Search

In this paper, we consider an orthogonal frequency-division multiplexing (OFDM) communication system over a power-line channel (PLC) that exhibits cyclic short-time variations. In order to follow the cyclic channel variation, a blind estimation method is proposed. Simulations on an indoor PLC network show that the proposed estimation method is suited for following the short-time variation of the channel, achieving a better bit-error rate (BER) than a base pilot-aided method. Index Terms--Blind estimation, nonlinear equalization, orthogonal frequency-division multiplexing (OFDM), power-line communication (PLC), time-variant channel.

In this paper, we analyze the effects of power lines parameters' uncertainties on the efficiency of a PowerLine Communication (PLC) system. An estimate of the upper and lower bounds' system response due to the uncertainties is given. The analyzed system is schematized as a connection of multiport devices described by their scattering parameter matrices and an adjoint method for sensitivity calculations, based on the use of a wavelet expansion, is used to determine the bounds. The wavelet based sensitivity analysis allows a direct determination of the time domain sensitivity avoiding computationally expensive convolution integral calculations; in addition a fast and accurate bounds calculation is obtained avoiding statistical Monte Carlo simulations. The bounds determination allows the estimate of noteworthy quantities for the tuning of currently used modulation schemes for PLC such as orthogonal frequency-division modulation. Index Terms--Power lines communication, sensitivity analysis, uncertain parameters.

A time-domain wavelet-based method for the analysis of transmission lines (TLs) characterized by the presence of skin and proximity effects is presented. The method is based on the time domain TL equations, expressed in terms of transient resistance and convolution operator. A formulation of the convolution in the wavelet domain is presented here, and is used to expand the TL equations in the wavelet domain. In this way, the TL equations, properly discretized in the space variable, are reduced to an algebraic system characterized by a sparse block matrix whose blocks are easily calculated. The solution of the system is conveniently performed using standard techniques, avoiding step by step procedures. The method can be applied to linear and nonlinear problems, characterized by conductors of different shapes. The results obtained are compared with standard techniques, showing both accuracy and CPU time reduction. Index Terms--Interconnects, proximity effect, skin effect, wavelet expansion.

A novel wavelet-based technique is presented for the extraction of two-port equivalents of common printed circuit board (PCB) discontinuities. The wavelet-transformed scattering parameters of the discontinuity can be included in a wavelet equivalent of TEM wave propagation paths along the PCB in order to obtain an overall equivalent of the whole structure under analysis. The wavelet representation minimizes the CPU time and computer storage requirements while maintaining excellent accuracy, thus, it proves to be a very useful modeling tool, especially in the design stage. Index Terms--Multiresolution analysis, printed circuits, wavelet transforms.

A boundary condition for the truncation of the grid in finite-difference time-domain (FDTD) formulations is presented. The proposed technique restricts the computational domain by truncating the FDTD grid close to the scattering objects. By using the equivalence theorem, the proposed procedure yields a coupling between the region outside the boundary (exterior problem) and inside the boundary (interior problem). A resulting hybrid FDTD-method of moments technique is used to obtain a numerical solution of the coupled problems. Numerical applications demonstrate the validity of the proposed procedure. Index Terms--Equivalence theorem, finite-difference time domain (FDTD), hybrid methods, method of moments (MoM).

A new numerical method for the time domain solution of Maxwell's equations in linear media is proposed in this paper. The field quantities derived from the spatial discretization of Maxwell's equations are expanded in the time domain by wavelets on the interval. This choice yields a new arrangement of the unknowns into a matrix (instead of the usual vector) and transforms the differential equations in time in a Sylvester matrix equation. The memory requirements are proportional to the number of spatial unknowns and the time evolution of the space quantities is obtained with better accuracy than in conventional marching-on-time techniques. Index Terms--Time domain numerical techniques, wavelet expansion.

An electromagnetic system that operates both as a rail and as a coil launcher is proposed, and its performance is analyzed. The device has a composite stator that consists of two rails properly slotted in order to allow the presence of a system of coils. The armature consists of a conductive slab sliding between the rails. Another system of slots is present in the armature where a system of short-circuited coils is placed. A source of constant voltage is connected to the rails and the current flowing in the armature through the sliding contacts produces a thrust force on it. If a proper system of currents able to produce a traveling wave of flux density in the region between the rails is used to feed the barrels, the induced currents that flow in the armature's short-circuited coils produce a further thrust force on the armature itself. The analysis of the behavior of this launcher is performed via a computer code based on an integral formulation and a tool that analyzes the phenomena related to the velocity skin effect; the interactions between the two systems of currents on the armature and the currents on the stator (rails and barrels) are investigated. The thermal behavior of the device has been taken into account by a simple adiabatic model since the short operating time allows one to neglect heat diffusion in the conductive parts of the system. A comparison between the velocities, respectively obtained by separately feeding rails and coils and the system with the combined feeding, has been performed. Preliminary results of the analysis show that this device, because of the presence of two systems of thrust forces acting on the armature, can be successfully used in the acceleration of heavy masses at relatively high velocities, the maximum achievable speed being limited by thermal and mechanical stresses. Index Terms--Coil launchers, induction launchers, rail launchers, sliding contacts.

This paper deals with a numerical modeling of the phenomena involved in sliding contacts in homopolar disk generators. Because of the velocity skin effect, the current distribution in sliding contacts is uneven and mainly concentrated in the rear part of the brushes. This can cause melting of the portion of the brush in the devices characterized by high currents. The knowledge of the distribution of the currents is important in order to avoid melting. An integral equivalent network formulation previously developed for the analysis of moving conductors has been modified to take into account the presence of sliding contacts. The numerical model presented is used to analyze the current distribution in a homopolar disk generator. Index Terms--Homopolar generator, integral equations, moving conductors, sliding contacts.

In this paper, a finite-difference time-domain (FDTD) model of an electrostatic discharge (ESD) event is developed. Analytical expressions for the field radiated during the ESD discharge phase have been determined to test the FDTD model of the strike arc. In order to take into account the electromagnetic field penetration through shielding structures, the conductive panels are efficiently modeled in the FDTD by the impedance network boundary conditions (INBCs). The FDTD-INBCs method avoids the huge amount of cells needed to model accurately the penetration in the traditional FDTD algorithm based on the utilization of the regular Yee grid. The method is applied to the analysis of ESD events in some configurations. Index Terms--Electrostatic discharge, finite-difference time-domain method, impedance network boundary conditions.

In this paper, a new method for the numerical solution of two-dimensional electromagnetic problems by wavelets on the interval and equivalence theorem is proposed. By adopting the equivalence theorem, the computational domain is restricted to the conductive regions while expanding the space variables by wavelets on the interval we obtain a storage of the unknowns in a matrix form instead of the usual vector form, yielding a savings of memory storage and CPU times. Index Terms--Equivalence theorem, wavelet expansion.

A hybrid method for nonlinear electromagnetic analysis that combines the Method of Moments (MOM) and the Finite Element Method (FEM) is presented. By using the equivalence principle, an interior and an exterior problem are used to study the actual problem. The MOM is used for the solution of the exterior problem where equivalent surface currents replace the solid bodies and generate the fields in the free space. The FEM is used to solve the interior problem defined into the solid bodies with given conditions at their boundaries. Considering that the actual fields are given by the sum of the fields in the two problems, a nonlinear matrix equation for the problem solution is obtained by forcing the continuity of the actual tangential electric and magnetic fields on the boundary of the nonlinear solid bodies. Two-dimensional numerical results are used to show the characteristics of the method by the analysis of test cases and by comparison with other numerical methods. Index terms - Eddy currents, nonlinear problems, hybrid formulation

A formulation of the method of moments for the analysis of nonlinear low frequency problems is presented. The examined system is subdivided in elementary volume elements whose nonlinearity is taken into account by means of the magnetizations M. Considering the nonlinear constitutive equation H = H(B) a nonlinear algebraic system of equations is obtained. By combining Gauss law and current continuity at the boundaries among regions with different conductivity a set of equations is obtained. Writing Ohm's law inside the conductive regions another integral equation set that allows the determination of the conduction current and surface charge unknowns is obtained. The method of moments is then applied to this system of equations. The use of pulse functions as subsection bases allows a quick matrix set up especially when regular volume shapes are selected. Calculated results are compared with results obtained with other methods. Index terms - Eddy currents, nonlinear problems, MOM formulation

The influence of the choice of discrete spaces on integral equation formulations for low frequency electromagnetics is studied by comparing two different approaches, where the other one employs standard finite element spaces and the other is based on constructing equivalent electric networks.

In this paper a frequency domain formulation of the method of moments taking into account the presence of ferromagnetic materials is presented. By combining the electric field definition, Ohm's law and Lorentz gauge inside every volume element into which the system is subdivided, a linear algebraic system of integral equations is obtained. Considering the nonlinear constitutive equation H=H(B) and the magnetization M a nonlinear algebraic system of equations is obtained. The unknowns of these systems are the time Fourier transforms of magnetizations and conduction currents. The solution of these systems for a finite number of harmonics gives the frequency domain solution of the problem. The use of pulse functions as subsection bases allows a quick matrix set up especially when regular volume shapes are selected. Calculated results are compared with results obtained with other methods.

In this paper an integral formulation for the analysis of electromagnetic fields distribution in systems with bodies in motion is presented. This formulation is based on the subdivision of conductive and ferromagnetic regions in elementary volumes in which a uniform distribution of current density J and magnetization M is assumed. By integrating in every volume Ohm's law in term of the magnetic vector potential produced by the conduction and magnetization currents we obtain an equivalent electric network whose lumped parameters vary during every time step interval as the moving parts of the system change their positions. The proposed model has been tested by comparison with analytical solutions.

In this paper a time domain formulation of the method of moments for the analysis of nonlinear materials is presented. The magnetization vector M and the nonlinear constitutive equation H=H(B) are used in order to express the magnetic vector potential as a function of conduction currents only. The Lorentz gauge is used in order to express the scalar potential as a function of the vector potential. By combining the electric field definition and Ohm's law an integral equation in the conduction current unknowns is obtained. The method of moments is then applied to this equation. The use of pulse functions as subsection bases allows a quick matrix set up especially when regular volume shapes are selected. Calculated results are compared with results obtained with other methods relating to benchmark problems.

In this paper a quasi-analytical procedure for the analysis of electromagnetic field distributions in bounded domains relating to linear materials is presented. The procedure allows the determination of analytical expressions for the magnetic fields by means of the superposition of analytical solutions relating to partially bounded domains i.e. half spaces or axially infinite geometries. The presence of ferromagnetic materials is taken into account by means of the determination of equivalent magnetization currents. The procedure can be used for the analysis of different geometries, and an application to the analysis of the induced voltage in a Non Destructive Testing system is shown.

The three-dimensional integration of the Biot-Savart's law for conductors in annular sector shape, in which only a constant distributed, theta-directed current flows, is reported. The obtained expressions allow a quick and accurate evaluation of the components of the vector potential of the stationary magnetic field and of the magnetic induction due to this type of conductor, in iron free-media. The relations presented in this paper, together with similar expressions, valid for other shapes of conductors and current distributions, can be usefully employed to evaluate the magnetic fields in complex, linear and iron-free structures, when the examined geometry can be well approximated with a series of current elements of elementary shape, of which solution of Laplace's equation is known in closed form.

A data-based estimation of the wind-power curve in wind turbines may be a challenging task due to the presence of anomalous data, possibly due to wrong sensor reads, operation halts, malfunctions or other. In this study, the authors describe a data-based procedure to build a robust and accurate estimate of the wind-power curve. In particular, they combine a joint clustering procedure, where both the wind speeds and the power data are clustered, with an Earth Mover Distance-based Extreme Learning Machine algorithm to filter out data that poorly contribute to explain the unknown curve. After estimating the cut-in and the rated speed, they use a radial basis function neural network to fit the filtered data and obtain the curve estimate. They extensively compared the proposed procedure against other conventional methodologies over measured data of nine turbines, to assess and discuss its performance. [ABSTRACT FROM AUTHOR]

The Remote Field Effect is used extensively for testing of thick tubular products where deep penetration into (primarily ferromagnetic) materials is required (1). The testing apparatus consists of two coils, one generating a magnetic field, the other is used as a pickup coil. The distance between the two coils is normally large (of the order of two coil diameters). The operating frequency is low, normally below 100 Hz. The method has been assumed in the past to rely on a special effect, and therefore the name Remote Field Effect. One of the most remarkable aspect of the method is its equal sensitivity to inner and outer defects in thick ferromagnetic tubes. This effect could not easily be explained by direct induction and therefore, many models have been proposed. The common model used is that of a wave propagating from the inside of the material to the outside and then back again (2). We propose a theoretical model that shows the effect to be merely that of induction at large distances (3-5). The field equations are solved directly using first an analytic Bessel function approach and then using an integral approach. Both of these models confirm the basic results associated with the remote field effect and also include the velocity of the coils. The models described here are general and applicable to other induction applications.