Search Results (52)

Blind source separation (BSS) of underwater acoustic mixed signals aims to improve signal clarity by separating noise components from aliased underwater signal sources. This enhancement directly increases target detection accuracy in underwater acoustic perception systems, particularly in scenarios involving multi-vessel interference or biological sound coexistence. Deep learning-based BSS methods have gained wide attention for their superior nonlinear modeling capabilities. However, existing approaches in underwater acoustic scenarios still face two key challenges: limited feature discrimination and inadequate robustness against non-stationary noise. To overcome these limitations, we propose a novel Robust Cross-Band Network (RCBNet) for the BSS of underwater acoustic mixed signals. To address insufficient feature discrimination, we decompose mixed signals into sub-bands aligned with ship noise harmonics. For intra-band modeling, we apply a parallel gating mechanism that strengthens long-range dependency learning so as to enhance robustness against non-stationary noise. For inter-band modeling, we design a bidirectional-frequency RNN to capture the global dependency relationships of the same signal across sub-bands. Our experiment demonstrates that RCBNet achieves a 0.779 dB improvement in the SDR compared to the advanced model. Additionally, the anti-noise experiment demonstrates that RCBNet exhibits satisfactory robustness across varying noise environments. Full article

Direct numerical simulations (DNSs) of the T106A low-pressure turbine were conducted for various turbulence intensities and length scales to investigate their effects on flow behaviour and transition. A source-term formulation of the synthetic eddy method (SEM) was implemented in the Nektar++ spectral/hp element framework to introduce anisotropic turbulence into the flow field. A single sponge layer was imposed, which covers the inflow and outflow regions just downstream and upstream of the inflow and outflow boundaries, respectively, to avoid acoustic wave reflections on the boundary conditions. Additionally, in the T106A model, mixed polynomial orders were utilized, as Nektar++ allows different polynomial orders for adjacent elements. A lower polynomial order was employed in the outflow region to further assist the sponge layer by coarsening the mesh and diffusing the turbulence near the outflow boundary. Thus, this study contributes to the development of a more robust and efficient model for high-fidelity simulations of turbine blades by enhancing stability and producing a more accurate flow field. The main findings are compared with experimental and DNS data, showing good agreement and providing new insights into the influence of turbulence length scales on flow separation, transition, wake behaviour, and loss profiles. Full article

This study presents an advanced NO_x reduction strategy for a 330 MW lignite-fired boiler using an immunological AI system: the SILO (Stochastic Immune Layer Optimizer) combustion optimizer inspired by artificial immune systems. The immunological AI optimizer adaptively models multi-variable interactions and fireball shape in real time, optimizing fuel–air mixing to reduce NO_x formation at the source. Unlike reactive secondary methods, the combustion optimizer reshapes the combustion process to reduce emissions while improving efficiency. Real-time temperature data from the AGAM acoustic system inform the combustion optimizer’s fireball modeling, ensuring combustion uniformity. A urea-based SNCR system serves as a secondary layer, controlled based on local furnace conditions to target thermal zones. Field results confirmed that SILO reduced NO_x emissions below 200 mg/Nm³, decreased urea consumption by up to 34%, and improved boiler efficiency by 0.29%. The architecture offers a scalable, DCS-integrated solution for aligning fossil-fueled operations with tightening emission standards. Full article

The propagation of high-frequency ultrasound waves will generate both physical and chemical effects as they propagate through a liquid medium, such as acoustic streaming, an acoustic fountain, and atomization. These phenomena are believed to be the main factors that contribute to the enhancement of mass transfer in the gas–liquid carbon dioxide (CO₂) absorption system. Computational Fluid Dynamic (CFD) simulation is one of the powerful tools that can be used to model the complex hydrodynamic behavior induced by the propagation of ultrasound waves in the liquid medium. In this study, the ultrasonic irradiation forces were simulated via the momentum source term method using commercial CFD software (ANSYS Fluent V19.1). In addition, a parametric study was conducted to investigate the influences of absorber height and ultrasonic power on the hydrodynamic mixing performance. The simulation results indicated that enhanced mixing and a higher intensification factor were achieved with increased fountain flow velocity, particularly at the lowest absorber height and highest ultrasonic power. Conversely, the energy efficiency was improved with the increase of absorber height and decrease of ultrasonic power. To determine the optimal combination of absorber height and ultrasonic power, this trade-off between the energy efficiency and intensification in the ultrasonic-assisted absorption system (UAAS) is a crucial consideration during process scale-up. Full article

Addressing the significant challenge of speech enhancement in ultra-low-Signal-to-Noise-Ratio (SNR) scenarios for Unmanned Aerial Vehicle (UAV) voice communication, particularly under edge deployment constraints, this study proposes the Edge-Deployed Band-Split Rotary Position Encoding Transformer (Edge-BS-RoFormer), a novel, lightweight band-split rotary position encoding transformer. While existing deep learning methods face limitations in dynamic UAV noise suppression under such constraints, including insufficient harmonic modeling and high computational complexity, the proposed Edge-BS-RoFormer distinctively synergizes a band-split strategy for fine-grained spectral processing, a dual-dimension Rotary Position Encoding ($\mathrm{RoPE}$) mechanism for superior joint time–frequency modeling, and $\mathrm{FlashAttention}$ to optimize computational efficiency, pivotal for its lightweight nature and robust ultra-low-SNR performance. Experiments on our self-constructed DroneNoise-LibriMix (DN-LM) dataset demonstrate Edge-BS-RoFormer’s superiority. Under a −15 dB SNR, it achieves Scale-Invariant Signal-to-Distortion Ratio (SI-SDR) improvements of 2.2 dB over Deep Complex U-Net (DCUNet), 25.0 dB over the Dual-Path Transformer Network (DPTNet), and 2.3 dB over HTDemucs. Correspondingly, the Perceptual Evaluation of Speech Quality (PESQ) is enhanced by 0.11, 0.18, and 0.15, respectively. Crucially, its efficacy for edge deployment is substantiated by a minimal model storage of 8.534 MB, 11.617 GFLOPs (an 89.6% reduction vs. DCUNet), a runtime memory footprint of under 500MB, a Real-Time Factor (RTF) of 0.325 (latency: 330.830 ms), and a power consumption of 6.536 W on an NVIDIA Jetson AGX Xavier, fulfilling real-time processing demands. This study delivers a validated lightweight solution, exemplified by its minimal computational overhead and real-time edge inference capability, for effective speech enhancement in complex UAV acoustic scenarios, including dynamic noise conditions. Furthermore, the open-sourced dataset and model contribute to advancing research and establishing standardized evaluation frameworks in this domain. Full article

The inherent energy constraints of sensor nodes render energy efficiency optimization a critical challenge in wireless sensor network deployments. This study presents an innovative acoustic source localization framework incorporating a two-level data aggregation technology, specifically designed to minimize energy expenditure while prolonging network lifetime. A mixed noise model is proposed to describe the characteristics of abnormal noise in real environments. Subsequently, the novel two-level data aggregation technology is proposed. The first level is implemented at individual sensors, where a large number of similar measurements may be collected. The second level data aggregation technology is performed at the cluster head nodes to eliminate the data redundancy between different sensor nodes. After the novel two-level data aggregation, most of the redundant data are eliminated and a significant amount of energy is saved. Then, a nonlinear iterative weighted least squares algorithm is applied to complete the final acoustic source location estimation based on the real remaining sensor measurements. Finally, through extensive simulation experiments, it was verified that the two-level data aggregation technology reduced energy consumption by at least 51% and 43%, respectively, and that the RMSE is less than 0.96. Full article

Railways are crucial in economic development and improving people’s livelihoods. Therefore, defect detection and maintenance of rails are particularly important. In order to accurately separate and identify the rail defect from the mixed signals by acoustic emission (AE) techniques, this paper proposes a novel time–frequency separation neural network (TFSNN) architecture to solve the problems existing in the blind source separation (BSS), such as in non-stationary signals and low stability in the convergence. Combined with the smoothed pseudo Wigner–Ville distribution (SPWVD), this method can increase the spectrogram resolution, suppress the noise interference, and effectively improve the extraction performance of crack signals. In addition, 1D-CNN and GRU structures were introduced in the TFSNN structure to exploit the dominant features from AE signals. A dense regressor was also subsequently used to estimate the separation weights. Simulation and experiments showed that compared with traditional algorithms like independent component analysis, shallow neural networks, and time–frequency blind source separation, the proposed algorithm can provide better separation performance and higher stability in rail crack detection. Full article

Efficiently extracting effective information from the massive experimental data from physical mechanics and accurately identifying the premonitory failure information from coal rock are key and difficult points of intelligent research on rock mechanics. In order to reveal the deterioration characteristics and the forewarning law of fractured coal rock, the digital image correlation method and the acoustic emission technology were adopted in this study to non-destructively detect the strain field, displacement field, and acoustic emission response in time and frequency domains. Additionally, by introducing the derivative functions of the multi-source information function for quantitative analysis, a comprehensive evaluation method was proposed based on the multi-source information fusion monitoring to forewarn red sandstone failure by levels during loading. The results show that obvious premonitory failure information, such as strain concentration areas, appears on red sandstone’s surface before macro-cracks can be observed. With an increase in the inclination angle of the prefabricated crack, the macroscopic failure mode gradually transforms from tensile splitting failure to tensile-shear mixed failure. Moreover, the dominant frequency signals of high frequency–low amplitude (HF–LA), intermediate frequency–low amplitude (IF–LA) and low frequency–low amplitude (LF–LA) are denser near the stress peak. The initial crack expansion time and failure limit time measured by multi-source information fusion are 20.72% and 26.71% earlier, respectively, than those measured by direct observation, suggesting that the forewarning of red sandstone failure by levels is realized with multi-source information fusion. Full article

The “cocktail party problem”, the challenge of isolating individual speech signals from a noisy mixture, has traditionally been addressed using statistical methods. However, deep neural networks (DNNs), with their ability to learn complex patterns, have emerged as superior solutions. DNNs excel at capturing intricate relationships between mixed audio signals and their respective speech sources, enabling them to effectively separate overlapping speech signals in challenging acoustic environments. Recent advances in speech separation systems have drawn inspiration from the brain’s hierarchical sensory information processing, incorporating top-down attention mechanisms. The top-down attention network (TDANet) employs an encoder–decoder architecture with top-down attention to enhance feature modulation and separation performance. By leveraging attention signals from multi-scale input features, TDANet effectively modifies features across different scales using a global attention (GA) module in the encoder–decoder design. Local attention (LA) layers then convert these modulated signals into high-resolution auditory characteristics. In this study, we propose two key modifications to TDANet. First, we substitute the fully trainable convolutional encoder with a deterministic hand-crafted multi-phase gammatone filterbank (MP-GTF), which mimics human hearing. Experimental results demonstrated that this substitution yielded comparable or even slightly superior performance to the original TDANet with a trainable encoder. Second, we replace the single multi-head self-attention (MHSA) layer in the global attention module with a transformer encoder block consisting of multiple MHSA layers. To optimize GPU memory utilization, we introduce a parameter sharing mechanism, dubbed “Reverse Cycle”, across layers in the transformer-based encoder. Our experimental findings indicated that these proposed modifications enabled TDANet to achieve competitive separation performance, rivaling state-of-the-art techniques, while maintaining superior computational efficiency. Full article

As a critical component in industrial production, pipelines face the risk of failure due to long-term corrosion. In recent years, acoustic emission (AE) technology has demonstrated significant potential in online pipeline monitoring. However, the interference of flow-induced noise seriously hinders the application of acoustic emission technology in pipeline corrosion monitoring. Therefore, a pattern-recognition model for online pipeline AE monitoring signals based on blind source separation (BSS) and a convolutional neural network (CNN) is proposed. First, the singular spectrum analysis (SSA) was employed to transform the original AE signal into multiple observed signals. An independent component analysis (ICA) was then utilized to separate the source signals from the mixed signals. Subsequently, the Hilbert–Huang transform (HHT) was applied to each source signal to obtain a joint time–frequency domain map and to construct and compress it. Finally, the mapping relationship between the pipeline sources and AE signals was established based on the CNN for the precise identification of corrosion signals. The experimental data indicate that when the average amplitude of flow-induced noise signals is within three times that of corrosion signals, the separation of mixed signals is effective, and the overall recognition accuracy of the model exceeds 90%. Full article

Most analyses of monophthong change have historically relied on static acoustic measures. It is unclear the extent to which dynamic measures can shed greater light on monophthong change than can already be captured using such static approaches. In this study, we conducted a real-time trend analysis of vowels in corpora collected from female Mainstream Australian English (MAusE) speakers under 30 years of age across three time periods: the 1960s, 1990s, and 2010s. Using three different methods for characterising the first and second formants (the target-based approach, discrete cosine transform (DCT), and generalised additive mixed model (GAMM)), we statistically examined differences for each of 10 monophthongs to outline change over the fifty-year period. Results show that all three methods complement each other in capturing the changing vowel system, with the DCT and GAMM analyses superior in their ability to provide greater nuanced detail that would be overlooked without consideration of dynamicity. However, if consideration of the vowel system as a whole is of interest (i.e., the relationships between the vowels), visualising the vowel space can facilitate interpretation, and this may require reference to static measures. We also acknowledge that locating the source of vowel dynamic differences in sound change involves reference to surrounding phonetic context. Full article

Ship-radiated noise is the main basis for ship detection in underwater acoustic environments. Due to the increasing human activity in the ocean, the captured ship noise is usually mixed with or covered by other signals or noise. On the other hand, due to the softening effect of bubbles in the water generated by ships, ship noise undergoes non-negligible nonlinear distortion. To mitigate the nonlinear distortion and separate the target ship noise, blind source separation (BSS) becomes a promising solution. However, underwater acoustic nonlinear models are seldom used in research for nonlinear BSS. This paper is based on the hypothesis that the recovery and separation accuracy can be improved by considering this nonlinear effect in the underwater environment. The purpose of this research is to explore and discover a method with the above advantages. In this paper, a model is used in underwater BSS to describe the nonlinear impact of the softening effect of bubbles on ship noise. To separate the target ship-radiated noise from the nonlinear mixtures, an end-to-end network combining an attention mechanism and bidirectional long short-term memory (Bi-LSTM) recurrent neural network is proposed. Ship noise from the database ShipsEar and line spectrum signals are used in the simulation. The simulation results show that, compared with several recent neural networks used for linear and nonlinear BSS, the proposed scheme has an advantage in terms of the mean square error, correlation coefficient and signal-to-distortion ratio. Full article

Today, the application of green materials in the building industry is the norm rather than the exception and reflects an attempt to mitigate the sector’s environmental impacts. Mass timber is growing rapidly in the construction field because of its long span, speed of installation, lightness and toughness, carbon sequestration capabilities, renewability, fire rating, acoustic isolation, and thermal resistance. Mass timber is close to overtaking steel and concrete as the preferred material. The endeavor of this research is to quantitatively assess the ability of this green material to leverage the abatement of carbon emissions. Life cycle assessment (LCA) is a leading method for assessing the environmental impacts of the building sector. The recently completed Adohi Hall mass timber building on the University of Arkansas campus was used as a case study in an investigation to quantify greenhouse gas (GHG) emissions throughout the construction phase only. The energy used in building operations is the most dominant source of emissions in the building industry and has galvanized research on increasing the efficiency of building operations, but reduced emissions have made the impacts of embodied carbon (EC) components more noticeable in the building life cycle. While most studies have focused on the manufacturing stage, only a few to date have focused on the construction process. Consequently, few data are available on the environmental impacts associated with the installation of mass timber as a new green material. The present study began with the quantification of the materials and an inventory of the equipment used for construction. Then, this study determined the EC associated with running the equipment for building construction. The GHG emissions resulting from the transportation of materials to the site were also quantified. Based on data collected from the construction site, the results of this study indicate that earthwork ranks first in carbon emissions, followed by mass timber installation and construction. In third place is ready-mix poured concrete and rebar installation, followed by Geopiers. A comparison of these results with those in the existing literature shows that the EC generally associated with the building construction phase has been underestimated to date. Furthermore, only emissions associated with the fuel usage of the main equipment were considered. Full article

The sound insulation of the façade and its elements is a very important characteristic, as it largely determines the degree of sound protection of the building’s interior from external noise sources. This feature, therefore, has a great influence on the acoustic comfort and health of the occupants. For this reason, it is very important that the way in which the sound insulation of the façade is quantified and represented corresponds to the way it is perceived. Although there have long been regulations describing how it should be measured and expressed through Single-Number Quantities (SNQs), there is much scientific debate about the appropriateness of current standardised methods for expressing sound insulation, in terms of whether they accurately represent human-perceived comfort. In this regard, much of the debate centres on the frequency range to be considered when expressing sound insulation, with no consensus as to whether the low-frequency bands (i.e., 50, 63, and 80 Hz) should be used for the calculation of façade sound insulation SNQs. In order to contribute to this knowledge, we conducted a listening test using a Two-Alternative Choice (2-AC) protocol on a sample of 100 participants to test whether participants’ annoyance with urban noise changed significantly with variations in window sound insulation only in the low-frequency range. The results of the experiment, analysed using Thurstonian models, showed that the influence of low frequencies is limited for the sound insulation of the tested window façade elements and urban-type noise of aircraft and mixed urban traffic at low speeds and only becomes relevant when the sound insulation of the elements is exceptionally low in the low-frequency range. Full article

Acoustic emission (AE) testing is a widely used nondestructive testing method for the early detection of failures in materials and structures. In this paper, an AE detection sensor combining optical fiber sensing with laser self-mixing interference (SMI) technology is proposed. A multi-coil optical fiber ring wound round a cylindrical acrylic skeleton was designed in order to sense the deformation caused by AE elastic waves, which was then demodulated using self-mixing interference technology. Finite element analyses were conducted in order to investigate the deformation of fiber under acoustic sources. AE signals induced via ball-dropping impact experiments were successfully detected by the proposed experimental system. The proposed SMI optical fiber AE sensing system has the advantages of being free from electromagnetic interference and having a simple structure, low implementation cost and high measurement resolution and sensitivity. Full article