Search

The southern China is located in high-temperature and high-humidity areas with long and high rainfall. Affected by the sea breeze, the salt on the surface of Composite insulators in the area is higher than others. If the insulator is damaged and exposed to continuous rain, there may be a consistent arc on the surface of the insulator due to electric field distortion caused by insulator’s damage. To explore the effect of continuous arc on the properties of silicone rubber materials operating on natural environment, we chose four groups of the silicon rubber samples with different contents of ATH and silica and coat them with kaolin in order to simulate natural pollution. Insulators were placed in the tracking test apparatus, the voltage of 2.5 kV, 3.5 kV, 4.5 kV, 6 kV and the corresponding current velocity were applied to simulate the actual aging process according to the relevant passage. Transverse relaxation time and residual mass were measured respectively, also hardness, lightness and static contact angle of the samples were carried out after the test. The result shows that transverse relaxation time and residual mass of silicone rubber decreased with the increase of accelerated aging time. If the residual mass decreases to a certain value, the silicone rubber material sample cannot continue to run while discharge is concentrated in a certain area on the surface of silicone rubber. After the aging process, lightness and static contact of the sample decreased, and hardness changed little. At the same time, it can find that the higher the relative content of ATH and silica, the better the performance of silicone rubber under the action of continuous arc when the rubber content reaches a certain proportion by comparing the properties of silicone rubber with different component.

The aging phenomenon of silicone rubber will occur during the operation of composite insulator, which is related to the operating period and operating environment. In order to evaluate the aging of silicone rubber of composite insulators in Dongguan, the changes in macroscopic properties such as hardness of aging silicone rubber were studied at first. And then three kinds of microscopic analysis methods (NMR, TGA and FTIR) were used to study the changes of silicone rubber. The results showed that the transverse relaxation time, ratio of residual mass and absorption peak of $Si(-CH_{3})_{2}$ also had similar changes with operating periods, which proved the feasibility and correctness of color difference analysis. Therefore, a method for evaluating the aging of silicone rubber of composite insulators suitable for Dongguan was presented.

© 2012 IEEE. Video-based facial expression recognition has received substantial attention over the past decade, while early expression detection (EED) is still a relatively new and challenging problem. The goal of EED is to identify an expression as quickly as possible after the expression starts and before it ends. This timely ability has many potential applications, ranging from human-computer interaction to security. The max-margin early event detector (MMED) is a well-known ranking model for early event detection. It can achieve competitive EED performance but suffers from several critical limitations: 1) MMED lacks flexibility in extracting useful information for segment comparison, which leads to poor performance in exploring the ranking relation between segment pairs; 2) the training process is slow due to the large number of constraints, and the memory requirement is also usually hard to satisfy; and 3) MMED is linear in nature, and hence may not be appropriate for data in a nonlinear feature space. To overcome these limitations, we propose an online multi-instance learning (MIL) framework for EED. In particular, the MIL technique is first introduced to generalize MMED, resulting in the proposed MIL-based EED (MIED), which is more general and flexible than MMED, since various instance construction and combination strategies can be adopted. To accelerate the training process, we reformulate MIED in the online setting and develop online multi-instance learning framework for EED (OMIED). To further exploit the nonlinear structure of the data distribution, we incorporate the kernel methods in OMIED, which results in the proposed online kernel multi-instance learning for early expression detection. Experiments on two popular and one challenging video-based expression data sets demonstrate both the efficiency and effectiveness of the proposed methods.

In this paper, we propose a new method for 3D hand scan registration. This method includes two stages: liner blending skinning(LBS) and non-rigid ICP. A template hand mesh model is used to register to the 3D hand scans, which have large pose and shape variations. To fit different types of 3D hand scans, the template model is driven by the skeletons follow the principle of LBS to eliminate pose variations, and then non-rigid ICP is used to deform the template so that shape variations can be eliminated. Finally, template and scan model completely overlap together, and the vertices of template and scan can be matched one by one. We believe that the ability to align a template to each scan is the basis for more in-depth research in future. For example, this work can be used as basis for training pose and shape parameters. With the help of these parameters a template can be used to model any 3D hand in real life just like SMPL[7] for human and SMAL[14] for animals.

Solar irradiance is a renewable energy that has been utilized by power plants. Direct normal irradiance (DNI) forecasting is a crucial approach to reduce the output uncertainty of solar power plants, and enhance reliability of power grids. According to statistics, the solar radiation of clear sky accounts for more than half of the total amount of radiation in a year. Hence, clear sky DNI forecasting models play a significant role in solar power systems. The existing clear sky models assume that the atmosphere is at rest and ignore the variation of aerosol optical depth in a day, which may reduce the accuracy of DNI forecasting. In this paper, a novel tree-based piecewise clear sky model is proposed to capture aerosol extinction coefficient varying with time in a day and forecast DNI when clouds are absent. Firstly, a linear model is deduced from a physical equation to represent the relationship of air mass and DNI approximately. Then, model trees with this linear mode at each leaf node are create. The database of the National Renewable Energy Laboratory (NREL) is used for experiments to evaluate the performance of the proposed model, compared with other clear sky models. The results of the experiments show that the proposed model provides more accurate forecast of DNI than other clear sky models.

The multilayer perceptrons (MLPs) are widely used in many fields, however, singularities in the parameter space may seriously influence the learning dynamics of MLPs and cause strange learning behaviors. Given that the singularities are the subspaces of the parameter space where the Fisher information matrix (FIM) degenerates, the FIM plays a key role in the study of the singular learning dynamics of the MLPs. In this paper, we obtain the analytical form of the FIM for unipolar activation function-based MLPs where the input subjects to the Gaussian distribution with general covariance matrix and the unipolar error function is chosen as the activation function. Then three simulation experiments are taken to verify the validity of the obtained results.

Robust stabilisation problem of discrete-time switched linear systems with bounds on the controls and states is investigated in this study. Firstly, the controller synthesis for systems with interval uncertainties which guarantees positivity and stability of the closed-loop system is studied. These results are then extended to the systems with polytopic uncertainties. All the conditions are formulated in terms of linear feasibility. Finally, numerical examples illustrate the effectiveness of the proposed methods.

A class of homogeneous polynomial functions is used to estimate the attractive stability region for an impulsive switched linear system with saturated control input. Under the arbitrary switching rule and the dwell time switching rule, local stabilization conditions are obtained in terms of linear matrix inequalities. The corresponding optimization problems are formulated to obtain a larger attractive stability region. Using the simple ellipsoid method to estimate the attractive stability region produces very conservative results, because the attractive stability region is normally irregular. To solve this problem, a polyhedron constructed using a level set of the homogeneous parameter-dependent quadratic Lyapunov function is used to estimate the attractive stability region. The polyhedron is closer to the attractive stability region than the simple ellipsoid. Finally, numerical examples are given to demonstrate the effectiveness of the proposed method.

This paper studies robust stability of positive switched systems PSSs with polytopic uncertainties in both discrete-time and continuous-time contexts. By using multiple linear copositive Lyapunov functions, a sufficient condition for stability of PSSs with dwell time is addressed. Being different from time-invariant multiple linear copositive Lyapunov functions, the Lyapunov functions constructed in this paper are time-varying during the dwell time and time-invariant afterwards. Then, robust stability of PSSs with polytopic uncertainties is solved. All conditions are solvable via linear programming. Finally, illustrative examples are given to demonstrate the validity of the proposed results.

The multilayer perceptrons (MLPs) have been widely used in many communication applications, however, the learning process of the multilayer perceptrons often becomes very slow, which is due to the existence of the singularities in the parameter space. As the singularities significantly affect the learning dynamics of MLPs, the standard gradient descent method is not Fisher efficient. In order to overcome this problem, natural gradient method and adaptive natural gradient method were proposed to accelerate the learning process. As is well known, step size in each iteration plays a key role in the performance of algorithms. In this paper, the modified adaptive natural gradient method is proposed where the step size in each iteration is adaptive modified. The aim of the proposed algorithm is to accelerate the convergence speed and increase the performance of MLPs. The simulation results verify the validity of the analytical results.

Wind power prediction is very important to guarantee security and stability of the wind farm and power system operation, and wind speed forecasting is the key to wind power prediction. Due to dramatic changes and shorter collection intervals in wind speed, it generates a larger numbers of samples, which affects modeling time and accuracy. Therefore, a short-term wind speed prediction method based on active learning with sample selection is proposed. The main objective of active learning is to collect training samples by minimizing the error of the prediction process while minimizing the number of training samples used. First the initial training set is selected, then selecting the sample with higher Euclidean distance adding to the training set and delete the samples with lower cross-validation error. Finally, the establishment of support vector machine model and neural network model to predict the results of comparison. Taking a wind farm in Jiangsu Province as the actual experimental data, simulation results show that this method can ensure the prediction accuracy, while effectively reducing the number of training samples.

This paper investigates a fault diagnosis scheme to detect whether a system is in normal or fault condition, and further diagnose the type and location of the fault. A complicated storage and transportation equipment system is proposed to experiment the presented fault diagnosis strategy. The fault tree technique is used to overcome the difficult to detect system states and huge consume for arranging large numbers of information collection points. In this paper, the established fault diagnosis algorithm which based on the fault tree can achieve diagnosing the equipment intelligently. Firstly, we build fault tree for every fault code based on the system structure and past working data. Once a fault is detected, the fault tree analysis will be performed to find out possible broken-down components automatically with the help of written MATLAB software. Under the guidance of provided reasonable diagnostic sequence, the users can maintain the equipment effectively without the help of professionals.

In this paper, a constructive fault tolerant control strategy is developed for a general class of uncertain non-holonomic systems. The terminal slide mode technique and the minimal dilation degree scheme are used to accommodate the actuator partial loss of effectiveness and additive actuation fault. Two step control procedures are proposed under the finite time convergence. The parking problem for a wheeled mobile robot is presented to illustrate the effectiveness of the fault tolerant control strategy and demonstrate the application of the general system.

Data centers have been widely used in many pervasive computing applications. This paper shows the cost-aware virtual machine allocation problem for off-grid green data centers is an integer programming problem which is NP-hard. The paper presents a cost-aware virtual machine allocation algorithm which attempts to utilize renewable energy sources and minimize the energy cost of fossil fuel while maintaining the quality-of-service requirements of the tasks. The simulation results show that the proposed algorithm is sensitive to the changes in the price of fossil fuel, and is able to achieve scalable performance.

In recent years, many graph-based methods have been introduced to detect saliency. These methods represent image regions and their similarity as vertices and edges in a graph. However, since they only represent pairwise relations between vertices, they give an incomplete representation of the relationships between regions. In this work, we propose a hypergraph based optimization framework for salient object detection to include not only the pairwise but also the higher-order relations among two or more vertices. In this framework, besides the relations among vertices, both the foreground and the background queries are explicitly exploited to uniformly highlight the salient objects and suppress the background. Furthermore, a probabilistic hypergraph is constructed based on local spatial correlation, global spatial correlation, and color correlation to represent the relations among vertices from different views. Extensive experiments demonstrate the effectiveness of the proposed method.

In this study, a time optimisation problem is solved for a class of linear time-invariant switched stochastic systems with multi-switching times using the calculus of variations. The objective of authors’ study is to minimise a cost functional defined on the system state, where the sequence of active subsystems is pre-specified and the switching times are the only control variables. The gradient of the cost functional with respect to the switching times is derived, which has an especially simple form and can be directly used in gradient descent algorithms to locate the optimal switching instants. The proposed method is applied on a two-tank system, which is a practical example of switched systems. Combining with another numerical example with three switching times case, the viability of the proposed method is further demonstrated.

As is well known, the parameter spaces of hierarchical systems such as multilayer perceptrons include singularities and the plateau phenomenon is ubiquitous in the process of learning. In the singular regions, the Fisher information matrix degenerates and the loss function is almost unchanged when the parameters arrive in the singular regions, which is called the plateau phenomenon. We wonder about whether the singularities and the plateau phenomenon exist in the parameter identification process of the linear and the ordinary nonlinear systems. In this paper, we can see that in some of the parameter identification of the nonlinear systems, the Fisher information matrix degenerates, the singularities exist and we can see the plateau phenomenon in the learning curves. A simulation example is provided to demonstrate the theoretical analysis in Section 3 .

In this paper, we solve an optimal control problem for switched stochastic systems using calculus of variations, where the objective is to minimize a cost functional defined on the state and the switching times are the sole control variables. In particular, we focus on the problem in which a pre-specified sequence of active subsystems is given. For one switching time case, the derivative of the cost functional with respect to the switching time is derived, which has an especially simple form and can be directly used in gradient descent algorithms to locate the optimal switching instant. Then, we propose an approach to deal with the problem with multi-switching times case. Finally, two numerical examples are given, highlighting the viability and advantages of the proposed methodology.

Optimal switch-time control is the study that investigates how best to switch between different modes. In this paper, we investigate the optimal switch-time control problem for discrete-time linear switched stochastic systems. In particular, under the assumption that the sequence of active subsystems is pre-specified, we focus on the problem where the objective is to minimize a cost functional defined on the states and the switching times are the only control variables. For systems with one switching time, using calculus of variations, we firstly derive the difference formulae of the cost functional with respect to the switching time, which can be directly used to find the optimal switching instant. Then, a method is presented to deal with the problem with multiple switching times case. Finally, the viability of the proposed method is illustrated through two numerical examples.

The multilayer perceptrons (MLPs) have strange behaviors in the learning process caused by the existing singularities in the parameter space. A detailed theoretical or numerical analysis of the MLPs is difficult due to the non-integrability of the traditional log-sigmoid activation function which leads to difficulties in obtaining the averaged learning equations (ALEs). In this paper, the error function is suggested as the activation function of the MLPs. By solving the explicit expressions of two important expectations, we obtain the averaged learning equations which make it possible for further analysis of the learning dynamics in MLPs. The simulation results also indicate that the ALEs play a significant role in investigating the singular behaviors of MLPs.

With the increasing grade of high voltage transmission lines, the frequent occurrence of corona discharge has brought great potential safety risk and power losses to power system. Therefore, the detection of corona discharge in transmission lines is of great significance to power system. Rectangular spiral antenna is widely used because of its smooth impedance and better circular polarization characteristics. In this paper, a planar rectangular spiral antenna is designed for the detection of radio frequency electromagnetic waves given out corona discharge of transmission lines. The basic structure of the antenna is determined by optimizing the length, spacing and turns of coil. The feasibility of the antenna is verified by experiments.

The accurate estimation of direct normal irradiance (DNI) under clear sky conditions plays an important role in the concentrated solar thermal plant. A hybrid model with adjustable inputs is proposed to calculate the clear-sky DNI, including a base clear-sky model and an error-correction model. The base clear-sky model is able to estimate the clear-sky DNI at any place with only the local date and location information, and the error-correction model serves as a supplementary to improve the calculating accuracy with available meteorological data. The error-correction model effectively integrates a linear part and a nonlinear part, and its inputs are adjustable according to the available meteorological observations. Several experiments have been conducted to evaluate the performance of the proposed model with data from three observation stations provided by the National Renewable Energy Laboratory open database. The results show that the hybrid model is able to provide great improvement over the base clear-sky model with 28%–70% on normalized root mean square error, and it also performs better than those using a linear or nonlinear error correction model. It is concluded that the performance of the hybrid model is comparable with other published methods in calculating the clear-sky DNI with concrete statistics.

Wind speed prediction is the key to wind power prediction, which is very important to guarantee the security and stability of the power system. Due to dramatic changes in wind speed, it needs high-frequency sampling to describe the wind. A large number of samples are generated and affect modeling time and accuracy. Therefore, two novel active learning methods with sample selection are proposed for short-term wind speed prediction. The main objective of active learning is to minimize the number of training samples and ensure the prediction accuracy. In order to verify the validity of the proposed methods, the results of support vector regression (SVR) and artificial neural network (ANN) models with different training sets are compared. The experimental data are from a wind farm in Jiangsu Province. The simulation results show that the two novel active learning methods can effectively select typical samples. While reducing the number of training samples, the prediction performance remains almost the same or slightly improved.

Summary In this paper, a novel with respect to economic dispatch of large-scale multi-area power systems is formulated and studied based on decomposition and coordination algorithm. More precisely, auxiliary problem principle (APP) is applied to overcome the defects of augmented Lagrangian relaxation technology with nonlinear inseparable quadratic terms. The main purpose of this paper is to investigate the intrinsic relationship between APP control parameters and convergence speed, and develop a general criterion that ensures that a suitable control parameter set referring to the implemented APP can be effectively estimated. Theoretical and experimental results of this paper are consistent with each other based on the verification of proposed numerical results. Copyright © 2013 John Wiley & Sons, Ltd.

Video-based facial expression recognition (FER) has recently received increased attention as a result of its widespread application. Using only one type of feature to describe facial expression in video sequences is often inadequate, because the information available is very complex. With the emergence of different features to represent different properties of facial expressions in videos, an appropriate combination of these features becomes an important, yet challenging, problem. Considering that the dimensionality of these features is usually high, we thus introduce multiview dimension reduction (MVDR) into video-based FER. In MVDR, it is critical to explore the relationships between and within different feature views. To achieve this goal, we propose a novel framework of MVDR by enforcing joint structured sparsity at both inter- and intraview levels. In this way, correlations on and between the feature spaces of different views tend to be well-exploited. In addition, a transformation matrix is learned for each view to discover the patterns contained in the original features, so that the different views are comparable in finding a common representation. The model can be not only performed in an unsupervised manner, but also easily extended to a semisupervised setting by incorporating some domain knowledge. An alternating algorithm is developed for problem optimization, and each subproblem can be efficiently solved. Experiments on two challenging video-based FER datasets demonstrate the effectiveness of the proposed framework.

Feature selection is a significant preprocessing procedure for many high-dimensional forecasting problems. In multivariate time series forecasting, the purpose of feature selection is to select a relevant feature subset from the original time series. One of the most common strategies for feature selection is mutual information (MI) criterion. K-nearest neighbors (k-NN) is a powerful means which has been widely used to estimate the MI between two high-dimensional variables directly from data set. In order to improve the effect of feature selection, a novel procedure is proposed in this paper to use MI and k-NN to perform feature selection in multivariate time series forecasting. This feature selection procedure not only reduces the irrelevance between the inputs and outputs variables, but also removes the redundant input variables. The procedure of this method is elaborated on a synthetic data set as well as on a real-world example.

The development of instrument technology enables total sky imagers with temporal and spatial high-resolution sky observations, which makes cloud classification based on ground-based sky images possible. However, it is difficult to classify clouds accurately just from a single image according to the International Cloud Classification System. Here, a method of cloud classification is proposed viewed the attenuation rate of solar direct normal radiation (DNI) as the judgment standard, based on a set of mainly statistical features describing the color and the texture of a sky image. To calculate the attenuation rate of DNI,a clear-sky model is firstly developed. And then, a neural network classifier is constructed to separate clouds into two classes: cloud with effect to DNI and cloud with no effect to DNI. The optimal architecture of the neural network is determined by means of two groups of training experiments. With the data from NREL Solar Radiation Research Laboratory, a test is carried out to evaluate the performance of the classifier. Results show that the proposed method can yield a success rate of about 91.67%. The method of cloud classification based on DNI is able to be applied in the photovoltaic system and may improve the accuracy of the short solar power forecasting.

Model structure selection with respect to short-term wind speed forecasting is relatively difficult due to the stochastic and intermittent nature of the wind speed distribution. In order to overcome the disadvantages in traditional approaches such as computing burden and low accuracy, a novel model structure selection technique about short-term wind speed forecasting is proposed in order to improve the computational efficiency and forecasting accuracy using the model variable selection, variable order estimation, model structure optimization techniques, and so on. The detailed and complete process flow associated to the theoretical analysis of the proposed model structure selection technique is described. Moreover, both the so-called overkill in the data filtering and so-called overfitting in the learning processing are avoided by a proper technique in the design of proposed approach. In order to verify the effectiveness of proposed strategy in a practical application, all the experimental results...

For modeling of multivariate time series, input variable selection is a key problem. Feature selection is to select a relevant subset to reduce the dimensionality of the problem without significant loss of information. This paper presents the estimation of mutual information and its application in feature selection problem. Mutual information is one of the most common strategies borrowed from information theory for feature selection. However, the calculation of probability density function (PDF) according to the definition of mutual information is difficult, especially for high dimensional variables. A k-nearest neighbor (k-NN) method based estimator is widely used to estimate the mutual information between two variables directly from the data set. Nevertheless, this estimator depends on smoothing parameter. There is no theoretically method to choose the parameter. This paper purposes to solve two problems: one is to employ resampling methods to help the mutual information estimator to improve feature selection and the other is to apply these methods to a wind power prediction problem.

The 3D display technology based on motion parallax has the advantages of no need to wear glasses and has no limitation on the viewing angle. However, the current problems of this method include 3D scenes and models can only be displayed within the screen; the system is sensitive to light. For the first problem, we proposed the concept of virtual bezel, which makes out system could produce a strong illusion that objects displayed outside the screen, and greatly increases the visual impact of 3D display. For the problem of light sensitivity, unlike other system using RGB camera, we use RGB-D images to detect and track the viewer's face, which makes our system work properly even at night.

Recently, a number of graph-based approaches have been proposed to detect salient regions in images. Although the graph is essential for these approaches, the graph construction method has not been studied in much detail. We propose a graph construction method that makes better use of multiple Gestalt principles. Specifically, spatial proximity, color similarity, and texture similarity between image regions are employed to choose different edges of the graph. Furthermore, the confliction among multiple Gestalt principles is solved using a primal rank support vector machines algorithm to compute the edge weights. Our experimental results on various salient region detection databases with comparisons to representative approaches demonstrate that the proposed graph construction method helps to detect salient objects in images.

We explicitly analyze the trajectories of learning near singularities in hierarchical networks, such as multilayer perceptrons and radial basis function networks, which include permutation symmetry of hidden nodes, and show their general properties. Such symmetry induces singularities in their parameter space, where the Fisher information matrix degenerates and odd learning behaviors, especially the existence of plateaus in gradient descent learning, arise due to the geometric structure of singularity. We plot dynamic vector fields to demonstrate the universal trajectories of learning near singularities. The singularity induces two types of plateaus, the on-singularity plateau and the near-singularity plateau, depending on the stability of the singularity and the initial parameters of learning. The results presented in this letter are universally applicable to a wide class of hierarchical models. Detailed stability analysis of the dynamics of learning in radial basis function networks and multilayer perceptrons will be presented in separate work.

This paper shows the cost-aware virtual machine allocation problem for 100% green data centers is an integer programming problem which is NP-hard. The paper presents a greedy algorithm for cost-aware virtual machine allocation which attempts to minimize the total cost of biofuel while maintaining the quality-of-service requirements of the tasks.

Reliable and accurate short-term wind speed forecasting is of great importance for secure power system operations. In this study, a novel two-step method to construct a multi-model ensemble, which consists of linear regression, multi-layer perceptrons and support vector machines, is proposed. The ensemble members first compete with each other in a number of training rounds, and the one with the best forecasting accuracy in each round is recorded. Then, after all the training rounds, the occurrence frequency of each member is calculated and used as the weight to form the final multi-model ensemble. The effectiveness of the proposed multi-model ensemble has been assessed on the real datasets collected from three wind farms in China. The experimental results indicate that the proposed ensemble is capable of providing better performance than the single predictive models composing it.

Radial basis function (RBF) networks are one of the most widely used models for function approximation and classification. There are many strange behaviors in the learning process of RBF networks, such as slow learning speed and the existence of the plateaus. The natural gradient learning method can overcome these disadvantages effectively. It can accelerate the dynamics of learning and avoid plateaus. In this letter, we assume that the probability density function (pdf) of the input and the activation function are gaussian. First, we introduce natural gradient learning to the RBF networks and give the explicit forms of the Fisher information matrix and its inverse. Second, since it is difficult to calculate the Fisher information matrix and its inverse when the numbers of the hidden units and the dimensions of the input are large, we introduce the adaptive method to the natural gradient learning algorithms. Finally, we give an explicit form of the adaptive natural gradient learning algorithm and compare it to the conventional gradient descent method. Simulations show that the proposed adaptive natural gradient method, which can avoid the plateaus effectively, has a good performance when RBF networks are used for nonlinear functions approximation.

Video-based facial expression recognition has received significant attention in recent years due to its widespread applications. One key issue for video-based facial expression analysis in practice is how to extract dynamic features. In this paper, a novel approach is presented using histogram sequence of local Gabor binary patterns from three orthogonal planes (LGBP-TOP). In this approach, every facial expression sequence is firstly convolved with the multi-scale and multi-orientation Gabor filters to extract the Gabor Magnitude Sequences (GMSs). Then, we use local binary patterns from three orthogonal planes (LBP-TOP) on each GMS to further enhance the feature extraction. Finally, the facial expression sequence is modeled as a histogram sequence by concatenating the histogram pieces of all the local regions of all the LGBP-TOP maps. For recognition, Support Vector Machine (SVM) is exploited. Our experimental results on the extended Cohn-Kanade database (CK+) demonstrate that the proposed method has achieved the best results compared to other methods in recent years.

This paper addresses the controller synthesis for a class of continuous-time switched positive linear systems. A sufficient condition for the existence of a set of state-feedback controllers which guarantee the positivity and stability of the closed-loop systems for switched positive linear systems are presented, including the requirement of positive or negative controllers. In addition, the controller synthesis problems with interval and polytopic uncertainties are also addressed. All the obtained conditions are solved in terms of Linear Programming. Finally, numerical examples are provided to illustrate the effectiveness of the result.

Gaussian regularization is an effective method to improve the generalization ability of neural networks. A Gaussian regularization RBF neural network (GRNN) which combines the advantages of RAN, and regularization is proposed in this paper. And a model using GRNN is presented to predict the ash fusion temperature (AFT) for some Chinese coals Compared with the traditional techniques, the GRNN prediction model has not only small training and testing error, but also a more compact network structure.

We study the dynamics of learning in a hierarchical model such as multilayer perceptron. Such a model includes singularities, which affects its dynamics seriously. The Milnor type attractors appear, because of the singularity. We will show its trajectories explicitly, and present the topological nature of the singularities.