Soft learning vector quantization and clustering algorithms based on non-Euclidean norms: multinorm algorithms

This paper presents the development of soft clustering and learning vector quantization (LVQ) algorithms that rely on multiple weighted norms to measure the distance between the feature vectors and their prototypes. Clustering and LVQ are formulated in this paper as the minimization of a reformulation function that employs distinct weighted norms to measure the distance between each of the prototypes and the feature vectors under a set of equality constraints imposed on the weight matrices. Fuzzy LVQ and clustering algorithms are obtained as special cases of the proposed formulation. The resulting clustering algorithm is evaluated and benchmarked on three data sets that differ in terms of the data structure and the dimensionality of the feature vectors. This experimental evaluation indicates that the proposed multinorm algorithm outperforms algorithms employing the Euclidean norm as well as existing clustering algorithms employing weighted norms.     

A thermodynamic analysis of a simple open-flow solar regenerator

Simple expressions have been developed for important variables of the flow in an open-flow solar regenerator (with applications in absorption cooling systems) through a first- and second-law analysis of the system. In particular, expressions for the regenerator figure of merit, the exergetic or rational efficiency and the entropy production rate have been given and applied to some available data with reasonably good results.     

Examination of heat transfer correlations and a model for flow boiling of R134a in small diameter tubes

Analysis of various existing correlations including a three-zone evaporation model is made using a comparison with recent experimental results obtained in this study. Flow boiling heat transfer experiments were conducted with two stainless steel tubes of internal diameter 4.26 mm and 2.01 mm. The working fluid was R134a and parameters were varied in the range: mass flux 100–500 kg/m²s; pressure 8–12 bar; quality up to 0.9; heat flux 13–150 kW/m². The local heat transfer coefficient was independent of vapour quality when this was less than about 40–50% in the 4.26 mm tube and 20–30% in the 2.01 mm tube. Local transient dryout was deduced when the quality was above these values. Furthermore, at high heat flux values the heat transfer coefficient decreased with vapour quality for the entire quality range indicating early occurrence of dryout.Existing correlations, which are based on large tube boiling processes, do not predict the present small diameter data to a satisfactory degree. A better agreement is observed with the recent, state-of-the-art, three-zone evaporation model. However, the model does not predict the effect of diameter and the partial dryout. Nevertheless, the observation suggests that the flow pattern based modelling approach performs at least as well as empirical correlations that are based on macroscale modelling. Aspects of the model that need further consideration are also proposed in this study.     

Quantifying motion in video recordings of neonatal seizures by robust motion trackers based on block motion models

This paper introduces a methodology for the development of robust motion trackers for video based on block motion models. According to this methodology, the motion of a site between two successive frames is estimated by minimizing an error function defined in terms of the intensities at these frames. The proposed methodology is used to develop robust motion trackers that rely on fractional block motion models. The motion trackers developed in this paper are utilized to extract motor activity signals from video recordings of neonatal seizures. The experimental results reveal that the proposed motion trackers are more accurate and reliable than existing motion tracking methods relying on pure translation and affine block motion models.     

Improving the accuracy and reliability of motion tracking methods used for extracting temporal motor activity signals from video recordings of neonatal seizures

This paper presents an approach for improving the accuracy and reliability of motion tracking methods developed for video based on block motion models. This approach estimates the displacement of a block of pixels between two successive frames by minimizing an error function defined in terms of the pixel intensities at these frames. The minimization problem is made analytically tractable by approximating the error function using a second-order Taylor expansion. The improved reliability of the proposed method is illustrated by its application in the extraction of temporal motor activity signals from video recordings of neonatal seizures.     

Heat and mass transfer in air-fed pressurised suits

Air-fed pressurised suits are used to protect workers against contamination and hazardous environments. The specific application here is the necessity for regular clean-up maintenance within the torus chamber of fusion reactors. The current design of suiting has been developed empirically. It is, therefore, very desirable to formulate a thermo-fluids model, which will be able to define optimum designs and operating parameters. Two factors indicate that the modelling should be as comprehensive as possible. Firstly, the overall thermo-fluids problem is three-dimensional and includes mass as well as heat transfer. The fluid field is complex, bounded on one side by the human body and on the other by what may be distensible, porous and multi-layer clothing.In this paper, we report firstly the modelling necessary for the additional mass and heat transport processes. This involves the use of Fick’s and Fourier’s laws and conjugate heat transfer. The results of an initial validation study are presented. Temperatures at the outlet of the suits were obtained experimentally and compared with those predicted by the overall CFD model. Realistic three-dimensional geometries were used for the suit and human body. Calculations were for turbulent flow with single- and two-component (species) models.     

Quantum neural networks (QNNs): inherently fuzzy feedforward neuralnetworks

This paper introduces quantum neural networks (QNNs), a class of feedforward neural networks (FFNNs) inherently capable of estimating the structure of a feature space in the form of fuzzy sets. The hidden units of these networks develop quantized representations of the sample information provided by the training data set in various graded levels of certainty. Unlike other approaches attempting to merge fuzzy logic and neural networks, QNNs can be used in pattern classification problems without any restricting assumptions such as the availability of a priori knowledge or desired membership profile, convexity of classes, a limited number of classes, etc. Experimental results presented here show that QNNs are capable of recognizing structures in data, a property that conventional FFNNs with sigmoidal hidden units lack.     

ELECTROHYDRODYNAMIC ENHANCEMENT OF NUCLEATE BOILING HEAT TRANSFER IN HEAT EXCHANGERS

The effect of a radial d.c. electric field on nucleate boiling heat transfer was investigated experimentally using a single-tube shell/tube heat exchanger. A dielectric liquid (Freon R. 114) was used in the shell and the tube was heated by circulating water through it.

It was found that the application of a sufficiently intense electric field to the boiling heat transfer surface resulted both in the elimination of boiling hysteresis and the enhancement of heat transfer for the range of superheat studied. An application of approximately 20 kV was more than sufficient to eliminate the hysteresis.

It was also observed that there appeared to be a transition between two situations. At lower superheat there is an appreciable improvement in the heat transfer coefficient due to both initial (0 to 10 kV) and incremental (10 to 20 kV) voltage increases. At higher superheat, however, the greater improvement (about four times) is obtained with the initial voltage application