34 resultados para 342.068
Resumo:
The annealing of ion implantation damage in silicon by rapid isothermal heating has been monitored by the time resolved reflectivity (TRR) method. This technique was applied simultaneously at a wavelength of 632. 8nm and also at 1152nm, where the optical absorption coefficient of silicon is less. The two wavelength method simplifies the interpretation of TRR results, extends the measurement depth and allows good resolution of the position of the interface between amorphous and crystalline silicon. The regrowth of amorphous layers in silicon, created by self implantation and implanted with electrically active impurities, was observed. Regrowth in rapid isothermal annealing occurs during the heating up stage of typical thermal cycles. Impurities such as B, P, and As increase the regrowth rate in a manner consistent with a vacancy model for regrowth. The maximum regrowth rate in boron implanted silicon is limited by the solid solubility.
Resumo:
A parallel processing network derived from Kanerva's associative memory theory Kanerva 1984 is shown to be able to train rapidly on connected speech data and recognize further speech data with a label error rate of 0·68%. This modified Kanerva model can be trained substantially faster than other networks with comparable pattern discrimination properties. Kanerva presented his theory of a self-propagating search in 1984, and showed theoretically that large-scale versions of his model would have powerful pattern matching properties. This paper describes how the design for the modified Kanerva model is derived from Kanerva's original theory. Several designs are tested to discover which form may be implemented fastest while still maintaining versatile recognition performance. A method is developed to deal with the time varying nature of the speech signal by recognizing static patterns together with a fixed quantity of contextual information. In order to recognize speech features in different contexts it is necessary for a network to be able to model disjoint pattern classes. This type of modelling cannot be performed by a single layer of links. Network research was once held back by the inability of single-layer networks to solve this sort of problem, and the lack of a training algorithm for multi-layer networks. Rumelhart, Hinton & Williams 1985 provided one solution by demonstrating the "back propagation" training algorithm for multi-layer networks. A second alternative is used in the modified Kanerva model. A non-linear fixed transformation maps the pattern space into a space of higher dimensionality in which the speech features are linearly separable. A single-layer network may then be used to perform the recognition. The advantage of this solution over the other using multi-layer networks lies in the greater power and speed of the single-layer network training algorithm. © 1989.
Resumo:
This paper discusses the inverter ratings of Brushless Doubly-Fed Machine (BDFM) adjustable speed drive (ASD) or generator (ASG) systems. Based on the per phase equivalent circuit model, the ratings of the two inverters in a bidirectional converter are evaluated individually. An approach to minimise the total inverter rating is presented, taking into account power factor constraints of the power grid. The effects of speed deviation and control winding excitation on the inverter ratings are discussed. Predictions of inverter ratings are presented with experimental verification. A design example is also provided in which the total inverter rating is minimised for a practical BDFM based ASG system. © 2005 IEEE.
Resumo:
A compact trench-gate IGBT model that captures MOS-side carrier injection is developed. The model retains the simplicity of a one-dimensional solution to the ambipolar diffusion equation, but at the same time captures MOS-side carrier injection and its effects on steady-state carrier distribution in the drift region and on switching waveforms. © 2007 IEEE.