switching probability, and deception attacks in the transmission channel are considered. Sufficient conditions for MASs to achieve consensus in mean square are obtained using the cumulative distribution function (CDF) and linear matrix inequality (LMI) technology. An effective algorithm is presented for obtaining event-based control gains. The advantages of the proposed control scheme are demonstrated through a simulation example.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Yang An, Hak Keung Lam, Sai Ho Ling
Summary: This paper develops a single-channel-based convolutional neural network to tackle multi-classification motor imagery tasks. The proposed method uses a single-channel learning strategy to extract effective information from each independent channel, making the information between adjacent channels not affect each other. It also proposes a data evaluation method and a mutual information-based regularization parameters auto-selection algorithm to generate effective spatial filters.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Operations Research & Management Science
Mingjie Gao, Ka-Fai Cedric Yiu
Summary: This paper investigates the convergence of the sample average approximation (SAA) solution for stochastic variational inequalities in regimes of moderate deviations. By using the delta method and exponential approximation, some results on moderate deviations are established. The results are applied to hypotheses testing, showing that the rejection region constructed by the central limit theorem has the probability of the type II error with exponential decay speed. Simulations and numerical results for the tail probabilities are also provided.
Article
Mathematics, Applied
M. I. N. G. J. I. E. Gao, Ka-fai cedric Yiu
Summary: This paper studies the moderate deviations and convergence rates for the optimal values and optimal solutions of sample average approximations. It gives an extension of the Delta method in large deviations and establishes a moderate deviation principle for the optimal value under Lipschitz continuity on the objective function. It also obtains a moderate deviation principle for the optimal solution and a Cramer-type moderate deviation for the optimal value when the objective function is twice continuously differentiable and the optimal solution of true optimization problem is unique.
SIAM JOURNAL ON OPTIMIZATION
(2023)
Article
Chemistry, Analytical
Wunna Tun, Kwok-Wai (Johnny) Wong, Sai-Ho Ling
Summary: This article presents a framework for HVAC fault detection using HVACSIM+ simulated data and GAF-2DCNN method. By converting time-series sensor data into informative 2D images and extracting features using 2DCNN, this method captures hidden temporal relationships in 1D signals. Experimental results demonstrate high accuracy and precision in HVAC fault detection using this method.
Article
Acoustics
Qi He, Mingjie Gao, Ka Fai Cedric Yiu, Sven Nordholm
Summary: In multimedia applications, it is common to use acoustic sensors collectively to enhance signals and locate sound sources. This article investigates the microphone array localization problem in a distributed acoustic network with TDOA measurements and proposes a mixed model to solve the problem. Experimental results demonstrate that the proposed model can successfully estimate sensor locations in noisy and reverberant environments, outperforming other relaxation methods.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2023)
Article
Engineering, Multidisciplinary
Jingzhen Liu, Ka-Fai Cedric Yiu, Xun Li, Tak Kuen Siu, Kok Lay Teo
Summary: This paper examines a continuous-time securities market and discusses how to minimize the variance of a portfolio's return given a random investment horizon and a targeted terminal mean return. It finds that the variance of an investment portfolio is no longer minimized when all assets are invested in a risk-free security. Additionally, the random investment horizon has a significant impact on the efficient frontier.
JOURNAL OF INDUSTRIAL AND MANAGEMENT OPTIMIZATION
(2023)
Article
Acoustics
Jonah Ong, Ba Tuong Vo, Sven Nordholm, Ba-Ngu Vo, Diluka Moratuwage, Changbeom Shim
Summary: This paper proposes a novel solution for online separation of an unknown and time-varying number of moving sources using only a single microphone array co-located with a single visual device. The approach exploits the complementary nature of simultaneous audio and visual measurements, accomplishing separation through a model-centric 3-stage process.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2022)