Reliability analysis in wind turbine engineering

Wind turbines are being designed in accordance with deterministic design rules. As wind turbines increase in size and power, these rules may not be adequate to ascertain a safe and well balanced design. Wind turbine industry can take advantage of the safety and reliability practices in e.g. aerospace, nuclear and offshore industry. These practices include System Reliability (such as FMECA, Event Sequence Analysis and Fault Tree Analysis) and Structural Reliability methods. The approach to introduce Probabilistic Safety Assessment (PSA) which incorporates these reliability methods into wind turbine engineering will be described. By means of two recent case studies, the applicability, benefits, and limitations of these methods will be illustrated.     

Iterative computational techniques in scattering based upon the integrated square error criterion

An iterative technique is developed to rigorously compute the electromagnetic time- and frequency-domain scattering problems. The method is based upon a wave-function expansion technique (this also includes the integral-representation techniques), in which the electromagnetic field equations and causality conditions are satisfied analytically, while the boundary conditions or the constitutive relations have to be satisfied in a computational manner. The latter is accomplished by an iterative minimization of the integrated square error. For the solution of an integral equation, it is shown how to obtain optimum convergence. Some numerical results pertaining to a number of representative problems illustrate the numerical advantages and disadvantages of the iterative method.     

New networks for perfect inversion and perfect reconstruction

The authors present a new network structure that realizes perfect inversion networks (PINs) and perfect reconstruction networks (PRNs). In some applications, such as transform source coders, it is important that the cascade of the forward and the inverse transform give the identity exactly (perfect inversion), although the coefficients and the intermediate results are quantized. In subband coders, for example, the split and merge filter banks should preferably have perfect reconstruction. It is advantageous if perfect reconstruction can be accomplished even when the coefficients and the intermediate results are quantized. The proposed network has a ladderlike shape and a predescribed symmetry between the forward and inverse network or between the split and merge bank. In some parts of this ladder network almost any function is allowed. Due to the prescribed symmetry, the property of perfect inversion or perfect reconstruction is structurally assured     

Plane Wave and Gaussian Beam Scattering by Long Dielectric Cylinders: 2.5D Simulations versus Measurements

For the development of millimeter wave imaging systems, it is important to be able to simulate some representative scattering configurations. Typically, Gaussian beams are used in active imaging systems. Since these beams only illuminate a spatially limited region, many objects can be treated as infinitely long 2D (in)homogenous cylinders. However, the incident Gaussian beams have a 3D character. Therefore, a dedicated 2.5D scattering simulator was developed. In this paper, simulation results obtained with this simulator are compared to measurements obtained from a bi-static microwave set-up and from a W-band millimeter wave set-up. Comparison of simulations and measurements proves that the 2.5D algorithm is a good simulation tool to study scattering of long inhomogeneous cylinders, illuminated by 3D plane waves or 3D Gaussian beams under different elevation angles.     

Removal of Microparticles by Ciliated Surfaces—an Experimental Study

Biological cilia are versatile hair‐like organelles that are very efficient in manipulating particles for, e.g., feeding, antifouling, and cell transport. Inspired by the versatility of cilia, this paper experimentally demonstrates active particle‐removal by self‐cleaning surfaces that are fully or partially covered with micromolded magnetic artificial cilia (MAC). Actuated by a rotating magnet, the MAC can perform a tilted conical motion, which leads to the removal of spherical particles of different sizes in water, as well as irregular‐shaped sand grains both in water and in air. These findings can contribute to the development of novel particulate manipulation and self‐cleaning/antifouling surfaces, which can be applied, e.g., to prevent fouling of (bio)sensors in lab‐on‐a‐chip devices, and to prevent biofouling of submerged surfaces such as marine sensors and water quality analyzers.     

2.5D Quantitative Millimeter Wave Imaging of a Hidden Object on a Simplified Human Body Model Using Value Picking Regularization

In this contribution, the authors provide a proof of principle for quantitative imaging of concealed objects on the human body using millimeter waves. A two-and-a-half-dimensional (2.5D) quantitative millimeter wave imaging algorithm is applied to reconstruct a hidden dielectric object on a clothed simplified human body model. At millimeter wave frequencies, the incident field is typically a fully three-dimensional (3D) Gaussian beam, illuminating only a limited spot on the body. Due to the large dimensions of the human body in terms of wavelengths, a 3D discretization is hardly feasible. Therefore, it is assumed that the electromagnetic properties of the body do not significantly change within the illuminated spot, along the longitudinal direction of a person. Hence, only the cross-section of a human body model is discretized. This 2.5D assumption however is still not sufficient to reduce the forward problem to a feasible size. Therefore, a priori knowledge on the illumination and on the scattering properties of the clothed human body is used to deduce a simplified model to describe the cross-section of the clothed human abdomen. The complex permittivity profile of a small dielectric object, hidden underneath clothing and representing some type of explosive, is reconstructed. The complex permittivity profiles of all other scatterers are assumed to be known. The presented quantitative inverse scattering algorithm is based on a Newton-type optimization, combined with an approximate line search and regularized by applying Stepwise Relaxed Value Picking regularization. The input data of the quantitative inverse scattering problem are synthetic scattering data since the authors are not aware of any amplitude and phase measurement data for concealed weapon detection yet made available to the inversion community at these high frequencies.     

A thyristor- and thermistor-based inrush current limiter for DC-link start-up

Large inrush currents can be harmful to equipment subjected to it and can also disturb other devices through voltage dips. Traditional ways of dealing with large inrush currents include the usage of parasitic elements in the supply, negative temperature coefficient (NTC) resistances and start-up relays. These solutions have several disadvantages among which are a non-negligible standby consumption, using a relay or a compromise between inrush current and load losses using an NTC. This paper proposes a solution which solves these disadvantages. The proposed circuit uses a semi-controlled bridge rectifier requiring a very low standby power consumption and is capable of withstanding grid interruptions. Moreover, it is designed to power on only when the load is not too heavy. This is accomplished by the use of positive temperature coefficient thermistors. The presented solution can be used for DC-link start-up circuits in applications with a power range from several watts up to a few kilowatts and it is particularly suitable for on-board battery chargers designed for electrical vehicles.     

Theoretical and numerical aspects of transmit SENSE

The ideas of parallel imaging techniques, designed to shorten the acquisition time by the simultaneous use of multiple receive coils, can be adapted for parallel transmission of a spatially selective multidimensional RF pulse. In analogy to data acquisition, a multidimensional RF pulse follows a certain trajectory in k-space. Shortening this trajectory shortens the pulse duration. The use of multiple transmit coils, each with its own time-dependent waveform and spatial sensitivity, compensates for the missing parts of k-space. This results in a maintained spatial definition of the pulse profile while its duration is reduced. This paper describes the basic equations of parallel transmission with arbitrarily shaped transmit coils ("Transmit SENSE") focusing on two-dimensional RF pulses. Results of numerical studies are presented demonstrating the theoretical feasibility of the approach.     

Scalable VLSI implementations for neural networks

This paper discusses research on scalable VLSI implementations of feed-forward and recurrent neural networks. These two families of networks are useful in a wide variety of important applications—classification tasks for feed-forward nets and optimization problems for recurrent nets—but their differences affect the way they should be built. We find that analog computation with digitally programmable weights works best for feed-forward networks, while stochastic processing takes advantage of the integrative nature of recurrent networks. We have shown early prototypes of these networks which compute at rates of 1–2 billion connections per second. These general-purpose neural building blocks can be coupled with an overall data transmission framework that is electronically reconfigured in a local manner to produce arbitrarily large, fault-tolerant networks.     

Industrial high‐rate (∼5 nm/s) deposited silicon nitride yielding high‐quality bulk and surface passivation under optimum anti‐reflection coating conditions

High‐quality surface and bulk passivation of crystalline silicon solar cells has been obtained under optimum anti‐reflection coating properties by silicon nitride (a‐SiN_x:H) deposited at very high deposition rates of ∼5 nm/s. These a‐SiN_x:H films were deposited using the expanding thermal plasma (ETP) technology under regular processing conditions in an inline industrial‐type reactor with a nominal throughput of 960 solar cells/hour. The low surface recombination velocities (50–70 cm/s) were obtained on p‐type silicon substrates (8·4 Ω cm resistivity) for as‐deposited and annealed films within the broad refractive index range of 1·9–2·4, which covers the optimum bulk passivation and anti‐reflection coating performance reached at a refractive index of ∼2·1. Copyright © 2005 John Wiley & Sons, Ltd.