CN114495975A - Abnormal sound source positioning method and device based on sound pressure amplitude ratio - Google Patents
Abnormal sound source positioning method and device based on sound pressure amplitude ratio Download PDFInfo
- Publication number
- CN114495975A CN114495975A CN202210101953.1A CN202210101953A CN114495975A CN 114495975 A CN114495975 A CN 114495975A CN 202210101953 A CN202210101953 A CN 202210101953A CN 114495975 A CN114495975 A CN 114495975A
- Authority
- CN
- China
- Prior art keywords
- abnormal
- sound source
- audio
- frequency
- sound pressure
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 102
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000005236 sound signal Effects 0.000 claims abstract description 23
- 230000004807 localization Effects 0.000 claims abstract description 15
- 238000001914 filtration Methods 0.000 claims description 9
- 238000001514 detection method Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 8
- 238000010801 machine learning Methods 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 230000005484 gravity Effects 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000003491 array Methods 0.000 description 4
- 239000002184 metal Substances 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000035939 shock Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/02—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/27—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The invention discloses an abnormal sound source positioning method and device based on a sound pressure amplitude ratio. After the sound waves are sequentially arranged through the sensors, the sequence and the characteristics are analyzed, the direction and the distance of noise are calculated, and then the position of the fault can be found accurately in the environment with more noise. Meanwhile, due to the non-contact property of the sound signals, the difficulty in vibration signal data acquisition can be effectively avoided. Sound source localization is achieved by using the difference in intensity of sound signals received by different microphones from the same sound source.
Description
Technical Field
The invention relates to a method and a device for detecting a fault position by using audio.
Background
Sound source localization is very important in processing sound signals, and is widely applied to aspects such as intelligent equipment, video conference systems, violation snapshot or fault diagnosis, and the like, and can automatically capture and aim at sound-producing objects. The direction of arrival of a sound source reaching the microphone array is obtained by processing the collected signals, and compared with a single microphone sensor, the microphone array formed by a plurality of microphone sensors has better advantages in the aspect of speech signal processing, has complementarity and can better eliminate background noise. Sound source localization is achieved herein by using the difference in intensity of sound signals received by different microphones from the same sound source.
In the current industrial manufacturing, the workload of equipment maintenance tests is increased rapidly along with the enlargement of the scale of a power grid, the economic cost and the social cost brought by equipment maintenance are highlighted day by day, and the development of the power grid and companies is restricted by the traditional equipment testing and detecting methods.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: aiming at the problem that a fault source is difficult to quickly find in the process of equipment overhaul test work, a sound source positioning method and a sound source positioning device based on a sound pressure amplitude ratio are provided to determine the specific position of a sound source in a space.
The technical scheme of the invention is as follows:
an abnormal sound source positioning method based on sound pressure amplitude ratio is characterized by comprising the following steps:
s1, normal state audio data acquisition and training: arranging a plurality of audio acquisition devices to acquire audio data of a target area, and firstly acquiring and training an audio model in a normal working state;
s2, abnormal audio recognition and extraction: the audio acquisition device acquires audio data of a target area in real time, compares the audio data with the trained audio model and identifies abnormal frequency;
s3, primary positioning of the sound source: comparing the abnormal frequency intensity acquired by each audio acquisition device, and selecting the audio acquisition device with the maximum intensity as a main device, wherein the abnormal sound source is closest to the main device;
s4, secondary positioning of the sound source position: and performing inverse Fourier transform on the abnormal frequency spectrum data, extracting the amplitude of the abnormal sound source audio data obtained by inverse Fourier transform on the time domain as the sound pressure amplitude of the abnormal audio, combining every two abnormal frequency sound pressure amplitudes obtained by each audio acquisition device, performing secondary positioning on the abnormal sound source by utilizing the relation that the pressure amplitude ratio is in inverse proportion to the distance square, and determining the actual coordinate of the abnormal sound source.
In S1: the target area is a cuboid space, a generator set is arranged in the cuboid space, a group of audio acquisition devices are respectively arranged at six vertexes of the cuboid, and each group of audio acquisition devices is composed of four microphones at different positions.
In S1: the method comprises the steps of obtaining a spectrogram through Fourier transform of collected audio data, finding out main frequencies of audio in a frequency domain through peak detection, carrying out the operation on a plurality of groups of audio data to obtain the main frequencies, and inputting the main frequencies as features into an isolated forest machine learning algorithm for training to obtain a corresponding model.
In S2: the method comprises the steps of identifying and extracting abnormal signals of collected audio data, carrying out Fourier transform on audio signals needing to be identified to obtain frequency domain information of the audio signals, carrying out peak detection to obtain main frequency, inputting the main frequency serving as characteristics into a trained machine learning model, judging the input characteristics by the model to obtain abnormal frequencies, and carrying out band-pass filtering on the audio signals to obtain time domain information of the abnormal frequencies.
In S3: and comparing the real-time collected audio data with the trained normal state model data, and if the newly added frequency is identified on the frequency domain of the audio data by comparing the normal state model and the intensity of the newly added frequency is greater than a specified threshold value, considering the newly added frequency as an abnormal frequency, and realizing the extraction of the abnormal frequency by filtering on the frequency domain.
In S4, constructing sound pressure amplitude ratio equation with other 3-way microphone of the main microphone and nearest adjacent 4-way microphone, and finding the optimal estimated position of sound source with equation residual and minimum
Wherein H is the equation residual sum minimum, r is the distance from the sound source to the microphone, E is the effective sound pressure,
the sound pressure amplitude ratio is the effective sound pressure ratio, and the value is obtained by the measured value of each microphone.
After S4, execute S5: and displaying the position coordinates of the abnormal audio signal in a two-dimensional image to display the position of the abnormal sound source.
The utility model provides an unusual sound source positioner based on sound pressure amplitude is than, includes audio acquisition device to the target area who waits to detect plans out a cuboid space as the center, arranges a set of audio acquisition device respectively at each summit of cuboid, and all audio acquisition devices pass through switch connection control module.
And building a support to install audio acquisition devices, wherein each audio acquisition device consists of 2-6 microphones, each group of microphones are arranged linearly, four groups of microphones positioned at the lower vertex of the cuboid are arranged linearly along the gravity direction, and four groups of microphones positioned at the upper vertex are arranged linearly along the horizontal direction.
The control module is an industrial control computer and is used for controlling automatic initialization, data acquisition, data transmission and calculation of the whole system, and meanwhile, the industrial personal computer is provided with an alarm system and a display module to broadcast the running state of the whole system.
The invention has the beneficial effects that:
most mechanical equipment can emit stable and regular noise under the normal working state, and when the equipment is aged or has other faults, the noise is obviously different from the normal working noise, and at the moment, the equipment can be overhauled by positioning an abnormal noise source. The abnormal sound of the equipment is mostly represented as audible sound (20-20 Khz), and if electric arcs and gas leakage occur, ultrasonic waves with higher frequency than the audible frequency are generated. Therefore, by measuring the ultrasonic wave component in a noisy environment, it is possible to determine whether or not gas leakage and arcing occur. The sound wave has directivity and is therefore easy to detect. After the sound waves are sequentially arranged through the sensors, the sequence and the characteristics are analyzed, the direction and the distance of noise are calculated, and then the position of the fault can be found accurately in the environment with more noise. Meanwhile, due to the non-contact property of the sound signals, the difficulty in acquiring vibration signal data can be effectively avoided. The sound source positioning is realized by utilizing the difference of the intensity of sound signals from the same sound source received by different microphones, the sound pressure amplitude ratio can be directly measured according to the algorithm of the text, and the position coordinate error can be flexibly changed according to the actual conditions such as calculation time, data processing amount and the like based on the system.
Description of the drawings:
FIG. 1 is a logic flow diagram of the present invention.
FIG. 2 is a schematic view of the apparatus of the present invention.
Detailed Description
Example 1:
the method for positioning the abnormal sound source of the generator set based on the sound pressure amplitude ratio comprises the following steps:
s1, acquiring and training audio data of the unit in a normal state: the audio acquisition module is a three-dimensional microphone matrix consisting of 8 acquisition cards and 84 element linear microphone arrays, acquires audio data of the generator set in normal operation through 32 microphones, and trains an audio model of the generator set in a normal state through deep learning;
s2, abnormal audio recognition and extraction: during actual monitoring, 32 microphones simultaneously acquire real-time running audio data of a unit, the real-time audio data of the unit acquired by the 32 microphones are received through an industrial personal computer, the real-time acquired audio data and trained unit normal state model data are compared, if a newly increased frequency is identified on a frequency domain of the audio data compared with a normal state model and the newly increased frequency intensity is greater than a specified threshold (the threshold is determined by a median value of the audio data on the frequency domain in the normal state), the newly increased frequency is regarded as an abnormal frequency, and the abnormal frequency is extracted through filtering processing on the frequency domain;
s3, determining the sound source position: firstly, comparing the intensity of the abnormal frequency of an abnormal sound source extracted by 32 paths of microphones in a frequency domain, and taking the microphone with the maximum intensity as a main microphone, wherein the abnormal sound source is closest to the position of the microphone, so as to carry out coarse positioning on the abnormal sound source; then, performing inverse Fourier transform on the abnormal frequency spectrum data after the last step of filtering, extracting the amplitude of the abnormal sound source audio data obtained by inverse Fourier transform on the time domain as the sound pressure amplitude of the abnormal audio, performing pairwise combination on the abnormal frequency sound pressure amplitudes acquired by the 32-path microphone, and performing secondary fine positioning on the abnormal sound source by utilizing the relation that the amplitude-to-amplitude ratio is in inverse proportion to the square of the distance, namely determining the actual coordinates (x, y, z) of the abnormal sound source;
and S4, visualizing the detected abnormal sound source position on the two-dimensional unit picture.
Example 2:
the method for positioning the abnormal sound source of the generator set based on the sound pressure amplitude ratio comprises the steps of audio data acquisition, identification, extraction and processing, abnormal audio and video signal alarm and abnormal sound source signal estimation positioning; and visualizing abnormal sound source signals. The method specifically comprises the following steps:
s1: the method comprises the steps that 8 groups of 4-element microphones, namely 32-path microphone arrays are used for collecting audio data of a generator set in real time, collected audio files are subjected to Fourier transform to obtain a spectrogram, main frequencies of audio in a frequency domain are found out through peak detection, a plurality of groups of audio files are subjected to the operation to obtain main frequencies, and the main frequencies are input into an isolated forest machine learning algorithm as features to be trained to obtain corresponding models; after an audio model of the unit in a normal state is trained, comparing audio data acquired on site with the trained audio model, and judging whether the unit generates an abnormal audio signal;
s2: the method comprises the steps that a characteristic project is constructed by collecting a large amount of normal generator set audio data, namely a main frequency training model in a spectrogram is obtained, meanwhile, abnormal frequencies are detected through an isolated forest algorithm, and time domain information of the abnormal frequencies can be extracted after inverse Fourier transform and filtering processing are carried out on the abnormal frequencies; the method comprises the steps of carrying out abnormal signal identification and extraction on collected audio data, carrying out Fourier transform on an audio signal to be identified to obtain frequency domain information of the audio signal, then carrying out peak detection to obtain main frequency, and finally inputting the main frequency serving as characteristics into a trained machine learning model. The model judges the input features, if the features are normal, the output is 1, and if the features are abnormal, the output is-1. And after the abnormal frequency is obtained, performing band-pass filtering on the audio signal to obtain time domain information of the abnormal frequency.
S3: and comparing the abnormal audio signals collected by the 32 paths of microphones to obtain the path of microphone with the maximum amplitude, and determining the path of microphone as the main microphone to realize coarse positioning. Performing secondary fine positioning based on the position, namely determining the coordinate position (x, y, z) of the abnormal audio signal;
for simplicity, the sound source localization problem on a two-dimensional plane is considered first. Suppose that four microphones are arranged at equal intervals on the X-axis with coordinates (-3a, 0) (-a, 0) (a, 0) (3a, 0), respectively. If the sound source is located at the S (x, y) point, the distances from the sound source to the four microphones can be respectively expressed as:
as can be seen from the related knowledge about sound waves, the sound pressure generated by a sound source at i microphones is given by,
wherein p is0As a sound sourceAt a reference distance, r0The voltage generated, omega, is the angular frequency of the simple harmonic vibration of the sound source,
c0at the speed of sound, and Δ ri-r0。
According to the characteristics of the microphone, at the ith microphone, the voltage output generated by the sound pressure is:
ei(t)＝pi(t)Hicosθi
wherein HiFor a parameter, theta, determined by the transfer characteristic of the ith microphoneiIs the angle of incidence of the sound wave at the ith microphone. If the transfer characteristics of the four microphones are identical (denoted by H), then the signal e is received for any two of themi(t) and ej(t) for a fixed time delay Δ t, divided by the difference in magnitudeij＝k(ri-rj) Except that the waveform is the same.
In the following, several parameters related to the localization of the sound source are derived. To this end, with e1(t) and e2(t) a ratio of
As can be seen from the above, the present invention,
e2(t+Δt12)＝p2(t+Δt12)Hcosθ2
＝p2(t)Hcosθ2exp{jk(t2-t1)}
＝p2(t)Hcosθ2exp{jk(r2-r1)}
in this way it is possible to obtain,
similarly, at e3(t) and e4Between (t), there are also similaritiesIn the context of (a) or (b),
suppose that the effective sound pressure of a sound source is E1，E2，E3，E4Effective sound pressure received by the sound pressure sensor, and distance from the sound source to the sensor1，r2，r3，r4. Due to the fact that
The following formula can be obtained:
wherein
And
the sound pressure amplitude ratio is the effective sound pressure ratio, and the value can be obtained by the measured value of each microphone;
the "coarse" positioning process includes: abnormal audio data sound pressure amplitude values are obtained by comparing 32 paths of microphones, the position of the microphone with the maximum amplitude value is used as the direction of an abnormal sound source to carry out coarse positioning on the abnormal sound source, and a main microphone is determined;
the "fine" positioning process includes: and constructing a sound pressure amplitude value ratio equation by using other 3 paths of microphones of the main microphone and the nearest adjacent 4 paths of microphones, searching the optimal estimated position of the sound source according to the equation residual error and the minimum, substituting the coordinate set of the area where the coarse positioning is positioned into the estimated abnormal audio position coordinate (x, y, z), and performing fine positioning on the abnormal sound source. The preliminary abnormal audio localization estimate according to the present study is given by:
wherein H is the equation residual sum minimum;
for example, in a test, a sound source is placed at a spatial coordinate (210, 170, 120) to emit abnormal frequency, the sound pressure amplitude of the acquired abnormal audio data is used to perform "coarse" positioning to obtain the number 3 of the main microphone, the nearest adjacent microphone is known to be the number 4 microphone according to the position relation of the microphones, and the amplitude value extracted by the first path of microphone of the number 3 microphone is recorded as E1The coordinate matrix of the space region determined by the coarse positioning is denoted as D, the coordinate of the main microphone is (265, 80.7, 179), and the coordinates of the remaining seven microphones are (265, 84.9, 179), (265, 89.1, 179), (265, 93.3, 179), (265, 93.3, 21), (265, 89.1, 21), (265, 84.9, 21), (265, 80.7, 21) in this order. The coordinates of the eight-path microphone are expanded into a matrix with the same size as the coordinate matrix D and are substituted into a formula
Wherein A represents the expanded microphone coordinate matrix, and the distance matrix r from the space coordinate to the 8-way microphone is calculated1,r2,r3,r4,r5,r6,r7,r8Substituting the distance matrix and the amplitude values extracted by each microphone into a formula
And (4) computing to obtain the minimum residual sum H of 227114.51, and finally indexing to find out the coordinate position of H, namely the computed abnormal audio position (213, 171, 121).
S4: and (x, y) of the position coordinates (x, y, z) of the abnormal audio signal is displayed on the two-dimensional image to show the position of the abnormal sound source.
Example 3: the sound source localization apparatus includes:
a first module: the equipment module, the industrial control computer, is a computer control machine specially designed for industrial field, and has important computer attribute and characteristic for detecting and controlling production process, electromechanical equipment and technological equipment. The bus structure is adopted, and the main function is to detect and control the production process, electromechanical equipment and process equipment.
The main function of the switch is to connect network devices and to transmit data by means of data exchange. A network switch is a device that expands the network and provides more connection ports in a sub-network to connect more network devices.
The display screen is a display applied to industrial control process or equipment, is greatly different from a civil or commercial display, and has special designs of dust prevention, shock prevention and the like.
And the alarm is arranged so as to monitor whether the generator set is abnormal in real time in order to prevent the occurrence of danger from causing unnecessary consequences and loss.
Build metal support around whole generating set in order to be used for fixed 8 microphone groups, the unit horizontal length 4000mm, wide 3000mm, high 250mm, so 2 metal support of length 4000mm build in the unit both sides, 1 metal support of length 3000mm builds in the unit rear side, 4 metal support of length 2500mm build in the vertical direction of unit. 7 in total.
The second module is an acquisition module, and is used for acquiring input sound signals of all microphones of the microphone array, as described herein, 8 acquisition cards and 8 4-element linear microphone arrays are selected, 32 microphone groups are arranged into a three-dimensional structure according to the actual size of the generator set, the microphones at the bottoms of 4 groups in the microphone arrays are 500mm away from the ground so as to avoid interference, and each group of microphones are provided with 12V 2-hole power supply inserts for power supply. In the three-dimensional model, the array can receive incident signals in all directions in space;
the third module, the monitoring module, the audio data that will gather the microphone array passes through the switch and transmits the industrial computer in a concentrated manner, by the operation such as the automatic initialization of industrial computer centralized control overall system, data acquisition and data transmission, the industrial computer mountable alarm system reports the running state of overall system in real time simultaneously. Broadcasting bare stream signals in real time by UDP/IP, and receiving signals by TCP/IP;
the fourth module, display module utilizes the display screen to show two-dimentional unusual audio signal, and the alarm lamp shows different colours and carries out audible-visual alarm through the alarm if discerning abnormal state then, and the alarm shows green when generating set normal operating, if detect out unusual audio frequency then the yellow light scintillation and buzzer sound production warning, if detect out unusual video like smog etc. then the yellow light scintillation and buzzer sound production warning.
The method is used on a generator set or equipment with larger noise, firstly, audio data of the equipment in normal operation is collected to train an audio model of the unit in a normal state, and the audio collected on site is compared with the audio model in the normal state during actual detection, so that whether the unit enters an abnormal state or not is judged. And if the abnormal state is identified, performing sound-light alarm through an alarm. If the abnormal state is identified, the abnormal sound source frequency extraction and the sound source positioning are continuously carried out. The method comprises the steps of firstly identifying newly added frequency on a frequency domain of audio data, extracting abnormal frequency through filtering according to the newly added frequency as abnormal frequency, and extracting amplitude of the filtered data on a time domain through inverse Fourier transform to serve as sound pressure amplitude of the abnormal audio. The sound pressure amplitude of abnormal audio data extracted by 32 paths of microphones is compared, the position of the microphone with the maximum amplitude is used as the direction of the abnormal sound source to carry out coarse positioning on the abnormal sound source, other 3 paths of microphones of the microphone and the nearest adjacent 4 paths of microphones form a sound pressure amplitude ratio equation, the optimal estimated position of the sound source is found according to the equation residual error and the minimum, and the coordinate set of the area where the coarse positioning is located is substituted into the estimated (x, y, z) to carry out fine positioning on the exceeded sound source. And finally displaying the abnormal sound source position on the two-dimensional image by the given (x, y, z) (on the premise that the three-dimensional detection coordinate system needs to be matched with the two-dimensional image display coordinate system).
Claims (10)
1. An abnormal sound source positioning method based on sound pressure amplitude ratio is characterized by comprising the following steps:
s1, normal state audio data acquisition and training: arranging a plurality of audio acquisition devices to acquire audio data of a target area, and firstly acquiring and training an audio model in a normal working state;
s2, abnormal audio recognition and extraction: the audio acquisition device acquires audio data of a target area in real time, compares the audio data with the trained audio model and identifies abnormal frequency;
s3, primary positioning of the sound source: comparing the abnormal frequency intensity acquired by each audio acquisition device, and selecting the audio acquisition device with the maximum intensity as a main device, wherein the abnormal sound source is closest to the main device;
s4, secondary positioning of the sound source position: and performing inverse Fourier transform on the abnormal frequency spectrum data, extracting the amplitude of the abnormal sound source audio data obtained by inverse Fourier transform on the time domain as the sound pressure amplitude of the abnormal audio, combining every two abnormal frequency sound pressure amplitudes obtained by each audio acquisition device, performing secondary positioning on the abnormal sound source by utilizing the relation that the pressure amplitude ratio is in inverse proportion to the distance square, and determining the actual coordinate of the abnormal sound source.
2. The abnormal sound source localization method based on sound pressure amplitude ratio according to claim 1, wherein in S1: the target area is a cuboid space, a generator set is arranged in the cuboid space, a group of audio acquisition devices are respectively arranged at eight vertexes of the cuboid, and each group of audio acquisition devices is composed of four microphones at different positions.
3. The abnormal sound source localization method based on sound pressure amplitude ratio according to claim 2, wherein in S1: the method comprises the steps of obtaining a spectrogram through Fourier transform of collected audio data, finding out main frequencies of audio in a frequency domain through peak detection, carrying out the operation on a plurality of groups of audio data to obtain the main frequencies, and inputting the main frequencies as features into an isolated forest machine learning algorithm for training to obtain a corresponding model.
4. The abnormal sound source localization method according to claim 3, wherein in S2: the method comprises the steps of identifying and extracting abnormal signals of collected audio data, carrying out Fourier transform on audio signals needing to be identified to obtain frequency domain information of the audio signals, carrying out peak detection to obtain main frequency, inputting the main frequency serving as characteristics into a trained machine learning model, judging the input characteristics by the model to obtain abnormal frequencies, and carrying out band-pass filtering on the audio signals to obtain time domain information of the abnormal frequencies.
5. The abnormal sound source localization method based on sound pressure amplitude ratio according to claim 4, wherein in S3: and comparing the real-time collected audio data with the trained normal state model data, and if the newly added frequency is identified on the frequency domain of the audio data by comparing the normal state model and the intensity of the newly added frequency is greater than a specified threshold value, considering the newly added frequency as an abnormal frequency, and realizing the extraction of the abnormal frequency by filtering on the frequency domain.
6. The sound pressure amplitude ratio-based abnormal sound source localization method of claim 5, wherein in step S4, the sound pressure amplitude ratio equation is constructed by using the other 3-way microphone of the main microphone and the nearest neighboring 4-way microphone, and the optimal estimated sound source position is found by using the equation residual and the minimum
Wherein H is the equation residual sum minimum, r is the distance from the sound source to the microphone, E is the effective sound pressure,
the sound pressure amplitude ratio is the effective sound pressure ratio, and the value is obtained by the measured value of each microphone.
7. The sound pressure amplitude ratio-based abnormal sound source localization method according to any one of claims 1 to 5, wherein S5 is performed after S4: and displaying the position coordinates of the abnormal audio signal in a two-dimensional image to display the position of the abnormal sound source.
8. The utility model provides an unusual sound source positioner based on sound pressure amplitude is than, includes audio acquisition device, its characterized in that: a cuboid space is planned by taking a target area to be detected as a center, a group of audio acquisition devices are respectively arranged at each vertex of the cuboid, and all the audio acquisition devices are connected with a control module through a switch.
9. The abnormal sound source localization apparatus according to claim 8, wherein: and building a support to install audio acquisition devices, wherein each audio acquisition device consists of 2-6 microphones, each group of microphones are arranged linearly, four groups of microphones positioned at the lower vertex of the cuboid are arranged linearly along the gravity direction, and four groups of microphones positioned at the upper vertex are arranged linearly along the horizontal direction.
10. The abnormal sound source localization apparatus according to claim 9, wherein: the control module is an industrial control computer and is used for controlling automatic initialization, data acquisition, data transmission and calculation of the whole system, and meanwhile, the industrial personal computer is provided with an alarm system and a display module to broadcast the running state of the whole system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210101953.1A CN114495975A (en) | 2022-01-27 | 2022-01-27 | Abnormal sound source positioning method and device based on sound pressure amplitude ratio |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210101953.1A CN114495975A (en) | 2022-01-27 | 2022-01-27 | Abnormal sound source positioning method and device based on sound pressure amplitude ratio |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114495975A true CN114495975A (en) | 2022-05-13 |
Family
ID=81477356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210101953.1A Pending CN114495975A (en) | 2022-01-27 | 2022-01-27 | Abnormal sound source positioning method and device based on sound pressure amplitude ratio |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114495975A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114964650A (en) * | 2022-08-01 | 2022-08-30 | 杭州兆华电子股份有限公司 | Gas leakage alarm method and device based on acoustic imaging |
| CN115604643A (en) * | 2022-12-12 | 2023-01-13 | 杭州兆华电子股份有限公司（Cn） | Automatic detection and positioning method for poor production of mobile phone charger |
| CN117854245A (en) * | 2023-12-25 | 2024-04-09 | 北京谛声科技有限责任公司 | Abnormal equipment monitoring method and system based on equipment operation audio |
-
2022
- 2022-01-27 CN CN202210101953.1A patent/CN114495975A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114964650A (en) * | 2022-08-01 | 2022-08-30 | 杭州兆华电子股份有限公司 | Gas leakage alarm method and device based on acoustic imaging |
| CN114964650B (en) * | 2022-08-01 | 2022-11-18 | 杭州兆华电子股份有限公司 | Gas leakage alarm method and device based on acoustic imaging |
| CN115604643A (en) * | 2022-12-12 | 2023-01-13 | 杭州兆华电子股份有限公司（Cn） | Automatic detection and positioning method for poor production of mobile phone charger |
| CN117854245A (en) * | 2023-12-25 | 2024-04-09 | 北京谛声科技有限责任公司 | Abnormal equipment monitoring method and system based on equipment operation audio |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114495975A (en) | Abnormal sound source positioning method and device based on sound pressure amplitude ratio | |
| JP6061693B2 (en) | Abnormality diagnosis apparatus and abnormality diagnosis method using the same | |
| CN101865789B (en) | Fault detecting device of near field acoustic holography sound image mode identification and detecting method thereof | |
| JP4368798B2 (en) | Sound source exploration system | |
| CN114414963A (en) | Acoustic imaging positioning system and method for intelligent monitoring of substation domain faults | |
| CN108731788A (en) | A kind of working at height arm low-frequency vibration vision inspection apparatus and method | |
| CN110006672A (en) | Rail vehicle fault monitoring method based on acoustic imaging technology | |
| JP5747340B2 (en) | Acoustic measuring device and acoustic measuring method | |
| JP2003111183A (en) | Sound source search system | |
| JP2002181913A (en) | Sound source probing system | |
| JPH1151300A (en) | Leakage monitoring device | |
| CN106066254B (en) | The detection in real time of factory's failure based on Internet of Things and positioning system and method | |
| CN107421622A (en) | A kind of monitoring of railway falling rocks and localization method | |
| CN105954796B (en) | A kind of method and apparatus for the hypocentral location for determining microseism | |
| CN110687506A (en) | Low-frequency noise source positioning device and method based on vector microphone array | |
| CN218037300U (en) | Abnormal sound source positioning device | |
| CN105203937B (en) | A kind of inside transformer discharge mode recognition methods and fault diagnosis system | |
| CN109019349A (en) | fault detection method, detector, computer storage medium and crane | |
| CN105954653B (en) | Method and system for locating breakdown point of GIS (gas insulated switchgear) field voltage withstand test | |
| CN109085248B (en) | Method, device and system for positioning impact source of pressure-bearing pipeline wall | |
| CN105785319B (en) | Airdrome scene target acoustical localization method, apparatus and system | |
| CN114151736B (en) | Ultrasonic three-array element alarm positioning instrument and method for monitoring natural gas leakage | |
| JP2011203048A (en) | Apparatus and method of locating sound source | |
| CN112130115A (en) | Transformer micro robotic fish positioning imaging method based on narrow-band ultrasonic beam formation | |
| CN112129813B (en) | Damage assessment method based on structural damage characteristic factor connection rule |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |