204 results about "Acoustic radiation" patented technology

The invention relates to the related field of underwater acoustic signal processing and noise recognition, in particular to an underwater acoustic radiation noise recognition method, and provides an underwater acoustic radiation noise real-time recognition system based on combination of a convolutional neural network (CNN) and a long short-term memory network (LSTM). The system can perform efficient data preprocessing, feature extraction and optimization processing after collecting underwater acoustic signals in real time, and performs rapid identification of noise types through the trained mixed deep learning model. Specifically, the CNN can extract local features from a spectrogram, the LSTM is good at capturing time sequence information, complex underwater acoustic signals can be better understood and distinguished through combination of the CNN and the LSTM, and the recognition precision is remarkably improved. In addition, the system is further provided with an automatic updating mechanism, the model can be retrained regularly according to newly-collected data, it is ensured that the model is always in the optimal state, and the system adapts to the constantly-changing underwater environment.

The invention relates to a wind power boosting rotor outfield vibration sound radiation simulation method, which comprises the following steps: based on fluid simulation software Star-ccm +, taking a wind power boosting rotor outer cylinder contour as a rigid wall surface, setting an inlet condition as a wind speed and an outlet condition speed as 0, and calculating outer cylinder contour pulsating pressure; based on structural finite element software Abaqus, a finite element model is established according to the outer cylinder structure of the wind power boosting rotor, pulsating pressure serves as load input, constraint conditions are added, and the vibration displacement of the outer cylinder is calculated; based on acoustic finite element software Actran, node position information of the structural finite element model is imported, an external field acoustic domain is established, and external cylinder external field vibration sound radiation is calculated by taking external cylinder vibration displacement as load input; according to the method, calculation is carried out based on a frequency domain, the real sound wave transmission condition and the actual sound pressure cannot be reflected, and certain limitation exists. Therefore, by adopting the vibration sound radiation time domain simulation analysis method, the vibration sound radiation condition in the time domain can be obtained, and the change of the actual sound pressure along with the time can be obtained.

A speaker unit with a speaker frame and two membranes arranged in the speaker frame. A first membrane radiates in a major acoustic radiation direction substantially perpendicular to a major plane of the speaker unit. A second membrane has a secondary acoustic radiation direction different from the major acoustic radiation direction. Two drive units are positioned within the speaker frame, attached to the two membranes, and positioned coaxial to each other at the same height in the speaker frame. An acoustic duct provides for a closed acoustic channel from the second membrane in the secondary acoustic radiation direction to a secondary surface of the speaker unit in a same plane as the major plane of the speaker unit.

A device and method for assembling objects by generating an axial acoustic radiation force in order to position the objects on pressure nodes and / or antinodes and generating a transverse acoustic radiation force in order to move the thus positioned objects on a porous membrane contact structure.

A rear panel, which is an external member, includes a frame portion including multiple sound radiation holes including a first sound radiation hole, a second sound radiation hole, and a third sound radiation hole. A first sloping portion is disposed within an area of the first sound radiation hole and is inclined with respect to a thickness direction of the frame portion towards a first direction non-parallel to the thickness direction, and a second sloping portion is disposed within an area of the second sound radiation hole and is inclined with respect to the thickness direction of the frame portion towards a second direction different from the first direction and non-parallel to the thickness direction, and the third sound radiation hole includes a surface parallel to the thickness direction of the frame portion.

The invention relates to a calculation method and device for a two-dimensional irregular waveguide sound field of an underwater structure and a medium, and belongs to the technical field of underwater sound radiation forecast.The calculation method for the two-dimensional irregular waveguide sound field of the underwater structure comprises the steps that the two-dimensional irregular waveguide sound field of the underwater structure is constructed; multi-connected domain complex waveguide sound field calculation is converted into single connected domain sound field calculation, the boundary of a waveguide area of a two-dimensional irregular waveguide sound field is determined based on the constructed coordinate system, the boundary comprises an irregular fluctuating seabed boundary, and integration is carried out along the boundary to determine an integral equation; discretizing the boundary of the waveguide region, determining a boundary integral equation based on the basic solution of the integral equation, calculating a coefficient matrix in the constructed linear equation set, and determining an unknown physical quantity on the boundary; the intra-domain integral equation is determined based on the boundary integral equation, the velocity potential function value is determined based on the intra-domain integral equation and the known boundary physical quantity, then the sound pressure of the waveguide region is determined, and the calculation precision of the underwater radiation sound field is improved.

The invention discloses a woodworking machine tool noise sound power level measurement method based on a sound intensity method, and belongs to the technical field of machine tool noise measurement, and the method specifically comprises the steps: determining a housing surface covered by a tool rotation plane projection as a main sound radiation region, and determining a main bearing pedestal installation surface as an infrasound radiation region; sound pressure, vibration acceleration signals and dynamic air pressure signals of a main sound radiation area are collected; resampling a dynamic air pressure signal rotation angle domain, extracting a reference pneumatic pulsating waveform synchronous with tool tooth sweeping, and establishing an airflow induced sound pressure model according to the reference pneumatic pulsating waveform and a phase relation; calculating a surface normal vibration speed according to the vibration acceleration, and establishing a structure radiation sound pressure model in combination with a sound wave equation; decomposing the sound pressure signal to obtain two types of component estimation values, and performing parameter optimization to obtain net sound intensity of a main sound radiation area; extracting vibration frequency band energy of an infrasound radiation area, calculating equivalent sound intensity by combining a sound radiation efficiency model, and outputting a total sound power level by space integration; the invention provides an effective method for accurately measuring and controlling the noise of the woodworking machine tool.

The invention discloses a sound barrier-based viaduct noise radiation prediction method and device and a medium, and relates to the technical field of noise radiation prediction, and the method comprises the steps: obtaining noise source data, and carrying out the noise feature extraction of a noise source state of the noise source data, so as to obtain a noise source feature matrix; obtaining dynamic environment data, and performing environment key parameter analysis on the dynamic environment data to determine environment parameters; based on the noise source feature matrix and the environmental parameters, a noise radiation path is obtained through mixed sound ray tracking; according to the noise radiation path, initial arrangement data of the sound barrier are determined through sound barrier interception configuration; and carrying out sound barrier interception optimization on the initial arrangement data of the sound barrier to obtain a multi-source sound barrier configuration strategy. According to the method, the technical problem that noise radiation prediction in a viaduct environment cannot meet sound barrier arrangement is solved.

The invention provides a rapid magneto-acoustic-electric tomography method without mechanical rotation and scanning. In order to solve the problems of low speed, poor stability, low imaging precision and the like of rotary scanning measurement in the existing magnetoacoustic-electric tomography (MAET), a magnetoacoustic-electric measurement theory suitable for annular sound source excitation and electrode detection is provided on the basis of sound radiation and a magnetoacoustic-electric coupling mechanism. An MAET imaging method based on annular multi-sound-source cyclic excitation and multi-electrode synchronous measurement is developed. The minimum configuration requirement of the MAET system is obtained based on the influence of the annular sound source, the number of electrodes and the positions of the electrodes on the imaging quality, and an optimal configuration scheme of rapid MAET imaging is provided. According to the invention, mechanical scanning without a transducer or phased scanning measurement of an array sound source is utilized, mechanical rotation of an object is not needed, sound source circulation excitation and multi-electrode synchronous measurement are realized through electronic control, and the imaging speed and stability are improved. The device is easy to be fused with a commercial ultrasonic CT system, and cross-scale multi-modal imaging combining ultrasonic and electrical impedance is realized.

The invention provides an underwater vehicle stern shell structure noise reduction method, and relates to the technical field of underwater vehicle noise reduction, and the method comprises the steps: obtaining the structure data of an underwater vehicle and a vibration signal in the operation process; determining a noise reduction target for the underwater vehicle according to the vibration signal; obtaining constraint conditions for the underwater vehicle; analyzing stern flow field characteristics and horseshoe vortex formation of the underwater vehicle on the basis of fluid dynamics analogue simulation; the method comprises the following steps: arranging a transition part of the underwater vehicle, vortex eliminating rectiblocks and reinforcing ribs; simulating vibration sound radiation of the underwater vehicle under fluid excitation by applying a finite element method in combination with sound-structure interaction analysis; optimizing the size of the transition part and the mounting position of the vortex elimination rectifying piece, and further optimizing the design parameters and the mounting position of the reinforcing rib by using a genetic algorithm; and outputting the target setting position of the reinforcing rib when the vibration sound radiation reaches the noise reduction target. According to the invention, systematic noise reduction analysis of the underwater vehicle structure can be realized.

The invention relates to the technical field of terminal equipment, and discloses a loudspeaker and electronic equipment. The loudspeaker comprises a frame, a magnetic circuit system and a vibration system, the magnetic circuit system and the vibration system are connected with the frame, and the magnetic circuit system is fixed at one end of the frame. The vibration system comprises a vibrating diaphragm, a dome and a voice coil. One end of the vibrating diaphragm is fixed at the other end of the basin frame, and the vibrating diaphragm protrudes towards the direction of the basin frame; the globe roof comprises a first globe roof and a second globe roof, the first globe roof is located on the side, away from the basin stand, of the vibrating diaphragm, the second globe roof is annularly arranged on the first surface of the first globe roof, and the second globe roof is connected with the other end of the vibrating diaphragm; the voice coil is fixed on the first surface of the first dome and is opposite to the magnetic circuit system; the voice coil is configured to generate a magnetic field effect with the magnetic circuit system to drive the first dome to vibrate and produce sound. Therefore, the dome is divided into two parts, and the diaphragm is used for transversely supporting the dome, so that the shape of the dome can be maximized, the area of the first dome can be increased, the effective sound radiation area of the loudspeaker can be increased, and the loudness of the loudspeaker can be increased.

The invention belongs to the technical field of automatic control, and particularly relates to a large-span ellipsoidal glass structure space noise reduction method based on sound field optimization, and the method comprises the steps: 1, transmitting a sample noise in an ellipsoidal glass structure through a sound wave generator, building a sound wave propagation model to describe the behavior of the sample noise in the ellipsoidal structure, an incident sound field is obtained; 2, analyzing the propagation characteristics of the sample noise in the ellipsoidal glass structure to obtain an acoustic structure coupling equation, and solving the acoustic structure coupling equation to obtain structural vibration displacement; 3, obtaining a sound radiation field of the ellipsoidal glass structure according to the structural vibration displacement; and step 4, combining the incident sound field and the sound radiation field, and performing spatial noise reduction by reflecting the sound pressure of the sound field. According to the method, the problems of poor adaptability and low control precision of a traditional noise reduction method in a non-uniform structure are effectively solved, and the noise reduction efficiency, the control sensitivity and the environment adaptability of a building space are remarkably improved.

The invention discloses an automatic radiation noise simulation method, device and equipment of acoustic finite element software and a storage medium, and relates to the technical field of noise simulation, and the method comprises the steps: obtaining a grid size and a convex hull grid distance according to a geometric model, a material and a frequency input by a user; performing mesh generation on the surface of the geometric model based on the mesh size to obtain a surface mesh; obtaining a convex hull grid based on the convex hull grid distance and the surface grid; generating a tetrahedral mesh in an area between the surface mesh and the convex hull mesh; based on the surface grid, the tetrahedral grid and the convex hull grid, acoustic radiation noise simulation is carried out, full-process autonomous generation from simulation input to grid calculation is realized, and the efficiency problem of grid division and boundary setting depending on artificial experience in the prior art is solved. The calculation precision is guaranteed, meanwhile, the pretreatment workload is reduced to the degree that only basic parameters need to be input, and the operation complexity of radiation noise simulation conducted by acoustic finite element software is reduced.

An ultrasonic control apparatus and method for a liquid jet are provided. An ultrasonic generator is equipped with two sets of ultrasonic transducers arranged perpendicularly to each other, and the two sets of ultrasonic transducers emit ultrasonic waves of a specific frequency. Two sets of reflectors are arranged correspondingly in front of the ultrasonic waves emitted by the two sets of ultrasonic transducers, where a distance exists between the ultrasonic transducers and the corresponding reflectors, two stable standing wave sound fields are formed by adjusting the distance. A liquid jet device is equipped with a nozzle that emits a continuous jet, where the nozzle is located above an intersection of pressure nodes of the two standing wave sound fields, and the two standing wave sound fields act on the liquid jet to adjust jet morphology under the action of acoustic radiation pressures.

The invention provides a static water low-noise hydrodynamic noise rotation test device which comprises a stepping motor and a speed reducer located below the stepping motor, a rotating shaft of a rotating arm is connected with the speed reducer through a damping device, and a sectional type variable-length suspender is used for connecting a test model and the rotating arm. The stepping motor and the speed reducer are installed on the supporting frame through the double-layer vibration isolation device, and the stepping motor and the speed reducer can push the rotating arm to rotate so as to drive the test model to do large-diameter rotary motion. The test device disclosed by the invention can be used for carrying out a rotation test in a relatively large water pool or a lake or on the sea, and controlling sound radiation of mechanical vibration noise to underwater at the same time, so that background noise is effectively reduced, and measurement of flow-induced vibration noise of a large-scale model is realized.

Various examples are provided related to methods and systems to reconstruct 3D image / map for the stiffness of soft tissues. In one example, a method includes obtaining particle velocities from generated mechanical waves in tissue of a subject; reconstructing a stiffness map of soft tissue in a region of interest (ROI) from the particle velocities; and generating an image of the stiffness map for rendering on a user device. In another example, a system includes an ultrasound scanner for shear wave elastography (SWE) and a computing or processing device that can obtain particle velocities; reconstruct a stiffness map of soft tissue in a ROI from the particle velocities; and generate an image of the stiffness map for rendering. The particle velocities can be generated from acoustic radiation forces produced by the ultrasound scanner.

The utility model discloses a rotor compressor silencer which comprises a silencer body, a plurality of convex petals and a plurality of concave petals are distributed on the silencer body, high-pressure pulsating gas flowing out of an air cylinder sequentially flows through the convex petals and the concave petals of the silencer, due to the change of the volumes of inner cavities of the convex petals and inner cavities of the concave petals, when the high-pressure pulsating gas flows through, propagation loss can be generated, and gas noise is reduced. A center hole is formed in the center of the silencer body, and the silencer is installed on an upper bearing of an upper shaft of a rotor compressor crankshaft in a sleeving mode through the center hole. At least one group of air outlet holes are formed in the convex flaps, and high-pressure pulsating gas is discharged after the direction of the high-pressure pulsating gas is changed; inwards-concave reinforcing ribs are arranged on the concave petals, so that the size of an inner cavity of the concave petals is further reduced, the silencing capacity of the concave petals is enhanced, meanwhile, the rigidity of the silencer is improved, and sound radiation of the silencer is reduced.

The invention discloses an acoustic processing method for directionally generating, grouping and managing sound frequency bands and eliminating and counteracting the sound frequency bands, which comprises the following steps of: S1, carrying out AD (Analog to Digital) sampling on analog sound signals to be processed to obtain digital sound signals to be processed; s2, performing dynamic frequency band division on the to-be-processed digital sound signal to obtain a plurality of frequency band signals; s3, mapping the plurality of frequency band signals into a plurality of sound source groups; s4, each sound source group generates a directive sound signal pointing to the target area by using the sound radiation unit of the sound source group, and after the directive sound signal is synchronously transmitted to the target area, an actual mixed sound signal is formed; and S5, detecting, eliminating and counteracting a sound wave counteracting effect generated by the actual mixed sound signal in the target area. According to the method, high-quality accurate directional propagation can be realized, and the adaptability to a complex environment is enhanced.

The invention provides an ultrasonic transducer and ultrasonic cleaning equipment, the ultrasonic transducer comprises a piezoelectric exciter and a vibration radiant panel, the vibration radiant panel is fixed on the piezoelectric exciter, a power supply is electrically connected with the piezoelectric exciter, so that the piezoelectric exciter can generate longitudinal vibration, and the longitudinal vibration of the piezoelectric exciter can excite bending vibration of the vibration radiant panel. The longitudinal vibration sandwich type piezoelectric ultrasonic transducer solves the defects that an existing longitudinal vibration sandwich type piezoelectric ultrasonic transducer is small in radiation area, weak in longitudinal displacement amplitude, long in longitudinal size and the like to a great extent, and has the advantages of being small in space needed by operation, large in sound radiation area and sound radiation power, small in size and the like. The ultrasonic transducer can be widely applied to the application requirements of miniaturization and thinning of the installation space of the ultrasonic transducer. Meanwhile, the longitudinal bending displacement amplitude conversion ratio of the ultrasonic transducer can be influenced by the length, and / or the width and / or the thickness of the vibration radiant panel, so that the longitudinal bending displacement amplitude conversion ratio of the ultrasonic transducer is further improved through the size of the vibration radiant panel.

A method for specifying desired quantities of the energy flux of the combined waves of acoustic radiation pressure to apply producing a mid-air haptic pushing force, which has the effect of simultaneously reducing the harmonic distortion present is described. Further, a method for communicating only the summaries of acoustic field contributions at the required locations in the form of summed portions of the relatively small matrix whose row and column count depend only on the control point count is described. Further, phased arrays of ultrasonic speakers can produce a relatively large amount of acoustic energy which is usually directed in a specific direction or focused to a particular point depending on the application of the array. Further, to allow the system to be driven more strongly than usual, the complex-valued linear system that governs the drive signal to each control point is solved twice. Further, to achieve mid-air haptics with high precision the radiation pressure generated must be modelled accurately.

The invention relates to a measurement and inversion reconstruction method for a helicopter rotor sound radiation ball in an acoustic wind tunnel, and relates to the technical field of helicopter aeroacoustics wind tunnel test technology and sound target characteristics, and the method comprises the steps: measuring the noise radiated by a helicopter rotor in the acoustic wind tunnel through a hemispherical far-field noise measurement device; noise signals of the rotor in different directions are obtained, and a rotor sound radiation ball model is constructed; based on a sound propagation theory in a flow field, constructing a propagation model from the surface of the sound radiation ball model to the position of a microphone, and based on the influence of background noise, establishing a sound propagation equation from the surface of the sound radiation ball model to the microphone in the hemispherical far-field noise measurement device; and converting the sound propagation equation into solving a least square problem, solving to obtain sound source distribution on the surface of the sound radiation ball model, and reconstructing noise at any position of the far field of the helicopter rotor. The method can be suitable for reconstructing the helicopter rotor sound radiation ball in complex environments such as acoustic wind tunnels and high background noise, and practical research and application are carried out.

This invention relates to the field of dynamic load section stress monitoring and feedforward adaptive intelligent correction technology for coal conveyor belts in thermal power plants. The invention provides an automatic monitoring and correction device for coal conveyor belts, comprising: a control host that extracts speckle decorrelation index and maps the transverse segregation gradient matrix through a data acquisition and detection module; a rheological state analysis module that introduces a fractional-order calculus dynamics model to analyze the creep stress field and outputs a deformation potential energy time series; a spatial feature prediction module that uses the Koopman operator to elevate the nonlinear system to a linear functional space, identifies outbreak nodes by tracking eigenvalue drift trajectories, and issues early warning commands; and an acoustic field adhesion breaking correction module that drives piezoelectric ceramic elements to construct a local acoustic radiation standing wave field, breaking van der Waals forces to release transverse shear potential energy. This invention solves the risk of belt misalignment and tearing caused by material segregation and adhesion under multi-coal blending conditions.

The present description relates to a printing nozzle (20) for a 3D printer comprising a duct (21) that has an opening (23), wherein the duct (21) contains a fluid material to be printed (30) which forms an interface (31) with the air at the opening (23), wherein the printing nozzle further comprises an electro-acoustic transducer (40) configured to generate and send first ultrasonic waves (50) through the fluid material to be printed (30) to the interface (31), resulting in the formation of a droplet (32) of the fluid material to be printed at the interface (31) as a result of acoustic radiation forces.

Focused acoustic radiation, referred to as tonebursts, are applied to a volume of liquid to generate a set of droplets. In one embodiment, a first toneburst is applied to temporarily raise a mound or protuberance on a free surface of the fluid. After the mound has reached a certain state, at least two additional toneburst can be applied to the protuberance to sequentially eject multiple bursts of multiple droplets. In one embodiment, the state of the mound can be maintained by a sustained acoustic signal, during which time multiple additional tonebursts can be applied to sequentially eject multiple bursts of multiple droplets from the mound.

The invention discloses a real-time intelligent ship underwater radiation noise forecasting method based on virtual and real data fusion. The method comprises the following steps: establishing an acoustic elasticity calculation model of a real ship; defining an excitation point, and randomly applying a group of normal frequency sweeping forces with different amplitudes; solving and calculating vibration sound radiation; respectively reading a vibration acceleration amplitude spectrum of a measuring point corresponding to each excitation point; reading a radiation noise result of the whole ship, and outputting the radiation noise result to a file sound source level-working condition; calculating results under different working conditions until the calculated working condition number is equal to the set working condition number M; forming a model training set by the output acceleration amplitude spectrums and the radiation noise sound source level results of the M working conditions; a sample number, a side point number and a frequency point number for BP neural network training are defined, a result of a training set is read, data are processed, and input and output matrixes for BP neural network training are generated; parameters of BP neural network training are defined, and real-time intelligent forecasting of the underwater radiation noise of the ship based on virtual and real data fusion is achieved.

The application claims a MEMS frequency accurate tracking and locking method based on a resonant phase-locked loop (R-PLL). The application relates to the technical field of MEMS (micro-electro-mechanical system) resonator control and frequency tracking, and the method takes the output of a voltage-controlled oscillator as an interference signal, takes the physical response (such as weak electromagnetic leakage or acoustic radiation) of a disturbed MEMS device as a feedback signal, compares the phase difference between the feedback signal and a reference signal through a phase detector, adjusts the interference frequency, and forms a closed-loop locking system. The application adjusts the interference source frequency through a phase detector and a filter, can track the target resonant frequency drift caused by factors such as temperature change in real time, makes the MEMS resonator keep in a phase-locked state with the target resonant frequency, and ensures that the interference energy is always efficiently coupled.

Provided is a decontamination device capable of confirming whether a mist circulation dispersion unit is working properly before and after decontamination operation or during decontamination and capable of ensuring the operation of the decontamination device. The decontamination device is provided with a mist supply unit, a mist circulation dispersion unit, and an ultrasonic wave detection unit. The mist supply unit changes a decontamination liquid used for decontaminating the inside of a work chamber into a decontamination mist and supplies the decontamination mist to the inside of the work chamber. The mist circulation dispersion unit makes a vibration plate provided with a plurality of ultrasonic wave transmitters vibrate ultrasonically to generate an ultrasonic wave-based acoustic flow from the surface of the plate to the vertical direction, and makes a pressing action based on acoustic radiation pressure of the acoustic flow act on the decontamination mist to make the decontamination mist circulate and disperse in the work chamber. The ultrasonic wave detection unit is provided with an ultrasonic wave receiver for detecting the operation of the ultrasonic wave transmitters, and detects the operation of the plurality of ultrasonic wave transmitters as a whole and / or each ultrasonic wave transmitter provided in the vibration plate.

This invention discloses a method for measuring the sound power level of woodworking machine tool noise based on the sound intensity method, belonging to the field of machine tool noise measurement technology. Specifically, it includes: determining the housing surface covered by the projection of the tool's rotation plane as the main sound radiation area, and the main bearing seat mounting surface as the secondary sound radiation area; acquiring sound pressure, vibration acceleration signals, and dynamic air pressure signals in the main sound radiation area; resampling the dynamic air pressure signal in the rotational angular domain, extracting the reference air pulse waveform synchronized with the tool tooth sweep, and establishing an airflow-induced sound pressure model based on this and the phase relationship; calculating the surface normal vibration velocity based on the vibration acceleration, and establishing a structural radiation sound pressure model based on the sound wave equation; decomposing the sound pressure signal to obtain estimates of two types of components, and obtaining the net sound intensity of the main sound radiation area after parameter optimization; extracting the vibration frequency band energy of the secondary sound radiation area, calculating the equivalent sound intensity based on the sound radiation efficiency model, and spatially integrating to output the total sound power level. This invention provides an effective method for the accurate measurement and control of woodworking machine tool noise.

This invention relates to a method for rapid three-dimensional acoustoelastic prediction and line spectrum identification of ship mechanical noise based on online equipment vibration monitoring. The method includes: establishing a refined three-dimensional acoustoelastic calculation model for the overall ship vibration and acoustic radiation calculation of the equipment-vibration isolator-base-shell; applying a unit sweep force at the equipment's center of gravity node to obtain the transfer function from the average vibration acceleration spectrum of each equipment's base surface to the radiated noise; multiplying the average vibration acceleration spectrum monitored online in real time with the transfer function to obtain the ship mechanical noise under real-time excitation of each equipment; performing incoherent superposition of the radiated noise from each equipment to achieve rapid three-dimensional acoustoelastic prediction of ship mechanical noise; extracting the line spectrum peak value, and determining the main source of the line spectrum peak value by calculating the energy contribution of each equipment to the line spectrum peak value. This method achieves high-precision online prediction and automatic identification of the line spectrum source of ship mechanical noise, with high computational efficiency and strong engineering applicability.

This application relates to the field of centrifugal pump fault diagnosis technology, specifically to a method and apparatus for centrifugal pump fault diagnosis based on acoustic radiation signal processing. The method involves acquiring the acoustic radiation signal of a centrifugal pump to obtain a test acoustic radiation signal; performing wavelet packet decomposition on the test acoustic radiation signal; processing the sub-frequency band information in different ways to obtain a reconstructed signal; extracting various features from the reconstructed signal; combining the features to obtain a dataset; inputting the dataset into a pre-trained centrifugal pump fault diagnosis model; and outputting the centrifugal pump acoustic radiation signal fault diagnosis result. This application utilizes a combination of wavelet packet decomposition and singular value decomposition for acoustic radiation signal feature extraction, effectively improving the accuracy and efficiency of centrifugal pump acoustic radiation signal fault identification. It can accurately identify centrifugal pump fault types even under strong background noise interference, which is significant for risk warning of centrifugal pump faults.