Search

In this paper, the non-fragile dissipative filtering problem is investigated for the two-dimensional (2-D) systems subjected to randomly occurring uncertainties and redundant channel protocol. For the redundant channel transmission, if a signal fails to be transmitted through a certain channel, the next channel is immediately activated to transmit the signal once again. The norm-bounded uncertainties are introduced in the considered system, which are governed by two stochastic variables obeying the Bernoulli distribution law. The aim of this paper is to design a dissipative filter in a non-fragile manner such that the augmented filtering error system is not only asymptotically stable in the mean-square sense but also satisfies a strict 2-D (Q , S , R) − α -dissipativity performance index. Based on the Lyapunov theory and stochastic analysis, sufficient conditions are first given to guarantee the asymptotic stability of the 2-D system in the mean-square sense. Then, the non-fragile filter is designed to ensure that the 2-D system satisfies the strict 2-D (Q , S , R) − α -dissipativity performance index, under which the expressions of the non-fragile filter gains are given by solving certain matrix inequalities. Finally, a simulation example is provided to show the effectiveness of the proposed dissipative filtering method. [ABSTRACT FROM AUTHOR]

In this paper, the robust H ∞ state estimator is designed for the two-dimensional (2-D) systems in Fornasini-Machesini second model under logarithmic quantization and Round-Robin protocol. The norm-bounded uncertainties are also involved in the considered system, which are governed by two stochastic variables satisfying the Bernoulli distribution. As a kind of static scheduling, the Round-Robin protocol assigns the communication access rights to the sensors in the chronological order. The aim of this paper is to design a state estimator so that the addressed estimation error system satisfies a predefined H ∞ performance index. Sufficient conditions are given to ensure the 2-D augmented error system to be robustly stable via intensive stochastic analysis, then the results are further extended to achieve the robust H ∞ stability for the augmented error system. Explicit expressions of the estimator gains are also given by solving certain matrix inequalities. Finally, a numerical example is presented to demonstrate the effectiveness of the proposed estimation scheme. [ABSTRACT FROM AUTHOR]

Highlights • This paper studies the consensus problem of discrete-time heterogeneous multi-agent systems with random link failures and switching topologies. • The properties of the infinite products of time-varying random row-stochastic matrices and graph theory are explored to analyze this problem. • A sufficient condition that only depends on the topology structures is obtained. Abstract In this paper, we consider the issue of consensus among the agents with heterogeneous dynamics under random link failures that obey Bernoulli distribution. The heterogeneous multi-agent systems (MASs) are composed of hybrid dynamic agents including first-order dynamic agents and second-order dynamic agents, and the communication topologies are assumed to be directed and time-varying. The failure control protocol is designed by randomly utilizing the previous one- or two-step states information sent by the neighbors. A criterion for the heterogeneous consensus is established by constructing augmented systems with nonnegative random coefficient matrices for the switching topologies. With the help of graph theory and stochastic matrix theory, a sufficient condition related to switching topologies is established to ensure the realization of heterogeneous consensus. Numerical examples are finally provided to validate the theoretical result. [ABSTRACT FROM AUTHOR]

This paper is concerned with the H ∞ state estimation issue for Takagi-Sugeno (T-S) fuzzy reaction-diffusion delayed neural networks (RDNNs) with randomly occurring gain uncertainties and semi-Markov jump parameters (sMJP). The considered gain perturbations are assumed to occur in a random manner and are modeled by a random variable with the Bernoulli distribution. Furthermore, different from the existing T-S fuzzy neural networks (NNs), as the first attempt, the reaction-diffusion phenomenon, the T-S fuzzy rules, and the sMJP are taken into account in the unified framework, which makes the proposed models more applicable. By utilizing the Lyapunov functional method and introducing a suitable free weight matrix, sufficient critered to guarantee the exponential stability and H ∞ performance of estimation error. In order to improve the tolerance of the proposed estimator to gain variations, a fuzzy resilient estimator design scheme is presented with the aid of some decoupling techniques. Finally, two numerical simulations verify the effectiveness and superiority of the proposed scheme. [ABSTRACT FROM AUTHOR]

In this paper, a new design problem of the H ∞ state estimator is developed for a kind of artificial neural networks (ANNs), where both infinite distributed delays and redundant channels are happening. These adopted redundant channels can effectively improve the reliability of networked systems from the viewpoint of engineering. Two series of stochastic variables satisfying Bernoulli distribution, are introduced to govern the infinite distributed delays and schedule the redundant channels. By utilizing both stochastic analysis and Lyapunov functional approach, we obtain a lot of sufficient conditions to ensure the desired H ∞ performance, while the mean-square stability is also satisfied for this investigated state estimation issues of ANNs. The needed estimator gains are designed making use of the matrix inequalities’ solution. In final, a simulation is showed to demonstrate the effectiveness and usefulness of the developed state estimator in this paper. [ABSTRACT FROM AUTHOR]

In this paper, the non-fragile extended dissipative synchronization issue is studied for Markov jump neural networks with inertial term. With regard to the inertial neural networks described by a second-order differential equation, an appropriate variable substitution is applied to transform the original system into a first-order differential system. Considering that the data transmissions are proceed in the shared band-limited digital communication networks when designing a controller to achieve synchronization. In order to alleviate the burden of networked communication, a controller based on an event-triggered strategy is used to achieve this goal. Furthermore, a non-fragile controller design scheme governed by Bernoulli distribution is proposed to deal with the uncertainty that may occur in the implementation process of the designed controller. Then, on the basis of the Lyapunov stability theory and an improved inequality technique, sufficient conditions are obtained to ensure that the error system meets the extended dissipative synchronization. Finally, two examples are provided to verify the effectiveness of the proposed results. [ABSTRACT FROM AUTHOR]

In this paper, the non-fragile extended dissipative synchronization issue is studied for Markov jump neural networks with inertial term. With regard to the inertial neural networks described by a second-order differential equation, an appropriate variable substitution is applied to transform the original system into a first-order differential system. Considering that the data transmissions are proceed in the shared band-limited digital communication networks when designing a controller to achieve synchronization. In order to alleviate the burden of networked communication, a controller based on an event-triggered strategy is used to achieve this goal. Furthermore, a non-fragile controller design scheme governed by Bernoulli distribution is proposed to deal with the uncertainty that may occur in the implementation process of the designed controller. Then, on the basis of the Lyapunov stability theory and an improved inequality technique, sufficient conditions are obtained to ensure that the error system meets the extended dissipative synchronization. Finally, two examples are provided to verify the effectiveness of the proposed results. [ABSTRACT FROM AUTHOR]

This paper deals with the problem of state estimation for Markov jump neural networks with random delays through redundant channels. First, a redundant channel is provided to increase the probability of successfully transmitting the measurements over the shared communication network, and two mutually independent Bernoulli sequences are used to reflect the data dropouts phenomena in the main channel and the redundant channel. In addition, the time-varying random delay is both mode-dependent and distribution-dependent, and its probabilistic characteristic obeys Bernoulli distribution. Then, by constructing a suitable Lyapunov-Krasovskii functional, the sufficient condition is obtained to guarantee the augmented estimation error system is mean-square stable with a prescribed H ∞ disturbance attenuation performance. Meanwhile, a mode-dependent estimator is designed by convex optimization method. Finally, the effectiveness of the proposed method is demonstrated by a simulation example. [ABSTRACT FROM AUTHOR]

This paper is concerned with the problem of dissipative filtering for discrete-time periodic Markovian jump systems with bounded nonlinearity. It is assumed that the data measurements from the plant to the filter are subject to randomly occurred packet dropouts satisfying Bernoulli distribution. The purpose of this paper is to design a filter such that the filtering error system is stochastically stable and strictly ( Q , S , R ) -dissipative. To eliminate the cross coupling between the Lyapunov matrix and system matrices, some slack matrix variables are introduced. Based on a mode-dependent and basis-dependent Lyapunov function, a sufficient condition of the desired dissipative filter in terms of linear matrix inequalities (LMIs) for discrete-time Periodic Markovian Jump Systems with bounded nonlinearity is derived. Finally, a numerical example is exploited to demonstrate the effectiveness of the proposed filter design method. [ABSTRACT FROM AUTHOR]

Residual networks usually require more layers to achieve remarkable performance in complex visual categorization tasks, such as pose estimation. However, the increasing number of layers leads to a heavy burden on training and forward inference as well as over-fitting. This paper proposed local binary residual block (LBB) to promote the very deep residual networks on the trainable parameters, FLOPs and accuracy. In each LBB, the 3 × 3 filters are binarized based on Bernoulli distribution under a sparse constraint, an activation function is prepared to trigger the non-linear response, and the linear 1 × 1 filters are learned in a real-valued way. After stochastic binarized initialization, the 3 × 3 filters in LBB need not be updated during training. The above architecture reduces at least 69.2% trainable parameters and 70.5% FLOPs compared to the original model. The LBB is derived from three observations: 1) Activated responses of one standard k × k convolutional layer can be approximated by combining binarized k × k filters with 1 × 1 filters; 2) Most computation in the very deep residual networks is spent on the 3 × 3 convolutions; and 3) 1 × 1 filters play an important role in cross-channel information integration. In addition, the LBB module is suitable for the very deep network framework, including stacked hourglass network and pyramid residual modules. Experiments are conducted on MPII and LSP dataset for pose estimation task; CIFAR-10, CIFAR-100 and ImageNet datasets for object recognition; ECSSD, HKU-IS, PASCAL-S, DUT-OMRON, DUTS for saliency detection. The results show that our model can accelerate the training and inference of the network with only a slight performance degradation. [ABSTRACT FROM AUTHOR]

This paper aims to iron out the dynamic event-triggered (ET) H ∞ state estimation issue for a class of discrete time-delay complex networks (CNs) with randomly occurring nonlinearities (RONs). In the signal transmission among the nodes, the effect of time-varying delays is examined. The RONs under consideration is modelled by a series of random variables obeying Bernoulli distribution. In the design of state estimators, a dynamic ET scheme is utilized with the hope of improving the energy utilization efficiency. A sufficient condition is first derived to ensure the exponential mean-square (EMS) stability and H ∞ performance index of the estimation error systems. Then, by using the matrix inequality technology, the desired state estimators are designed. Lastly, a numerical simulation example is given to show the usefulness of the proposed estimator design algorithm. [ABSTRACT FROM AUTHOR]

This paper considers the resilient tracking control problem for nonlinear unknown cyber-physical systems (CPSs) subject to jamming attacks in the wireless transmission channel. First, a novel model-free adaptive control (MFAC) framework against jamming attacks is established. A Bernoulli distribution process is used to describe the happen behavior of jamming attacks. Second, a n -steps predictive compensation algorithm is designed in the controller side to reduce the effect of jamming attacks. Then, the model-free adaptive controller is designed such that the tracking error of the nonlinear system is stochastic stable, and all the analysis are based on input/output data. Simulation results demonstrate the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

This paper investigates the synchronization control issue for a set of switched inertial neural networks in the discrete-time domain, in which the persistent dwell-time switching law is employed to depict the switchings among system parameters. Thereinto, the foregoing networks having the second-order differential equations are degraded to the first-order differential ones through applying the variable transformation. In addition, in order to cope with random gain fluctuations caused by noise or harsh environments, in the controller, two random variables obeying the Bernoulli distribution are employed to simulate the occurrence of gain fluctuations. Based on Lyapunov stability theory, persistent dwell-time concept and stochastic analysis theory, some sufficient criteria are derived under which the synchronization error system is exponentially mean-square stable with a prescribed l 2 - l ∞ property. Finally, a numerical example, including some illustrative simulations, is given to present the feasibility of the derived analytical results. [ABSTRACT FROM AUTHOR]

The paper is concerned with the finite-time sliding mode control (SMC) problem for a class of discrete networked singular Markovian jump systems (SMJSs) subject to packet losses (PLs) via the delay-fractioning approach. A random variable obeying the Bernoulli distribution is used to depict the phenomenon of PLs, which is frequently encountered in the practical engineering especially during the networked transmission. A key issue to be discussed is how to guarantee both the singular stochastic finite-time boundedness (SSFTB) of the resultant sliding mode dynamic systems and the reachability of the pre-selected sliding surface. Accordingly, a sufficient condition is obtained to ensure the SSFTB of sliding motion based on delay-fractioning method and linear matrix inequality technique. Furthermore, a new SMC mechanism is proposed to compensate the influence induced by LPs and ensure the reachability condition of pre-selected sliding surface simultaneously. Finally, the effectiveness of the proposed SMC approach is tested by a simulation example. [ABSTRACT FROM AUTHOR]

In this paper, the particle filtering problem is investigated for a class of networked nonlinear systems with random one-step sensor delay and missing measurements. The phenomena of missing measurements and random one-step sensor delay are modeled by two random variables, both obeying the Bernoulli distribution. Here, we derive an explicit expression for the likelihood function when the possible occurrence of one-step sensor delay and measurement loss is taken into consideration. Based on this likelihood function, we propose a novel particle filtering algorithm to treat the nonlinear estimation problem in the simultaneous presence of random sensor delay and measurement loss. Finally, a simulation example is given to illustrate the feasibility and advantages of the proposed filtering scheme compared with traditional particle filtering algorithm. [ABSTRACT FROM AUTHOR]

This paper investigates the design and analysis problem of H ∞ filter for a class of nonlinear systems based on T-S fuzzy models with both multiplicative noises and redundant channels, which are governed by a set of Bernoulli distributed white sequences. The packet dropout’s probability of the i -th channel depends on random Bernoulli variables. The aim of this study is to design and analyze an H ∞ filter that can stabilize the T-S fuzzy filtering error dynamics. By utilizing both Lyapunov functional approach and stochastic analysis technique, we establish some sufficient conditions such that the addressed system is asymptotically stable in the mean square with a given H ∞ performance. The needed filtering parameters are obtained by making use of the matrix inequalities’ solution. In the end, an example is given to demonstrate the effectiveness and usefulness of the proposed filtering approach. [ABSTRACT FROM AUTHOR]

This paper is concerned with the state estimation problem for a class of on–off nonlinear stochastic coupling networks with time-delay. The on–off stochastic coupling scheme is governed by a set of Bernoulli random variables with individual switching probability. By introducing the stochastic coupling variable into the structure of the extended Kalman filter, the estimator is developed for each node to guarantee an optimized upper bound on the state estimation error covariance despite the stochastic coupling uncertainties and linearization errors, where the gain matrices are derived by the solutions to two Riccati-like difference equations. A numerical example involving tracking four interacting robots is used to verify the effectiveness of the proposed estimator. [ABSTRACT FROM AUTHOR]

This paper deals with the problem of H ∞ filter design for a class of nonlinear networked systems based on T–S fuzzy model with multiple stochastic time-varying delays, and sensor faults and packet dropouts are considered simultaneously. A sequence of stochastic and independent variables, which obey the Bernoulli distribution, are introduced to depict stochastic time-varying delays. The possible of sensor failure can be described by unrelated random variables taking values on an interval, and the packet dropouts are described as a set of Bernoulli distributed white noises. The approach of piecewise quadratic Lyapunov function is applied to reduce the conservatism. The filter parameters are obtained by solving a set of linear matrix inequalities. Finally, a simulation example is provided to illustrate the effectiveness of the proposed filter design approach. [ABSTRACT FROM AUTHOR]

This paper investigates the quantized H ∞ filtering problem for networked control systems (NCSs) with multiple influencing factors such as randomly occurring nonlinearities and sensor saturation. The dynamic quantization strategy is employed, and stochastic variable is considered satisfying the Bernoulli distribution. The quantized H ∞ filter is designed to ensure the exponentially mean-square stable of the system and to guarantee a prescribed H ∞ disturbance attenuation level. Finally, a simulation example is given to demonstrate the effectiveness of the filter design method. [ABSTRACT FROM AUTHOR]

In this paper, the problem of fault–tolerant H ∞ control for a class of networked control systems with randomly occurring missing measurements is investigated. More precisely, a stochastic variable satisfying the Bernoulli random distribution is introduced to describe the missing measurements. Moreover, network-induced time-varying delay is considered in the network model. By utilizing stochastic analysis, delay-dependent criteria are developed to ensure the stability of the faulty networked control system with the prescribed H ∞ performance. Based on the derived results, the fault tolerant controller is further designed to compensate for the system failure effects. Finally, a numerical example is provided to illustrate the effectiveness of our proposed design methods. [ABSTRACT FROM AUTHOR]

This paper investigates the non-fragile synchronization problem of dynamical networks with randomly occurring nonlinearities and controller gain fluctuations. More precisely, these randomly occurring phenomena are modeled by stochastic variables satisfying the Bernoulli distribution. By utilizing the Lyapunov–Krasovskii functional method, sufficient synchronization criteria are first established in terms of linear matrix inequalities (LMIs). Based on the obtained results, a set of non-fragile synchronization controllers is further designed to ensure that the synchronization of the dynamical networks can be achieved. Finally, a numerical example is provided to illustrate the applicability and effectiveness of our theoretical results. [ABSTRACT FROM AUTHOR]

This paper focuses on the state estimation problem for a class of discrete-time switching neural networks with persistent dwell time (PDT) switching regularities and mode-dependent time-varying delays in H ∞ sense. The considered switching regularity is more general that extends the frequently studied dwell-time (DT) and average dwell-time (ADT) switching. The random packet dropouts, which are governed by a Bernoulli distributed white sequence, are considered to exist together for the estimator design of underlying switching neural networks. The desired mode-dependent estimators are designed such that the resulting estimation error system is exponentially mean-square stable and achieves a prescribed H ∞ level of disturbance attenuation. Finally, the effectiveness and the superiority of the developed results are demonstrated through a class of synthetic oscillatory networks. [ABSTRACT FROM AUTHOR]

This paper considers linear quadratic (LQ) optimal control problems based on state observer for a class of discrete-time system over lossy data network. Both sensor data and control commands are transmitted by the same network. And the packets may be randomly lost according to the Bernoulli distribution. In the infinite horizon expected total cost we derive expressions for computing the corresponding controller gain. And the hold-input compensation strategy is applied to the closed system in which the previous control signal is used when packet dropped. Moreover, illustrative example is presented to demonstrate the effectiveness of the proposed results. [ABSTRACT FROM AUTHOR]

Abstract: This paper investigates the problem of dissipativity-based state estimator design for Markov jump discrete-time neural networks, where the communication links between the neural network and estimator are assumed to be imperfect. The phenomenon of missing data is modeled by a stochastic variable following the Bernoulli random distribution. The focus is on the design of a Markov switching estimator such that the resulting closed-loop system is dissipative. Some sufficient conditions for the existence of admissible estimator are obtained in terms of linear matrix inequalities. Finally, a numerical example is employed to demonstrate the effectiveness of our proposed approach. [Copyright &y& Elsevier]

Abstract: Variational Bayes learning or mean field approximation is widely used in statistical models which are made of mixtures of exponential distributions, for example, normal mixtures, binomial mixtures, and hidden Markov models. To derive variational Bayes learning algorithm, we need to determine the hyperparameters in the a priori distribution; however, the design method of hyperparameters has not yet been established. In the present paper, we propose two different design methods of hyperparameters which are applied to the different purposes. In the former method, the hyperparameter is determined for minimization of the generalization error. In the latter method, it is chosen so that candidates of hidden structure in training data are extracted. It is experimentally shown that the optimal hyperparameters for two purposes are different from each other. [Copyright &y& Elsevier]