CN103278801A - Noise imaging detection device and detection calculation method for transformer substation - Google Patents
Noise imaging detection device and detection calculation method for transformer substation Download PDFInfo
- Publication number
- CN103278801A CN103278801A CN 201310191316 CN201310191316A CN103278801A CN 103278801 A CN103278801 A CN 103278801A CN 201310191316 CN201310191316 CN 201310191316 CN 201310191316 A CN201310191316 A CN 201310191316A CN 103278801 A CN103278801 A CN 103278801A
- Authority
- CN
- China
- Prior art keywords
- noise
- microphone acoustic
- audio
- acoustic pickup
- transformer station
- 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
Images
Landscapes
- Circuit For Audible Band Transducer (AREA)
Abstract
The invention discloses a noise imaging detection device and a noise imaging detection calculation method for a transformer substation. The noise imaging detection device consists of a microphone sound pick-up array, an audio preprocessing module, a multipath audio acquisition card, a digital signal processing control module, a short-term energy spectrum estimating unit, a sound mixing unit, a client, a display, an input keyboard and an acousto-optic alarm. According to the detection calculation method, the accurate time when a noise source reaches a sound pick-up is acquired by establishing a three-dimensional sound pick-up distribution model, and fault noise can be positioned rapidly and accurately via audio preprocessing, sound mixing processing, short-term energy spectrum estimation and application of three-dimensional spatial time difference positioning algorithm program calculation, so that technical support is provided for a high-voltage transformer substation to timely control the equipment running situation in an unattended machine room, and theoretical basis is provided for the accurate judgment of equipment fault points.
Description
Technical field
The present invention relates to noise imaging detection techniques field, relate in particular to a kind of transformer station's noise imaging method for detecting device and method for detecting based on the TDOA algorithm.
Background technology
Transformer station's high pressure machine room is an environment that has multiple noise.When some fault appears in equipment, send sometimes some improper sounds, these improper sounds just are mingled in neighbourhood noise.If can accurately locate these improper sounds, with regard to the accurate failure judgement equipment of energy, thereby fix a breakdown, reduce the loss that equipment failure brings transformer station.
But transformer station's high pressure machine room is a place that environment is more special simultaneously, under the substation operation state, be not allow the close equipment of people, belong to hazardous location, duration of work forbids that personnel approach, therefore, equipment failure can't be predicted, can't locate after equipment failure, very difficult to the detection of its accident defect.And between regular turn(a)round, the phenomenon of the failure when technician can't reappear high pressure again, the breakthrough that never is resolved of this problem, can only rely on and manually infer by rule of thumb, and large tracts of land is exchange device more.Therefore, resemble this class hazardous area of high voltage substation, especially need to be without the people intelligent sonic location system at the scene.At present, the intelligent degree of the lookup method in the extraordinary noise source of transformer station is very low, can't search fast and accurately noise source, deal with problems.Therefore research auditory localization under noisy environment just seems particularly important, and also has using value in other industry.
Auditory localization is to utilize acoustics and electronic installation to accept and process acoustic field signal, determines a kind of technology of sound generation source, and the microphone array auditory localization algorithm used now is substantially based on three kinds of auditory localization principles.(1) the auditory localization principle of imitative ears.(2) the auditory localization principle based on time of arrival poor (TDOA).(3) positioning principle based on Amplitude ratio of sound pressure.Through research for many years, some actual available positioning systems have been arranged in the world.But up to now, advanced country's equipment use great majority are the development of five sixties and improved products mostly.
Domestic each enterprise after entering 21 century, study colleges and universities etc. in one's power and obtained certain achievement in the research aspect the processing of microphone acoustic pickup array voice signal.2004 old can, Wang Zengfu proposes a kind of method of using Amplitude ratio of sound pressure to carry out auditory localization.The voltage signal amplitude that the method receives from each acoustic pickup of array and corresponding acoustic pickup are to relation sound source distance to be measured, provided and take the constraint condition expression formula that Amplitude ratio of sound pressure is parameter, set up the algorithm that utilizes these constraint conditions to carry out auditory localization.Within 2004, quiet right grade of woods proposes a kind of improved near field auditory localization and Speech separation algorithm based on microphone acoustic pickup array.It realizes auditory localization and speech Separation in conjunction with dualbeam two-dimensional localization and near field minimum variance beam-forming technology in the array near field range.The robot perception systematic study based on Multi-source Information Fusion towards the anti-terrorism field that Hebei University of Technology in 2006 completes is that integrated vision sensor, hearing transducer and olfactory sensor are in the tracker of many sensory information integration technology of one.3 microphone acoustic pickup arrays that auditory system adopts isosceles triangle to distribute, the azimuth-range that judges sound source by calculating time delay between each microphone acoustic pickup and geometric relationship.But, our country's research in this respect or in the experimental study stage, also do not enter the practical stage.
Use at present the noise orientator, adopt one-dimensional linear array or two-dimensional array acquisition noise more, it is larger that one-dimensional array picks up limitation to the sound field directional information, and precision is also lower; It is larger that two-dimensional array is affected at the measuring distance time error by incident direction, the sensitivity of microphone is low, system running speed slow, inefficiency, the also Give maintenance work that requires that can't meet in real work brings very large puzzlement, and therefore the relevant substitute products of first chance more not on market are badly in need of a set of more advanced, the noise positioning system accurately and efficiently of research and development.
In the method for localization method use microphone acoustic pickup array to speaker location and tracking based on time of arrival poor (TDOA), it is simple that employing time delay localization method has calculating, and the advantages such as accurate positioning, get more and more people's extensive concerning.Often use in actual applications classical broad sense cross-correlation method or its improve algorithm ask the microphone acoustic pickup between time delay.Yet often have incoherent ground unrest and reverberation in actual environment, they will have a strong impact on the quality that the microphone acoustic pickup receives signal.In the situation that consider noise and reverberation, with classical or improved method computation delay, often lost efficacy.
summary of the invention
The invention provides a kind of intelligentized, accuracy is high, search efficiently the transformer station's noise imaging arrangement for detecting based on the TDOA algorithm and detecting computing method
.
Realize that the scheme that above-mentioned purpose adopts is: a kind of transformer station noise imaging arrangement for detecting is by microphone acoustic pickup array, the audio frequency pretreatment module, the multichannel voice frequency capture card, the digital signal processing control module, the short-time energy spectral estimation unit, the audio mixing unit, client, display, input keyboard and audible-visual annunciator form, and wherein, microphone acoustic pickup array is connected with the audio input end of audio mixing unit with described audio frequency pretreatment module respectively, the audio output of audio frequency pretreatment module successively with described multichannel voice frequency capture card, the digital signal processing control module is connected with the short-time energy spectral estimation unit, the audio output of described audio mixing unit is connected with the short-time energy spectral estimation unit, the control signal output terminal of short-time energy spectral estimation unit is connected with described multichannel voice frequency capture card, and described digital signal processing control module is by bus and client, display, input keyboard is connected with audible-visual annunciator, in described multichannel voice frequency capture card, the active audio frequency filter unit is set, described active audio frequency filtering by continuous time integrated filter and tone filter form.
Described microphone acoustic pickup array arranges according to three-dimensional rectangle cube model structure.
Described audio mixing unit adopts the Multi-channel audio sound mixing device.
Described audio frequency pretreatment module, comprise self-adaptation amplifying circuit, pre-filtering circuit, minute frame circuit and windowing circuit, and described pre-filtering circuit is low-pass filter circuit, and the frame of described minute frame circuit setting moves as 10ms.
Described multichannel voice frequency capture card adopts 10 A/D converters, and this number converter comprises analog input multiplexer, automatic zero set (AZS) comparer, clock generator, 10 successive approximation registers and output register.
Described three-dimensional rectangle cube model at least arranges 6 microphone acoustic pickups.
A kind of transformer station noise imaging detecting computing method comprise the following steps:
1) according to transformer station's device location to be detected, set up rectangle cube mathematical model, based on this model, set up three-dimensional system of coordinate, at least 6 microphone acoustic pickups of each fixed position setting of this model, except the base plane of model, have three acceptance points on other plane at least;
2) sound signal of microphone acoustic pickup collection is carried out to self-adaptation amplification, pre-filtering, minute frame, windowing by described audio frequency pretreatment module;
3) sound signal of microphone acoustic pickup collection is carried out to stereo process by described audio mixing unit, obtain simulation panorama noise signal;
4) position in the time of simulating the panorama noise signal and estimate the identical received energy of different microphone acoustic pickups by described short-time energy spectral estimation unit within the shortest time, the impact that when making up noise source and arriving different microphone acoustic pickup, sound intensity brings;
5) will gather, filter by the multichannel voice frequency capture card through the sound signal of self-adaptation amplification, pre-filtering, minute frame, windowing and the sound signal after the processing of short-time energy spectral estimation unit, and be converted to digital audio and video signals;
6) described digital audio and video signals is reached to described client by described digital signal processing control module, advance end-point detection and filter out the effective noise signal data, utilize three dimensions Localization Estimate Algorithm of TDOA operation program to be calculated, must locate detecting origin of target noise location;
7) adopt the image superimposing technique to take transformer station model as showing bottom, by microphone acoustic pickup array ultrasonogram, be that the dynamic superpose layer is realized the real-time Overlapping display of noise data information, accomplish that transformer station's static object and noise respective objects are mutually identical, realize noise imaging.
Described three dimensions Localization Estimate Algorithm of TDOA operation program, its algorithm expression way comprises:
Three dimensions time difference positioning equation group:
Arrange the abbreviation equation:
In formula: (x
i, y
i, z
i)
t, i=0,1,2,3 ... for the locus of each microphone acoustic pickup, wherein i=0 means main microphone acoustic pickup, i=1, and 2,3 mean auxiliary microphone acoustic pickup, (x
i, y
i, z
i)
tfor the locus of target, r
imean the distance between sound source and i microphone acoustic pickup, △ r
imean that sound source arrives (x to i microphone acoustic pickup and sound source
0, y
0, z
0)
trange difference between the microphone acoustic pickup;
The matrix expression of three dimensions time difference positioning equation group:
If rank (A)=3, the sound source position estimated value is:
The estimated value of target location is can be obtained fom the above equation:
Transformer station of the present invention noise imaging arrangement for detecting and method for detecting, in the actual conditions that take into full account the high voltage substation machine room, set up three-dimensional three-dimensional acoustic pickup distributed model, based on this model, at the unmanned front end, obtain the correct time that noise source arrives acoustic pickup, by ethernet line, data transmission is processed to the server of center duty room, obtain accurate location and the distribution situation of noise source.Sound being obtained and processes separately of this detecting system novelty processed, and that the system front end acoustic pickup has is portable, process the characteristics such as rapid, real-time.This detecting system has fundamentally solved accurately early warning of big-and-middle-sized transformer station's high pressure machine room equipment failure, the difficult problems such as abort situation can not accurately judge, can provide technical guarantee for the machine operation of grasping in time unmanned machine room, for the accurate judgement of equipment failure point provides theoretical foundation.This detecting system also has following characteristics:
(1) acquisition and processing separately, has guaranteed the accuracy of noise source location;
(2) time of arrival, acquisition module adopted the devices at full hardware circuit to realize, had guaranteed the precision of location;
(3) clock timing are carried out clock alignment with center-side, guarantee the accuracy of clock;
(4) action status information that can print record is as noise source position coordinates, noise source time of occurrence, noise source rack information of living in etc.;
(5) making of equipment set device is small and exquisite, exquisite, and on-the-spot installation wiring is simpler, convenient.
The accompanying drawing explanation
Below in conjunction with accompanying drawing, the present invention is described further:
Fig. 1 is the structural representation of transformer station's noise imaging arrangement for detecting in the present invention.
Fig. 2 is in the present invention
threethe structural representation of dimension rectangle cube model.
Fig. 3 is the FB(flow block) of transformer station's noise imaging method for detecting in the present invention.
Embodiment
As Fig. 1, shown in Fig. 2, transformer station of the present invention noise imaging arrangement for detecting, by microphone acoustic pickup array 1, audio frequency pretreatment module 2, multichannel voice frequency capture card 4, digital signal processing control module 6, short-time energy spectral estimation unit 5, audio mixing unit 3, client 7, display 8, input keyboard 9 and audible-visual annunciator 10 form, wherein, microphone acoustic pickup array 1 is connected with the audio input end of audio mixing unit 3 with described audio frequency pretreatment module 2 respectively, the audio output of audio frequency pretreatment module 2 successively with described multichannel voice frequency capture card 4, digital signal processing control module 6 is connected with short-time energy spectral estimation unit 5, the audio output of described audio mixing unit 3 is connected with short-time energy spectral estimation unit 5, the control signal output terminal of short-time energy spectral estimation unit 5 is connected with described multichannel voice frequency capture card 4, described digital signal processing control module 6 is by bus and client 7, display 8, input keyboard 9 is connected with audible-visual annunciator 10.
In transformer station's noise imaging arrangement for detecting, described microphone acoustic pickup array 1 setting of the structure according to three-dimensional rectangle cube model 12, for the collection of transformer station's noise audio signal, described three-dimensional rectangle cube model 12 at least arranges 6 microphone acoustic pickups 13.For the specific space of high voltage substation, set up rectangle cube mathematical model, set up three-dimensional system of coordinate 11 based on this model, be referred to as three-dimensional rectangle cube model 12, this detecting system designs 6 microphone acoustic pickups 1 altogether
3, except base plane, supreme on other plane have three acceptance points.Like this, the sound source receiving trap that any four points of Different Plane form, can locate three-dimensional sound source position.According to the TDOA algorithm model, least twice calculates, the locus that just can accurately orient noise source.During practical application, can calculate respectively with four points of Different Plane, the result of repeatedly calculating can obtain a spheroid, by asking the center of gravity of this sphere, just can accurately locate noise source, improves precision.
Described audio mixing unit 3 adopts the Multi-channel audio sound mixing device, for the stereo process of sound signal, and output simulation panorama noise signal.The aerial velocity of propagation of sound is 340 meter per seconds approximately, must obtain accurately the mistiming that noise source arrives each receiving trap, could accurately utilize algorithm and model of the present invention to position, therefore, the real-time reception of sound and fast processing are the keys of whole system success or failure.The detecting real-time of sound and reception in the present invention, adopted the audio mixing unit 3 based on hardware circuit to realize.The input end of audio mixing unit 3 is 6 each and every one microphone acoustic pickups 13.Owing to having adopted simulation audio mixing technology, after any one receiving trap receives the sound of noise source, response can be made the very first time in audio mixing unit 3, starts collecting unit 6 each and every one microphone acoustic pickups 13 are gathered at synchronization simultaneously.The response time of this audio mixing unit 3 can reach below Microsecond grade the soonest, meets actual detecting needs.Its audio mixing principle of work is, the sound that noise source is sent arrive at first which microphone acoustic pickup 13 depend on noise source from which microphone acoustic pickup 13 more close to.In actual use, timing while from which point coming on earth, this can't predict in advance.Therefore, will use audio mixing unit 3 here, no matter which microphone acoustic pickup 13 is noise source arrive, the audio frequency after audio mixing all can start timing by trigger automatically, accurately calculates the time that arrives each receiver.
Described audio frequency pretreatment module 2, comprise self-adaptation amplifying circuit, pre-filtering circuit, minute frame circuit and windowing circuit, for amplification, filtering, minute frame and the windowing process of transformer station's noise signal, described pre-filtering circuit is low-pass filter circuit, and the frame of described minute frame circuit setting moves as 10ms.Sound is propagated in air, can decay along with propagation distance.To the location of noise source, be the sound receiver with the diverse location that is distributed in space in fact, receive the sound that noise source is sent.Difference due to locus, will inevitably cause the error degree difference of sound signal, it is exactly that the amplitude of the voice signal that it can receive acoustic pickup is amplified in certain scope in order to address this problem that self-adaptation is amplified, for sampling and the accurate judgement in later stage are given security time of arrival.The pre-filtering purpose of voice signal is to suppress each frequency domain components medium frequency of input signal to exceed the institute of s/2 important, and wherein s is sample frequency, to prevent frequency alias, disturbs.Like this, prefilter must be a low-pass filter, establishes its cutoff frequency and is
h, for most audio codecs,
h=3400Hz, sample frequency s=8KHz.Sound signal is the spectral characteristic held stationary at short notice, there is smooth performance in short-term, therefore sound signal can be divided into to the very little time period (approximately 10～30 ms) when actual treatment, be referred to as " frame ", least unit as Audio Signal Processing, non-overlapped part between frame and frame is called frame moves, and the process that sound signal is divided into some frames is called to a minute frame.Minute frame is little can clearly describe the time varying characteristic of voice signal but calculated amount is large; Minute frame is large can reduce calculated amount but between consecutive frame, change little, easy lossing signal feature.Generally get frame length 20ms, frame moves as 1/3～1/2 of frame length, and the present invention gets frame and moves as 10ms.Before data are processed, in order to keep the stationarity in short-term of voice signal, utilize window function to reduce the Gibbs effect caused by truncation.A frame speech data s (n) who has taken out is carried out to windowing process, and the windowing formula is to take advantage of s (n) with certain window function w (n), thereby forms windowing voice Sw (n).
Described multichannel voice frequency capture card 4 adopts 10 A/D converters (ADC), and this number converter comprises the analog input multiplexer, automatic zero set (AZS) comparer, clock generator, 10 successive approximation registers (SAR), output register.For sound signal is converted to digital signal.In this multichannel voice frequency capture card 4, the active audio frequency filter unit also is set, described active audio frequency filtering by continuous time integrated filter and tone filter form.Described ADC also provides the sleep pattern of selection able to programme to save power consumption.
Described short-time energy spectral estimation unit 5 adopts accurate noise to estimate to come the characteristic of acquisition noise, and obtain the expectation the estimation voice, position when estimating different audio frequency receiving the identical received energy of microphone acoustic pickups 13 within the shortest time, for the acoustic environment at high s/n ratio, can distinguish noiseless.Now, the energy of non-fault noise is very little, and noisy energy value increases to a certain numerical value very significantly, can distinguish thus the starting point and ending point of noise signal, simultaneously to multichannel voice frequency capture card output single channel analogue audio frequency.In the quick audio treatment technology, the most important thing is the estimation of short-time energy spectrum, short-time energy spectral estimation unit 5 of the present invention can adopt accurate noise to estimate to come the characteristic of acquisition noise, obtain the estimation voice of expectation by good enhancing algorithm, position in the time of can within the shortest time, estimating the identical received energy of different audio frequency receiving elements, be beneficial to the impact that when making up noise source and arriving different receiving trap, sound intensity brings, and have that applicable signal to noise ratio (S/N ratio) scope is large, method simple, be easy to the advantage such as processing in real time.At first speech recognition in actual application, all requires the input audio signal of system is judged, accurately finds out the starting point and ending point of voice segments.Do like this, can make the data that gather is really the data of voice signal, thereby reduces data volume and operand and reduce the processing time.The problem of differentiating the starting point and ending point of voice segments mainly is summed up as the problem of difference voice and noise.If neighbourhood noise and system input noise are very little, can guarantee that the input signal-to-noise ratio of system is very high, as long as calculate the short-time energy of input signal, just can come voice segments and noise background difference so.But, be difficult in actual applications guarantee so high signal to noise ratio (S/N ratio), thereby can not only rely on short-time energy to differentiate the terminal of voice segments.On the other hand, the differentiation of the terminal of some voice segments can run into special difficulty, for example, when beginning and the end of voice segments are all weak fricative or weak plosive situation, when the voice segments end is the situation of nasal sound, the short-time energy of these sounds is very little, often with ground unrest in identical level.The short-time average zero-crossing rate of voiceless sound and above lifted phoneme in these cases, can utilize short-time average zero-crossing rate to be judged, because will exceed several times than the average zero-crossing rate of ground unrest.
Described digital signal processing control module 6 is for the processing of digital audio and video signals, and is transferred to client 7, according to the command information of client 7, short-time energy spectral estimation unit 5 and multichannel voice frequency capture card 4 controlled simultaneously.
Referring to Fig. 3, transformer station of the present invention noise imaging detecting computing method comprise the following steps:
1) set up rectangle cube mathematical model 1: according to transformer station's device location to be detected, set up rectangle cube mathematical model, set up three-dimensional system of coordinate based on this model, at at least 6 microphone acoustic pickups of each fixed position setting of this model, except the base plane of model, have three acceptance points on other plane at least;
2) the audio frequency pre-service 2: the sound signal to the collection of microphone acoustic pickup is carried out self-adaptation amplification, pre-filtering, minute frame, windowing by described audio frequency pretreatment module;
3) stereo process 3: the sound signal to the collection of microphone acoustic pickup is carried out stereo process by described audio mixing unit, obtains simulation panorama noise signal;
4) short-time energy spectrum estimates 4: the position in the time of simulating the panorama noise signal and estimate the identical received energy of different microphone acoustic pickups by described short-time energy spectral estimation unit within the shortest time, the impact that when making up noise source and arriving different microphone acoustic pickup, sound intensity brings;
5) sound signal mould/number conversion 5: will gather, filter by the multichannel voice frequency capture card through the sound signal of self-adaptation amplification, pre-filtering, minute frame, windowing and the sound signal after the processing of short-time energy spectral estimation unit, and be converted to digital audio and video signals;
6) detecting origin of target noise location 6: described digital audio and video signals is reached to described client by described digital signal processing control module, advance end-point detection and filter out the effective noise signal data, utilize three dimensions Localization Estimate Algorithm of TDOA operation program to be calculated, must locate detecting origin of target noise location;
Described three dimensions Localization Estimate Algorithm of TDOA operation program, be based on the description of small echo Time Delay Estimation Algorithms
This algorithm for estimating is a kind of improvement algorithm on the broad sense Time Delay Estimation Method.Continuous wavelet transform is a kind of digital conversion mode, when being applied in digital signal processing by the method, from the system responses angle, in fact the output of signal after a series of band-pass filter, from the spectrum analysis angle, it is to be on the frequency band that a system selectivity is identical by signal decomposition that small echo changes.Experimental results show that the Time Delay Estimation Algorithms based on wavelet transformation, cancelled the assumed condition of generalized correlation method, enlarged the range of application that time delay is estimated, and improved precision.Analyze theoretically, the microphone acoustic pickup is no less than 3 just can carry out target localization to sound source, we by problem reduction, be at first in two dimensional surface to the location of sound source, on this basis, then algorithm is generalized to the three dimensions situation.In three dimensions, because two sensors are determined a pair of hyperboloid, so at least need four sensors to position target.Suppose that positioning system consists of 4 sensors, the locus of each sensor is (x
i, y
i, z
i)
t, i=0,1,2,3.Wherein i=0 means master reference, i=1, and 2,3 mean auxiliary sensor.The locus of target is (x, y, z)
t, r
imean the distance between sound source and i sensor, △ r
imean that sound source arrives (x to i sensor and sound source
0, y
0, z
0)
trange difference between sensor by the Representation Equation is:
Three dimensions time difference positioning equation group:
Arrange the abbreviation equation:
The matrix expression of three dimensions time difference positioning equation group:
If rank (A)=3, the sound source position estimated value is:
The estimated value of target location is can be obtained fom the above equation:
B. embed the three dimensions Localization Estimate Algorithm of TDOA operation program based on the small echo Time Delay Estimation Algorithms in described client, its algorithm expression way comprises:
Three dimensions time difference positioning equation group:
Arrange the abbreviation equation:
In formula: (x
i, y
i, z
i)
t, i=0,1,2,3 ... for the locus of each microphone acoustic pickup, wherein i=0 means main microphone acoustic pickup, i=1, and 2,3 mean auxiliary microphone acoustic pickup, (x
i, y
i, z
i)
tfor the locus of target, r
imean the distance between sound source and i microphone acoustic pickup, △ r
imean that sound source arrives (x to i microphone acoustic pickup and sound source
0, y
0, z
0)
trange difference between the microphone acoustic pickup;
The matrix expression of three dimensions time difference positioning equation group:
if rank (A)=3, the sound source position estimated value is:
The estimated value of target location is can be obtained fom the above equation:
7) noise imaging, demonstration, warning 7: adopt the image superimposing technique to take transformer station model as showing bottom, by microphone acoustic pickup array ultrasonogram, be that the dynamic superpose layer is realized the real-time Overlapping display of noise data information, accomplish that transformer station's static object and noise respective objects are mutually identical, realize noise imaging.When noise reflection converting equipment occurs when abnormal, by audible-visual annunciator and alarm.
Claims (8)
1. transformer station's noise imaging arrangement for detecting, is characterized in that: be by microphone acoustic pickup array, the audio frequency pretreatment module, the multichannel voice frequency capture card, the digital signal processing control module, the short-time energy spectral estimation unit, the audio mixing unit, client, display, input keyboard and audible-visual annunciator form, and wherein, microphone acoustic pickup array is connected with the audio input end of audio mixing unit with described audio frequency pretreatment module respectively, the audio output of audio frequency pretreatment module successively with described multichannel voice frequency capture card, the digital signal processing control module is connected with the short-time energy spectral estimation unit, the audio output of described audio mixing unit is connected with the short-time energy spectral estimation unit, the control signal output terminal of short-time energy spectral estimation unit is connected with described multichannel voice frequency capture card, and described digital signal processing control module is by bus and client, display, input keyboard is connected with audible-visual annunciator, in described multichannel voice frequency capture card, the active audio frequency filter unit is set, described active audio frequency filtering by continuous time integrated filter and tone filter form.
2. transformer station as claimed in claim 1 noise imaging arrangement for detecting, it is characterized in that: described microphone acoustic pickup array arranges according to three-dimensional rectangle cube model structure.
3. transformer station as claimed in claim 1 noise imaging arrangement for detecting, is characterized in that: described audio mixing unit employing Multi-channel audio sound mixing device.
4. transformer station as claimed in claim 1 noise imaging arrangement for detecting, it is characterized in that: described audio frequency pretreatment module, comprise self-adaptation amplifying circuit, pre-filtering circuit, minute frame circuit and windowing circuit, described pre-filtering circuit is low-pass filter circuit, and the frame of described minute frame circuit setting moves as 10ms.
5. transformer station as claimed in claim 1 noise imaging arrangement for detecting, it is characterized in that: described multichannel voice frequency capture card adopts 10 A/D converters, and this number converter comprises analog input multiplexer, automatic zero set (AZS) comparer, clock generator, 10 successive approximation registers and output register.
6. transformer station as claimed in claim 2 noise imaging arrangement for detecting, it is characterized in that: described three-dimensional rectangle cube model at least arranges 6 microphone acoustic pickups.
7. transformer station's noise imaging is detected computing method, and it is characterized in that: it comprises the following steps:
1) according to transformer station's device location to be detected, set up rectangle cube mathematical model, based on this model, set up three-dimensional system of coordinate, at least 6 microphone acoustic pickups of each fixed position setting of this model, except the base plane of model, have three acceptance points on other plane at least;
2) sound signal of microphone acoustic pickup collection is carried out to self-adaptation amplification, pre-filtering, minute frame, windowing by described audio frequency pretreatment module;
3) sound signal of microphone acoustic pickup collection is carried out to stereo process by described audio mixing unit, obtain simulation panorama noise signal;
4) position in the time of simulating the panorama noise signal and estimate the identical received energy of different microphone acoustic pickups by described short-time energy spectral estimation unit within the shortest time, the impact that when making up noise source and arriving different microphone acoustic pickup, sound intensity brings;
5) will gather, filter by the multichannel voice frequency capture card through the sound signal of self-adaptation amplification, pre-filtering, minute frame, windowing and the sound signal after the processing of short-time energy spectral estimation unit, and be converted to digital audio and video signals;
6) described digital audio and video signals is reached to described client by described digital signal processing control module, advance end-point detection and filter out the effective noise signal data, utilize three dimensions Localization Estimate Algorithm of TDOA operation program to be calculated, must locate detecting origin of target noise location;
7) adopt the image superimposing technique to take transformer station model as showing bottom, by microphone acoustic pickup array ultrasonogram, be that the dynamic superpose layer is realized the real-time Overlapping display of noise data information, accomplish that transformer station's static object and noise respective objects are mutually identical, realize noise imaging.
8. transformer station as claimed in claim 7 noise imaging is detected computing method, it is characterized in that: described three dimensions Localization Estimate Algorithm of TDOA operation program, and its algorithm expression way comprises:
Three dimensions time difference positioning equation group:
Arrange the abbreviation equation:
In formula: (x
i, y
i, z
i)
t, i=0,1,2,3 ... for the locus of each microphone acoustic pickup, wherein i=0 means main microphone acoustic pickup, i=1, and 2,3 mean auxiliary microphone acoustic pickup, (x
i, y
i, z
i)
tfor the locus of target, r
imean the distance between sound source and i microphone acoustic pickup, △ r
imean that sound source arrives (x to i microphone acoustic pickup and sound source
0, y
0, z
0)
trange difference between the microphone acoustic pickup;
The matrix expression of three dimensions time difference positioning equation group:
If rank (A)=3, the sound source position estimated value is:
The estimated value of target location is can be obtained fom the above equation:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201310191316 CN103278801A (en) | 2013-05-22 | 2013-05-22 | Noise imaging detection device and detection calculation method for transformer substation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201310191316 CN103278801A (en) | 2013-05-22 | 2013-05-22 | Noise imaging detection device and detection calculation method for transformer substation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103278801A true CN103278801A (en) | 2013-09-04 |
Family
ID=49061376
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201310191316 Pending CN103278801A (en) | 2013-05-22 | 2013-05-22 | Noise imaging detection device and detection calculation method for transformer substation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103278801A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105676182A (en) * | 2016-02-26 | 2016-06-15 | 南方电网科学研究院有限责任公司 | Positioning method and system for wind noise source |
| CN106153363A (en) * | 2015-04-07 | 2016-11-23 | 中国人民解放军海军工程大学 | A kind of mechanical equipment fault automatic identifying method based on acoustic image monitoring |
| CN108344504A (en) * | 2018-05-02 | 2018-07-31 | 莆田学院 | Noise detection system and analysis method in 8 road car of one kind |
| WO2019036962A1 (en) * | 2017-08-23 | 2019-02-28 | 深圳企管加企业服务有限公司 | Machine room noise prompt system based on internet of things |
| CN110488896A (en) * | 2019-08-16 | 2019-11-22 | 深圳供电局有限公司 | A kind of control system and its control method of building environment |
| CN111540347A (en) * | 2020-05-12 | 2020-08-14 | 山东科华电力技术有限公司 | Cable tunnel monitoring method and system based on audio |
| CN112305501A (en) * | 2020-10-21 | 2021-02-02 | 珠海格力电器股份有限公司 | Method and device for determining noise source, storage medium and electronic device |
| CN112562619A (en) * | 2020-10-27 | 2021-03-26 | 东风汽车集团有限公司 | Noise reduction method for robot automatic stamping line |
| CN112945597A (en) * | 2021-01-28 | 2021-06-11 | 河海大学 | Transformer substation high-voltage equipment operation state monitoring method based on voice recognition method |
| CN112946578A (en) * | 2021-02-02 | 2021-06-11 | 上海头趣科技有限公司 | Novel double-ear positioning method |
| CN113267330A (en) * | 2021-05-14 | 2021-08-17 | 国网重庆市电力公司电力科学研究院 | GIS equipment mechanical fault detection system and method based on acoustic imaging |
| CN114001816A (en) * | 2021-12-30 | 2022-02-01 | 成都航空职业技术学院 | Acoustic imager audio acquisition system based on MPSOC |
| CN114397010A (en) * | 2021-12-29 | 2022-04-26 | 南京中科声势智能科技有限公司 | Transient signal acoustic imaging method based on wavelet decomposition |
| WO2022133739A1 (en) * | 2020-12-22 | 2022-06-30 | 贵州电网有限责任公司 | Time difference-based sound source positioning method and apparatus for head-mounted ar glasses |
| CN116338583A (en) * | 2023-04-04 | 2023-06-27 | 北京华控智加科技有限公司 | Method for determining noise source inside equipment based on distributed microphone array |
| CN117420503A (en) * | 2023-12-15 | 2024-01-19 | 国网天津市电力公司电力科学研究院 | Positioning system and method for transformer internal inspection equipment |
-
2013
- 2013-05-22 CN CN 201310191316 patent/CN103278801A/en active Pending
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106153363A (en) * | 2015-04-07 | 2016-11-23 | 中国人民解放军海军工程大学 | A kind of mechanical equipment fault automatic identifying method based on acoustic image monitoring |
| CN105676182A (en) * | 2016-02-26 | 2016-06-15 | 南方电网科学研究院有限责任公司 | Positioning method and system for wind noise source |
| WO2019036962A1 (en) * | 2017-08-23 | 2019-02-28 | 深圳企管加企业服务有限公司 | Machine room noise prompt system based on internet of things |
| CN108344504A (en) * | 2018-05-02 | 2018-07-31 | 莆田学院 | Noise detection system and analysis method in 8 road car of one kind |
| CN110488896A (en) * | 2019-08-16 | 2019-11-22 | 深圳供电局有限公司 | A kind of control system and its control method of building environment |
| CN111540347A (en) * | 2020-05-12 | 2020-08-14 | 山东科华电力技术有限公司 | Cable tunnel monitoring method and system based on audio |
| CN112305501A (en) * | 2020-10-21 | 2021-02-02 | 珠海格力电器股份有限公司 | Method and device for determining noise source, storage medium and electronic device |
| CN112562619A (en) * | 2020-10-27 | 2021-03-26 | 东风汽车集团有限公司 | Noise reduction method for robot automatic stamping line |
| CN112562619B (en) * | 2020-10-27 | 2024-03-26 | 东风汽车集团有限公司 | Robot automatic stamping line noise reduction method |
| WO2022133739A1 (en) * | 2020-12-22 | 2022-06-30 | 贵州电网有限责任公司 | Time difference-based sound source positioning method and apparatus for head-mounted ar glasses |
| CN112945597A (en) * | 2021-01-28 | 2021-06-11 | 河海大学 | Transformer substation high-voltage equipment operation state monitoring method based on voice recognition method |
| CN112946578A (en) * | 2021-02-02 | 2021-06-11 | 上海头趣科技有限公司 | Novel double-ear positioning method |
| CN112946578B (en) * | 2021-02-02 | 2023-04-21 | 上海头趣科技有限公司 | Binaural localization method |
| CN113267330A (en) * | 2021-05-14 | 2021-08-17 | 国网重庆市电力公司电力科学研究院 | GIS equipment mechanical fault detection system and method based on acoustic imaging |
| CN113267330B (en) * | 2021-05-14 | 2023-03-14 | 国网重庆市电力公司电力科学研究院 | GIS equipment mechanical fault detection system and method based on acoustic imaging |
| CN114397010A (en) * | 2021-12-29 | 2022-04-26 | 南京中科声势智能科技有限公司 | Transient signal acoustic imaging method based on wavelet decomposition |
| CN114001816B (en) * | 2021-12-30 | 2022-03-08 | 成都航空职业技术学院 | Acoustic imager audio acquisition system based on MPSOC |
| CN114001816A (en) * | 2021-12-30 | 2022-02-01 | 成都航空职业技术学院 | Acoustic imager audio acquisition system based on MPSOC |
| CN116338583A (en) * | 2023-04-04 | 2023-06-27 | 北京华控智加科技有限公司 | Method for determining noise source inside equipment based on distributed microphone array |
| CN116338583B (en) * | 2023-04-04 | 2023-09-01 | 北京华控智加科技有限公司 | Method for determining noise source inside equipment based on distributed microphone array |
| CN117420503A (en) * | 2023-12-15 | 2024-01-19 | 国网天津市电力公司电力科学研究院 | Positioning system and method for transformer internal inspection equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103278801A (en) | Noise imaging detection device and detection calculation method for transformer substation | |
| Argentieri et al. | A survey on sound source localization in robotics: From binaural to array processing methods | |
| US9432770B2 (en) | Method and device for localizing sound sources placed within a sound environment comprising ambient noise | |
| CN109839612A (en) | Sounnd source direction estimation method based on time-frequency masking and deep neural network | |
| WO2018045973A1 (en) | Sound source localization method for robot, and system | |
| CN103308889B (en) | Passive sound source two-dimensional DOA (direction of arrival) estimation method under complex environment | |
| CN110875060A (en) | Voice signal processing method, device, system, equipment and storage medium | |
| CN104991573A (en) | Locating and tracking method and apparatus based on sound source array | |
| Liu et al. | Continuous sound source localization based on microphone array for mobile robots | |
| CN111429939B (en) | Sound signal separation method of double sound sources and pickup | |
| CN109669159A (en) | Auditory localization tracking device and method based on microphone partition ring array | |
| CN106373589B (en) | A kind of ears mixing voice separation method based on iteration structure | |
| CN102305925A (en) | Robot continuous sound source positioning method | |
| CN106226739A (en) | Merge the double sound source localization method of Substrip analysis | |
| CN108109617A (en) | A kind of remote pickup method | |
| CN109188362A (en) | A kind of microphone array auditory localization signal processing method | |
| CN107167770A (en) | A kind of microphone array sound source locating device under the conditions of reverberation | |
| CN109859749A (en) | A kind of voice signal recognition methods and device | |
| CN103760520A (en) | Monolingual sound source DOA estimation method based on AVS and sparse representation | |
| CN109212481A (en) | A method of auditory localization is carried out using microphone array | |
| CN203148562U (en) | Transformer substation noise imaging monitoring device | |
| CN105607042A (en) | Method for locating sound source through microphone array time delay estimation | |
| CN103901400A (en) | Binaural sound source positioning method based on delay compensation and binaural coincidence | |
| Paulose et al. | Acoustic source localization | |
| Rathsam et al. | Analysis of absorption in situ with a spherical microphone array |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130904 |
|
| WD01 | Invention patent application deemed withdrawn after publication |