Search

Highlights • Falling film flow mode transition on 2D finned tubes is studied experimentally. • Large tube spacing delayed the appearance of sheet mode under increasing flow rate. • Large tube spacing prolonged the sheet mode under decreasing flow rate. • The hysteresis in mode transition was enhanced by the increase in h /(p − t). • New correlations for flow mode transition on 2D finned tubes were proposed. Abstract This study proposed a correlation of Reynolds number (Re) for falling film mode transitions on horizontal enhanced tubes with 2D integral fins on the basis of an analysis of experimental data from current and previous studies considering the effect of tube spacing and fin structures. Transitional Re for flow mode transitions are analyzed on the basis of observing water falling film flow on four types of 2D finned tubes (ranges of fin height h , fin pitch p , and fin tip t of these tubes are 0.503–1.050 mm, 1.04–2.80 mm, and 0.31–1.27 mm) and smooth tubes under various tube spacings (s , including 6, 10, and 15 mm). Results show that large tube spacing delayed the appearance of sheet mode under increasing flow rate but prolonged the sheet mode under decreasing flow rate. This hysteresis phenomenon was enhanced by the increase in h /(p − t). The proposed correlation, which linked the transitional Re with three dimensionless groups (i.e., Galileo number (Ga), s / ξ and h /(p − t)) and a constant, predicted seven groups of transitional Re for the column–sheet and sheet within ±14.7%. This correlation could be applicable to the 2D finned tubes with 19 ≤ Ga 1/4 ≤ 407, 2.19 ≤ s / ξ ≤ 5.49, and 0.33 ≤ h /(p − t) ≤ 27.67. [ABSTRACT FROM AUTHOR]

Horizontal falling film heat exchanger is widely applied to industrial area. In this work, falling film flow mode of tube bundles test bed was built for solving the basic problem of falling film heat transfer. Three tube spacing( s =6mm, 10mm, 15mm,) and three types of tube(smooth, two(2D) and three (3D) dimensional finned enhanced)were tested using water. Changing flow rate to obtain the effect of tube spacing on falling film Reynolds number(Re) at the same flow mode, the effect of tube spacing on flow mode transitions were investigated for three types tube. The result showed: (1)Increasing of tube spacing( s =6mm, 10mm, 15mm) has small effect for the Re of smooth, and lacking of enough surface tension appeared instable flow modes( s =20mm, 30mm). (2)For the Re of 2D enhanced tube, it decreased with tube spacing increased occurred in droplet and droplet-column modes, but for the column and column-sheet modes, Re was increasing (3)The tube spacing has great effect for the Re of 3D enhanced tube, and Re increased with tube spacing. The research gives guidelines to develop new efficient heat transfer equipment. [ABSTRACT FROM AUTHOR]

The performance of differential microphone arrays (DMAs) depends on many factors such as the number of sensors and the array geometry. This paper develops an approach that exploits nonuniform linear geometries and the minimum-norm filter to improve the robustness of DMAs against white noise. Unlike the conventional way that forms an N th-order DMA by subtractively combining the outputs of two DMAs of order N - 1 , this approach works in the short-time Fourier transform (STFT) domain and applies a complex weight to the output of each sensor and then sum the weighted outputs to form the beamforming output in every STFT subband. The minimum-norm filter is obtained by maximizing the white noise gain of the beamformer subject to the so-called fundamental constraints. The nonuniform linear arrays are created by adjusting the interelement spacing according to some rule. We show that the use of nonuniform linear geometries can significantly improve the robustness of DMAs, particularly at low frequencies. We also show that the diagonal loading technique can help improve the robustness of DMA beamformers, though the improvement is not significant. [ABSTRACT FROM AUTHOR]

Human intention adaptation is usually realized by changing impedance/admittance parameters according to human interaction in physical human-robot interaction (pHRI). However, determining which admittance parameters need to be updated and how to update depends on the context of pHRI and system stability. This paper aims to provide a systematic study for all three admittance parameters considering human intention adaptation and system passivity for ensuring an intuitive and safe pHRI. For intuitiveness, a general human intention framework based on the robot's roles is summarized as a premise for designing human intention-oriented updating rules for admittance parameters. For the changing trends of these parameters (i.e., increasing or decreasing), only those impacting system passivity are studied. A power envelope regulation (PER) concept is then proposed to impose constraints on variable admittance parameters inferred from human intention for maintaining safe interaction. Our initial results suggest that the passivity condition is not a sufficient condition for system stability in pHRI, whereas preserving the power envelope inferred from the passivity condition can maintain passivity as well as avoid mismatch of admittance parameters arising from unpredicted human interaction. As a result, drastic changes in admittance controller dynamics, which usually result in instability, are restrained. The effectiveness of our approaches not only to maintain intuitive interaction but also to preserve passivity is validated via numerical simulation and experiments. [ABSTRACT FROM AUTHOR]

It is important to study the falling-film pattern of a horizontal tube bundle in order to set up a heat and mass transfer model accurately. The falling-film pattern of a horizontal tube bundle is simulated in this paper. The technique is based on computational flow dynamics (CFD) for the two-phase flow of gas and water. The experimental results were used to validate the mathematical model. It indicates that the simulation results accord with experimental data well. The simulated results show that the flow pattern varies with different flow rates. Under the different flow rates, it observes the droplet, droplet-columnar, columnar, columnar-sheet and sheet flow patterns. The critical value is 0.0125 kg/s between droplet and columnar, and the critical value is 0.02 kg/s between columnar and sheet. [ABSTRACT FROM AUTHOR]

In the wastewater treatment plants (WWTPs), soft sensors are viewed as a simple signal estimator for hard-to-measure quantities. However, the presence of unreliable data, coupled with increasing demands for measurement quality assurance, has rendered inadequate such a simplistic view. In this paper, a probabilistic self-validating soft-sensor is proposed with the capability of performing self-diagnostics, self-reconstruction and online uncertainty measurement. In this framework, data collecting for soft-sensor modeling (easy-to-measure data) is validated by a Variational Bayesian Principal Component Analysis (VBPCA) model before performing a soft-sensor model construction. By integrating Relevant Vector Machine (RVM) as a predictive model, not only prediction values are obtained, but also the credibility of information for easy-to-measure and hard-to-measure quantities can be generated. The performance of the proposed soft-sensor is validated through two simulation studies of WWTPs with different process characteristics. The results suggest that the proposed strategy significantly improves the prediction performance. [ABSTRACT FROM AUTHOR]

Demand control of the air distribution systems is crucial to the satisfactory thermal environment under variable heat gain conditions in an energy efficient manner. In this study, the airflow patterns, ventilation performance and local thermal comfort have been experimentally investigated in mock-up meeting and office rooms with a variable air volume (VAV) system. The Active Diffuser, which had a constant inlet air velocity, was compared with some commonly used diffusers: Rectangular; Swirl-Radial; Swirl-Compact and Multi-Nozzle Diffusers. In the test, there were four different heat gain levels: high heat gain (46 W/m2) in a meeting room setting; high (46 W/m2), medium (24 W/m2) and low (11 W/m2) heat gain in an office room setting, respectively. The corresponding specific supply airflow rates in the high, medium and low heat gain cases were 4.3 L/s m2, 2.1 L/s m2, 1 L/s m2, respectively. Visualized figures showed the effect of VAV on airflow patterns were quite different among different diffusers. The indexes including air velocity, ventilation effectiveness, effective draft temperature, draught rating and so on were therefore analysed. By using the Active Diffuser, the airflow distribution and most of these indexes were more uniform and more stable under the VAV. Although the Rectangular, Swirl-Compact, Swirl-Radial and Multi-Nozzle Diffusers might perform better in one or two cases regarding some indexes, their airflow patterns and performances of indexes were unstable under the VAV. Thus, the Active Diffuser was recommended for the VAV system for the application where heat gains vary significantly. • The performance of five diffusers was compared in the varied airflow rate conditions • Four different heat gain cases in meeting and office room settings were studied • The airflow patterns of commonly used diffusers were unstable for the VAV system • The Active Diffuser performed better than other diffuser regarding to local thermal comfort • The Active Diffuser with constant inlet air velocity was recommended for the VAV system [ABSTRACT FROM AUTHOR]

• Precision row effect factor was provided for R134a condensed on an array of 15 horizontal plain tubes. • Row effect factors were initially determined with an experimental error within ±2.0%. • Effect of saturated temperature and heat flux on the row effect was investigated. • New test facility for the accurate determination of row effect on film condensation was constructed. The effect of condensate inundation on the film condensation heat transfer of R134a on an array of 15 horizontal plain tubes was experimentally investigated with a new experimental method. This method could accurately determine the row effect. A test facility, which has a test section and contains two arrays of 17 horizontal plain tubes, was constructed to realize the new experimental method. The outer diameter of the plain tube was 19.05 mm, and the tube length in the test section was 500 mm. The test condition was conducted at saturation temperatures of 38.7 °C, 40.6 °C, and 43.5 °C and heat flux of 14.8, 18.0, and 21.6 kW/m2. The experimental error of the determined row effect factor was within ±2.0%, which was only 14.0% of that of the common experimental method (±14.5%). Experimental results revealed that the film condensation row effect factor for the plain tube initially decreased rapidly and then increased gradually with the number of tube rows. In comparison with the saturated temperature, the row effect was more sensitive to heat flux condition for n range from 8 to 15. The precision row effect can be used as benchmark data to perfect and develop the film condensation theory. [ABSTRACT FROM AUTHOR]

Dereverberation technology is needed in a wide range of speech applications as reverberation often greatly degrades the quality and intelligibility of the speech signal of interest captured by microphones. The commonly used weighted-prediction-error method generally requires long prediction-error filters to remove the reverberation components, which makes it computationally expensive. To deal with this issue, this paper proposes a computationally efficient dereverberation algorithm based on tensor decomposition in which the long prediction-error filter is decomposed into a group of short sub-filters through multiple Kronecker products. Consequently, the high dimensional cross-correlation matrix that needs to be inverted in the dereverberation algorithm is then converted into a set of low dimensional matrices, which leads to significant reduction in the computational complexity. Simulation results demonstrate the properties of the proposed algorithm. • A tensor decomposition based speech dereverberation algorithm is presented in the framework of the weighted-prediction-error (WPE) algorithm. • The prediction-error filter is decomposed into a set of short sub-filters via multiple Kronecker products to reduce the computational cost. • The proposed algorithm is computationally more efficient than the WPE algorithm and has comparable dereverberation performance with WPE. [ABSTRACT FROM AUTHOR]

The widely studied subspace and linear filtering methods for noise reduction require the noise correlation matrix to be invertible. In certain application scenarios, however, this matrix is either rank deficient or very ill conditioned, so this requirement cannot be fulfilled. In this paper, we investigate possible solutions to this important problem based on subspace techniques for single-channel time-domain noise reduction. The eigenvalue decomposition is applied to both the speech and noise correlation matrices to separate the null and nonnull subspaces. Then, a set of optimal and suboptimal filters are derived from the nullspace of the noise signal. Through simulations, we observe that the proposed filters are able to significantly reduce noise without introducing much distortion to the desired signal. In comparison with the conventional Wiener approach, the developed filters perform significantly better in improving both the signal-to-noise ratio (SNR) and the perceptual evaluation of speech quality (PESQ) score when the noise correlation matrix is rank deficient. [ABSTRACT FROM AUTHOR]

Due to their small size, differential microphone arrays (DMAs) are very attractive. Moreover, they have been effective in combating noise and reverberation. Recently, a new class of DMAs of different orders have been developed with the MacLaurin’s series and the frequency-independent patterns. However, the MacLaurin’s series does not approximate well the exponential function, which appears in the general definition of the beampattern, when the intersensor spacing is not small enough. To circumvent this problem, we propose in this paper to approximate the exponential function with the Jacobi–Anger expansion. Based on this approximation and the frequency-independent Chebyshev patterns, we derive first-, second-, and third-order DMAs. Furthermore, in order to improve the robustness of DMAs against white noise amplification, we propose to use more microphones combined with minimum-norm filters. It is also shown that the Jacobi–Anger expansion is optimal from a mean-squared error perspective. Simulations are carried out to evaluate the performance of the proposed DMAs. [ABSTRACT FROM AUTHOR]

With the advent of information technology, online platforms have become an vital intermediary for acquiring green information and purchasing green products. This study uses two-sided market theory to estimate the size of the green market for green subsidies under various targets. First, we compare the differences in the equilibrium market size and platform profits for three scenarios (no subsidy, maximum platform profits, and maximum social welfare). In addition, we measure the basic conditions for start-up subsidy schemes. Finally, we employ numerical analysis to explore how other critical factors, including cross-network externalities, product pricing, and research and development (R&D) costs, impact the market size in three scenarios. The result demonstrates how online platforms use cross-side network externalities to design asymmetric pricing strategies or subsidy policies to develop the market for green products. It has managerial implications for online platform owners or politicians in subsidizing the expansion of green products. • Two-sided market theory can is used in the scenario of green product diffusion. • Online platforms is an vital intermediary for acquiring green information and purchasing green products. • Online platforms offer subsidies for green products to increase profits. • Green product subsidy policy needs to change according to market size. • Subsidies to consumers are more beneficial to expand the market. [ABSTRACT FROM AUTHOR]

Biological mechanism of miR-210-3p in endometrial carcinoma (EC) remains unclear. Here, our purpose is to study effects of miR-210-3p on malignant progression of EC. Bioinformatics analysis showed miRNA and mRNA are abnormally expressed in EC tissues. Quantitative real-time fluorescence polymerase chain reaction (qRT-PCR) was utilized to compare miR-210-3p mRNA level in EC cells and tissues. qRT-PCR and western blot were used to measure RUNX1T1 and NCAM1 at mRNA and protein levels, and western blot for p-AKT and AKT proteins related to PI3K/AKT signaling pathway. Furthermore, EC cell behaviors were assayed via Cell Counting Kit-8, cell colony formation assay, wound healing, transwell and flow cytometry experiments. Interaction between RUNX1T1 and miR-210-3p was verified through dual-luciferase assay. Immunohistochemistry was used to analyze RUNX1T1 expression in clinical samples MiR-210-3p was considerably upregulated and RUNX1T1 was significantly under-expressed in EC. Overexpression of miR-210-3p stimulated cell proliferation, migration, invasion, and restrained cell apoptosis in EC. Dual-luciferase assay proved that RUNX1T1 was a target gene of miR-210-3p. The level of RUNX1T1 in EC was downregulated after overexpressing miR-210-3p. Rescue assay showed that overexpression of RUNX1T1 had an inhibitory impact on tumor-relevant cell behaviors, whereas overexpression of miR-210-3p rescued such inhibition. Overexpression of RUNX1T1 reduced p-AKT expression, which was restored with concomitantly overexpressed miR-210-3p. In general, miR-210-3p behaves as an oncogene in EC by down-regulating the expression of RUNX1T1. This study elucidates a new functional mechanism in EC, and indicates miR-210-3p an underlying target. • MiR-210-3p was considerably upregulated and RUNX1T1 was significantly under-expressed in endometrial carcinoma. • Overexpression of miR-210-3p stimulated tumor-relevant cell functions of EC. • MiR-210-3p was an oncogene of EC by down-regulating the expression of RUNX1T1. [ABSTRACT FROM AUTHOR]

Abstract: Conventional multichannel noise reduction techniques are formulated by splitting the processed microphone observations into two terms: filtered noise-free speech and residual additive noise. The first term is treated as desired signal while the second is a nuisance. Then, the objective has typically been to reduce the nuisance while keeping the filtered speech as similar as possible to the clean speech. It turns out that this treatment of the overall filtered speech as the desired signal is inappropriate as will become clear soon. In this paper, we present a new study of the multichannel time-domain noise reduction filters. We decompose the noise-free microphone array observations into two components where the first is correlated with the target signal and perfectly coherent across the sensors while the second consists of residual interference. Then, well-known time-domain filters including the minimum variance distortionless response (MVDR), the space–time (ST) prediction, the maximum signal-to-noise ratio (SNR), the linearly constrained minimum variance (LCMV), the multichannel tradeoff, and Wiener filters are derived. Besides, the analytical performance evaluation of these time-domain filters is provided and new insights into their functioning are presented. Numerical results are finally given to corroborate our study. [Copyright &y& Elsevier]

• Condensing, falling, and acting condensates are identified and defined. • Effect of different categories of condensates on film condensation is investigated. • Film condensation model considering different categories of condensates is built. • Precise inundation effect coefficient is determined by homologous method. The condensates generated in the process of film condensation, such as condensing, falling, and acting condensates, are identified and defined to obtain precise inundation effect model of 3D finned tube bundle. Furthermore, the influence of three categories of condensates on the film condensation under different heat fluxes and condensing temperatures is experimentally investigated via the homologous method. A test bench with a test section contains two arrays of 8 horizontal tubes is initially constructed. Experimental results show that the inundation effect coefficient decreases from 1.00 to 0.78 as acting condensate Re increases from 75 to 1860. The influence of heat flux on three categories of condensates is more significant than that of condensing temperature. As tube row depth increases from 1 to 6, the proportion of condensing condensate Re to acting condensate Re decreases from 100% to 15.0%, whereas the proportion of falling condensate Re to acting condensate Re increases from 0 to 85.0%. Furthermore, the condensation heat transfer model of single 3D finned tube and inundation effect model are established, and the deviations of results between experiments and models are almost within ±3.0%. The deviations of inundation effect coefficients between experiment and model proposed in this study are 92.5% and 50.0% less than those between experiment and models using routine method and considering single condensate with homologous method, respectively. Moreover, the inundation effect model proposed in this study reveals that different proportions of condensing or falling condensates to acting condensate lead to different inundation effect coefficient at the same acting condensate Re. The research results will lay the foundation for further revealing the mechanism of film condensation on 3D finned tube bundle and guide the thermal design of 3D finned tubes applied in shell-and-tube condensers. [ABSTRACT FROM AUTHOR]

• New test facility for inundation effect coefficient of 3D finned tube is built. • Precise inundation effect coefficient is determined by homologous method. • Inundation effect on an array of 6 horizontal new 3D finned tubes is investigated. • Inundation effect model of new 3D finned tube is established. • Inundation effect coefficient in this article is 10-30% higher than that in others. To accurately evaluate the condensation heat transfer performance of new 3D finned tubes in shell-and-tube condenser, the inundation effect of the film condensation of R-134a on an array of 6 horizontal new 3D (three-dimensional) finned tubes is experimentally investigated with a new experiment method of homologous method. And the precise inundation effect coefficient of film condensation is determined. A novel test facility, whose test section contains two arrays of 8 horizontal tubes to realize the homologous method, is developed and constructed. The outer diameter of the tube and effective tube length in the test section are 19.05 and 500 mm, respectively. The condensation heat transfer coefficient of new 3D finned tube and inundation effect coefficient of new 3D finned tube bundle are determined under different conditions. The results show that the inundation effect coefficient of new 3D finned tube bundle decreases from 1.00 to 0.78 as Re increases from 75 to 1860. And the effect of heat flux on the inundation effect is more sensitive than the condensing temperature. Furthermore, for new 3D finned tube, the condensation heat transfer and inundation effect models are established, and the deviations of results between experiment and models are almost within ±6.0%. In comparison with literatures, the average inundation effect coefficient of new 3D finned tube bundle is 10.0-25.0% higher than that of 3D finned tube bundle in literatures, within 3.0% lower than that of two-dimensional finned tube bundle, and 15.2% higher than that of plain tube bundle. And the average condensation heat transfer coefficient of new 3D finned tube bundle in this article is 5.0-25.0% higher than that of 3D finned tube bundle in literatures, 10.0-30.0% higher than that of 2D finned tube bundle, and more than 10 times that of plain tube bundle. The precise inundation effect coefficients can be used as benchmark data to build precise and universal inundation effect model of 3D finned tube, and the models established in this study will help guide the thermal design of the new 3D finned tube applied in shell-and-tube condenser. [ABSTRACT FROM AUTHOR]

• New correlations for flow mode transition considering the liquid distribution conditions were first proposed. • A database of the transitional Reynolds number with six new designed liquid distributors was presented. • Effect of liquid distribution conditions on falling film flow mode transitions was experimentally determined. • The flow mode transitions and hysteresis in it were found strongly affected by liquid distribution conditions. • A test bed with six new designed liquid distributors was established. This study investigated the effect of nonuniform liquid distribution on the evolution of falling film flow mode on an array of horizontal tubes. A falling film flow mode observation system with water as the working fluid was designed and evaluated. In the test system, six liquid distributors, namely, DIS1, DIS2, DIS3, DIS4, DIS5 and DIS6, were designed to realise different liquid distribution conditions. The test tube array was composed of smooth copper tubes in which their outer diameter, tube length and tube spacing were 19.05, 280 and 10.0 mm, respectively. Results showed that the falling film flow mode transitions were strongly affected by liquid distribution conditions, and the falling film flow mode transitions under nonuniform liquid distribution conditions could not be predicted well by using the classic models that have been established under uniform liquid distribution conditions. For the droplet to droplet–column (D → DC), droplet–column to column (DC → C), column to column–sheet (C → CS) and column–sheet to sheet (CS → S) flow mode transitions, the transitional Reynolds number (Re tr) for uniform condition (Re tr,u) were 105, 145, 352 and 436, and the differences between Re tr for the six nonuniform and uniform conditions were 50.4%, 38.7%, 18.1%, 5.93%, 23.4% and 37.1%; 22.1%, 37.0%, 23.3%, 55.7%, 51.9% and 55.3%; 10.7%, 3.53%, 10.7%, 12.4%, 16.7% and 29.6% and 94.1%, 58.2%, 13.7%, 26.2%, 29.6% and 96.8%, respectively. A semiempirical model was established to predict the evolution of falling film flow mode under nonuniform liquid distribution conditions. This study serves as reference for improving the flow pattern determination criteria of falling condensate on horizontal tube bundles. [ABSTRACT FROM AUTHOR]

• CdSe quantum disks, whose photoluminescence peak coincide with the EQE peak of the CIGS solar cells, have been synthesized and embedded into the PMMA to form the LDS layer for CIGS solar cells. • The LDS layer increases the EQE of CIGS devices by up to about 162% in the range 300–460 nm. • The placement of the LDS layer upon the CIGS solar cell improved the electrical power output of the bare cell to 114%. As the luminescent down-shifting (LDS) material, Cadmium selenide (CdSe) quantum disks (QDs) have been synthesized for improving the short-wavelength response of Cu(In, Ga)Se 2 (CIGS) solar cells. The peak of the photoluminescence spectra of the as-prepared QDs is coinciding with the peak of the external quantum efficiency (EQE) of CIGS solar cell. The absorption and photoluminescence spectra of the as-prepared CdSe QDs show that the photons of wavelength below 460 nm, which could not be converted into electricity by CIGS solar cell, would be absorbed by the QDs and generating longer wavelength photons that could contribute to charge carrier generation by the cell. The QDs were embedded into the polymethylmethacrylate (PMMA) forming the colloidal fluorescent mixture. Subsequently, the mixture was applied to the surface of the solar cell as the LDS layer. It is found that the EQE of CIGS devices increases up to about 25% in the range of 300–460 nm due to LDS. The placement of the LDS layer upon the CIGS solar cell improved the electrical power output of the bare cell to 114%. [ABSTRACT FROM AUTHOR]

We present the generation of optofluidic lasers in a Fabry-Pérot (FP) microcavity with low pumping threshold. The FP microcavity is constructed with one micro-concave mirror and another planar mirror facing each other to form a stable cavity. The two mirrors are coated with dielectric layers with reflectivity ≥99.9%. Laser behaviors are tested with gain medium, Rhodamine 6 G (R6G), dissolved in ethanol and ion water, respectively; a lowest pumping threshold as small as 0.13 μJ/mm 2 is achieved in the ethanol solution, corresponding to a theoretically calculated quality value Q 0 about 4×10 5 . Meanwhile experimental results demonstrates that the lowest concentration of dye medium that can emit laser signal in the FP microcavity is 3 μM. [ABSTRACT FROM AUTHOR]

The performance of adaptive beamformers may suffer from significant degradation in the presence of steering vector errors, statistics estimation errors, and reverberation. To address this issue, robust beamforming methods, which were originally studied in the narrow-band cases, are studied and compared in this paper for processing acoustic and speech signals, which are broadband in nature. We study two types of methods. In the first type, the robustness of the beamformer is improved by adding a norm constraint and/or a steering vector uncertainty constraint to the optimization problem. It is worth noticing that the norm constraint also helps to control the sidelobes of the beampattern, which makes the beamformers able to suppress interference and multipath effect, thereby improving the robustness of the beamformers with respect to reverberation. Another type of methods improve the robustness using the spatial smoothing technique in which the noise covariance matrix is implicitly estimated first by subtracting an estimate of the desired signal covariance matrix with a delay-and-sum beamformer from the observation signal covariance matrix. Simulations and experiments are performed to investigate the performance of the studied robust adaptive beamformers in acoustic environments. The results show that the robust beamformers outperform their non-robust counterparts in terms of: (1) better performance in reverberation and different noise levels; (2) resilience against steering vector and noisy signal covariance matrix estimation errors; and (3) better predicted speech quality and intelligibility measured using the PESQ and STOI measures. [ABSTRACT FROM AUTHOR]