Sensor Fusion System Using Recurrent Fuzzy Inference

In robotic and manufacturing systems, it is difficult to measure the state of systems accurately because of many uncertain factors and noise, and it is very important to estimate the state of systems. We must measure the phenomena of systems by multiple sensors and estimate the state of systems by acquiring information of sensors. However, we can not acquire all of sensor information synchronically, because each sensor has particular sensor information and measuring time. For estimating the state of systems by multiple sensors, a multi-sensor fusion system fusing various sensory information is needed. In this paper, we propose a Recurrent Fuzzy Inference (RFI) with recurrent inputs and apply it to a multi-sensor fusion system for estimating the state of systems. The membership functions of RFI are expressed by Radial Basis Function (RBF) with insensitive ranges. The shape of the membership functions can be adjusted by a learning algorithm. The learning algorithm is based on the steepest descent method and incremental learning which can add new fuzzy rules. The effectiveness of the multi-sensor fusion system using RFI will be shown through a numerical experiment of moving robot and estimation of surface roughness in grinding process.     

Directed Evolution of a Cp*RhIII-Linked Biohybrid Catalyst Based on a Screening Platform with Affinity Purification

Directed evolution of Cp*Rh^III-linked nitrobindin (NB), a biohybrid catalyst, was performed based on an in vitro screening approach. A key aspect of this effort was the establishment of a high-throughput screening (HTS) platform that involves an affinity purification step employing a starch-agarose resin for a maltose binding protein (MBP) tag. The HTS platform enables efficient preparation of the purified MBP-tagged biohybrid catalysts in a 96-well format and eliminates background influence of the host E. coli cells. Three rounds of directed evolution and screening of more than 4000 clones yielded a Cp*Rh^III-linked NB(T98H/L100K/K127E) variant with a 4.9-fold enhanced activity for the cycloaddition of acetophenone oximes with alkynes. It is confirmed that this HTS platform for directed evolution provides an efficient strategy for generating highly active biohybrid catalysts incorporating a synthetic metal cofactor.     

Decay of material properties of degraded MEA caused by repeated cold starts under subzero conditions in polymer electrolyte fuel cells

This study investigated the characteristics of cell performance degradation, decline of component performance, and changes in the properties of membrane electrode assembly materials caused by repeated cold starts under a subzero condition of ?30 °C. It was made clear that functional decay appeared mainly at the cathode due to increased proton conductive impedance and reduction of reactivity of the electrode catalyst. Among the cathode components, an increase in proton conductive impedance in the cathode electrolyte was dominant. Furthermore, the application of ion chromatography and a newly developed proton‐induced gamma‐ray emission method to measure fluorine in the off‐gas drain revealed that decomposition of the electrolyte was dominant in the cathode catalyst layer. A decrease in fluorine in the cathode electrolyte measured by fluorine‐19 nuclear magnetic resonance confirmed this decomposition. A hypothesis is also presented concerning the cause of the performance degradation. © 2012 Wiley Periodicals, Inc. Heat Trans Asian Res; Published online in Wiley Online Library ( wileyonlinelibrary.com/journal/htj ). DOI 10.1002/htj.20394     

Preparation of Lanthanum Nickel Oxide-Coated Ni Sheet Anodes and Their Application to Electrolytic Production of (CF3)3N in (CH3)4NF·4.0HF Melt

A new process for electrolytic production of a perfluorinated compound, (CF₃)₃N, using lanthanum nickel oxide-coated Ni sheet anode in the (CH₃)₄NF·4.0HF melt at room temperature, was developed. Thin films of the lanthanum nickel oxides were prepared on Ni sheets by sol-gel coating method using polyvinlylpyrrolidone(PVP). The main components of the thin films were La₂O₃, LaNiO₃, and La₂NiO₄ at 500, 750 and 1000 °C, respectively. The anode performance in the (CH₃)₄NF·4.0HF melt depends greatly on the main component of the thin film, and the LaNiO₃-coated Ni sheet anode gives the best anode performance. The potential of LaNiO₃-coated Ni sheet anode remains constant at 5.9 V during electrolysis at 20 mA·cm⁻² in the (CH₃)4NF·4.0HF melt for 100 h. This is because LaNiO₃ and NiF₃, and/or Ni₂F₅, the latter of which was formed during electrolysis, in the film give a high electronic conductivity to the surface film during electrolysis. The maximum mole fraction of (CF₃)₃N (21.4%) was obtained at 20 mA·cm⁻² in (CH₃)₄NF·4.0HF melt using the LaNiO₃-coated Ni sheet.     

DRYING OF OPTICALLY SEMITRANSPARENT MATERIALS BY COMBINED RADIATIVE-CONVECTIVE HEATING

The objective of this work has been to basically elucidate the drying characteristics of an optically semitransparent material by combined radiative and convective heating. The experiments were conducted for a graphite suspension, a slurry of surplus activated sludge and a wet silica sand. The time-change of the drying rate as well as of the surface temperature of the brimfully wet material layer were measured under the step heating conditions using an infrared lamp bundle and a blast of hot air heated by an electric heater.

The experimental data obtained show satisfactory agreement with the calculated results from unsteady heat and mass transfer equations derived on the basis of a semitransparent drying model during the preheating and the constant drying rate periods.     

Soft Margins for AdaBoost

Recently ensemble methods like ADABOOST have been applied successfully in many problems, while seemingly defying the problems of overfitting.ADABOOST rarely overfits in the low noise regime, however, we show that it clearly does so for higher noise levels. Central to the understanding of this fact is the margin distribution. ADABOOST can be viewed as a constraint gradient descent in an error function with respect to the margin. We find that ADABOOST asymptotically achieves a hard margin distribution, i.e. the algorithm concentrates its resources on a few hard-to-learn patterns that are interestingly very similar to Support Vectors. A hard margin is clearly a sub-optimal strategy in the noisy case, and regularization, in our case a mistrust in the data, must be introduced in the algorithm to alleviate the distortions that single difficult patterns (e.g. outliers) can cause to the margin distribution. We propose several regularization methods and generalizations of the original ADABOOST algorithm to achieve a soft margin. In particular we suggest (1) regularized ADABOOST_REG where the gradient decent is done directly with respect to the soft margin and (2) regularized linear and quadratic programming (LP/QP-) ADABOOST, where the soft margin is attained by introducing slack variables.Extensive simulations demonstrate that the proposed regularized ADABOOST-type algorithms are useful and yield competitive results for noisy data.     

Gustatory disturbance due to cerebrovascular disorder

In the presence of denaturant and thiol catalyst, a disulfide-containing protein denatures and converts to a mixture of scrambled isomers, which can be purified and structurally characterized. Scrambled isomers adopt a different conformation and a varied extent of unfolding. Their relative concentration (composition) signals the state of unfolding of the denatured protein and is determined by the denaturing condition. In this report, tick anticoagulant peptide (TAP) (60 amino acids and 3 disulfides) has been denatured in the presence of urea, guanidine hydrochloride, guanidine thiocyanate, organic solvents, and at elevated temperature. The recoveries of scrambled TAP were analyzed. The results demonstrate that each denaturing condition generates a unique structure (composition of scrambled species) of denatured TAP. Among various species of scrambled TAP, the beads-form species contains the smallest disulfide loop and appears to represent the most extensively unfolded state. The yield of the beads-form species as a fraction of the total denatured TAP is invariably determined by the strength of the denaturing condition.     

Effect of Alkali Ions on the Crystallization of 3BaO·5SiO2 Glass

In this study we have investigated the effects of alkali additions on the spherulitic crystallization process in 3BaO·5SiO₂ glass. Both Li⁺ and Na⁺ increased the growth rate of the spherulites and reduced the crystallization temperature from that in the base glass. Li⁺ and Na⁺ also reduced the temperature at which the microstructural transformation from spherulites to laths took place. Dielectric loss measurements suggest that the effect of the alkali additions on crystallization temperature is due to their concentration in grain boundaries, thereby lowering the viscosity of the remaining glass, and increasing mobility rates.