Showing papers on "Adaptive algorithm published in 1982"

Robustness of adaptive control algorithms in the presence of unmodeled dynamics

Abstract: This paper reports the outcome of an exhaustive analytical and numerical investigation of stability and robustness properties of a wide class of adaptive control algorithms in the presence of unmodeled dynamics and output disturbances. The class of adaptive algorithms considered are those commonly referred to as model-reference adaptive control algorithms, self-tuning controllers, and dead-beat adaptive controllers; they have been developed for both continuous-time systems and discrete-time systems. The existing adaptive control algorithms have been proven to be globally assymptotically stable under certain assumptions, the key ones being (a) that the number of poles and zeroes of the unknown plant are known, and (b) that the primary performance criterion is related to good command following. These theoretical assumptions are too restrictive from an engineering point of view. Real plants always contain unmodeled high-frequency dynamics and small delays, and hence no upper bound on the number of the plant poles and zeroes exists. Also real plants are always subject to unmeasurable output additive disturbances, although these may be quite small. Hence, it is important to critically examine the stability robustness properties of the existing adaptive algorithms when some of the theoretical assumptions are removed; in particular, their stability and performance properties in the presence of unmodeled dynamics and output disturbances. A unified analytical approach has been developed and documented in the recently completed Ph.D. thesis by Rohrs [15] that can be used to examine the class of existing adaptive algorithms. It was discovered that all existing algorithms contain an infinite-gain operator in the dynamic system that defines command reference errors and parameter errors; it is argued that such an infinite gain operator appears to be generic to all adaptive algorithms, whether they exhibit explicit or implicit parameter identification. The practical engineering consequences of the existence of the infinite-gain operator are disastrous. Analytical and simulation results demonstrate that sinusoidal reference inputs at specific frequencies and/or sinusoidal output disturbances at any frequency (including d.c.) cause the loop gain of the adaptive control system to increase without bound, thereby exciting the (unmodeled) plant dynamics, and yielding an unstable control system. Hence, it is concluded that none of the adaptive algorithms considered can be used with confidence in a practical control system design, because instability will set in with a high probability.

Exponential convergence and persistent excitation

Abstract: The notion of robustness of an adaptive algorithm is discussed. A number of simple examples are given of adaptive algorithms behaving in a nonrobust manner, and then conditions for robustness are presented which demand use of persistently exciting signals.

Robustness of self-tuning controllers

Abstract: The theory of adaptive control usually assumes that the controlled system is linear and of known finite order. The robustness problem examined here is to find the extent to which these conditions may be violated while retaining stability of the adaptive algorithm. Explicit frequency-domain circle-type results are given for certain classes of nonlinear or high-order systems.

Adaptive Spatial/Temporal/Spectral Filters For Background Clutter Suppression And Target Detection

Abstract: This paper presents some theoretical analysis and applications of adaptive filtering techniques for the detection of dim targets in the presence of highly structured background clutter. The computer algorithms carry out spatial, temporal, and multispectral filtering processes of forward looking infrared (FLIR) images taken at various times and in different spectral bands. These images were obtained from the Air Force TABILS data base. The basic criterion to drive these adaptive processes is based upon the minimum-mean-square error algorithm. The solution to the problem is to find a set of filter coefficients that will achieve the minimum mean square error at the output of the adaptive filter. The resulting Wiener-Hopf equation involves inverting the covariance matrix. Direct inversion of the covariance matrix, however, is a time-consuming process. To relax the computation complexity, a rapidly convergent adaptive algorithm based upon iteratively estimating the inverse of the covariance matrix was developed. The filtering operation consists of updating the inverse matrix in a sample-by-sample manner. This algorithm lends to a simplified hardware structure because it eliminates the matrix inversion process.

On a new adaptive lattice algorithm for recursive filters

Abstract: A new and simplified adaptive algorithm is presented for lattice structured recursive filters implementing the two-multiplier form based on Itakura and Saito. A comparison is also presented between the results obtained from this new algorithm and the algorithm introduced recently by Parikh, Ahmed, and Stearns.

A speech analysis algorithm which eliminates the influence of pitch using the model reference adaptive system

Abstract: A new adaptive algorithm based upon a least square criterion with a weighting factor is presented and shown to be quite useful for estimating ARMA parameters together with input in speech analysis. The estimator of both the input pulse train for voiced speech and the input white noise for unvoiced speech are easily obtained from the prediction errors by using this new adaptive algorithm. When these estimated inputs are used as the input of the model to be estimated, the influence of the pitch can be eliminated from the estimated ARMA parameters. By using this method the accuracy of formant and antiformant estimators is shown experimentally in comparison with LPC and cepstrum estimators.

Robust adaptive array antennas

Abstract: In this paper we present a new type of adaptive array antenna. The array optimizes well‐known cost functionals such as mean‐square error or output power, subject to strict inequality constraints on any part; or all the antenna pattern shape, e.g., controlling side‐lobe levels, null placement, directional control, etc. Moreover, the adaptive algorithm determines array gains which are ’’robust’’ in the sense that the antenna pattern will satisfy the pattern shape constraints even if the array weights are implemented with error, as might occur with digitally controlled weights, for example. The antenna pattern also exhibits a robustness property against wavefront distortion, mutual coupling uncertainties, and array geometry errors. Finally, by appropriate choice of constraints the array can have a broadband capability.

Constrained LMS adaptive algorithm

Abstract: A new least-mean-squares (LMS) adaptive algorithm is developed in the letter. This new algorithm solves a specific variance problem that occurs in LMS algorithms in the presence of high noise levels and when the input signal is bandlimited. Quantitative results in terms of an accuracy measure of a finite impulse response (FIR) system identification are presented.

A model reference adaptive control system for discrete multivariable systems with time delays

Abstract: This paper presents a new method for designing a model reference adaptive control system for multivariable plants with time delays in the input and output variables. An adaptive algorithm guarantees the asymptotic stability of the error between the plant output and the reference sequence using no anticipative value of the plant output and unknown plant parameter. The validity of the theoretical result is illustrated by a numerical example.

Applications of output error recursive estimation algorithms for adaptive signal processing

Abstract: Several problems in signal processing can be approached as recursive identification problems for AR, ARMA or ARMAX models. The output error with extended estimation model adaptive algorithm is proposed and evaluated for real-time estimation of the parameters of the signal model. The applications of these techniques for the adaptive signal processing are presented. The theoretical aspects related to the convergence properties of this algorithm are also discussed.

A globally stable APCS in the presence of bounded unmeasured noises and disturbances

Abstract: This paper presents a globally stable adaptive, predictive control system for discrete time multivariable linear processes with pure time-delays under the effect of measurement and process noise plus unmeasured disturbances. This solution, which extends a previous result by the author et al. [1], is based on a different approach compared to recent results in stochastic adaptive control literature. The proof of stability relies on convergence properties of the a posteriori error and the estimated parameters, which are derived from an adaptive algorithm that includes a criterion for stopping or continuing parameter adaptation. The main features of the present result are: i) the only restriction on noise and disturbances acting on the system is that they be bounded; ii) the estimated parameters converge towards the actual process parameters as much as necessary to obtain the stability results and no persistent excitation is required, and iii) the global stability results guarantee boundedness of the process input and output and minimization of the control or tracking error in terms of the absolute value.

On the high resolution and unbiased frequency estimates of sinusoids in white noise--A new adaptive approach

Abstract: A new adaptive algorithm is proposed to give an unbiased and high resolution frequency estimate of sinusoids in white noise. Base on the normalized Least-Squares (LS) lattice algorithm and the inverse power iteration method, the eigenvector associated with the minimum eigenvalue of the signal covariance matrix is estimated in the algorithm. The zeros of the "eigenvector polynomial" thus obtained are all on the unit circle and at the angles of the sinusoid frequencies. It is an adaptive realization of the Pisarneko's "harmonic retrieval" method. But differing from the adaptive method proposed by Thompson where the eigenvector is obtained through a gradient search in a constrained optimization formulation, in the new method the eigenvector is computed by an inverse power iteration. It enjoys all the advantages of the normalized LS lattice such as fast computations, low round-off noise and an easy stability check, etc. as well as fast convergence rate of the inverse power iteration method. Computer simulation results are shown.

Fast least-squares (LS) in the voice echo cancellation application

Abstract: The existing echo cancellation methods are primarily based on the LMS adaptive algorithm. Despite the fact that the LMS echo canceller works better than its predecessor-the echo suppressor, its performance can be substantially improved if the Recursive LS (RLS) algorithm is used instead. However the αp2operations (p: filter order) per sample required prevents the RLS algorithm from being used in this and many other applications where the filter order is relatively high. The computational complexity of the RLS has recently been brought down to αp by exploiting the shifting structure of the signal covariance matrix. Two fast algorithms, namely the LS lattice and the "fast Kalman", are used here. Comparisons between the two fast LS algorithms and the LMS gradient algorithm are made and the performance difference is demonstrated. Two important problems in voice echo cancellation: the flat delay estimation and the near-end speech detection, are approached novelly through a minimum-mean-squared-error flat delay estimator and a likelihood near-end speech detector. Simulation results are very satsifactory.

Carrier recovery circuit

Abstract: The invention provides a carrier recovery circuit which reduces the amount of calculation processing recovered to renew the tap coefficients of an adaptive filter to achieve improvement performance at low signal-to-noise ratio and make high speed pull in possible. Tap coefficients of variable coefficient filter 12 of adaptive line enhancer 1 are blocked into a plurality of blocks, and switches of switch circuit 14 individually corresponding to the blocks are selected by coefficient renewal control circuit 3 and renewal of the tap coefficients of the selected block is executed by adaptive algorithm circuit 13. Since all tap coefficients need not necessarily be renewed at a time in each symbol cycle, only the tap coefficients of the selected block are renewed. Thus although the adaptive speed of the adaptive emission line emphasizer decreases, the adaptive line enhancer can be constructed with a narrower band. Further, by varying the number of blocks to be selected in response to frame synchronizing information, prior to establishment of frame synchronization the number of blocks to be selected can be reduced and high speed pull in is realized.

Asymptotic behavior of an adaptive estimation algorithm with application to M-dependent data

Abstract: Theoretical results are presented concerning the asymptotic behavior of the parameter estimates generated by an adaptive algorithm in stationary dependent random situations. Proofs are exhibited in the general case and derived for the case of statistical dependence in the input for a finite number of lags. It is found that the estimate mean-error norm converges to an asymptotic bound, giving a finite bias, generally nonzero, and that the asymptotic mean-norm square error is bounded and can be arbitrarily reduced by decreasing the adaptation factor.

An adaptive algorithm for estimating the acceleration of highly maneuvering targets

Abstract: A "current model" concept and modified Rayleigh distribution are proposed to describe the "current" probability distribution of target maneuvering accelaration. By noticing the relation between the state (accelaration) estimate and the mean-value of the state noise, a maneuvering acceleration mean-value and variance adaptive Kalman filtering algorithm is suggested. In three dimensional case, a nonlinear state model and a direct method of estimating tangential and normal accelerations of maneuvering targets are developed. Some computational results are presented.

Model Reference Adaptive Control with Inexact Model Matching Part I

Abstract: A new scheme for designing model reference adaptive control systems using only input and output measurements is presented. With the proposed scheme, the requirement on the commonly assumed a priori knowledge of (i) the relative degree n* of the plant and (ii) sign of the high frequency gain, can be eliminated. Besides the elimination of requirement of the two restrictive assumptions, the proposed algorithm has the advantage of using fewer adaptive gains and of improving the speed of convergence. The same method has also been employed to address the important problem of using a reduced order model.

Recursive identification techniques

Abstract: Some basic results on recursive identification techniques and their properties are reviewed. The link between adaptive algorithms, recursive identification, and off-line identification is stressed. The fundamental character of the prediction and its gradient with respect to the adjustable parameters is pointed out.

Experimental performance data for slowed-down adaptive algorithms†

Abstract: Considerable savings in hardware and computation time can be achieved very simply by slowing down the adaptive algorithm for an adaptive digital filter either by updating weights loss often than every sample (typically every n samples, where n is the number of fitter weights), or by updating only one weight per sample cyclically repeating every n samples. This technique has the advantage of being useful with any of the many adaptive algorithms including the LMS non-recursive algorithm and the Stearns-White recursive algorithm. Experiments verify that the slowed down algorithms not only save hardware and computation time, but actually improve filter performance.

An Adaptive Spread Spectrum Receiver for Multipath/Scatter Channels

Abstract: Transmitted signal spectrum spreading is used to enhance multipath diversity performance. An adaptive DPSK spread spectrum multiple access receiver structure is presented for which optimal demodulation is obtained in slowly time-varying complex Gaussian scatter/multipath channels under convergence of the adaptive algorithm. The receiver performs a cross-correlation function equalization of a multipath degraded spread spectrum signal using a Stochastic Approximation technique. A potential of coherenit maximal-ratio multipath combining performance can be obtained under general slowly varying channel conditions.

Adaptive echo cancellation with dispersion and delay in the adjustment loop

Abstract: We analyze an echo canceler structure which has dispersion and/or delay in the tap adjustment loop. Such a structure has the potential of providing a saving in the analog filtering requirements, when the cancellation is performed in the sampled-data domain. We derive the optimal coefficient vector, in the mean-square error sense, and we derive conditions for a unique solution. It is observed that this problem is somewhat similar, but not equivalent, to the mean-square equalizer problem. Similarities and differences between the two are discussed. We also propose an adaptive algorithm which is well suited for this canceler structure. It is shown that the computational complexity of this algorithm is significantly greater than that of the conventional updating algorithm.

A stable hybrid adaptive algorithm with periodic sampling and gain adjustment

Abstract: In this article a hybrid algorithm for model reference adaptive control of single-input single output systems is presented. The control structure involves a continuous time as well as a discrete time part, instead of being all discrete or all continuous as in previous approaches. The system is sampled periodically at a frequency F, and knowledge of bounds on the plant parameters enables us to determine a bound F* such that the closed loop system is stable whenever F > F*.

Comparison of some algorithms for tap weight evaluation in adaptive echo cancellers

Abstract: The performance of some adaptive algorithms for coefficient updating of a digital echo canceller are compared. The examined algorithms are the least mean square one (LMS), the normalized LMS and the simplified versions employing the sign information. Their performance are evaluated on the basis of the echo return loss enhancement (ERLE) steady state value and convergence speed. For the algorithms employing a linear function of the error, as gradient estimate, the resuIts show that convergence speed is dependent on the echo canceller tap number and that its trend is exponential. Algorithms employing the error sign as gradient estimate are the slowest if the same variance of residual echo must be obtained. Furthermore some consideration are made in comparison with an algorithm designed in order to minimize the mean square error evaluated over a M sample block.

An Adaptive Identifier for Discrete-Time Linear Systems

Abstract: It is shown that the Geodwin-Ramadge-Caines method for adaptive control of discrete-time systems may be used for the identification of system parameters. An expression for the optimal gain constant is derived. Then, the algorithm may be modified to allow for improved convergence and maintenance of linear relationships of the unknown parameters of the system.