3-D high frequency computation of transformer R, L parameters

High frequency electrical machine modeling based on lumped parameter equivalent circuit requires that eddy currents effects be taken into account for R and L coefficient derivation. Substituting hysteretic materials (complex permeability) to conductive materials makes it possible to accurately approximate proximity losses and shielding effect while considering realistic meshes for finite element methods computation. Three-dimensional magnetodynamic simulations of transformer test-cases using anisotropic complex permeability for magnetic core and windings have shown a good agreement with the standard model results.     

Development and numerical validation of a new model for walls with phase change materials implemented in TRNSYS

This paper presents a model of a wall with variable properties dedicated to modelling phase change materials (PCMs) in building envelopes. The model is implemented in the TRNSYS simulation tool and referred to as Type 3258. The 1-D conduction heat transfer equation is solved using an explicit finite-difference method coupled with an enthalpy method to consider the variable PCM thermal capacity. This model includes temperature-dependent thermal conductivity and PCM-specific effects like hysteresis and supercooling. The stability conditions are discussed and the algorithm implemented in TRNSYS is described. A numerical validation performed on wall test cases proposed by the International Energy Agency is presented, showing that the developed model is in agreement with reference models. The paper also discusses the impact of temporal and spatial discretization on the model performance. Modelling problems encountered when using an effective heat capacity method (compared to an enthalpy method) and when representing supercooling are also discussed.     

Using a MODFLOW grid, generated with GMS, to solve a transport problem with TOUGH2 in complex geological environments: The intertidal deposits of the Venetian Lagoon

The tides of the Venetian Lagoon generally vary between −0.5 and +0.7 m asl. Occasionally, they may reach a maximum of 1.5 m (acqua alta) and a minimum of −0.8 m asl (acqua bassa). Intertidal areas, called “barene,” exist all along the coast of the Lagoon. These areas are characterized by canals that concentrate the flow of water (and the deposition of sands) during the rising and waning of the tides, and that inundate and drain the vegetated areas found between canals (where organic-rich clays are deposited). Therefore, since the area is subject to subsidence, in time, sand dykes (the original canals) become juxtaposed to clayey dykes (the original vegetated areas). In addition, the sands form a continuous hydrogeologic network within the clays, very similar to that of a vascular system that effectively drains the whole “barena” deposits. In order to be effective, measures for monitoring, confining, or remediating the transport of pollutants through this kind of environment must explicitly take into account the geologic complexity. The same complexity must be included in the numerical models that support remediation efforts. At the moment, there appears to be no off-the-shelf graphical interface that is able to manage such complexity for TOUGH2. To attempt to solve this problem we have used a calibrated USGS-MODFLOW model, of the barena of “Passo a Campalto” in the Venetian Lagoon, developed with the GMS graphical interface. The model is made of 42 layers, which, apart from the first layer, are 0.5 m thick; the first layer has the thickness distribution of a dump found on top of the barena deposit at Passo a Campalto. Each layer consists of 100×60 square cells, for a total of 252,000 cells, only about half of which are active. Using a FORTRAN routine, we translate this grid, with all the hydrogeologic boundary conditions, into a TOUGH2 input file, and we provide the additional necessary information for running a TOUGH2 simulation. The results are promising, in that we were able to produce TOUGH2 grids with very complex geology and to run the models with success. For visualization, the results can be imported back into GMS as 3D scatter point sets, or they can be plotted with any adequate plotting software such as MatLab. Developing conceptual and numerical models with an elaborate graphical interface such as GMS effectively allows setting up complex problems while concentrating on their physics.     

Speech recognition in living rooms: Integrated speech enhancement and recognition system based on spatial,spectral and temporal modeling of sounds

Research on noise robust speech recognition has mainly focused on dealing with relatively stationary noise that may differ from the noise conditions in most living environments. In this paper, we introduce a recognition system that can recognize speech in the presence of multiple rapidly time-varying noise sources as found in a typical family living room. To deal with such severe noise conditions, our recognition system exploits all available information about speech and noise; that is spatial (directional), spectral and temporal information. This is realized with a model-based speech enhancement pre-processor, which consists of two complementary elements, a multi-channel speech–noise separation method that exploits spatial and spectral information, followed by a single channel enhancement algorithm that uses the long-term temporal characteristics of speech obtained from clean speech examples. Moreover, to compensate for any mismatch that may remain between the enhanced speech and the acoustic model, our system employs an adaptation technique that combines conventional maximum likelihood linear regression with the dynamic adaptive compensation of the variance of the Gaussians of the acoustic model. Our proposed system approaches human performance levels by greatly improving the audible quality of speech and substantially improving the keyword recognition accuracy.     

Impact of bilayer character on High K gate stack dielectrics breakdown obtained by conductive atomic force microscopy

The time to breakdown distribution of bilayer gate stack dielectrics is measured at nanometric scale using an atomic force microscope in conduction mode under ultra-high vacuum. The bilayer consists of a SiON interfacial layer and a HfSiON High-K layer. Thanks to the small tip/sample contact area the time to breakdown distribution of the single interfacial layer is measured separately. It is found that the Weibull parameters of the Interfacial layer distribution are the same as those of the high percentile part of the bilayer bimodal distribution. This experimentally confirms the validity of former dielectric breakdown model assumptions. Considering the fields in each layer an accurate evaluation of acceleration factors and voltage scaling of the bimodal distribution are given.     

Weight Distribution of the Hermitian Forms Codes

 We recall the construction of the hermitian-forms codes and some results about hermitians forms and exponential sums. With the help of these results, we give the weight distribution of the hermitian-forms codes and give some examples for which experimental computations have been made. Received: September 8, 1993; revised version: October 4, 1995     

Cluster-based dynamic variance adaptation for interconnecting speech enhancement pre-processor and speech recognizer

A conventional approach to noise robust speech recognition consists of employing a speech enhancement pre-processor prior to recognition. However, such a pre-processor usually introduces artifacts that limit recognition performance improvement. In this paper we discuss a framework for improving the interconnection between speech enhancement pre-processors and a recognizer. The framework relies on recent proposals for increasing robustness by replacing the point estimate of the enhanced features with a distribution with a dynamic (i.e. time varying) feature variance. We have recently proposed a model for the dynamic feature variance consisting of a dynamic feature variance root obtained from the pre-processor, which is multiplied by a weight representing the pre-processor uncertainty, and that uses adaptation data to optimize the pre-processor uncertainty weight. The formulation of the method is general and could be used with any speech enhancement pre-processor. However, we observed that in case of noise reduction based on spectral subtraction or related approaches, adaptation could fail because the proposed model is weak at representing well the actual dynamic feature variance. The dynamic feature variance changes according to the level of speech sound, which varies with the HMM states. Therefore, we propose improving the model by introducing HMM state dependency. We achieve this by using a cluster-based representation, i.e. the Gaussians of the acoustic model are grouped into clusters and a different pre-processor uncertainty weight is associated with each cluster. Experiments with various pre-processors and recognition tasks prove the generality of the proposed integration scheme and show that the proposed extension improves the performance with various speech enhancement pre-processors.