相关参考文献
注意：仅列出部分参考文献，下载原文获取全部文献信息。
Article
Computer Science, Artificial Intelligence
Chaouki Aouiti et al.
Summary: This article deals with the global dissipativity of Quaternion-Valued Fuzzy Cellular Fractional-Order Neural Networks (QVFCFONNs). The model is solved by separating it into four real-valued parts and using Lyapunov functionals and Linear Matrix Inequalities (LMIs) approach. New sufficient conditions are derived to ensure the global dissipativity for the considered network model.
NEURAL PROCESSING LETTERS
(2023)
Article
Computer Science, Interdisciplinary Applications
Xinna Li et al.
Summary: This paper discusses the issue of prescribed-time synchronization in networks of piecewise smooth systems. It presents two new lemmas for nonlinear piecewise smooth systems, which allow for arbitrary preset stable time and are independent of control parameters and initial values. The paper also introduces a feedback controller and a novel event-triggered controller to achieve prescribed-time synchronization. The conditions for prescribed-time synchronization are obtained using set-valued analysis technique and generalized Lyapunov functional approach, in terms of linear matrix inequalities (LMIs). Two application examples are provided to validate the obtained results and the effectiveness of the control schemes.
MATHEMATICS AND COMPUTERS IN SIMULATION
(2023)
Article
Computer Science, Artificial Intelligence
Qiankun Song et al.
Summary: This paper investigates the stability issue of quaternion-valued neural networks with neutral delay, proportional delay, and leakage delay. By utilizing the principle of homeomorphism, techniques of matrix inequality, and Lyapunov stability theory, a main stability criterion is derived in the form of a quaternion-valued linear matrix inequality to ensure the unique existence and global stability of the equilibrium point for the considered networks. An illustrative example and simulations are provided to demonstrate the effectiveness of the theoretical result.
Article
Physics, Multidisciplinary
Hao Pu et al.
Summary: This article studies the fixed-time and predefined-time synchronization of fully complex-valued discontinuous delayed neural networks using the non-separation approach. Two novel types of complex-valued controllers are designed to achieve specific control objectives. Theoretical criteria are established based on the fixed/predefined-time stability theory to achieve synchronization, and the settling time for fixed-time synchronization is explicitly calculated. The effectiveness of the theoretical results is supported by numerical simulation results and two numerical examples are provided.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2023)
Article
Mathematics
Wanlu Wei et al.
Summary: This article investigates fixed-time synchronization and preassigned-time synchronization for quaternion-valued neural networks with time-varying delays and discontinuous activation functions. A direct analytical method is proposed to design quaternion-valued controllers for fixed-time synchronization, based on the quaternion-valued signum function and related properties. Moreover, preassigned-time synchronization is explored using quaternion-valued control design with preassigned synchronization time and finite control gains. Compared to existing methods, the direct method without separation presented in this article simplifies theoretical analysis, and the proposed quaternion-valued control schemes are simpler and more effective than traditional designs. Two numerical examples are provided to support the theoretical results.
ACTA MATHEMATICA SCIENTIA
(2023)
Article
Mathematics, Interdisciplinary Applications
Zhengqiu Zhang et al.
Summary: This paper focuses on the existence-uniqueness (EU) and globally asymptotic stability (GAS) of equilibrium solutions (ESS) for a class of quaternion-valued fuzzy BAM neural networks (QVFBAMNNS) with delays. A sufficient condition ensuring the EU of ESS is achieved by utilizing the Homeomorphism theorem (HT) and the properties of unary high degree inequality. The GAS of ESS for the above networks is assured by applying the integral inequality approach (IIA) without utilizing V'(t) <= 0. Studying the EU of ESS using the properties of high degree inequality and studying the GAS for neural networks (NNS) using IIA are new research approaches.
CHAOS SOLITONS & FRACTALS
(2023)
Article
Computer Science, Artificial Intelligence
Jingjing Wang et al.
Article
Computer Science, Artificial Intelligence
Jinrong Yang et al.
Summary: This paper investigates the synchronization problem of bidirectional associative memory memristive neural networks (BAMMNNs) with time-varying delays and proposes a new solution method. By designing special control strategies, fixed-time synchronization (FXTS) conditions for the drive-response systems are obtained based on Lyapunov function and inequality techniques. The settling time (ST) is estimated using the improved fixed-time stability lemma. Numerical simulation is conducted to verify the conclusions.
Article
Computer Science, Artificial Intelligence
Tao Peng et al.
Summary: A control strategy based on Lyapunov functions is proposed for the fixed-time synchronization of a time-delay quaternion-valued neural network (QVNN). This strategy is further extended to the prescribed-time synchronization of the QVNN. Another improved two-step switching control strategy is also proposed based on this flexible control strategy. Numerical examples demonstrate the advantages of the proposed methods.
Article
Mathematics
Hayrengul Sadik et al.
Summary: In this paper, the fixed-time synchronization problem of a type of reaction-diffusion fuzzy neural networks with stochastic perturbations is investigated using simple control schemes. Generalized fixed-time stability results for stochastic nonlinear systems are introduced, and generic fixed-time stability criteria are established. The upper bounds of settling time are directly calculated using special functions. A type of controller is designed for the fixed-time synchronization, which is simpler and more applicable. A numerical example with Matlab simulations is provided to validate the theoretical results.
Article
Engineering, Electrical & Electronic
Sebastian Miron et al.
Summary: The main goal of this article is to dispel the misconception of quaternion as an exotic and impractical signal representation in signal and image processing, and provide an overview of its key aspects, models, and methods through literature examples. The article concludes by discussing the remaining challenges and open problems in quaternion signal and image processing.
IEEE SIGNAL PROCESSING MAGAZINE
(2023)
Article
Computer Science, Artificial Intelligence
Sunny Singh et al.
Summary: This article investigates the anti-synchronization problem for quaternion-valued inertial neural networks (QVINNs) with unbounded time delays. Various conditions are established to guarantee the anti-synchronization of QVINNs using different control strategies and Lyapunov theory. The authors utilize a non-reduction order method and non-separation approach for QVINNs, making the analysis more concise and easier to deal with. The approach adopted in this study is more realistic and has not been considered by any researcher before for QVINNs with unbounded time delays.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Interdisciplinary Applications
Hai Zhang et al.
Summary: This article discusses the quasi-synchronization (QS) issues for Caputo delayed Cohen-Grossberg neural networks (CGNNs). A novel lemma is established using suitable fractional differential inequality. Two different adaptive controllers are designed, taking advantage of adaptive control schemes with reducing control cost and high tracking accuracy. The conditions of QS are obtained by applying the proposed lemma, inequality techniques, and Lagrange's mean value theorem with appropriate Lyapunov functions. Two numerical examples in different dimensions are shown to test the correctness of the gained theorems.
MATHEMATICS AND COMPUTERS IN SIMULATION
(2023)
Article
Computer Science, Artificial Intelligence
Yanlin Zhang et al.
Summary: This paper investigates the fixed-time synchronization (FXTSYN) of unilateral coefficients quaternion-valued memristor-based neural networks (UCQVMNNs) with mixed delays. A direct analytical approach is suggested to obtain FXTSYN of UCQVMNNs utilizing one-norm smoothness in place of decomposition. The set-valued map and the differential inclusion theorem are used to deal with drive-response system discontinuity issues. Innovative nonlinear controllers and the Lyapunov functions are designed to accomplish the control objective. Criteria of FXTSYN for UCQVMNNs are given using inequality techniques and the novel FXTSYN theory, and the accurate settling time is obtained explicitly. Numerical simulations are presented to validate the obtained theoretical results as accurate, useful, and applicable.
Article
Computer Science, Artificial Intelligence
Fanchao Kong et al.
Summary: This paper aims to study the fixed-time stabilization of a class of delayed discontinuous reaction-diffusion Cohen-Grossberg neural networks. A new fixed-time stability lemma is given by providing some relaxed conditions and based on inequality techniques. The periodic solutions of the formulated networks are stabilized at zero in a fixed-time based on the new fixed-time stability lemma, which is a new topic on reaction-diffusion networks.
Article
Computer Science, Artificial Intelligence
Md Arzoo Jamal et al.
Summary: This article focuses on the problem of fixed-time stability of a Cohen-Grossberg bidirectional associative memory neural network with destabilizing impulsive effects. We obtain a novel sufficient condition for the impulsive dynamical systems to be fixed-time stable for destabilizing impulses. Unlike the usual Lyapunov inequality, we apply a new Lyapunov inequality to obtain results under impulsive perturbations. Based on the average impulsive interval and the comparison principle, we derive the results of this paper. Two types of continuous controllers are studied, and the settling-time functions obtained in this article depend on the parameters of the impulsive sequences. Finally, numerical examples validate the efficiency of our obtained theoretical results.
Article
Mathematics
Amir Rezaie et al.
Summary: This paper addresses the fixed-time stability problem of chaotic systems with internal uncertainties and external disturbances. New sliding-mode surfaces are introduced to design fixed-time controllers that enable the convergence of the trajectories of the chaotic system to the origin in finite time, independently of the initial conditions. The proposed approach is assessed using a simulation study of a PMSM system and the Matouk system, and it shows robustness to external disturbances and boundedness of the settling time to a constant value for any initial condition.
Article
Mathematics
Enli Wu et al.
Summary: This paper studies the fixed-time synchronization problem of complex-valued coupled networks with hybrid perturbations. A novel quantized controller is designed to overcome the difficulties caused by complex parameters, variables, and disturbances. By using the Lyapunov stability theorem and properties of the Wiener process, sufficient conditions are provided for selecting control parameters to achieve fixed-time synchronization, and an upper bound for the setting time is obtained. Numerical simulations confirm the theoretical results and effectiveness of the control strategy.
Article
Acoustics
Eric Guizzo et al.
Summary: This paper presents a novel semi-supervised architecture called RH-emo, which extracts quaternion embeddings from real-valued monoaural spectrograms for speech emotion recognition tasks. Experimental results on four popular datasets demonstrate that the embeddings created with RH-emo significantly improve test accuracy and greatly reduce the resource demand of models compared to traditional real-valued CNN architectures.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2023)
Article
Engineering, Electrical & Electronic
Huisheng Zhang et al.
Summary: Widely linear modeling is an important technique for capturing second-order statistics of quaternion signals. However, existing algorithms based on this modeling approach are computationally expensive. In this letter, a fast estimation technique is proposed using a real augmented representation, resulting in significant reduction in computations. The proposed algorithm is mathematically equivalent to the original one and can be applied to other widely linear quaternion algorithms. Theoretical analysis and simulation results validate the effectiveness of the proposed technique.
IEEE SIGNAL PROCESSING LETTERS
(2023)
Article
Computer Science, Artificial Intelligence
Jingxin Liu et al.
Summary: Based on the framework of differential inclusion, a quaternion-valued neural network is designed in this paper to solve nonsmooth nonconvex optimization problems in quaternion domain. The network modeling avoids penalty factor calculation and guarantees convergence through nonsmooth analysis. The research results demonstrate the feasibility and effectiveness of this method in solving nonconvex optimization problems over quaternion domain.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)
Article
Computer Science, Theory & Methods
Ruoxia Li et al.
Summary: This article presents a dissipativity and synchronization control method for a class of delayed quaternion-valued memristor-based recurrent neural networks. By utilizing Lyapunov functionals, analytical techniques, and novel control algorithms, sufficient conditions are established to achieve dissipativity and synchronization of fuzzy quaternion-valued memristive systems. Additionally, the lexicographical order method is introduced to determine the magnitude of different vectors, and simulation results are provided to demonstrate the effectiveness of the proposed method.
FUZZY SETS AND SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Wanlu Wei et al.
Summary: This article focuses on fixed-time synchronization and preassigned-time synchronization of quaternion-valued neural networks, with the development of pure power-law control schemes to achieve synchronization. The designed controllers in this paper are simpler and more effective, simplifying the traditional design and providing finite control gains. An example is used to illustrate the feasibility and validity of the obtained results.
Article
Mathematics
Lu Pang et al.
Summary: This paper primarily focuses on the fixed-time synchronization of fuzzy-based impulsive complex networks. By developing fixed-time stability and proposing a T-S fuzzy control strategy, simplified criteria are obtained for achieving fixed-time synchronization of fuzzy-based impulsive complex networks, along with an estimation of the synchronized time.
Article
Automation & Control Systems
Fei Tan et al.
Summary: This article examines the secure exponential synchronization problem of multiplex Cohen-Grossberg neural networks under stochastic deception attacks. It proposes an attack resistant controller and an exponential adaptive quantitative controlling algorithm to achieve synchronization with minimal error under malicious attacks.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2022)
Article
Computer Science, Artificial Intelligence
Yali Cheng et al.
Summary: This paper discusses the problem of achieving fixed-time synchronization in fractional-order complex-valued neural networks with time-varying delay. By utilizing the properties of fractional calculus and fractional-order comparison principle, an improved lemma is proposed to derive the fixed-time synchronization conditions. Based on sliding model control and Lyapunov stability theorem, an effective sliding mode surface is constructed, and a suitable sliding model control is designed to make the synchronization error converge to zero within a fixed time.
Article
Computer Science, Artificial Intelligence
Yu Zhou et al.
Summary: The study introduces a quaternion capsule neural network (Q-CapsNet) with a region attention mechanism for facial expression recognition (FER) in color images, addressing FER problems such as internal dependencies between color channels, latent geometry information disentanglement, and emotional feature extraction from vital facial regions. Evaluation on public color FER datasets demonstrates that the proposed network outperforms comparative models and state-of-the-art FER methods in terms of performance.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2022)
Article
Computer Science, Theory & Methods
Ruoxia Li et al.
Summary: This article discusses dissipativity and synchronization control for a class of delayed quaternion-valued memristor-based recurrent neural networks. By utilizing Lyapunov functionals, analytical techniques, and novel control algorithms, sufficient conditions for achieving dissipativity and synchronization of fuzzy quaternion-valued memristive systems are established. The introduction of the lexicographical order method aids in determining the magnitude of different vectors, and simulation results demonstrate the effectiveness of the proposed method.
FUZZY SETS AND SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Shao Shao et al.
Summary: This paper presents a method for pre-specified time synchronization of switched coupled neural networks (SCNNs) using smooth controllers. In contrast to traditional finite-time synchronization methods, the synchronization time obtained in this method is independent of control gains, initial values or network topology, and can be pre-set according to task requirements.
Article
Mathematics
Wenqiang Yang et al.
Summary: This paper discusses the use of QIC to achieve FIXTS of NNs and designs a fixed-time controller to ensure synchronization in finite time. After confirming system stability, it is suitable for image protection – the encryption effect and decryption quality depend on the encryption algorithm and NNs synchronization error.
Article
Automation & Control Systems
Cheng Hu et al.
Summary: This article addresses the issue of FXT and PAT synchronization in discontinuous dynamic networks by improving FXT stability and developing simple control schemes. The improved FXT stability and simple control design effectively solve the synchronization problem in complex networks. The numerical examples confirm the effectiveness of the improved FXT stability and synchronization strategies.
IEEE TRANSACTIONS ON CYBERNETICS
(2021)
Article
Automation & Control Systems
Xiaoyang Liu et al.
IEEE TRANSACTIONS ON CYBERNETICS
(2020)
Article
Computer Science, Information Systems
Jianying Xiao et al.
INFORMATION SCIENCES
(2020)
Article
Computer Science, Artificial Intelligence
Zihan Li et al.
NEURAL PROCESSING LETTERS
(2020)
Article
Computer Science, Artificial Intelligence
Ruoyu Wei et al.
Article
Physics, Multidisciplinary
Hui Deng et al.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2019)
Article
Automation & Control Systems
Eduardo Bayro-Corrochano et al.
ROBOTICS AND AUTONOMOUS SYSTEMS
(2018)
Article
Automation & Control Systems
Andrey Polyakov
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2012)
