adaptive quantization is used in many ADPCM systems to improve performance. We present an analysis dealing with the effect of adaptive quantizers on the stability of ADPCM systems as compared with ADPCM with nonadapti...
详细信息
adaptive quantization is used in many ADPCM systems to improve performance. We present an analysis dealing with the effect of adaptive quantizers on the stability of ADPCM systems as compared with ADPCM with nonadaptive quantization. Stability of the system is noted to equate with recovery from encoder internal state discrepancies. A sufficient condition for stability that leads to a relation between adaptation rate and stability is presented. This indicates that a price is paid for adaptation in terms of more stringent restrictions on the predictor to guarantee stability. The theorem further indicates that stability is related to the maximum rate of quantizer step size decrease and can be viewed as a theoretical justification for the shape of the Jayant ''one-word memory'' multiplier curve for adaptive quantization.
A model-free approach to linear quadratic (LQ) performance assessment is presented in this paper. First a test for LQ optimality, based on closed-loop signals, is presented. Its result indicates whether the controll...
详细信息
A model-free approach to linear quadratic (LQ) performance assessment is
presented in this paper. First a test for LQ optimality, based on closed-loop
signals, is presented. Its result indicates whether the controller should be
adjusted. From the same test it is possible to determine the closed-loop pole
positions which would be obtained by using the LQ optimal controller. These may then
be used in a direct adaptive control scheme to adjust the controller. Both state and
output feedback are addressed for discrete-time, single-input-single-output (SISO),
linear, time-invariant processes. (Author abstract) 11 Refs.
In this paper, we address the problem of risk-sensitive filtering and smoothing for discrete-time Hidden Markov Models (HMM) with finite-discrete states. The objective of risk-sensitive filtering is to minimise the ex...
详细信息
In this paper, we address the problem of risk-sensitive filtering and smoothing for discrete-time Hidden Markov Models (HMM) with finite-discrete states. The objective of risk-sensitive filtering is to minimise the expectation of the exponential of the squared estimation error weighted by a risk-sensitive parameter. We use the so-called Reference Probability Method in solving this problem. We achieve finite-dimensional linear recursions in the information state, and thereby the state estimate that minimises the risk-sensitive cost index. Also, fixed-interval smoothing results are derived. We show that L(2) or risk-neutral filtering for HMMs can be extracted as a limiting case of the risk-sensitive filtering problem when the risk-sensitive parameter approaches zero.
In this paper, we address the problem of exponential stability of filters and fixed-lag smoothers for discrete-time and discrete-state Hidden Markov Models (HMMs). By appealing to a generalised Perron-Frobenius result...
详细信息
ISBN:
(纸本)9783952426906
In this paper, we address the problem of exponential stability of filters and fixed-lag smoothers for discrete-time and discrete-state Hidden Markov Models (HMMs). By appealing to a generalised Perron-Frobenius result for nonnegative matrices, we demonstrate exponential forgetting for both the recursive filters and smoothers, and obtain overbounds on the rate of forgetting. Simulation studies are carried out to substantiate the results.
It is suggested that modern controller design packages often fall short of offering what is truly practical: lower order controllers, discrete-time controllers operating in a sampled-data loop, and finite word length ...
详细信息
It is suggested that modern controller design packages often fall short of offering what is truly practical: lower order controllers, discrete-time controllers operating in a sampled-data loop, and finite word length (FWL) realizations of controllers with the FWL property minimally impacting closed-loop performance. Several methods for achieving these objectives are considered. The methods deal with controller complexity, transfer function matching, stability robustness, signal spectrum matching, fractional representations of open-loop unstable controllers, discrete time controllers, and realizing digital controllers.< >
It has been shown that the set of all nonlinear plants that can be stabilized by a known linear controller which also stabilizes a linear nominal model of the plant can be parametrized by a stable operator known as th...
详细信息
ISBN:
(纸本)9783952426906
It has been shown that the set of all nonlinear plants that can be stabilized by a known linear controller which also stabilizes a linear nominal model of the plant can be parametrized by a stable operator known as the Youla-Kucera parameter. This paper extends previous work by allowing the model of the nominal plant in the above scenario to be nonlinear, and hence explores the possibilities of converting the closed loop plant identification problem to one of open loop identification. The ideas rely on a concept of differential coprimeness for nonlinear fractional system descriptions.
The use of Kalman filtering techniques to reconstruct down-sampled speech signals is investigated. We compare a full-sampled, coarsely quantised speech coding system with a down-sampled, finely quantised speech coding...
详细信息
The use of Kalman filtering techniques to reconstruct down-sampled speech signals is investigated. We compare a full-sampled, coarsely quantised speech coding system with a down-sampled, finely quantised speech coding system transmitting at the same bit-rate. Testing reveals that the down-sampled system can outperform the full-sampled system. Sample rate, sample accuracy and decoding delay are traded to achieve improvements at a fixed bit-rate. (C) 1997 Elsevier Science B.V.
In this paper, a non-linear approach to the design of model reference adaptive control is presented. The essence of the idea is to generate a full non-linear model of the plant dynamics and the parameter adaptation dy...
详细信息
ISBN:
(纸本)9783952426906
In this paper, a non-linear approach to the design of model reference adaptive control is presented. The essence of the idea is to generate a full non-linear model of the plant dynamics and the parameter adaptation dynamics as a gradient descent algorithm with respect to a Riemannian metric. It is shown how a Riemannian metric can be chosen so that the modelled plant dynamics do in fact match the true plant dynamics. The approach is demonstrated by a case study of a simple single-pole and no zero, discretetime plant.
This paper derives a general expression for the mean square error in estimating the fundamental frequency of a multiharmonic signal from a finite sequence of noisy measurements. The distinguishing feature of this expr...
详细信息
This paper derives a general expression for the mean square error in estimating the fundamental frequency of a multiharmonic signal from a finite sequence of noisy measurements. The distinguishing feature of this expression is that it is applicable at values of signal-to-noise ratio (SNR) within the threshold region, in contrast to earlier expressions (the Cramer-Rao bounds) that are valid only at high SNRs. Theoretical performance curves are thereby calculated (mean square error versus SNR) that establish the existence of a threshold effect. Until now, the existence of a threshold effect was demonstrable only by simulation. Examples are given comparing various multiharmonic estimation scenarios to the single tone case under comparable conditions. The theoretical performance curves in these examples are corroborated by Monte Carlo simulation.
We examine a number of crucial questions that arise in the windsurfer approach to adaptiverobust control. Considerations are limited to the case where the plant is stable and has no zeros on the imaginary axis. The k...
详细信息
We examine a number of crucial questions that arise in the windsurfer approach to adaptiverobust control. Considerations are limited to the case where the plant is stable and has no zeros on the imaginary axis. The key conclusion is that, given a strictly proper stable model of a strictly proper stable plant, we can improve the performance robustness of the closed-loop system through the windsurfer approach if the plant and the existing model have no unstable zeros within the designed closed-loop bandwidth and if the deterioration in performance robustness caused by increasing the closed-loop bandwidth results in a sufficiently high signal-to-noise ratio for a certain closed-loop output error. Situations that may cause the iterative identification and control design process to terminate prematurely are identified. A simulation example is used to illustrate the results discussed.
暂无评论