Deep multilayer multiple kernel learning

Multiple kernel learning (MKL) approach has been proposed for kernel methods and has shown high performance for solving some real-world applications. It consists on learning the optimal kernel from one layer of multiple predefined kernels. Unfortunately, this approach is not rich enough to solve relatively complex problems. With the emergence and the success of the deep learning concept, multilayer of multiple kernel learning (MLMKL) methods were inspired by the idea of deep architecture. They are introduced in order to improve the conventional MKL methods. Such architectures tend to learn deep kernel machines by exploring the combinations of multiple kernels in a multilayer structure. However, existing MLMKL methods often have trouble with the optimization of the network for two or more layers. Additionally, they do not always outperform the simplest method of combining multiple kernels (i.e., MKL). In order to improve the effectiveness of MKL approaches, we introduce, in this paper, a novel backpropagation MLMKL framework. Specifically, we propose to optimize the network over an adaptive backpropagation algorithm. We use the gradient ascent method instead of dual objective function, or the estimation of the leave-one-out error. We test our proposed method through a large set of experiments on a variety of benchmark data sets. We have successfully optimized the system over many layers. Empirical results over an extensive set of experiments show that our algorithm achieves high performance compared to the traditional MKL approach and existing MLMKL methods.     

Development of broadband forward‐wave directional couplers loaded by periodic patterned ground structure

In this paper two wideband Forward‐Wave Directional Couplers (FWDCs) with 0 dB and 3 dB coupling level are proposed. Using periodic patterned ground structure in a microstrip coupled lines by a new unit cell; even‐ and odd‐mode characteristic impedances of the couplers are equal over a wide frequency range. Moreover, it provides a constant phase difference between even and odd‐modes. The proposed cell is modeled using the equivalent circuit model and a design procedure is introduced for designing FWDCs for an arbitrary value of coupling level. The introduced couplers are numerically investigated and a prototype of both couplers is made. It is shown that for 0 dB coupling level, the measured coupling is 0.85 dB with 1 dB flatness over fractional bandwidth of 96% bandwidth. In case of 3 dB coupling, the measured coupling level is 3.5 dB at 7.42 GHz with 1 dB flatness over fractional bandwidth of 67.1%.     

A mathematical framework to study fast walking of human

The body of a walking human is an elaborated dynamic system that operates adaptively in various conditions such as fast walking. Due to dynamic redundancies, the individual motor control strategies for speeding up the walking can be different among normal subjects. However, in reality, we see that the pattern of motion is quite similar among people and it is only the profile of hip joint motion along its path which determines the speed. The objective of the current paper is to develop a mathematical framework to investigate time optimal motion of a human during walking. To this end, a nine-link planar biped model is used. The motion is considered to take place in sagittal plane and to follow a normal pattern of motion. The solution is obtained using a phase plane method to solve minimum time problem which is subjected to inequality constraints of variable maximum joint torques and stability conditions. The solution method can be used to find the maximum possible speed of a human with specific body characteristics and to obtain a hip joint trajectory which could produce that speed. The proposed method can be utilized to study quantitative effect of different parameters such as joint strength in fast walking.     

A comparative study between the augmented Lagrangian method and the complementarity approach for modeling the contact problem

The increasing demand for real-time high-fidelity multibody dynamics simulations in several modern fields such as robotics and computer game industries has motivated many researches to propose novel approaches to model multibody systems with several contacts. The possibility of different contact conditions in a system with several contacts yields a combinatorial problem of potentially large size. Rigid contact model which is the most common model used for real-time simulations yields a non-smooth dynamic formulation. The solution of such a system can be governed using different methods. In this paper a comparison between the complementarity approaches and the augmented Lagrangian based formulations to deal with non-smooth contact models is presented via numerical examples, and the advantages and shortcomings of each method are discussed.     

A fast and noise resilient cluster-based anomaly detection

Clustering, while systematically applied in anomaly detection, has a direct impact on the accuracy of the detection methods. Existing cluster-based anomaly detection methods are mainly based on spherical shape clustering. In this paper, we focus on arbitrary shape clustering methods to increase the accuracy of the anomaly detection. However, since the main drawback of arbitrary shape clustering is its high memory complexity, we propose to summarize clusters first. For this, we design an algorithm, called Summarization based on Gaussian Mixture Model (SGMM), to summarize clusters and represent them as Gaussian Mixture Models (GMMs). After GMMs are constructed, incoming new samples are presented to the GMMs, and their membership values are calculated, based on which the new samples are labeled as “normal” or “anomaly.” Additionally, to address the issue of noise in the data, instead of labeling samples individually, they are clustered first, and then each cluster is labeled collectively. For this, we present a new approach, called Collective Probabilistic Anomaly Detection (CPAD), in which, the distance of the incoming new samples and the existing SGMMs is calculated, and then the new cluster is labeled the same as of the closest cluster. To measure the distance of two GMM-based clusters, we propose a modified version of the Kullback–Libner measure. We run several experiments to evaluate the performances of the proposed SGMM and CPAD methods and compare them against some of the well-known algorithms including ABACUS, local outlier factor (LOF), and one-class support vector machine (SVM). The performance of SGMM is compared with ABACUS using Dunn and DB metrics, and the results indicate that the SGMM performs superior in terms of summarizing clusters. Moreover, the proposed CPAD method is compared with the LOF and one-class SVM considering the performance criteria of (a) false alarm rate, (b) detection rate, and (c) memory efficiency. The experimental results show that the CPAD method is noise resilient, memory efficient, and its accuracy is higher than the other methods.     

Robust Stability and Stabilization Of TCP‐Networked Control Systems with Multiple Delay System Modeling

This paper presents a new model for networked control systems (NCSs) under transmission control protocol (TCP) as a multiple‐delay system by considering both sensor to controller and controller to actuator delays. An analytical TCP model has been considered for the network part, and an active queue management (AQM) controller is designed to regulate the desired queue length, which ensures holding the network induced delay and its variation within their lower bounds. The model is assumed to possess structured uncertainties due to the stochastic nature of the network. Robust stability and stabilization conditions are derived in terms of linear matrix inequalities (LMIs) by applying the Lyapunov‐Krasovskii stability criterion. Illustrative examples are presented and it has been shown that the proposed method will obtain less conservative results compared to the existing approaches in the literature.     

Structural Properties of Networked Control Systems With Bandwidth Limitations and Delays

The structural properties of networked control systems with both bandwidth limitations and delays are investigated. Sufficient conditions are given for controllability (stabilizability) and reconstructibility (detectability). Our results enhance previous works by capturing bandwidth limitations and delays simultaneously. The adopted modeling framework could be readily used in control and estimation methods, including optimal and predictive schemes. It also facilitates the use of scheduling optimization algorithms in conjunction with the control scheme presented.     

Robust model predictive control of biped robots with adaptive on-line gait generation

International Journal of Control, Automation and Systems - In this paper, an on-line gait control scheme is proposed for the biped robots for walking up and down the stairs. In the proposed...     

Multi-objective optimization of laminated composite shells for minimum mass/cost and maximum buckling pressure with failure criteria under external hydrostatic pressure

In the present study, Multi-objective optimization of composite cylindrical shell under external hydrostatic pressure was investigated. Parameters of mass, cost and buckling pressure as fitness functions and failure criteria as optimization criterion were considered. The objective function of buckling has been used by performing the analytical energy equations and Tsai-Wu and Hashin failure criteria have been considered. Multi-objective optimization was performed by improving the evolutionary algorithm of NSGA-II. Also the kind of material, quantity of layers and fiber orientations have been considered as design variables. After optimizing, Pareto front and corresponding points to Pareto front are presented. Trade of points which have optimized mass and cost were selected by determining the specified pressure as design criteria. Finally, an optimized model of composite cylindrical shell with the optimum pattern of fiber orientations having appropriate cost and mass is presented which can tolerate the maximum external hydrostatic pressure.

     

Data-Driven Texture Rendering on an Electrostatic Tactile Display

In this article, we present a data-driven texture rendering method applied to a tactile display based on electrostatic attraction. The proposed method was examined in two steps. First, accelerations occurring due to sliding a tool on three different surfaces were measured, and then the collected data were replayed on an electrostatic tactile display. The proposed data-driven texture rendering method was evaluated against a conventional method in which a standard input such as a square wave was used for texture representation. Second, data from the Penn Haptic Texture Toolkit were used to generate virtual textures on the same tactile display. Psychophysical experiments were carried out for both steps, during which subjects rated similarities among the rendered virtual textures and the real samples. Confusion matrices were created, and multidimensional scaling (MDS) analysis was performed to create a perceptual space for further examination and to extract underlying dimensions of the textures. The results show that the virtual textures generated using the data-driven method were similar to the real textures. Roughness and stickiness were the primary dimensions of texture perception. Together with the supporting results from the MDS analysis, this study showed that the data-driven method is a viable solution for realistic texture rendering with electrostatic attraction.