Support vector regression methodology for prediction of input displacement of adaptive compliant robotic gripper

The prerequisite for new versatile grippers is the capability to locate and perceive protests in their surroundings. It is realized that automated controllers are profoundly nonlinear frameworks, and a faultless numerical model is hard to get, in this way making it troublesome to control utilizing tried and true procedure. Here, a design of an adaptive compliant gripper is presented. This design of the gripper has embedded sensors as part of its structure. The use of embedded sensors in a robot gripper gives the control system the ability to control input displacement of the gripper and to recognize specific shapes of the grasping objects. Since the conventional control strategy is a very challenging task, soft computing based controllers are considered as potential candidates for such an application. In this study, the polynomial and radial basis function (RBF) are applied as the kernel function of Support Vector Regression (SVR) to estimate and predict optimal inputs displacement of the gripper according to experimental tests and shapes of grasping objects. Instead of minimizing the observed training error, SVR _poly and SVR _rbf attempt to minimize the generalization error bound so as to achieve generalized performance. The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by the SVR approach compared to other soft computing methodology.     

Statistical hierarchical clustering algorithm for outlier detection in evolving data streams

Machine Learning - Anomaly detection is a hard data analysis process that requires constant creation and improvement of data analysis algorithms. Using traditional clustering algorithms to analyse...     

A New and Fast HPLC Method for Determination of Rutin, Troxerutin, Diosmin and Hesperidin in Food Supplements Using Fused-Core Column Technology

A new and fast high-performance liquid chromatography (HPLC) method using fused-core column for separation of rutin, troxerutin, diosmin, and hesperidin has been developed and used for determination of these flavonoids in food supplements. Efficient separation of flavonoids and internal standard methylparaben was achieved on the fused-core column Ascentis Express RP-Amide (100?×?3.0 mm), particle size 2.7 μm, with mobile phase acetonitrile/water solution of acetic acid pH?3 (30:70, v/v) at a flow rate of 1.0 mL?min^?1 and at temperature 50 °C. The detection wavelengths were set at 283 nm for hesperidin and at 255 nm for rutin, troxerutin, diosmin, and internal standard methylparaben. Under the optimal chromatographic conditions, good linearity with correlation coefficients in the range (r?=?0.9991–0.9998; n?=?7) for all flavonoids was achieved. Commercial samples of food supplements were extracted with 100 % dimethyl sulfoxide using ultrasound bath for 10 min and then diluted to methanol. A 5-μL sample volume of the filtered solution was directly injected into the HPLC system. Accuracy of the method defined as a mean recovery of flavonoids from food supplement matrix was in the range 96.2–104.4 % for all flavonoids. The intraday method precision was satisfactory, and relative standard deviations of sample analysis including preparation and determination of different food supplements were in the range 0.5–3.5 % for all flavonoids. The developed method has shown high sample throughput during sample preparation process, modern separation approach, and short time (5 min) of analysis.     

The effect of temperature and oxygen content on the flowing liquid metal corrosion of structural steels in the Pb–Bi eutectic

We performed corrosion tests of 1000 h each on approximately 20 types of structural steels (austenitic, ferritic and martensitic) in convection loops with flowing Pb–Bi at 500, 450 and 400 °C and a temperature gradient of 100 °C. These experiments were performed in liquid Pb–Bi with different oxygen concentrations (from approximately 1 × 10⁻⁶ to 2 × 10⁻⁵ wt.%) to ascertain at what oxygen concentration and up to what temperature the oxygen technology can create protective oxide or spinel layers to reduce or prevent corrosion. The results showed that the structural materials contemplated for building an ADS system, including 9% Cr–1% Mo (W) martensitic steels and similar steels with a higher Si content (2–3%), can be used with their surface unpassivated at up to 450 °C and suffer only minimal corrosion (up to 5 μm/year). At higher temperatures, their surface must be passivated prior to and regularly during the operation; however, no technology to perform such passivation in the presence of Pb–Bi is known that this time. In addition, we measured the impact of various alloying elements, such as Fe, Cr, Ni, Mn, Si, Al and Mo, on the corrosion of such steels and searched for potential ways to passivate their surface or create protective oxide or spinel layers during operation by varying the amount of oxygen in liquid Pb–Bi.     

Quasicrystal–crystal structural transformation in Al–5 wt.% Mn alloy

The atomic structure of Al–5 wt.%Mn (Al–5Mn) alloy, prepared by rapid solidification, and pre-annealed at 623 and 773 K for 5 and 1 h, respectively, were characterized by X-ray powder diffraction (XRD) and extended X-ray absorption fine structure (XAFS) techniques. The sample in as-quenched stage was found crystalline, consisting of metastable α-Al (Al–Mn solid solution) and icosahedral quasicrystalline I-Al₆Mn phases. Five hours annealing at 623 K proved thermal stability of both the phases. Pre-annealing at 773 K/1 h on the other hand leads to α-Al phase decomposition and structural transformation of metastable I-Al₆Mn to stable orthorhombic Al₆Mn phase. The EXAFS results indicate that Mn atoms are located preferably on the outer shell of icosahedrons. During the I-Al₆Mn→o-Al₆Mn transformation the total Al atoms coordinating one Mn were found to be constant (∼10). Based on the results, only distance/symmetry changes in atomic arrangement around Mn atoms were suggested.     

Adaptive neuro fuzzy selection of heart rate variability parameters affected by autonomic nervous system

Heart rate variability (HRV) parameters can be used as specific indicator of autonomic nervous system (ANS) behavior. ANS, with its main two branches, sympathetic and parasympathetic, may be considered as a coordinated neuronal network which controls heart rate continually. Many parameters define heart rate variability in different domains such as time, frequency or nonlinear. An excessively high computational complexity can occur when developing models for medical applications when the best set of inputs to use is not known. To build a model that can predict a specific process output, it is desirable to select a subset of variables that are truly relevant or the most influential to this output. This procedure is typically called variable selection, and it corresponds to finding a subset of the full set of recorded variables that exhibits good predictive abilities. In this study an architecture for modeling complex systems in function approximation and regression was used, based on using adaptive neuro-fuzzy inference system (ANFIS). Variable searching using the ANFIS network was performed to determine how the ANS branches affect the most relevant HRV parameters. The method utilized may work as a basis for examination of ANS influence on HRV activity.     

Easy to use and rapid isolation and detection of a viral nucleic acid by using paramagnetic microparticles and carbon nanotubes-based screen-printed electrodes

A method that is easy to use, rapid, with a low cost of detecting viral nucleic acid in a biological sample represents the essential tool in targeted therapy. In this study, we report the use of paramagnetic microparticles covered by streptavidin and modified by an oligonucleotide probe with a specific viral sequence labeled by biotin to detect human immunodeficiency virus (HIV) and influenza virus subtype H5N1. The viral nucleic acids were primarily detected by adsorptive transfer stripping technique coupled with square wave voltammetry using carbon paste, hanging mercury drop or carbon nanotubes-based screen-printed working electrodes. Detection limits were estimated for both sequences down to picograms per 3 μl. To isolate the viral sequences, paramagnetic microparticles covered with biotin-labeled oligonucleotides were used. We calculated the yield of isolation for H5N1 and/or HIV sequences, which was defined as “isolated concentration of viral nucleic acid sequence”/“given viral nucleic acid sequence” × 100. We estimated the yield for both sequences as 59%. Moreover, we studied the influence of human serum, dsDNA and non-complementary sequence of nucleic acids on isolation of viral nucleic acids. We also used carbon nanotubes-based screen-printed electrodes coupled with micro-flow instrument to detect viral nucleic acids. We were able to isolate and detect nanogram amounts of nucleic acids.