ARMA Time Series Modeling: an Effective Method

The ability to generate rational models of time series plays an important role in such applications as adaptive filtering, spectral estimation, digital control, array processing, and forecasting. A method for effecting an autoregressive moving average (ARMA) model estimate is presented which possesses a number of admirable properties: 1) it has an elegant algebraic structure, 2) its modeling performance in spectral estimation applications has been empirically found to typically exceed that of such contemporary techniques as the periodogram, the Burg method, and the Box-Jenkins method on a variety of problems, 3) it is implementable by computationally efficient algorithms, and 4) it is based on pseudomaximum likelihood concepts. Taken in combination, these properties mark this method as being an effective tool in challenging applications requiring high modeling performance in a real time setting.     

Test Results with an Experimental Direction-Finding System

A direction-finding system was built that can simultaneously estimate the arrival angles of two incident signals. It is conceptually straightforward to extend the system to cases with greater numbers of signals. The system is designed to work with uncorrelated signals whose form is unknown, e.g., two communications signals residing in the same bandwidth. Pulsed as well as continuous signals can be handled. The technique involves the determination of the received signal's correlation matrix as an intermediate step. Pattern range tests were performed to determine angle estimate bias and accuracy.     

Investigations of Stability & Dynamic Performances of a Current-Injected Regulator

The stability and dynamic performances of a buck/boost regulator employing a current-injected control are examined. Small-signal models for the power state, the multiloop error processor, and the duty-cycle pulse modulator are developed. The error-processor model which incorporates the current-injected loop, the dc loop, and the compensation network permits evaluation of the effects of each individual control loop and their combined efforts toward shaping the performance characteristics of the closed-loop system. Comparisons are made between this modeling approach and earlier approaches. Some important yet subtle dissimilarities are discussed. This model predicts the constant-frequency 50-percent duty-cycle instability which is inherent to the current-injected control.     

Transformer Induced Instability of the Series Resonant Converter

It is shown that the common series resonant power converter is subject to a low frequency oscillation that can lead to the loss of cyclic stability. This oscillation is caused by a low frequency resonant circuit formed by the normal L and C components in series with the magnetizing inductance of the output transformer. Three methods for eliminating this oscillation are presented and analyzed. One of these methods requires a change in the circuit topology during the resonance cycle. This requires a new set of steady state equations which are derived and presented in a normalized form. Experimental results are included which demonstrate the nature of the low frequency oscillation before cyclic stability is lost.     

Plasma-physical aspects of the solar cycle

Mass motions below the photosphere drive the solar cycle which is associated with variations in the magnetic field structure and accompanying phenomena. In addition to semi-empirical models, dynamo theories have been used to explain the solar cycle. The emergence of magnetic field generated by these mechanisms and its expansion into the corona involves many plasma physical processes. Magnetic buoyancy aids the expulsion of magnetic flux. The corona may respond dynamically or by continually adjusting to a quasi-static force-free or pressure-balanced equilibrium. The formation and disruption of current sheets is significant for the overall structure of the coronal magnetic field and the physics of quiescent prominences. The corona has a fine structure consisting of magnetic loops. The structure and stability of these are important as they are one of the underlying elements which make up the corona.     

Simulation of Permanent Magnet Generator/Rectifier Combination

Permanent magnet generators are often used as the pilot exciters of three-stage generating units to provide an efficient and reliable source of electrical power. No external power supply is then necessary, and the problems associated with brushwear and with arcing at the rubbing contacts are eliminated. A technique is presented here by which the performance of a permanent magnet generator, when combined with a rectifier-fed d.c. load at its output terminals, may be accurately and efficiently computed. The machine model takes into account the magnetization characteristic of the permanent magnet rotor and involves only parameters which can be determined from terminal measurements on the machine at standstill. Both computed and measured characteristics for an experimental machine are presented, and a comparison of these characteristics shows that the model can predict accurately all the major characteristics of the machine, as well as its detailed internal performance, when supplying a typical resistiveinductive load.     

Tracking the Frequency Translation of a Sum of Orthogonal Sinusoids

The maximum likelihood approach is used to derive a method for estimating and tracking the frequency translation of a signal consisting of a sum of orthogonal sinusoids corrupted by additive white noise. The likelihood function is reduced to an equivalent statistic expressed in terms of the squared magnitude of the finite Fourier transform of the received signal. A function that generates an error signal for a frequency translation tracking loop is derived, and a method of generating the error signal using the discrete Fourier transform (DFT) of the received signal weighted by a linear ramp is suggested. Two noise-free examples are presented.     

Carrier Tracking by Smoothing Filter Improves Symbol SNR

The potential benefit of using a smoothing filter to estimate a carrier phase over use of phase-locked loops (PLL) is determined. Numerical results are presented for the performance of three possible configurations of an all-digital coherent demodulation receiver. These are residual carrier PLL, sideband-aided residual carrier PLL, and finally sideband aided with Kalman smoother. The average symbol SNR after losses due to carrier phase estimation is computed for different total power SNRs, symbol rates, and symbol SNRs. It is found that smoothing is most beneficial for low symbol SNRs and low symbol rates. Smoothing gains up to 0.7 dB over sideband-aided residual carrier PLL, and the combined benefit of smoothing and sideband aiding relative to residual carrier loop is often in excess of 1 dB.