Adaptive Filtering Technique Research Articles

Modified recursive maximum likelihood adaptive filter for nonlinear aircraft flight-path reconstruction

Nonlinear aircraft flight-path reconstruction is basically a state estimation problem that can be solved with adaptive filtering techniques, as it involves the estimation of flight trajectories and unknown parameters such as biases, scale factors, and noise statistical uncertainties of flight instrumentation systems. Among many algorithms, the recursive maximum likelihood (RML) method is a popular scheme in adaptive filtering for nonlinear state-parameter estimation problems. However, the RML algorithm is sensitive to initialization errors of system parameters. Divergence may occur at large values of these errors. The objective of the present study is to develop a modified recursive maximum likelihood (MRML) adaptive filter that is less sensitive to the effects of initialization errors. The new algorithm revises the conventional RML adaptive filter by including the effect of the parameter estimator in the prediction error vector computation. Numerical results are presented for a nonlinear aircraft model. System states and a variety of parameters including measurement noise standard deviations were estimated with the conventional RML adaptive filter and compared with corresponding estimates of the new MRML adaptive filter. Numerical simulations were carried out with different a priori estimates of parameters and system state vector elements. The results indicate that the MRML adaptive filter, as developed here, produces estimates that are both more accurate and less sensitive to parameter initialization errors than those obtained with the conventional RML adaptive filter.

Adaptive Filtering for Unbalance Vibration Suppression.

A method is introduced which suppresses the unbalance vibration of a rotor suspended by magnetic bearings, using two adaptive filtering techniques. The first filter separates the rotor displacement signal into rotation-synchronous and rotation-asynchronous components. The former signal can be cancelled by the output of the second filter, which allows estimation of the unbalance disturbance force using the impulse response of the plant of rotor with control system measured on-line under high-speed rotation. Both simulation and experimental results are shown to describe the basic and practical effectiveness of the proposed method. The system works well even when using only the initial part of the plant response.

Implementation of discrete orthogonal transforms based on adaptive filtering systems

Abstract A number of approaches for implementing orthogonal analysers using adaptive filtering techniques are investigated in this paper. A general connection between adaptive filtering and orthogonal transforms is presented. In particular, the LMS-based discrete orthogonal analyser is investigated and sufficient conditions for implementation are derived. We then propose an orthogonal analyser using the adaptive Sampling Matrix Inversion (SMI) algorithm, which is applicable for all orthogonal transforms. Finally, an orthogonal analyser using an adaptive Howells-Applebaum loop is analysed and shown to be capable of performing the Fourier analysis for continuous time signals.

Nonlinear multivariate image filtering techniques

In this paper, nonlinear multivariate image filtering techniques are proposed to handle color images corrupted by noise. First, we briefly review the principle of reduced ordering (R-ordering) and then define three R-orderings by selecting different central locations. Considering noise attenuation, edge preservation, and detail retention, R-ordering based multivariate filters are designed by combining the R-ordering schemes. To implement color image filtering more effectively, we develop them into a locally adaptive version. The output of the adaptive filter is the closest sample to a central location that is a weighted linear combination of the mean, the marginal median, and the center sample. As a result, we study an adaptive hybrid multivariate (AHM) filter consisting of the mean filter, the marginal median filter, and the identity filter. The performance of the two adaptive filtering techniques is compared with that of some nonadaptive ones. The examples of color image filtering show that the adaptive multivariate image filtering gives a rather good performance improvement.

The lattice filter in ground‐roll suppression

Seismic reflections are sometimes masked by Ray‐leigh‐type surface waves that are termed ground roll in seismic literature. An adaptive lattice filter is used to recover reflected signals contaminated by ground roll. Experiments on synthetic and field data showed that the adaptive lattice filter technique is very effective in ground‐roll elimination. In addition, the filter works as a whitening operator, compresses the signal, and increases the signal‐to‐noise ratio.

A personal computer-based visual evoked potential stimulus and recording system.

A system for recording electroretinograms and visual evoked cortical potentials has been constructed with the use of a personal computer and a digital signal processing card. The system is based on widely available commercial hardware. It has been designed to be capable of performing routine visual electrophysiology as well as allowing the development of novel visual stimuli and signal detection techniques. The system enables both transient and steady-state stimulation rates. Pattern stimuli can be presented in pattern-reversal, pattern-onset, pattern-offset or motion-onset modes. In addition to conventional signal averaging, the digital signal processing card can also provide on-line Fourier analysis and is facilitating the development of adaptive filtering techniques for the detection of steady-state visual evoked cortical potentials. This versatile system is in regular clinical use for the measurement of electroretinograms and visual evoked cortical potentials.

The covering problem and μ-dependent adaptive algorithms

The paper presents a family of techniques, called adaptive covering algorithms, which solve a particular covering problem-how to best cover a target shape using a set of simply parameterized elements. The algorithms, inspired by adaptive filtering techniques, provide a computationally simple, robust, and efficient alternative to more traditional methods such as Bayesian approaches, convex hulls, and multi-layer perceptrons. The paper develops a theoretical understanding of the adaptive covering algorithms by relating their behavior via weak convergence techniques to the evolution of a deterministic ordinary differential equation (ODE). In the process, the authors give new convergence results for a class of step size-dependent recursive algorithms. Stability and instability of the ODE can be interpreted in terms of the local stability/instability of the algorithm. In terms of the covering problem, candidate coverings tend to improve as more data is gathered whenever the ODE is stable. Several examples are given which demonstrate the ideas and which verify that the analysis accurately predicts the true behavior of the algorithms.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Spells of Low-Frequency Oscillations and Weather Regimes in the Northern Hemisphere

Abstract The low-frequency variability in the midlatitudes is described through an analysis of the oscillatory phenomena. In order to isolate nearly periodic components of the atmospheric flow, the multichannel version of the singular spectrum analysis (M-SSA) is developed and applied to an NMC 32-year long set of 700-hPa geopotential heights. In the same way that principal component analysis identifies the spatial patterns dominating the variability, M-SSA identifies dynamically relevant space–time patterns and provides an adaptive filtering technique. Three major low-frequency oscillations (LFOs) are found, with periods of 70 days, 40–45 days, and 30–35 days. The 70-day oscillation consists of fluctuations in both position and amplitude of the Atlantic jet, with a poleward-propagating anomaly pattern. The 40–45-day oscillation is specific to the Pacific sector and has a pronounced Pacific/North American (PNA) structure in its high-amplitude phase. The 30–35-day mode is confined over the Atlantic region,...

Automatic identification of voice band telephony coding schemes using neural networks

A new method for identifying the type of coding present on a telecommunications channel is described. The method employs a combination of least mean squares adaptive filter techniques and neural network methods. The method is explained, and simulation results presented.

Adaptive filtering of nonlinear systems with memory by quantized mean field annealing (digital subscriber loop example)

A technique for adaptive filtering of nonlinear systems with memory that combines quantized mean field annealing (QMFA) and conventional recursive-least-squares/fast-transversal-filter (RLS/FTF) adaptive filtering is developed. This technique can efficiently handle large-order nonlinearities with or without memory. The nonlinear channel is divided into a memory nonlinearity followed by a dispersive linear system. QMFA is applied to obtain the coefficients and the order of the memory of the nonlinearity, and RLS/FTF is applied to determine the weights of the dispersive linear system. Statistical thermodynamic analysis that provides theoretical measures for making annealing algorithms computationally efficient. The method is applied to a full duplex digital subscriber loop. Simulations show a performance improvement of over 40 dB compared to ordinary RLS/FTF and steepest descent algorithms, and the solution is robust. >

Lift on an oscillating cylinder in smooth and turbulent flow

Lift forces were measured at six sections of a circular cylinder in smooth and turbulent flows over the Reynolds number range 1.1 × 105 to 5.5 × 105. Introduction of turbulence of 18% longitudinal intensity produced a return to organized vortex shedding, characterized by large lift forces with a broad spectral peak centred near St = 0.23.With the cylinder in forced oscillation of amplitudes up to 3% of diameter, motion-correlated forces were observed. Adaptive digital filter techniques were used to remove inertial components of force transducer signal.

Suppression of dominant topographic overprints in gravity data by adaptive filtering: Southern Wyoming province

In areas of the world with a protracted tectonic history, the gravity field may be dominated by anomalies due to individual structural features developed during the latest tectonic event. These individual tectonic features often have a topographic expression. Adaptive filtering techniques can be used to suppress the gravity effects of such features. Iterative space domain techniques are particularly well suited to these types of problems. One such technique, the stochastic gradient algorithm, allows easy updating of filter coefficients at each spatial step through the data. This approach eliminates windowing and spectral estimation problems associated with cross‐spectral admittance schemes. In this study, adaptive filtering of Bouguer gravity and topography in the southern Wyoming Province was successful in suppressing gravity signatures of Laramide structural features. The regional gravity trends due to isostatic effects were significantly absent from the residual anomaly map. A large, continuous, positive gravity anomaly stretching from the southern Wind River Range to the northern end of the Laramie Range was another outstanding feature observed in the residual anomaly. This positive anomaly is about the same magnitude, size and shape as anomalies due to greenstone belts observed in other Archean cratonal areas. Scattered occurrences of greenstone lithologies are known throughout the area of the anomaly, but a continuous feature is absent in outcrop. Due to the unusual structural history of the area, it is possible that such a continuous belt does exist, but was buried at shallow crustal levels by Laramide thrusting, or later collapse of a Laramide uplift due to a change in the regional stress field.

Dynamic non-invasive measurements of arterial diameter and wall thickness

Non-invasive measurements of arterial diameter and wall thickness are critical in characterizing the onset and development of vascular disease. A precise dynamic method was proposed and tested for this purpose. A non-invasive method of measuring the variations in diameter and thickness of human arteries throughout the cardiac cycle was developed, using a high-precision ultrasonic echo-tracking system. An adaptive filtering technique was used to suppress artefacts caused by the layered tissue structure of the vessel wall. Based on decorrelation of microstructure noise, this technique improved the detectability of the wall interfaces, which allowed a determination of thickness and diameter. The accuracy and reproducibility of the method were tested by measurements of plastic films with known thicknesses. The discrepancies between standard micrometer and pulse-echo measurement were consistently less than 5 microns for film thicknesses ranging from 220 to 800 microns. The difference between two successive measurements was less than 2 microns. The identity of the measured vascular interfaces was checked in two ways. First, experiments on fixed bovine carotid arteries showed that the identified echogenic interfaces corresponded to the actual anatomical structure, as obtained by acoustic microscopy. Second, the radial artery thickness and diameter were extrapolated to obtain the change in wall volume over one cardiac cycle. The volume was found to be nearly constant, indicating incompressibility. This method will make it possible to obtain new information on atherogenesis and other vascular diseases.

An adaptive bit time estimator

A bit time estimator which uses adaptive filtering techniques is presented. The filter weights of an adaptive linear predictor are shown to provide a reliable estimate of the bit time T of a random binary square wave contaminated with additive white Gaussian noise, with little or no a priori information. The quality of this estimator is then evaluated via the least mean square algorithm, and a comparison is made between it and a more conventional estimator based on a zero crossing detector. This comparison shows that an adaptive estimator based on a linear predictor is generally superior. >

Applications of Astrom-Gray-Markel recursions

The Astrom-Gray-Markel recursions algorithm computes the power of an ARMA model from its power spectral density. Two applications of this algorithm are shown in this paper. The first application is the derivation of some ARMA identities relating the ARMA coefficients, the reflection coefficients and the tap weights. They provide an insight about the AR spectral shape without actually plotting the spectrum. The second is an analytical evaluation of the cross-spectral correlation (CSC) between two ARMA models without any numerical integrations. The CSC is useful to identify time-varying signals of the same type when adaptive filtering techniques can be applied to these signals, as an alternative to a time-domain correlation which may require long durations of the signals.

Automatic tuning of continuous-time integrated filters using an adaptive filter technique

An adaptive filter technique for tuning continuous-time integrated filters is presented. This technique is based on the model-matching configuration and tunes both the poles and zeros of the transfer function. Circuit details of an experimental prototype are given. The experimental prototype consists of an integrated third-order filter that is automatically tuned by off-chip circuitry realizing the adaptive tuning system. Both experimental and simulation results are presented to confirm the viability of the proposed approach.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Acoustic echo cancellation with a transform domain adaptive filter

The adaptive filtering technique is one of the major approaches to control the acoustic echoes in teleconferencing systems or hands‐free telephony systems. A brief overview of the performance of different adaptive algorithms for acoustic echo cancellation [cf, S. M. Kuo and H. Zhao, J. Acoust. Soc. Am. Suppl. 1 86, S56 (1989)] will be presented. A transform domain adaptive filter (TDAF) algorithm is proposed to construct an acoustic echo canceler that has a convergence rate faster than that of the conventional LMS adaptive transversal filter with colored signals. The proposed approach does not introduce extra delay during the process so that it can be used in the application where delay is not allowed. Different orthogonal transforms such as the discrete Fourier transform (DFT), discrete cosine transform (DCT), and discrete Hartley transform (DHT) are used for signals with different eigenvalue spread. Real‐time implementations of this class of canceler with digital signal processors (DSP) will also be discussed.

Recent Technical Developments in the S-VHS VCR for Broadcasting and Professional Applications

S-VHS frequency spectrum allocations for luminance and chrominance signals on video recording, resultant performance, and comparison with the U-SP format are reviewed briefly. Technical considerations leading to selection of oxide tape and ferrite heads for S-VHS recording are discussed, and Y/C video signal transmission and its benefits are described. Treated in more detail are some of the innovative video processing techniques recently developed to achieve significant picture performance improvements in professional S-VHS VCRs and peripheral products. These specific techniques include a unique three-line adaptive comb filter to separate luminance and chrominance components from a composite NTSC video signal; a chrominance enhancement technique developed by Faroudja Laboratories and adapted for use with S-VHS, which significantly restores chroma bandwidth lost during record/playback without degradation of signal-to-noise ratio; an adaptive comb filter technique that supplements phase-shift video processing to improve adjacent track chrominance crosstalk cancellation; and a Y/C digital noise-reduction system designed to significantly improve chrominance as well as luminance signal-to-noise ratio.

On a Higher Order Recursive Technique for Frequency Estimation

This paper considers the problem of frequency estimation using recursive schemes. The adaptive notch filter technique is reviewed and an algorithm based on the extended instrumental variable method is proposed. A new method for constructing the instrumental variables and a higher order recursive formulation are considered. Use of hypothesis testing analysis in this algorithm produces an adaptive procedure for constructing the instruments and estimating the frequencies.

Active noise control on systems with time-varying sources and acoustical elements

An active noise control system has been developed based on a system identification concept using a recursive adaptive filter to model the direct and feedback elements of an acoustic system. An independent random noise source is used with a second adaptive filter to provide a model of the output transducer and error path transfer functions. This system has the ability to continuously cancel narrow band and broadband noise. The system identification configuration combined with adaptive filter techniques enables the system to respond quickly and accurately to changes in time-varying sources, such as frequency or amplitude, and changes in the acoustical system, such as temperature. Noise reduction results will be presented. The system capabilities will be demonstrated on a variety of time-varying sources and time-varying acoustical elements. The ability of the system to track these changes in real-time as well as implications for actual applications will be discussed.