Improving object classification robustness in RGB-D using adaptive SVMs

Nowadays object recognition is a fundamental capability for an autonomous robot in interaction with the physical world. Taking advantage of new sensing technologies providing RGB-D data, the object recognition capabilities increase dramatically. Object recognition has been well studied, however, known object classifiers usually feature poor generality and, therefore, limited adaptivity to different application domains. Although some domain adaptation approaches have been presented for RGB data, little work has been done on understanding the effects of applying object classification algorithms using RGB-D for different domains. Addressing this problem, we propose and comprehensively investigate an approach for object recognition in RGB-D data that uses adaptive Support Vector Machines (aSVM) and, in this way, achieves an impressive robustness in cross-domain adaptivity. For evaluation, two datasets from different application domains were used. Moreover, a study of state-of-the-art RGB-D feature extraction techniques and object classification methods was performed to identify which combinations (object representation - classification algorithm) remain less affected in terms of performance while switching between different application domains.     

Evaluation of quadratic cost functionals for neutral systems: the frequency-domain approach

As shown in Grabowski (1983), the problem of evaluation of the quadratic performance index for delay-differential systems of neutral type can be effectively solved by the use of Lyapunov functionals. In this paper another solution to that problem, based on the frequency-domain approach, is proposed. With the use of Plancherel's theorem and other tools of complex analysis, this task is reduced to a Riemann-Hilbert problem, an explicit solution to which is found by elementary methods. Our results can be regarded as a generalization of those due to Walton and Marshall (1984, 1987). Two simple examples of application are also provided.     

A two-pronged approach for springback variability assessment using sparse polynomial chaos expansion and multi-level simulations

In this study, we show that stochastic analysis of metal forming process requires both a high precision and low cost numerical models in order to take into account very small perturbations on inputs (physical as well as process parameters) and to allow for numerous repeated analysis in a reasonable time. To this end, an original semi-analytical model dedicated to plain strain deep drawing based on a Bending-Under-Tension numerical model (B-U-T model) is used to accurately predict the influence of small random perturbations around a nominal solution estimated with a full scale Finite Element Model (FEM). We introduce a custom sparse variant of the Polynomial Chaos Expansion (PCE) to model the propagation of uncertainties through this model at low computational cost. Next, we apply this methodology to the deep drawing process of U-shaped metal sheet considering up to 8 random variables.     

Sharpening filter for false color imaging of dual-energy X-ray scans

Dual-energy X-ray imaging has a vast range of application in security. Luggage inspection is an essential process for an airplane or court house security as well as securing mass events. An image of a content of some package may help to figure out if there is any dangerous object inside and avoid possibly threatening situation. As the raw X-ray images are not always easy to analyze and interpret, some image processing methods like an object detection, a frequency resolution increase or a pseudocoloring are being used. Since color can be a powerful tool to improve the usefulness of an information display, we propose pseudocoloring improvement by modifying material-based approach with edge detection to fill and sharpen color layers over the image making it easier to read and analyze. We demonstrate the effectiveness of the methods using real data, acquired from a professional dual-energy X-ray scanner.     

Decision trees for uplift modeling with single and multiple treatments

Most classification approaches aim at achieving high prediction accuracy on a given dataset. However, in most practical cases, some action such as mailing an offer or treating a patient is to be taken on the classified objects, and we should model not the class probabilities themselves, but instead, the change in class probabilities caused by the action. The action should then be performed on those objects for which it will be most profitable. This problem is known as uplift modeling, differential response analysis, or true lift modeling, but has received very little attention in machine learning literature. An important modification of the problem involves several possible actions, when for each object, the model must also decide which action should be used in order to maximize profit. In this paper, we present tree-based classifiers designed for uplift modeling in both single and multiple treatment cases. To this end, we design new splitting criteria and pruning methods. The experiments confirm the usefulness of the proposed approaches and show significant improvement over previous uplift modeling techniques.     

Optimization of reinforcement layout of scissor‐type bridge using differential evolution algorithm

Scissors mechanisms are commonly used in safety engineering during the construction of temporary structures, owing to their inherent advantages of foldability, transformability, and reusability. We effectively utilized these scissors mechanism features to develop a lightweight, deployable emergency Mobile Bridge (MB) based on optimization, and control of the folding structure. Here, we discuss the problems of optimal reinforcement layout for the MB by formulating and solving three optimization problems, namely: (a) the load capacity maximization problem, (b) the weight minimization problem, and (c) coupling the load capacity maximization problem and the weight minimization problem. The potential benefits resulting from the application of reinforcement were evaluated using a combination of finite element analysis and an optimization algorithm based on the differential evolution method. The results demonstrate the significant positive influence of the additional reinforcing members. In particular, the limit load was increased by over 10 times, while the weight was decreased to approximately half. The proposed methodology enabled the development of a substantially improved version of the MB characterized by a higher load capacity and lower weight in comparison to the initial bridge design.     

Power system stability enhancement by WAMS-based supplementary control of multi-terminal HVDC networks

This paper deals with supplementary control of a MTDC network designed for the stability enhancement of a AC power system. The proposed control is a WAMS-based control modulating the real and reactive power at the terminals of the DC network. Relevant control formulas have been derived for a linear multi-machine system model with the application of the direct Lyapunov method. Validity and robustness of the proposed control has been verified by computer simulation for a multi-machine test system using a nonlinear model and detailed modeling of power system components. The proposed control is robust and insensitive to changes in the network configuration and loading conditions in the AC power system. In the case, when more of the MTDC networks and/or the HVDC links are used in one interconnected power system the proposed stabilizing control produces additive damping i.e. each controlled network element contributes to the positive damping. Some practical aspects have also been discussed. The proposed WAMS-based stabilizing control of the MTDC network is innovative by both its main concept and the derivation of control formulas using the direct Lyapunov method.     

Towards the improvement of thermal efficiency in lignite‐fired power generation: Concerning the utilization of Polish lignite deposits in state‐of‐the‐art IGCC technology

Integrated coal Gasification Combined Cycle (IGCC) is the most advanced technology for coal‐fired power generation. The two‐stage entrained flow gasification process allows for the use of a wide range of coal, as long as the gasification temperature is above the ash melting point of a used fuel. In this gasification technology, lignite, which often has a low ash melting point, can be preferably utilized. However, ash fluidity is also another importance, because the behaviour of molten slag can diminish a stable ash discharge from a gasifier. As the eligibility of coal ash properties is a considerable factor, water physically and chemically kept in lignite (30 – 60% in mass) attributes to deteriorating gasification efficiency, because it causes significant heat loss and increasing oxygen consumption. Developing a thermal evaporative lignite drying method will be a necessary attempt to apply lignite to the coal gasification process. For those preceded objectives, coal and ash properties and drying characteristics of several grades of Polish lignite, extracted from Belchatow and Turow deposits, have been experimentally investigated in a preliminary study evaluating the applicability and consideration for its utilization in state‐of‐the‐art clean coal technology, IGCC. This paper particularly discusses the eligibility of Polish lignite from the perspective of the fusibility and fluidity of ash melts and the fundamental drying kinetics of lignite in superheated steam in the light of water removal. The viscosity of ash melts is measured at high temperature up to 1700 °C. In the drying tests, the significant influence of structural issues, because of the provenance and origin of lignite on the drying characteristics, was found by applying the method of sensitivity analysis of physical propensity. This paper concludes that the investigated Polish lignite has characteristics favourable for utilization in IGCC technology, once the precautions related to its high moisture have been carefully addressed. Copyright © 2016 John Wiley & Sons, Ltd.     

Online version of the theorem of Thue

A sequence S is nonrepetitive if no two adjacent blocks of S are the same. In 1906 Thue proved that there exist arbitrarily long nonrepetitive sequences over 3 symbols. We consider the online variant of this result in which a nonrepetitive sequence is constructed during a play between two players: Bob is choosing a position in a sequence and Alice is inserting a symbol on that position taken from a fixed set A. The goal of Bob is to force Alice to create a repetition, and if he succeeds, then the game stops. The goal of Alice is naturally to avoid that and thereby to construct a nonrepetitive sequence of any given length.We prove that Alice has a strategy to play arbitrarily long provided the size of the set A is at least 12. This is the online version of the theorem of Thue. The proof is based on nonrepetitive colorings of outerplanar graphs. On the other hand, one can prove that even over 4 symbols Alice has no chance to play for too long. The minimum size of the set of symbols needed for the online version of Thue?s theorem remains unknown.     

Performance indices of a common rail-system CI engine powered by diesel oil and biofuel blends

Conventional fuels used for supplying internal combustion piston engines include petrols and diesel oils produced from petroleum. These are a non-renewable energy source. The environmental policy of the European Union is geared towards increasing the share of renewable fuels in the overall energy consumption. An alternative fuel originating from a renewable source, which could be used for feeding self-ignition internal combustion engines are the fatty acid methyl esters (FAME) of plant oils. The paper reports selected results of testing a 1.3 MULTIJET SDE 90 PS self-ignition engine with the Common Rail reservoir feed system supplied with mixtures of diesel oil and rape oil fatty acid methyl esters (FAME). Tests were carried out on an engine test bed equipped with an eddy-current brake. The purpose of the tests was to determine the economic–energy and ecological indices of engine operation. The concentrations of exhaust gas gaseous components were measured using a MEXA-1600DEGR analyzer, while the particulate concentrations, with a MEXA-1230PM analyzer. In addition, the variations of working medium pressures in the engine chamber and of fuel pressure upstream the injector were recorded as a function of crankshaft rotation angle using the AVL IndiSmart 612 indication system for this purpose. The physicochemical properties of fuels used in the tests were determined using a fuel analyzer. The obtained testing results made it possible to determine and assess the operation indices of the engine fed with mixtures of diesel oil and rape oil fatty acid methyl esters (FAME) with slightly higher ester contents than the requirements of the currently applicable diesel oil standard.