Acoustics

Particle pollution has been recognized as a significant form of environmental pollution. More specifically, the inhalation of very small (ultrafine) airborne particulate matter that is emitted from the burning of fossil fuels poses one of the most serious threats to human health. High-efficiency retention of these particles is one of the most challenging environmental problems, since conventional techniques such as electrostatic precipitators, bag filters, or cyclones have low collection efficiency in the respirable range (0.1 μm–1.0 μm). Acoustically induced agglomeration of ultrafine particles is a promising technique to increase the size of small particles before they enter a conventional filter. During this process, high-intensity acoustic fields are applied to the flue gas stream, inducing interactions among suspended particles that give rise to collisions and agglomeration. The preconditioned aerosol can then be filtered within conventional filters with higher collection efficiency. The present work reports the results of a numerical investigation of the effect of ultrasound preconditioning on the particle size distribution as a function of parameters related to the ultrasound system design, such as the sound frequency and intensity, and the initial distribution characteristics. Particle agglomeration is modeled via the solution of the population balance equation with the Multi-Monte Carlo method. Results show that acoustic agglomeration can shift the particle size distribution toward larger diameters and reduce the number concentration of particles, thus leading to increased capture efficiency of conventional filters. Full article

Timber façade assemblies are increasingly used in residential construction for reasons including prefabrication, construction speed, and lower environmental footprint. However, adequate acoustic performance in lightweight assemblies requires additional materials, which affect both thermal performance and embodied carbon emissions. This study investigates the trade-offs between airborne sound insulation (R_w), thermal transmittance (U_value), and embodied carbon in timber façade assemblies. 382 assemblies, organized into seven families, were compiled from the literature. Correlation and Pareto-based multi-objective analyses were applied to evaluate the relationships among the three performance metrics. The findings indicate a positive correlation between acoustic and thermal performance across families, with R² values ranging from 0.69 to 0.88. Nevertheless, no correlation was found between acoustic insulation and embodied carbon, indicating that acoustic improvements can be achieved without proportional increases in environmental impact. Pareto analysis identified 45 Pareto-optimal assemblies representing the best achievable trade-offs between the three criteria. For façade assemblies with overall thickness ranging from 287.5 to 477 mm, increasing insulation thickness from 120 to 300 mm in 40 mm increments improves acoustic and thermal performance by 1 dB and 0.02–0.04 W/m² K, respectively, at an embodied carbon cost of 2–3 kg CO₂ eq/m². A catalogue of seven representative assemblies is included in the article as a practical design reference. Full article

The present study investigates the flow-acoustic interactions within the slat and flap coves of a 30P30N high-lift configuration. Velocity fluctuations in the slat cove exhibit a dominant tonal feature at a Strouhal number based on the slat chord $\left(c_s\right)$, $S{t}_s=f{c}_s/U_{\infty }=1.75$ ($f\approx$ 1500 Hz), corresponding to the second acoustic mode predicted by the feedback model. Strong coherence between wall-pressure sensors in the slat cove and shear layer velocity fluctuations confirms this coupling. The same tonal signature appears in the flap cove wall-pressure spectra, with high coherence between slat and flap surface-pressure signals, indicating propagation of acoustic waves from the slat to the flap. However, no tonal content is observed in the flap cove velocity spectra, suggesting that while acoustic waves propagate between coves, they do not excite shear-layer unsteadiness within the flap cove. Full article

Continuous monitoring of gas-induced changes in coal is important for carbon dioxide storage, coalbed methane recovery, and underground mine safety. Conventional ultrasonic monitoring primarily relies on direct-wave velocities, which may exhibit limited sensitivity to subtle, distributed changes within the coal microstructure. This study evaluates coda wave interferometry (CWI) for monitoring the response of an anthracite coal specimen to helium (He) and carbon dioxide (CO₂) injection under controlled triaxial loading with an axial-to-confining stress ratio of 2:1, with confining stress held 1.0 MPa above the gas pressure in every test so that the effective confining stress was constant at 1.0 MPa and the stages differ only in the gas present. Gas was introduced at nominal injection pressures of 2.5, 5.5, and 12.5 MPa. Ultrasonic waveforms were recorded continuously for 5 h during the CO₂ experiments and 7 h during the He experiments. P- and S-wave velocities, and their fractional changes (dv/v), were calculated from Akaike Information Criterion-based arrival picks, while coda-derived relative velocity changes (δv/v) were estimated by the CWI stretching method over a 300–600 µs coda window. All six gas–pressure conditions were imposed sequentially on a single specimen, which was vented, degassed, and reconditioned between successive runs. CO₂ exhibited slower upstream-pressure dissipation than He, a response consistent with sorptive retention and adsorption-induced modification of the coal pore structure. Direct-wave velocities captured pronounced mechanical changes at low and intermediate injection pressures but showed limited sensitivity during the 12.5 MPa CO₂ experiment. In contrast, CWI detected a persistent negative δv/v trend at 5.5 MPa and a progressive negative trend at 12.5 MPa. Although adsorption was not measured independently, the contrasting He and CO₂ responses demonstrate that CWI can complement direct-wave analysis by detecting subtle, distributed changes associated with coupled mechanical and gas–coal interactions. Full article

Acoustic vehicle classification from roadside microphones supports source-specific traffic noise monitoring, but classifiers trained on one dataset generalize poorly to recordings from different environments. This study compares pre-trained audio transformers (BEATs, AST) against a CNN baseline for cross-dataset vehicle classification into five categories (car, truck, motorcycle, bus, background), training on the IDMT-Traffic dataset (Germany) and testing on the MELAUDIS dataset (Australia). Two domain adaptation methods (DANN, ArcFace) are applied to the best-performing transformer. BEATs achieves a balanced F1 of 0.46, a 77% improvement over the CNN baseline (0.26). AST achieves F1 = 0.43, a 65% improvement. Both pre-trained transformers substantially outperform the baseline, with self-supervised pre-training producing slightly more transferable representations than supervised pre-training. ArcFace metric learning achieves the highest cross-dataset F1 (0.50), modestly outperforming standard fine-tuning. DANN degrades cross-dataset performance but achieves the highest microphone robustness score (F1 = 0.80). Classification of the underrepresented bus class (53 training samples) improves from F1 = 0.38 to 0.54 with ArcFace. Microphone robustness is strong across all models (BEATs F1 = 0.75), confirming that recording environment dominates the domain shift over hardware variation. The choice of domain adaptation strategy should be guided by the expected type of domain shift. Full article

In this paper, a modal decomposition method is employed to analyze the properties of Rayleigh–Lamb waves propagating in an isotropic plate containing an open surface-breaking crack. The mode conversion coefficients (MCCs) are calculated as functions of crack depth and ultrasonic frequency for both an incident Lamb wave. Compared with the incident A0 mode, the MCC of the S0 mode exhibits a less smooth dependence on the phase of the crack modulated. The experimental, finite element method (FEM) and analytical results confirm that crack characteristics can be quantitatively evaluated from the Lamb wave mode conversion indued by the crack. This finding is beneficial for interpreting the detected ultrasonic signals in scenarios involving crack measurements between different waves, such as the modulation of the probing wave’s MCC by the material response in a tensile test machine. Full article

Temperature-dependent characterization of nanomaterials is commonly performed through electrical measurements, and many temperature-sensing applications similarly rely on electrical readout signals. Although these approaches are simple and widely used, they require direct sample contact and may lose sensitivity when the electrical response is weak, saturated, or only weakly dependent on temperature. In this study, a photoacoustic (PA) measurement system was used to evaluate the temperature-dependent response of AgNW networks. PA measurement system detects ultrasonic waves generated through transient thermoelastic expansion after optical absorption, which enables sensitive probing of temperature-dependent optothermal responses beyond conventional electrical measurements. Silver nanowire (AgNW) films with different surface coverage were fabricated through one to four repeated spin-coating cycles using a 1% AgNW ethanol dispersion containing nanowires with diameters of 20–40 nm and lengths of 10–20 µm. The photoacoustic (PA) signals of the fabricated films were measured at various temperatures using a customized PA measurement system and were compared with sheet resistance values obtained using a four-point probe. Both sheet resistance and PA signals increased with increasing temperature, regardless of AgNW surface coverage. Notably, AgNW films with higher surface coverages showed only weak temperature dependence in sheet resistance but pronounced temperature-dependent PA response. These results indicate that PA analysis is particularly useful for characterizing AgNW networks with higher surface coverage, including those used in stretchable electrodes. All samples showed linear relationships for both sheet resistance and PA signals, with coefficients of determination (R²) exceeding 0.9. Overall, these findings suggest that PA measurement may provide complementary information on temperature-dependent behavior that is not fully reflected in sheet-resistance measurements. Full article

Speech is a fundamental component of human communication and has become increasingly important with the widespread adoption of voice-based technologies, text messaging systems, Chatbots, and Large Language Models (LLMs). While AI systems are going through exponential breakthroughs, they often prioritize the dominant variant of a language. Using the TEDx Talks Brazilian Accents (TTBAcc) dataset, this research provides an acoustic analysis of two Brazilian Portuguese accents: Nordestino and Paulistano and an unsupervised accent clustering to find distinctive characteristics and investigate factors that can potentially support the monitoring of emerging and endangered accents. The acoustic analysis compares rhotics and vowels produced by speakers of both accents. The analysis of rhotics provided limited evidence of meaningful difference between accents. In contrast, the vowels exhibit significant differences in the fundamental frequency ($f_0$), the first and second formant frequencies ($F_1$ and $F_2$) and intensity, suggesting that these acoustic features may be more effective to distinguish between the investigated accents. For accent clustering, $f_0$, length, volume, and formant frequencies ($F_1$ and $F_2$) are used as inputs. Among the evaluated approaches, K-Means, with $K=38$ achieves the best performance, obtaining a Rand Score of 0.775 ± 0.028 and 0.524 ± 0.032 homogeneity, the best performing model. The optimal number of cluster exceeds the number of classified accents, which may indicate either an over-segmentation of data or a finer-grained phonetic variation captured in acoustic features. These findings highlight the potential of using acoustic features for investigating the variation in regional accents, and provides a basis for further research in monitoring emerging and endangered accents. Full article

Broadband, low-frequency sound absorption within a compact device remains a central unsolved problem in noise control engineering, arising from fundamental trade-offs among resonator volume, absorption bandwidth, panel thickness, and frequency tunability that no passive, linear, time-invariant system can simultaneously circumvent. This review establishes a unified mass–spring–damping (MSD) framework applied systematically across the full spectrum of resonance-based absorber technologies. From first principles, we derive the mass–stiffness coupling result (the mass-disappearing result of Shen and Liu): fixing the resonance frequency imposes $K=M{\omega }_{\mathrm{res}}^2$, so acoustic mass and stiffness cannot be adjusted independently; the half-absorption bandwidth ${\Pi }_1=\eta /\left(M{\omega }_{\mathrm{res}}\right)+V{\omega }_{\mathrm{res}}/\left(c_0{S}_{\mathrm{tar}}\right)$ then depends explicitly on the cavity volume V (system stiffness) and on the damping coefficient $\eta$, rather than on mass as an independent lever. This explains why neck extension, space-coiling, and membrane loading—which merely add mass while leaving the cavity stiffness unchanged—fail to broaden the absorption band at fixed volume, and refocuses the design effort on stiffness reduction and damping control. Five non-dimensional performance metrics are introduced that collapse the scattered literature into a single, scale-independent language for rigorous comparison across all absorber families: normalised half-absorption bandwidth ${\Pi }_1$, volume efficiency ${\Pi }_2$, integral absorption criterion ${\Pi }_3$ tied to the Rozanov causality bound, quality factor $Q=1/{\Pi }_1$, and frequency-thickness ratio ${\Pi }_4$. A two-degree-of-freedom acoustic–structural coupling model yields closed-form effective stiffness and damping, revealing how structural loss augments acoustic damping, how modal veering produces split absorption peaks, and how the anti-resonance frequency becomes a designable parameter. A critical distinction is drawn between mathematical negative stiffness (a fitting artefact) and physical negative stiffness via repulsive magnets, bistable elements, or negative-capacitance piezoelectric shunts, which genuinely reduces cavity stiffness, lowers resonance frequency, and widens bandwidth beyond the passive causality bound. The shunt electromechanical diaphragm further demonstrates $\alpha >0.9$ at nine tonal frequencies spanning three octaves without mechanical modification. Finally, embedding MSD equations and ${\Pi }_1$–${\Pi }_4$ bounds as hard physical priors in AI/LLM-assisted design frameworks is identified as the key step toward provably physically consistent absorber synthesis. Full article

Hybrid structural materials (HSMs), such as nanoparticle-treated fabrics, have been shown to enhance acoustic and ballistic performance in multifunctional protective structures. They offer a promising means for low-frequency (<~1000 Hz) noise mitigation, which remains a critical challenge in aerospace and defense applications. The measurement and modeling of their sound transmission loss (TL) behavior using a coupled acoustic–poroelastic approach is explored in this study. A colloid-based soaking and drying process is used to impregnate nanoparticles into the fabric. Previous studies using SEM imaging have established that at low (<~20 wt.%) treatment levels, the nanoparticles agglomerate in the interstitial spaces between yarn crossover points, whereas at higher levels, they begin to coat the yarn bundle tops. TL was measured experimentally using normal-incidence impedance tube tests. Further, parameters such as static flow resistivity, porosity, flexural modulus, and density required to model the neat and treat samples as fluid-filled porous solids using the Biot–Allard model were obtained from experiments for a limited set of neat and treated cases. Static flow resistivity was measured using an air permeability tester as per ISO 9237, and a modified version of the Peirce’s cantilever beam test was used to obtain the flexural modulus for neat and treated samples. Porosity was estimated using digital image analytics. The poroelastic fabric model was implemented in finite element simulations, and the predicted TL was compared with experiments including those for uncalibrated treated cases. The model shows close alignment with measured TL at low frequencies (<~600 Hz) for all cases but deviates closer towards the theoretical mass law at higher frequencies, where flanking effects and the influence of the hierarchy of pores are expected to be dominant in experiments. Further studies are underway to incorporate such higher-order effects to improve predictions at higher frequencies. The development of this model provides a means to capture the influence of nanoparticle addition on the acoustic performance of Kevlar, enabling fast and efficient virtual design iterations. The approach helps optimize HSMs for noise mitigation in multifunctional applications for the aerospace, defense, and infrastructural sectors. Full article

Digital modeling of traditional Chinese musical instruments significantly lags behind that of their Western counterparts, limiting advances in cultural preservation and related research. This paper addresses score-to-audio generation for the Chinese bamboo flute (dizi), aiming to synthesize expressive audio with human-like performance nuances directly from Musical Instrument Digital Interface (MIDI) scores. Two core challenges remain in existing score-to-audio synthesis methods: first, the scarcity of large-scale paired MIDI-audio training data for traditional Chinese instruments; second, the inability of standard MIDI to encode instrument-specific expressive techniques, such as vibrato, pitch bends, and ornamentations. To address these challenges, we propose a scalable workflow that generates large-scale synthetic MIDI-audio pairs through rule-based score randomization, automated digital audio workstation (DAW) rendering with commercial sample libraries, and standardized feature extraction. Performance technique information is directly encoded into the MIDI stream using out-of-range MIDI note numbers, achieving more effective conditioning than external control methods. Our system is built upon the Musical Instrument Digital Interface–Differentiable Digital Signal Processing (MIDI-DDSP) framework. Large-scale synthetic data is used for pre-training to establish timbral consistency, while real recordings from the University of Rochester Multi-Modal Music Performance (URMP) flute corpus are employed for fine adjustment to achieve expressive dynamic variations. Synthetic data alone yields stable but less expressive outputs, whereas real data alone risks overfitting. The combined strategy achieves a realistic timbre with controllable dynamics and performance techniques. Full article

Research on acoustic vases embedded in historical buildings has progressively developed through archaeology, architectural history, and acoustics, producing extensive documentation on their distribution, geometry, installation and possible acoustic role, with experimental evidence from in situ measurements, laboratory tests, and numerical analyses. However, documentation and measurement procedures remain highly heterogeneous, limiting reproducibility and comparative interpretation. Within this context, the present work proposes a preliminary methodological framework for future investigations. It reviews the principal European documentation, census, experimental, and numerical approaches identified by the authors. Particular attention is given to the recording system developed by Palazzo-Bertholon and Valière and to the experimental methodologies adopted in previous studies, including procedures, acoustic descriptors, and instrumentation. A possible extension of the “fiche d’information” is discussed to make geometrical, spatial, and installation-related data more explicit for comparative analyses, acoustic modelling, and measurement planning. A documented case study is used to illustrate the proposed organization of information, followed by a simplified acoustic simulation showing how structured data may support model construction and interpretation. The proposed procedure therefore combines simulation and in situ measurements, with standard and advanced investigation levels. The study is aimed at improving documentation quality, consistency, and comparability rather than providing definitive evidence of acoustic effectiveness. Full article

To investigate the evolution of acoustic emission (AE) parameters during the flexural failure of reinforced concrete (RC) beams subjected to freeze–thaw and corrosion, four RC beams were fabricated and assigned to four conditioning regimes: no deterioration, freeze–thaw only (75 cycles), corrosion only (4.8% mass loss), and combined. Three-point bending tests were conducted, combining AE and digital image correlation (DIC) techniques. The damage process was divided into four stages: micro-crack initiation, stable crack propagation, unstable crack propagation, and failure. The evolution of AE parameters including ring count, energy, amplitude, peak count, and duration was analyzed. Each parameter is positively correlated with load level and rises as the damage stage advances. The slope of cumulative parameters reflects crack development more reliably than instantaneous values. The effect of corrosion on these parameters is significantly greater than that of freeze–thaw. For corroded beams, AE parameter levels are higher during the micro-crack initiation stage but lower during the stable crack propagation stage. The overall AE activity decreases with increasing deterioration degree. High-amplitude events increase with damage progression, but fewer high-amplitude events are observed at the failure stage of severely deteriorated beams. This study reveals the correspondence between AE parameters and damage stages, providing an experimental basis for damage assessment using AE techniques. Full article

In the context of temporary urban environments such as festivals and construction sites, acoustic discomfort inside textile-based party tents remains a critical yet under-addressed issue. This study, developed in collaboration with industrial partners from the construction and textile sectors, aimed to design and assess lightweight, flexible sound-absorbing solutions capable of improving acoustic comfort while reducing environmental impact. A set of nine composite textile samples were fabricated using combinations of recycled and technical materials. Their acoustic behavior was characterized through standardized laboratory tests (ISO 10140-2 and ISO 354:2004) and full-scale prototypes under simulated real-world conditions. Simultaneously, a Life Cycle Assessment (LCA) was conducted according to ISO 14040:2006 and EN 15804:2020 standards to quantify environmental impacts from cradle to gate. The results demonstrate that it is possible to achieve an effective balance between acoustic performance and environmental sustainability. Integrating both types of assessment into the material selection process enables more informed, responsible design decisions for temporary textile structures, aligning construction practices with growing demands for functional and sustainable solutions. Full article

Wave propagation along acoustic black hole waveguides (ABH) consisting of arrays of functionally graded perforated rings (FGPR) is investigated by developing their linearized electro-acoustic analog models. These models simulate the interactions between the different components of the ABH/FGPR as influenced by the design parameters of the perforated rings and coupled acoustic cavities. The developed electric analogs consist of arrays of resistors simulating the flow resistance through the perforations, inductances to quantify the inertia of the gas moving inside these perforations, and capacitances to describe the storage behavior of the cavities between neighboring rings. The dynamic interactions between these electric components are described by a state-space model in terms of the incoming and outgoing flow velocities as well as the pressures before and after each FGPR. In this manner, it is possible to determine the spatial distributions of the flow velocities and pressures along the ABH which are necessary to demonstrate the effectiveness of the proposed ABH/FGPR in generating the ABH effect by monotonically decelerating the flow along the waveguide and bringing it to a complete stop. The predictions of the developed electro-acoustic analog models are validated against the predictions of the classical “Transfer Matrix Method (TMM)” and experimental results established by Petrover and Baz (2025). Furthermore, these predictions are also validated against the predictions of a developed MATLAB SimScape Model of the electro-acoustic analogs. The comparisons between the predicted and measured results show close agreements and validate the accuracy of the developed electro-acoustic analog models. Full article

Personal sound zones (PSZs) in cars allow occupants in different seats to enjoy distinct audio content without mutual interference. However, achieving sufficient acoustic isolation in real-world vehicle cabins remains challenging. In-vehicle experiments were conducted to enhance acoustic contrast using four headrest loudspeakers mounted on the rear-right seat to create dark zones at the front seats. Three arrangements of four headrest loudspeakers on a single seat were evaluated: longitudinal horizontal, lateral horizontal and vertical configurations. Compared with distributing four headrest loudspeakers across two seats, concentrating them on one seat yielded better acoustic contrast. When evaluated in the car using control points distributed on a horizontal plane, the vertical configuration exhibited superior performance among the tested arrangements, achieving an overall acoustic contrast of 12.4 dB over the 88–4525 Hz frequency band. This represents improvements of 3.9 dB over the two-loudspeaker setup and 0.9 dB over the two-seat-distributed four-loudspeaker arrangement. These results indicate that the vertical loudspeaker arrays on a single seat provide an effective approach for improving PSZ isolation in automotive environments. However, this enhanced acoustic isolation comes at the cost of increased array effort at low frequencies compared to the two-seat distributed configuration, as quantified in the study. Full article

Rotating machines in industrial environments often generate fundamental-frequency noise and its harmonics, which may affect workers’ physical and mental health. This study investigates the active control of single-frequency noise in an open sound field. A multi-harmonic active noise control (ANC) algorithm based on local secondary path (LSP) equalization is adopted to reduce computational complexity. A semi-anechoic chamber test platform is built to experimentally investigate the influence of the distance between the error microphone and the secondary source (Des), as well as collinear and non-collinear arrangements, on noise-reduction level, spatial attenuation range, and the upper frequency limit of noise control. The experimental results show that, under a 100 Hz tonal noise condition, the noise-reduction level first increases and then decreases as Des increases. This phenomenon can be explained by the phase relationship between the primary and secondary sound fields. As Des changes, the propagation delay of the secondary sound field also changes, thereby altering the interference pattern at the monitoring positions. When the phase difference approaches destructive interference, higher noise reduction is achieved, whereas further increases in Des gradually reduce the cancellation effectiveness because of phase mismatch. A relatively high noise-reduction performance is maintained when Des ranges from 200 cm to 500 cm. Secondary destructive interference may also appear at certain distances. In addition, the collinear arrangement shows a significantly higher upper frequency limit for noise reduction in comparison with the non-collinear arrangement. The results provide useful experimental evidence and practical guidance for optimizing the layout of ANC systems in complex acoustic environments such as vehicles and industrial plants. Full article

The empirical calibration of remote microphone probes, used to acquire unsteady pressure fluctuations in a wide range of fluid-dynamic applications, often introduces spurious resonance into the estimated frequency response, i.e., the transfer function, of the probe over the multiple steps it requires. To prevent this spurious resonance from affecting the unsteady pressure measurements, the transfer functions tend to be manually post-processed. Yet, such a procedure can constitute an additional source of uncertainty that hampers the accuracy of the results. A semi-empirical calibration method based on Bayesian inference was previously developed to tackle this problem: ASSIST (BAyesian proceSsing of SpurIous reSonance in calibraTion data). Through this technique, spurious resonance is removed and replaced with a physically correct alternative in a much less operator-reliant manner. However, its application requires a thorough understanding of the model and the related assumptions. This paper provides the knowledge required to apply ASSIST to the calibration of remote microphone probes. It covers the acquisition of the calibration data, the set-up of the method, the analysis of the results, and the iterative tuning of the input parameters to achieve the optimal fit. These steps are demonstrated on an open-source acoustic finite-element method simulation dataset, allowing the analytic line-cavity model to be compared with a 3D model and the impact of the relevant parameters constituting the method to be discussed. This framework is finally used to propose a novel development of the technique that reduces the calibration process of remote microphone probes to one single step, removing the source of the spurious resonance and enabling their in-situ calibration on non-sealing surfaces. Full article

Due to fluctuations in flow rate, pressure, and pump operating states, as well as environmental disturbances such as temperature variations and structural vibrations, pipeline leakage signals exhibit significant nonstationary characteristics. The traditional fixed sensor is limited by the layout position, resulting in suboptimal detection performance. For micro-leakage, it is even more difficult to achieve detection. With the advantages of small size and strong passing ability, the acoustic inner detector is well-suited to the task of comprehensive pipeline detection. Therefore, this paper carried out unsteady micro-leakage detection based on acoustic internal inspection signals. The unsteady micro-leakage simulation experiment of pipeline was carried out, and the leakage acoustic signal was collected for method verification. This paper investigates the integration of variational mode decomposition (VMD), random forest (RF) and least squares support vector machine (LSSVM) for signal processing and leakage classification. An unsteady micro-leakage detection method based on acoustic internal inspection signals was proposed, which is well-suited to the leakage detection task of pipelines. Experimental results indicated that the proposed method achieved a recognition accuracy of 95.31%, outperforming conventional leakage detection methods. Full article