Compressive Sensing For Sparsely Excited Speech Signals

Compressive sensing (CS) has been proposed for signals with sparsity in a linear transform domain. We explore a signal dependent unknown linear transform, namely the impulse response matrix operating on a sparse excitation, as in the linear model of speech production, for recovering compressive sensed speech. Since the linear transform is signal dependent and unknown, unlike the standard CS formulation, a codebook of transfer functions is proposed in a matching pursuit (MP) framework for CS recovery. It is found that MP is efficient and effective to recover CS encoded speech as well as jointly estimate the linear model. Moderate number of CS measurements and low order sparsity estimate will result in MP converge to the same linear transform as direct VQ of the LP vector derived from the original signal. There is also high positive correlation between signal domain approximation and CS measurement domain approximation for a large variety of speech spectra.

Self-excited triggered vacuum gap

The self-triggering action of the TVG under DC conditions using the same power source as the main circuit is described. Because of the low impedance of the power source compared with the thyratron pulse generator, the TVG can be triggered with voltages as low as 85 V. It is found that circuit conditions influence the minimum stable arc currents that can be sustained.

Vibrations of a periodic rail-sleeper system excited by an oscillating stationary transverse force

The rail-sleeper system is idealized as an infinite, periodic beam-mass system. Use is made of the periodicity principle for the semi-infinite halves on either side of the forcing point for evaluation of the wave propagation constants and the corresponding modal vectors. It is shown that the spread of acceleration away from the forcing point depends primarily upon one of the wave propagation constants. However, all the four modal vectors (two for the left-hand side and two for the right-hand side) determine the driving point impedance of the rail-sleeper system, which in combination with the driving point impedance of the wheel (which is adopted from the preceding companion paper) determines the forces generated by combined surface roughness and the resultant accelerations. The compound one-third octave acceleration levels generated by typical roughness spectra are generally of the same order as the observed levels.

Some Studies on Hydrogen-Oxygen Diffusion Flame

The opposed-jet diffusion flame has been considered with four step reaction kinetics for hydrogenoxygen system. The studies have revealed that the flame broadening reduces and maximum temperature increases as pressure increases. The relative importance of different reaction steps have been brought out in different regions (unstable, near extinction and equilibrium). The present studies have also led to the deduction of the oveall reaction rate constants of an equivalent single step reaction using matching of a certain overall set of parameters for four step reaction scheme and equivalent single step reaction.

Ab initio molecular dynamics simulations of excited hydrogen halides and methyl halides

Using excited-state ab initio molecular dynamics simulations employing the complete-active-space self-consistent-field approach, we study the mechanism of photodissociation in terms of time evolution of structure, kinetic energy, charges and potential energy for the first excited state of hydrogen halides and methyl halides. Although the hydrogen halides and methyl halides are similar in the photodissociation mechanism, their dynamics are slightly different. The presence of the methyl group causes delay in photodissociation as compared to hydrogen halides.

Low-temperature Er3+ emission in Ge-S-Ga glasses excited by host absorption

The photoluminescence (PL) of a series of (GeS2)(80)(Ga2S3)(20) glasses doped with different amounts of Er (0.17, 0.35, 0.52, 1.05 and 1.39 at.%) at 77 and 4.2 K has been studied. The influence of the temperature on the emission cross-section of the PL bands at -> 1540, 980 and 820 nm under host excitation has been defined. A quenching effect of the host photoluminescence has been established from the compositional dependence of the PL intensity. It has been found that the present Er3+-doped Ge-S-Ga glasses posses PL lifetime values about 3.25 ms. (C) 2007 Elsevier B.V. All rights reserved.

Prediction of flame liftoff height of diffusion/partially premixed jet flames and modeling of mild combustion burners

In this article, a new flame extinction model based on the k/epsilon turbulence time scale concept is proposed to predict the flame liftoff heights over a wide range of coflow temperature and O-2 mass fraction of the coflow. The flame is assumed to be quenched, when the fluid time scale is less than the chemical time scale ( Da < 1). The chemical time scale is derived as a function of temperature, oxidizer mass fraction, fuel dilution, velocity of the jet and fuel type. The present extinction model has been tested for a variety of conditions: ( a) ambient coflow conditions ( 1 atm and 300 K) for propane, methane and hydrogen jet flames, ( b) highly preheated coflow, and ( c) high temperature and low oxidizer concentration coflow. Predicted flame liftoff heights of jet diffusion and partially premixed flames are in excellent agreement with the experimental data for all the simulated conditions and fuels. It is observed that flame stabilization occurs at a point near the stoichiometric mixture fraction surface, where the local flow velocity is equal to the local flame propagation speed. The present method is used to determine the chemical time scale for the conditions existing in the mild/ flameless combustion burners investigated by the authors earlier. This model has successfully predicted the initial premixing of the fuel with combustion products before the combustion reaction initiates. It has been inferred from these numerical simulations that fuel injection is followed by intense premixing with hot combustion products in the primary zone and combustion reaction follows further downstream. Reaction rate contours suggest that reaction takes place over a large volume and the magnitude of the combustion reaction is lower compared to the conventional combustion mode. The appearance of attached flames in the mild combustion burners at low thermal inputs is also predicted, which is due to lower average jet velocity and larger residence times in the near injection zone.

Puzzles of excited charm meson masses

We attempt a comprehensive analysis of the low lying charm meson states which present several puzzles, including the poor determination of masses of several non-strange excited mesons. We use the well-determined masses of the ground states and the strange first excited states to 'predict' the mass of the non-strange first excited state in the framework of heavy hadron chiral perturbation theory, an approach that is complementary to the well-known analysis of Mehen and Springer. This approach points to values for the masses of these states that are smaller than the experimental determinations. We provide a critical assessment of these mass measurements and point out the need for new experimental information. (c) 2007 Elsevier B.V. All rights reserved.

Theory of laminar flame propagation with non-normal diffusion.

A study has been made of the problem of steady, one-dimensional, laminar flame propagation in premixed gases, with the Lewis number differing from (and equal to) unity. Analytical solutions, using the method of matched asymptotic expansions, have been obtained for large activation energies. Numerical solutions have been obtained for a wide range of the reduced activation temperature parameter (n {geometrically equal to} E/RTb), and the Lewis number δ. The studies reveal that the flame speed eigenvalue is linear in Lewis number for first order and quadratic in Lewis number for second order reactions. For a quick determination of flame speeds, with reasonable accuracy, a simple rule, expressing the flame speed eigenvalue as a function of the Lewis number and the centroid of the reaction rate function, is proposed. Comparisons have been made with some of the earlier works, for both first and second order reactions.