Output Dimension Research Articles

Iterative Water-Filling for Gaussian Vector Multiple-Access Channels

This paper proposes an efficient numerical algorithm to compute the optimal input distribution that maximizes the sum capacity of a Gaussian multiple-access channel with vector inputs and a vector output. The numerical algorithm has an iterative water-filling interpretation. The algorithm converges from any starting point, and it reaches within 1/2 nats per user per output dimension from the sum capacity after just one iteration. The characterization of sum capacity also allows an upper bound and a lower bound for the entire capacity region to be derived.

Comparative study on state analysis of BOP in NPPs

Under the deregulation environment, the performance enhancement of balance of plant (BOP) or the turbine cycle in nuclear power plants is being highlighted. To analyze performance level of the turbine cycle, we use the performance test procedures provided from an authorized institution such as American Society of Mechanical Engineers. However, through plant investigation, it was proved that the requirements of the performance test procedures about the reliability and quantity of sensors was difficult to be satisfied. As a solution of this weakness, a state analysis method that is the expanded concept of signal validation, was proposed on the basis of the statistical and the neural network approaches. The authors recommended the statistical linear regression model by analyzing correlation among turbine cycle parameters as a reference state analysis method. Its advantage is that its derivation is not heuristic, it is possible to calculate model uncertainty, and it is easy to apply to an actual plant. The error of the statistical linear regression model is below 3% under normal as well as abnormal system states. Additionally the auto-associative neural network (AANN) model was recommended since the statistical model is impossible to apply to the validation of all of the sensors and is sensitive to the outlier that is the signal located out of statistical distribution. Because there are a lot of sensors need to be validated in turbine cycle, principal component analysis (PCA) and wavelet analysis (WA) were applied as a pre-processor for the reduction of input dimension and for the enhancement of training accuracy, and the results were compared. Both are effective to reduce the number of the input and output dimension, but WA is superior to PCA in the compensation of the outliers. The outlier localization capability of WA enhanced the robustness of the AANN. In case of the AANN with WA, the training accuracy of the model was nearly 100% and the trained neural network restored the degraded signals to the values within /spl plusmn/3% of the true signals.

Plant Ecological Strategies: Some Leading Dimensions of Variation Between Species

▪ Abstract An important aim of plant ecology is to identify leading dimensions of ecological variation among species and to understand the basis for them. Dimensions that can readily be measured would be especially useful, because they might offer a path towards improved worldwide synthesis across the thousands of field experiments and ecophysiological studies that use just a few species each. Four dimensions are reviewed here. The leaf mass per area–leaf lifespan (LMA-LL) dimension expresses slow turnover of plant parts (at high LMA and long LL), long nutrient residence times, and slow response to favorable growth conditions. The seed mass–seed output (SM-SO) dimension is an important predictor of dispersal to establishment opportunities (seed output) and of establishment success in the face of hazards (seed mass). The LMA-LL and SM-SO dimensions are each underpinned by a single, comprehensible tradeoff, and their consequences are fairly well understood. The leaf size–twig size (LS-TS) spectrum has obvious consequences for the texture of canopies, but the costs and benefits of large versus small leaf and twig size are poorly understood. The height dimension has universally been seen as ecologically important and included in ecological strategy schemes. Nevertheless, height includes several tradeoffs and adaptive elements, which ideally should be treated separately. Each of these four dimensions varies at the scales of climate zones and of site types within landscapes. This variation can be interpreted as adaptation to the physical environment. Each dimension also varies widely among coexisting species. Most likely this within-site variation arises because the ecological opportunities for each species depend strongly on which other species are present, in other words, because the set of species at a site is a stable mixture of strategies.

Repeatability of echocardiographic measurements in the human fetus.

To determine the repeatability of cross-sectional, M-mode and Doppler echocardiographic measurements in the human fetus. This was a prospective echocardiographic study of 10 normal fetuses, involving measurement of 32 different echocardiographic variables in each. The intra- and interobserver error of measurements were quantified. The median (range) gestational age was 23 (17-34) weeks. Cross-sectional echocardiography: for left ventricular end-diastolic volume the intraobserver coefficient of variation was 13% and the interobserver limits of agreement were a ratio of 0.46-2.19. For left ventricular ejection fraction the repeatability was +/- 8.8% and the interobserver limits of agreement were +/- 20%. M-mode echocardiography: the coefficient of variation for left ventricular end-diastolic dimension was 10%, and the interobserver limits of agreement were a ratio of 0.74-1.70. The intraobserver repeatability of left ventricular fractional shortening was +/- 8.8%, and interobserver limits of agreement +/-15%. Limits of agreement for M-mode-derived left ventricular end-diastolic volume were wide, intraobserver coefficient of variation 32% and interobserver limits were a ratio of 0.44-4.36. Doppler echocardiography: the intra- and interobserver errors were high for Doppler variables such as acceleration time, acceleration slope, stroke volume, cardiac output and vessel dimension. There was less error associated with measurements such as maximum Doppler velocity, velocity time integral, ejection time and heart rate. The repeatability of most echocardiographic measurements in the fetus is poor. This applies particularly to volumetric data such as ventricular volumes and volume flow estimations. Interobserver errors are consistently higher than intraobserver errors, confirming that when sequential measurements are clinically important, the same observer should be used.

Left ventricular systolic and diastolic function during early neonatal period using transthoracic echocardiography.

To examine the effects of ductal closure on left ventricular (LV) systolic and diastolic function during the early neonatal periods, 45 normal term neonates delivered after uncomplicated pregnancies (mean 39 weeks) were studied using two-dimensional and Doppler echocardiography. We measured ductus arteriosus size, arterial blood pressures, ascending aortic size, LV dimensions, and transmitral flow velocity patterns and calculated LV output and rate-corrected fiber shortening fraction (mVcfc) at 2, 12, 24, and 120 hours after birth. The inner diameter of the ductus arteriosus was 4.3 +/- 0.7mm at 2 hours, 2.1 +/- 0.6 mm at 12 hours, and had closed in 42 of 45 neonates at 24 hours. LV output and LV end-diastolic dimension showed the highest level at 2 hours of age. However, the mVcfc did not change from 2 to 120 hours of age. The peak velocity during early diastole (peak E) was significantly greater at 2 hours than at 12 hours. The peak velocity during atrial contraction (peak A) remained unchanged during this period. The normalized peak filling rate at isovolumic relaxation time did not change over 120 hours. The present study demonstrated changes in LV systolic function and LV diastolic filling during the early neonatal period. LV systolic and diastolic function was preserved under the hemodynamic changes associated with the early neonatal period.

A Note on the Equivalence of NARX and RNN

This paper presents several aspects with regards the application of the NARX model and Recurrent Neural Network (RNN) model in system identification and control. We show that every RNN can be transformed to a first order NARX model, and vice versa, under the condition that the neuron transfer function is similar to the NARX transfer function. If the neuron transfer function is piecewise linear, that is f(x):=x if uxu , 1 and f(x):=sign(x) otherwise, we further show that every NARX model of order larger than one can be transformed into a RNN. According to these equivalence results, there are three advantages from which we can benefit: (i) if the output dimension of a NARX model is larger than the number of its hidden unit, training an equivalent RNN will be faster, i.e. the equivalent RNN is trained instead of the NARX model. Once the training is finished, the RNN is transformed back to an equivalent NARX model. On the other hand, (ii) if the output dimension of a RNN model is less than the number of its hidden units, the training of a RNN can be speeded up by using a similar method; (iii) the RNN pruning can be accomplished in a much simpler way, i.e. the equivalent NARX model is pruned instead of the RNN. After pruning, the NARX model is transformed back to the equivalent RNN.

ℋ∞output feedback control for linear discrete time-varying systems via the bounded real lemma

In this paper we develop a solution to the discrete-time ℋ ∞ output feedback control problem for linear time-varying (LTV) systems. The solution is developed along the strategy set up by Doyle et al. in 1989 and the main ingredient in its derivation is the extension of the well known bounded real lemma to a (discrete) time-varying context, developed by van der Veen and Verhaegen in 1995. This approach contributes to the conceptual simplicity, and hence to the accessibility, of the solution. Apart from that, we treat the infinite-horizon case for LTV system of non-uniform state dimension, and varying input and output dimension. Both situations can easily occur in practice, for example in multirate sampled data control systems. The algorithm that can be derived from the solution presented in then applied to the ℋ ∞ output feedback of a dynamical system changing from one operation point in its operation envelope to another.

A wearable multichannel tactile display of voice fundamental frequency.

This paper describes a wearable sensory aid that provides the deaf with tactually encoded information about intonation. Fundamental frequency is represented as both place and rate of vibration in a linear array of solenoids. Pitch extraction is accomplished through low-pass filtering and peak detection. A microcomputer is used to measure pitch period, which in turn determines which of the solenoids is actuated. By comparing consecutive periods, the system discriminates against random, noise-related inputs. The device is switchable between 1-, 8-, and 16-channel operation. The electronics package is contained in a case that may be worn on a belt. The solenoid array is worn on the forearm. The system is powered by five, rechargeable lithium cells and runs for at least 6 hours between charges. Proposed developments include the incorporation of digital pitch extraction methods and the option to use the spatial output dimension to encode speech parameters other than fundamental frequency.

In utero right ventricular output in the fetal lamb: the effect of heart rate.

1. The effect of heart rate on right ventricular output was examined in six lambs during a period extending from 126 to 139 days of gestation. The fetuses had been surgically instrumented at least four days previously with a main pulmonary artery flow probe, right ventricular dimension transducers and left and right atrial pacing electrodes. 2. During spontaneous variations in heart rate, rate was correlated positively with right ventricular output (P less than 0.0001) and end-diastolic dimension (P less than 0.0001) among the lambs considered as a group, but no significant effect of rate on stroke volume was found. When individual responses were examined, output increased significantly with rate in sixteen out of seventeen observations. 3. With left atrial pacing, heart rate did not affect output. With right atrial pacing, rate correlated negatively with output (P less than 0.0001). With pacing from either site, rate correlated negatively with end-diastolic dimension (P less than 0.0001) and stroke volume (P less than 0.0001). 4. The introduction of a longer period interval during each pacing rate inhibited the rate-related decrease in dimension and allowed the ventricle to fill to the same end-diastolic dimension. The systole following these longer intervals had a greater stroke volume than did the preceding systoles with smaller end-diastolic dimension. The faster the preceding paced rate, the greater was the increase in stroke volume (P less than 0.001). 5. Right ventricular dimensions and volumes were measured in vitro, and the relationship was found to be linear using regression analysis. 6. This study demonstrates that experimentally induced variations in heart rate produce changes in end-diastolic volume and contractility which prominently affect right ventricular stroke volume. As a consequence, rate has, over a broad range, either no significant effect on output or a negative one. With spontaneous variations in rate, additional changes in contractility and venous return occur which affect stroke volume and end-diastolic volume and enhance right ventricular output. These relationships are similar to those in the adult heart, and demonstrate the absence of a maturational change in the effects of rate on ventricular function from the fetus to the adult.

Ascending aortic blood velocity and acceleration using Doppler ultrasound in the assessment of left ventricular function.

We have used a 2.2 MHz continuous-wave Doppler blood velocity meter (Bach-Simpson BVM 202) to measure ascending aortic blood velocity and acceleration, and have obtained from the velocity signal a noninvasive measure of stroke volume and cardiac output by combining the Doppler technique with M-mode echocardiography. In two separate studies we have systematically altered the loading conditions of the heart with lower body pressure; and the inotropic state of the heart with dobutamine (5 micrograms . kg-1 . min-1), and documented the changes in mean velocity (MV), maximum acceleration (MA), stroke volume (SV), cardiac output (CO) and left ventricular end-diastolic dimension (EDD) (M-mode echocardiography). Application of lower body pressure to subjects in a 30 degrees head-up tilt position caused a systematic increase in preload, as shown by a 9% increase in EDD, which raised SV by a maximum of 33% (p less than or equal to 0.001) and CO by 32% (p less than or equal to 0.01), thus showing a classical Starling response; whilst there was relatively little increase in MA. Conversely, infusion of dobutamine, an inotropic agent, caused a 29.2% increase in MA (p less than or equal to 0.01) with minimal increase in SV. Thus, the ability to measure ascending aortic blood velocity allows noninvasive monitoring of changes in both inotropic state and Starling function, with considerable ease and rapidity.

Echocardiographic hemodynamic study during ultrafiltration sequential dialysis.

4 patients on regular dialysis were studied by the echocardiographic method during ultrafiltration and dialysis performed sequentially according to two different protocols. Blood pressure, heart rate, cardiac output, stroke volume, systolic and diastolic dimension of the left ventricle, systolic and diastolic volumes of the left ventricle, ejection fraction, shortening fraction and total peripheral vascular resistance index were measured. During ultrafiltration there is an increase of the total peripheral vascular resistance index. Myocardial contractility improves only during dialysis. Physiopathologic implications are discussed.