The robust fuzzy control problem for a category of network-type re-entrant manufacturing systems (RMSs) is explored in this paper. A continuum model is first employed to represent the RMSs with external disturbances and time-varying communication delays in terms of a nonlinear hyperbolic partial differential equation (PDE) model. The nonlinear model is then reformulated in the framework of a T-S fuzzy model. A novel Lyapunov-Krasovskii-type stability theorem is proposed for the concerned PDEs with delays, extending the classical Lyapunov-Krasovskii framework to delayed PDE systems. Using the proposed stability theorem, LMI-based conditions are developed for stability analysis and control synthesis of the closed-loop RMSs under external disturbances and time-varying delays, which demonstrates effective alignment of production output with market demand despite environmental perturbations. At last the results of a simulation validate the developed control approach.

This article proposes a quantum spatial graph convolutional neural network (QSGCN) model that is implementable on quantum circuits, providing a novel avenue to processing non-Euclidean type data based on the state-of-the-art parameterized quantum circuit (PQC) computing platforms. Four basic blocks are constructed to formulate the whole QSGCN model, including the quantum encoding, the quantum graph convolutional layer, the quantum graph pooling layer, and the network optimization. In particular, the trainability of the QSGCN model is analyzed through discussions on the barren plateau phenomenon. Simulation results from various types of graph data are presented to demonstrate the learning, generalization, and robustness capabilities of the proposed quantum neural network (QNN) model.

Abstract Recognizing cars based on their features is a difficult task. We propose a solution that uses a convolutional neural network (CNN) and image binarization method for car make and model classification. Unlike many previous works in this area, we use a feature extraction method combined with a binarization method. In the first stage of the pre-processing part we normalize and change the size of an image. The image is then used to recognize where the rear-lamps are placed on the image. We extract the region and use the image binarization method. The binarized image is used as input to the CNN network that finds the features of a specific car model. We have tested the combinations of three different neural network architectures and eight binarization methods. The convolutional neural network with parameters of the highest quality metrics value is used to find the characteristics of the rear lamps on the binary image. The convolutional network is tested with four different gradient algorithms. We have tested the method on two data sets which differ in the way the images were taken. Each data set consists of three subsets of the same car, but is scaled to different image dimensions. Compared to related works that are based on CNN, we use rear view images in different position and light exposure. The proposed method gives better results compared to most available methods. It is also less complex, and faster to train compared to other methods. The proposed approach achieves an average accuracy of 93,9% on the first data set and 84,5% on the second set.

This paper proposes a subgraph-based hierarchical Q-learning network (SgHQN) approach to solve the optimal resource scheduling problem for complex industrial networks. In the industrial network, each connection between two individual stations has limited communication bandwidth, while each station has limited computing and storage capability and is only accessible to its local information. The resource packages flowing within the industrial network are treated as agents that have different sizes and different levels of decision-making priority. This makes the industrial resource scheduling problem on the industrial network a multi-level decision-making problem with information asymmetry. Specifically, the resource packages with lower decision-making priority have knowledge of the decisions made by those with higher priority, but not vice versa. To solve this resource scheduling problem with information asymmetry, an SgHQN model is developed by exploiting partial observations. It is found that the proposed SgHQN can be used to solve resource scheduling problems for general industrial networks. Numerical experiments simulating industrial scheduling scenarios demonstrate the effectiveness and advantages of our method.

This paper analyzes the security problem of cloud control systems (CCSs), where a malicious false data injection (FDI) attacker may deteriorate the system performance by tampering with the measurements being transmitted. The CCS defender allocates defense budgets among nodes to ensure the safe operation of CCSs. The strategic interactions between the FDI attacker and the CCS defender are modeled as a Stackelberg game, and the optimal strategies for both sides are analyzed in the sense of Nash equilibrium. Numerical examples illustrate the main results of this paper.

Heuristic approaches are often used to efficiently find solutions to complex engineering problems such as 3D chip layout design. When considering numerous constraints, multiple evaluation functions can be used. It is not straightforward how to combine these various functions to obtain a satisfying solution. This paper proposes a particle swarm optimization (PSO) algorithm as a method of finding the right blending for a given set of heuristics. Such an approach not only modifies some parameters, but redefines the entire design process by defining it as a game. The designer plays the role of the game master who proposes heuristic evaluation functions. Design components are autonomous intelligent agents moving in a virtual world. The PSO algorithm develops blending factors for given functions to find a universal design strategy for a given design problem. A case study of a 3D chip layout design is presented and discussed. The consequences and benefits of such an approach are also shown.

In this work we present a constructive genetic programming method with penalty function for hw/sw cosynthesis of embedded systems. The genotype is a tree which contains in its nodes system construction options. Unlike existing solutions in this approach individuals which violate time constrains are investigated during evolution process. Therefore the algorithm is even more able to escape local minima of optimizing parameters.

Based on the classical network model, this paper investigates the PD and PI control for synchronization of the complex dynamic directed network with delayed nodes. We obtain some sufficient conditions for global synchronization. In particular, using derivative and integral protocols are better than using traditional protocols in achieving synchronizaiton. Finally, some simulation examples are given to theoretical results.

This paper extends our previous work on synchronizing a network of Kuramoto-oscillator digraph with a pacemaker. Compared with the minimal driven or pinned nodes of two linear coupling models, those nodes forced by the pacemaker are relative conservative in our Kuramoto model with nonlinear couplings. Comparative analyses of the results and dynamics are made among these three models. For two given directed networks, discussion are given to test whether less nodes forced by the pacemaker are feasible to achieve phase agreement or frequency synchronization in a Kuramoto-oscillator network.

Securing Internet of Things is currently one of the most challenging tasks standing before the research community. Finding adequate combination of strong cryptography, lightweight computation and minimalistic data transmission with scarce energy resources and no maintenance times is very hard. This paper introduces new mechanism for IEEE 802.1X authentication scheme that has been previously extended for the IEEE 802.15.4 networks. The new solution has been designed to minimize the number of transmitted packets during the authentication process through maximization of the Extensible Authentication Protocol Maximum Transfer Unit. The results have been promising for the certificate based EAP Methods with reduction of around 10% of transferred packets for the EAP-TLS-RSA based methods and around 7% for the EAP-TLS-ECDSA based methods.

With the rapid development of sensor networks in the last decade, the cooperative design for networked observers has received an increasing attention from engineering community. This paper aims at developing a unified framework for cooperative design of networked observers to stabilize LTI plants. Apart from the traditional centralized design of MIMO system, the proposed cooperative design approach only utilizes the local information of each sensor. For undirected networks, this paper obtains a sufficient and necessary condition for the existence of the parameters that lead to the stabilization of the LTI plant. In particular, we give the detailed design procedures for the parameters of networked observers, including feedback gains and the coupling strength. The numerical simulation is also given to validate the proposed theoretical results.

We describe a method for construction of specific types of Neural Networks composed of structures directly linked to the structure of the molecule under consideration. Each molecule can be represented by a unique neural connectivity problem (graph) which can be programmed onto a Cellular Neural Network. The idea was to translate chemical structures like small organic molecules or peptides into a self learning environment which is CNN based. In the case of small molecules, each cell of the CNN stands for one atom of the molecule under consideration. But in contrast to the standard CNN architecture where each cell is connected to the neighboring cells, only those cells of the feature net are connected for which there also exists a chemical bond in the molecule under consideration. This implies that the feature net topology varies from molecule to molecule. In the case of peptides, the amino acids that form the building blocks of the peptide are reflected by the CNN cells wherein the amino acid sequence defines the network topology. Unlike the standard CNN used for image processing, there are no input values like the input image that are fed into the feature net. Instead, all information about the input molecule is supplied to the feature net by means of the topology. The output of several feature nets is fed into a supervisor neural network which computes the final output value. The combination of several feature nets and a supervisor networks constitutes the Molecular Graph Network (MGN). The designed networks are used for selection of molecules representing wanted properties such as activity against specific diseases, interactions with other compounds, toxicity etc. and possibly being candidates to be tested further as new drugs.

In order to recognize early symptoms of melanoma, the fatal cancer of the skin, systems for computer aided melanoma diagnosis have been developed for years. In this work we analyze an ensemble-based binary classifier for discriminating melanoma from dysplastic nevus utilizing wavelet-based features of the dermatoscopic skin lesion images. The multiresolution decomposition of the dermatoscopy images is done through wavelet packets. We search for the optimal wavelet base maximizing the quality of the classifier in terms of AUC (Area Under Curve) for models optimized by some common quality measures: accuracy, precision, Fl-score, FP-rate, specificity, BER and recall. Within the statistics of our experiments reverse bi-orthogonal wavelet rbio 3.1 makes the best wavelet model of melanoma.

Strategy selection is a fundamental problem for the evolutionary game process in structured populations. This paper aims at investigating the strategy selection problem by introducing a continuous evolutionary game dynamics on complex networks. It is shown that the population preference for strategies keeps unchanged under random drift in the evolutionary process. However, because of selection, the population will favor one strategy over the other based on the population structure and game payoffs. In particular, for the prisoners dilemma game, our results show that the cooperation is never favored in complete networks. However, it is greatly promoted by cycle networks. The above results are consistent with those in the traditional discrete evolutionary game dynamics. It should be especially pointed out that the proposed framework provides a potential effective tool for analyzing and controlling the evolutionary process in populations.

Shape grammars are generative systems dedicated to specific needs of designers. In the last few years, they have received increased interest especially in building reconstruction and building model generation. We propose to combine the formalism with computational intelligence methods and apply to the 3D layout problem which require efficient search of large and discontinuous spaces. The approach is illustrated by the example of a designing 3D ICs layouts. The presented results have been generated with a use of a dedicated application PerfectShape.

Over the last decade, complex networks have emerged to be a promising research field in the area of circuits and systems. This mini-review paper introduces the special session that deals with theory and applications of complex networks and provides brief review of their advances and challenges. The paper further promotes some important research topics in the field with emphasis on the multidisciplinary research interests.

This paper aims at further investigating the pinning impulsive control of complex dynamical networks. In detail, we introduce a novel approach for analyzing the synchronization stability of complex networks with impulsive signals. Moreover, we prove that one random selective impulsive controller can always pin a directed strongly connected complex network to its homogeneous solution under suitable coupling strength and impulsive signal. A simple example is then given to validate the above theoretical results.

This paper further investigates the dynamical properties of a class of basic biochemical networks modules - single substrate and single product with no inhibition (SSN) module. It is well known that the biological networks can be modelled by using different kinetics according to their reaction types. This paper introduces the SSN model under the law of mass action and proves that it can admit uniquely globally asymptotically stable positive equilibrium. Moreover, we prove that the SSN module under Hill kinetics can admit uniquely asymptotically stable positive equilibrium. It indicates some potential applications in the understanding of some biological networks modules and the design of some specific network controllers.

We describe a method for the construction of specific types of neural networks composed of structures directly linked to the structure of the molecule under consideration. Each molecule can be represented by a unique neural connectivity problem (graph) which can be programmed onto a cellular neural network. A composite network can further successfully perform classification and regression on real-world chemical data sets. The method can be regarded as a statistical learning procedure that turns the molecular graph, representing the 2D formula of the compound, into an adaptive whole molecule composite descriptor. By translating the molecular graph structure into a dynamical system, the algorithm can compute an output value that is highly sensitive to the molecular topology. This system can be trained by gradient descent techniques which rely on the efficient calculation of the gradient by backpropagation.

We describe the use of ensemble methods to build proper models for time series prediction. Our approach extends the classical ensemble methods for neural networks by using several different model architectures. We further suggest an iterated prediction procedure to select the final ensemble members.

In this paper we investigate the influence of system non-uniformity on the existence and stability of synchronous motion in an array of bi-directionally coupled electronic circuits. In computer simulations we find the level of non-uniformity for which synchronous behavior is sustained. We also present several examples of attractors, which appear when the synchronous motions is no longer stable.

Different methods for removing noise contaminating time series are presented, which all exploit the underlying (deterministic) dynamics. All approaches are embedded in a probabilistic framework for stochastic systems and signals, where the two main tasks, state and orbit estimation, are distinguished. Estimation of the true current state (without noise) is based on previously sampled elements of the time series, only, and corresponds to filtering. With orbit estimation, the entire measured time series is used to determine a less noisy orbit. In this case not only past values but also future samples are used, which, of course, improves performance.

----

We propose a new approach to so-called shadowing problem i.e. finding a system trajectory which reproduces with a given accuracy the measured time series. Our approach is based on the nonlinear observability problem known in control engineering. Solution of the shadowing problem opens new possibilities for nonlinear signal processing e.g., section-wise approximation of the measured signal by pieces of trajectories from a chosen nonlinear dynamical system (model); signal restoration when the measured signal has been corrupted e.g., by quantization; signal coding and compression.

This paper presents a new technique for signal coding and compression based on chaos control techniques. Given signal (e.g. speech or any other) recorded in a time interval is cut into smaller samples, further each of the sections is stored as a piece of a chaotic trajectory supplied by a chosen chaos generator (coder!). Effective techniques (based on a chaos control procedure) for finding initial conditions of the chaos generator to produce a desired shape of the output signal with a desired accuracy are described. In effect the signal to be coded/compressed can be stored as a set of points-initial conditions of the chosen chaos generator. Recovery of the original signal is straightforward: the stored initial conditions are applied to the coding chaos generator and the solutions taken in appropriate intervals are simply pieced together. Advantages and drawbacks of the proposed technique are discussed.

The existence of a countable infinity of unstable periodic orbits is one of the main features of chaotic behavior and strange attractors. We exploit this property from the signal processing point of view. Processing of a chaotic time series measured from some unknown process involves the reconstruction of system dynamics, uncovering the unstable periodic orbits embedded in the attractor and the calculation of eigenvalues. Such a characterisation obtained via numerical data processing is further used for the suppression of chaos, i.e. control in terms of stabilisation of chosen orbit.>

We consider a class of autonomous, continuous time, chaotic dynamical systems, the state equations of which can be represented in so-called Lur'e form. It is shown that the dynamic behaviour of such a system can be influenced in such a way that its state will reproduce any chosen periodic orbit being a solution. This kind of control can be achieved via injection of a single continuous time signal representing the output of the system associated with an unstable periodic orbit embedded in the chaotic attractor. The method is applicable to any chaotic system of the Lur'e type with single nonlinearity.>