CN107743292B - A kind of failure automatic detection method of voicefrequency circuit - Google Patents
A kind of failure automatic detection method of voicefrequency circuit Download PDFInfo
- Publication number
- CN107743292B CN107743292B CN201711145812.5A CN201711145812A CN107743292B CN 107743292 B CN107743292 B CN 107743292B CN 201711145812 A CN201711145812 A CN 201711145812A CN 107743292 B CN107743292 B CN 107743292B
- Authority
- CN
- China
- Prior art keywords
- radian
- time series
- audio
- important
- sampled point
- 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.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 32
- 230000005236 sound signal Effects 0.000 claims abstract description 17
- 238000005070 sampling Methods 0.000 claims description 18
- 239000000523 sample Substances 0.000 claims description 11
- 230000006835 compression Effects 0.000 claims description 9
- 238000007906 compression Methods 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 230000015556 catabolic process Effects 0.000 claims description 3
- 239000013074 reference sample Substances 0.000 claims description 3
- 238000012360 testing method Methods 0.000 abstract description 4
- 238000000034 method Methods 0.000 description 10
- 238000013461 design Methods 0.000 description 5
- 230000005611 electricity Effects 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000033772 system development Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R29/00—Monitoring arrangements; Testing arrangements
- H04R29/004—Monitoring arrangements; Testing arrangements for microphones
-
- 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)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Health & Medical Sciences (AREA)
- Otolaryngology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Signal Processing For Digital Recording And Reproducing (AREA)
- Circuits Of Receivers In General (AREA)
Abstract
The invention belongs to airborne field of embedded software, are related to the failure automatic detection method of voicefrequency circuit, solve the test problems of airborne audio collection fault.The present invention compared to existing detection method have the advantages that quickly, facilitate transplanting, low cost and facilitate realization automate, specifically include the time series modeling of (1) audio signal；(2) radian representation of concept audio data waveform variation tendency is defined；(3) audio data is compressed using trend importance；(4) the trend deviation and amplitude deviation of audio signal are calculated；(5) define similarity with judge audio collection circuit whether failure.
Description
Technical field
The invention belongs to airborne embedded softwares, and in particular to the Acquisition Circuit fault detection side in audio collecting system
Method.
Background technique
Audio Data Acquisition System is the important component of avionics system, is responsible for being acquired cabin sound.Audio
Acquisition system is mainly based upon the realization of audio collection circuit, and can the failure of audio collection circuit whether is related to effectively complete
The acquisition of pairs of cabin sound, thus it is most important for the fault detection of audio collection circuit.
The existing fault detection method to audio collection circuit is mainly by the physics electricity to audio collection digital circuit
It is flat to be detected and be manually entered voice and whether observe output consistent with input.Wherein, physical level detection method is by setting
Special hardware circuit is counted, tests the height of output level, then by test result and by trouble-free audio data collecting electricity
Result when road compares, whether obtaining fault；Artificial detection method is defeated to audio data collecting circuit by manpower
Enter voice, then the output voice after conversion is observed in such a way that human ear is listened, according to the observation result audio data
Acquisition Circuit whether failure.
Physical level detection method needs to be integrated into audio collection circuit for the special circuit of voicefrequency circuit design is detected
On.Since detection circuit is more complicated, thus power consumption is higher, more to the resource occupation of system；Detection circuit is once integrated again
It finishes, if wanting to increase detection content or be improved to existing detection circuit, entire audio collecting system need to be set again
Meter, expends exploitation and maintenance cost is higher.
The method of artificial detection pass through when detecting artificial observation by audio collection circuit output whether with input one
It causes.Due to uncertainty of the human ear when working long hours, not can guarantee when there is a large amount of audio collection circuit to need to detect
The accuracy of detection；And under can not accomplishing that automatic detection, detection efficiency are relatively low due to artificial detection.
Summary of the invention
In order to solve the problems in background technique, the present invention proposes a kind of failure automatic detection method of voicefrequency circuit, should
Not only testing cost is low, detection accuracy is high and realizes the event of audio collection circuit fast and automatically changed for fault detection method
Barrier detection, while there is versatility for the audio collecting system of different field.
Basic realization principle of the invention is:
The present invention is by inputting reference audio signal to audio collection circuit, and by the digital audio and video signals of output and reference
Audio signal carries out similitude comparison, judges that can audio collection circuit keep the consistency of audio, and then judge voicefrequency circuit
With the presence or absence of failure.
The specific technical solution of the present invention is:
1) audio collection circuit is started to work；
2) digital audio time series and reference audio time series are obtained；
3) radian conversion is carried out to digital audio time series and reference audio time series respectively:
4) it repeats step 3) to execute n-1 times, respectively obtains the corresponding digital audio arc of corresponding digital audio time series
Spend time series S '1={ (r1,t1),(r2,t2),...,(rn-1,tn-1) and reference audio time series corresponding reference
Audio radian time series S '0={ (l1,t1),(l2,t2),...,(ln-1,tn-1), wherein n > 1；
Wherein riFor tiMoment corresponding radian；Wherein liFor tiMoment corresponding radian；
5) compression factor factor N, trend deviation threshold Θ are set, respectively to digital audio radian time series and with reference to sound
Frequency radian time series is compressed, and the important radian time series set S " of digital audio is obtained1={ (r1,t1),(r2,
t2),...,(rm,tm), the important sampled point serial number collection P of digital audio and the important radian time series set S " of reference audio0
={ (l1,t1),(l2,t2),...,(lk,tk), the important sampled point serial number collection Q of reference audio；
Wherein, m is that digital audio radian time series passes through compressed radian number, and k is the important radian of reference audio
Time series passes through compressed radian number；
6) minmal sequence number set Φ=P ∪ Q is calculated；
7) calculating the important radian time series of digital audio needs increased sampled point serial number set Φ-P；And reference audio
Important radian time series needs increased sampled point serial number set Φ-Q；
8) according to sampled point serial number set Φ-P and digital audio radian time series S ' need to be increased1, after obtaining expansion
The important radian time series Ω of digital audio1={ (r1,t1),(r2,t2),...,(rK,tK) and when digitized audio samples point
Between sequence Ψ1={ (y1,t1),(y2,t2),...,(yK,tK), wherein K=| Φ |, yKFor digitized audio samples point time series
The audio amplitude of middle k-th sampled point；
9) according to reference sample point serial number set Φ-Q and reference audio radian time series S ' need to be increased0, expanded
Fill the important radian time series Ω of rear reference audio0={ (l1,t1),(l2,t2),...,(lK,tK) and reference audio sampling
Point time series Ψ0={ (x1,t1),(x2,t2),...,(xK,tK), wherein K=| Φ |, xKFor the reference audio sampled point time
The audio amplitude of k-th sampled point in sequence；
10) the important radian time series of digital audio and digitized audio samples point time after the expansion obtained to step 8)
Sequence is normalized；
11) the important radian time series of reference audio and reference audio sampled point time after the expansion obtained to step 9)
Sequence is normalized；
12) according to the radian time series and sampling point sequence calculating sound after step 10) and step 11) normalized
Frequency sampling radian time series trend similarity, specific formula for calculation is:
Wherein Δ yi=yi+1-yi, Δ xi=xi+1-xi, Δ ti=ti+1-ti；
13) audio sample time series is calculated according to the sampling point sequence after step 10) and step 11) normalized
Amplitude deviation, specific formula for calculation are:
14) similarity is calculated, specific formula for calculation is:
δ=α × η+β × ε, wherein α and β meet: alpha+beta=1 and α, β are >=0；
15) breakdown judge；
Similarity threshold is set as δ0；0≤δ0≤1；
If δ >=δ0, then it is determined as dissmilarity；Export failure；Conversely, it is similar, then export fault-free.
The following are the preferred embodiments of above-mentioned technical proposal:
Above-mentioned steps 2) comprise the concrete steps that:
Reference audio signal input audio Acquisition Circuit is exported using T as the digital audio and video signals in period, and is denoted as
Digital audio time series, expression are:
S1={ (y1,t1),(y2,t2),...,(yn,tn),
Wherein, yiFor the amplitude of sampling instant digital audio and video signals, tnFor the sampled point time information of digital audio and video signals；
Reference audio signal is sampled by the period of T, obtains reference audio time series, expression is:
S0={ (x1,t1),(x2,t2),...,(xn,tn),
Wherein, xiFor the amplitude of sampling instant reference audio signal, tnFor the sampled point time information of reference audio signal.
Above-mentioned steps 3) be specifically:
For the radian rad of digital audio time series ith sample pointyiIt indicates are as follows:
For the radian rad of reference audio time series ith sample pointxiIt indicates are as follows:
Above-mentioned steps 5) comprise the concrete steps that:
The compression of digital audio radian time series:
A1: the trend deviation of digital audio radian time series is calculated；Specific formula for calculation is:
θy=| rj-ri|, meet 0 < j-i≤N, wherein i, j ∈ [1, n], computer capacity is [1, n-1]；
A2: the important radian time series set S " of the corresponding digital audio of digital audio radian time series is established1=
{(r1,t1),(r2,t2),...,(rm,tm) and the important sampled point serial number collection P of digital audio；
A radian is selected, according to the calculating of step A1) the radian trend deviation and judges whether to meet deviation threshold item
Part θy>=Θ, if satisfied, then record the radian while recording the sequence number that the radian corresponds to sampled point；If not satisfied, then carrying out
The calculating of next radian, until cambered all calculate of institute is completed, the finally obtained important radian time series of digital audio
Set S "1={ (r1,t1),(r2,t2),...,(rm,tm) and the important sampled point serial number collection P of digital audio；
The compression of reference audio radian time series:
B1: the trend deviation of reference audio radian time series is calculated；Specific formula for calculation is:
θx=| lj-li|, meet 0 < j-i≤N, wherein i, j ∈ [1, n], computer capacity is [1, n-1]；
B2: the important radian time series set S " of the corresponding reference audio of reference audio radian time series is established0=
{(l1,t1),(l2,t2),...,(lk,tk) and the important sampled point serial number collection Q of reference audio；
A radian is selected, according to the calculating of step B1) the radian trend deviation and judges whether to meet deviation threshold item
Part θx>=Θ, if satisfied, then record the radian while recording the sequence number that the radian corresponds to sampled point；If not satisfied, then carrying out
The calculating of next radian, until cambered all calculate of institute is completed, the finally obtained important radian time series of reference audio
Set S "0={ (l1,t1),(l2,t2),...,(lk,tk) and the important sampled point serial number collection Q of reference audio.
The present invention has the advantage that effect:
1, the method that uses of the present invention can quickly and effectively detect audio collection circuit with the presence or absence of failure, be convenient for and
When solve failure, ensure that the successful acquisition of audio data.
2, the portability with higher of the similarity detection method based on time series that the present invention uses, improves inspection
The versatility of survey method, and conveniently detection content is improved, the cost of system development is reduced, automation is easy to implement
Detection.
3, the method that the importance measures based on trend difference used in the present invention screen sampled audio signal point,
Guarantee that required sampled data can be efficiently reduced in the case where detection accuracy, reduces the consumption to system resource, mention
The high efficiency of detection.
Detailed description of the invention
Fig. 1 is workflow of the invention.
Specific embodiment
Basic realization principle of the invention:
The sampled data for carrying out similarity system design is screened by introducing radian concept, selection can reflect that audio becomes
The crucial sampled point of change.First by the way that audio data time series is converted to radian time series, calculated according to sampling instant
The trend deviation of radian section calculates trend deviation according to the trend deviation threshold of setting one by one, when trend deviation is more than or equal to threshold
When value, extracts the sampled point and the corresponding radian time is key point, realize the compression for being compared data to radian sequence,
Required data volume is calculated so that reducing in the case where keeping certain precision, improves computational efficiency.The present invention is expected by setting
Compressibility factor parameter limits maximum compression ratio.By the key point that will be formed after screening, corresponding weight is formed
Want radian time series and important sampled point time series.And then the similarity system design of sampled signal will be converted to corresponding
The similarity system design of important radian time series and important sampling point sequence.In order to complete similarity system design, need two arcs
Degree time series and important sampled point time series are extended for the time series of equal amount section.The present invention passes through to two radians
The corresponding sampling point moment for including in time series asks minimum comprising collection, i.e. original reference signals and after Acquisition Circuit samples
The important sampled point time series formed seeks union, based on and the sampled point concentrated complete to two radian time serieses and
The expansion of important sampled point time series.After the time series after all expansions is normalized according to amplitude again, meter
Calculate corresponding trend difference value and amplitude difference value.It is fixed by introducing respective weights coefficient to trend difference and amplitude difference
Justice similarity, it is according to the threshold decision of setting whether similar by calculating the similarity of time series, from which further follow that audio
Data acquisition circuit whether there is failure.
This method the specific implementation process is as follows:
The failure of step 1) audio collection circuit detects automatically to be started；
Step 2) reference audio signal input audio Acquisition Circuit is exported using T as the digital audio and video signals in period, is expressed as
Corresponding digital audio time series S1={ (y1,t1),(y2,t2),...,(yn,tn), wherein yiFor sampling instant audio signal
Amplitude, tiInformation at the time of for sampled point；N > 1；
Step 3) samples reference audio signal by the period of T, obtains reference audio time series S0={ (x1,
t1),(x2,t2),...,(xn,tn), wherein xiFor the amplitude of sampling instant audio signal, tiInformation at the time of for sampled point.
Step 4) carries out radian conversion to digital audio time series and reference audio time series respectively:
For the radian rad of digital audio time series ith sample pointyiIt indicates are as follows:
For the radian rad of reference audio time series ith sample pointxiIt indicates are as follows:
Step 5) is to digital audio time series S1It is executed n-1 times according to step 4), obtains the corresponding digital radian time
Sequence S '1={ (r1,t1),(r2,t2),...,(rn-1,tn-1), wherein riFor tiMoment corresponding radian；
Step 6) is to reference audio time series S0It is executed n-1 times according to step 4), obtains reference audio radian time sequence
Arrange S '0={ (l1,t1),(l2,t2),...,(ln-1,tn-1), wherein liFor tiMoment corresponding radian；
Compression factor factor N, trend deviation threshold Θ is arranged in step 7).
The trend deviation θ of step 8) calculating radian time series=| rj-ri|, meet 0 < j-i≤N,
Wherein i, j ∈ [1, n], computer capacity are [1, n-1].
Step 9) according to trend deviation formula calculate one by one trend deviation and judging whether meet deviation threshold condition θ >=
Θ if satisfied, the point is then increased to important radian time series set, and records the sequence number of corresponding sampled point；If discontented
Foot, then calculate next radian, and calculating process meets step 8) requirement；
Step 10) is to digital audio radian time series S '1Circulation executes step 8) and step 9), obtains corresponding number
The important radian time series S " of audio1={ (r1,t1),(r2,t2),...,(rm,tm) and the important sampled point sequence of digital audio
Number collection P；Wherein, m is that digital audio radian time series passes through compressed radian number；
Step 11) is to reference audio radian time series S '1Circulation executes step 8) and step 9), is referred to accordingly
The important radian time series S " of audio0={ (l1,t1),(l2,t2),...,(lk,tk) and the important sampled point sequence of reference audio
Number collection Q；Wherein, k is that the important radian time series of reference audio passes through compressed radian number；
Step 12) calculates minmal sequence number set Φ=P ∪ Q；
Step 13), which calculates the important radian time series of digital audio, needs increased sampled point serial number set Φ-P and reference
The important radian time series of audio needs increased sampled point serial number set Φ-Q；
Step 14) is according to need to increase sampled point serial number set Φ-P and digital audio radian time series S '1, expanded
Fill the important radian time series Ω of rear digital audio1={ (r1,t1),(r2,t2),...,(rK,tK) and digitized audio samples
Point time series Ψ1={ (y1,t1),(y2,t2),...,(yK,tK), wherein K=| Φ |, yKFor the digitized audio samples point time
The audio amplitude of k-th sampled point in sequence；
Step 15) is according to need to increase reference sample point serial number set Φ-Q and reference audio radian time series S '0, obtain
The important radian time series Ω of reference audio after to expansion0={ (l1,t1),(l2,t2),...,(lK,tK) and reference audio
Sampled point time series Ψ0={ (x1,t1),(x2,t2),...,(xK,tK), wherein K=| Φ |, xKFor reference audio sampled point
The audio amplitude of k-th sampled point in time series；
The important radian time series of digital audio and digitized audio samples after the expansion that step 16) obtains step 14)
Point time series is normalized；
The important radian time series of reference audio and reference audio sampling after the expansion that step 17) obtains step 15)
Point time series is normalized；
Step 18) is according to the radian time series and sampling point sequence meter after step 16) and step 17) normalized
Audio sample radian time series trend similarity is calculated, specific formula for calculation is:
Wherein Δ yi=yi+1-yi, Δ xi=xi+1-xi, Δ ti=ti+1-ti；
Step 19) calculates the audio sample time according to the sampling point sequence after step 16) and step 17) normalized
Sequence amplitude deviation, specific formula for calculation is:
Step 20) calculates similarity, and specific formula for calculation is:
δ=α × η+β × ε, wherein α and β meet: alpha+beta=1 and α, β are >=0；
Step 21) breakdown judge；
Similarity threshold is set as δ0；0≤δ0≤1；
If δ >=δ0, then it is determined as dissmilarity；Export failure；Conversely, it is similar, then export fault-free.
Claims (4)
1. a kind of failure automatic detection method of voicefrequency circuit, which comprises the following steps:
1) audio collection circuit is started to work；
2) digital audio time series and reference audio time series are obtained；
3) radian conversion is carried out to digital audio time series and reference audio time series respectively；
4) it repeats step 3) to execute n-1 times, when respectively obtaining the corresponding digital audio radian of corresponding digital audio time series
Between sequence S '1={ (r1,t1),(r2,t2),...,(rn-1,tn-1) and reference audio time series corresponding reference audio
Radian time series S '0={ (l1,t1),(l2,t2),...,(ln-1,tn-1), wherein n > 1；
Wherein riFor tiMoment corresponding radian；Wherein liFor tiMoment corresponding radian；
5) compression factor factor N, trend deviation threshold Θ, respectively to digital audio radian time series and reference audio arc are set
Degree time series is compressed, and the important radian time series set S " of digital audio is obtained1={ (r1,t1),(r2,t2),...,
(rm,tm), the important sampled point serial number collection P of digital audio and the important radian time series set S " of reference audio0={ (l1,
t1),(l2,t2),...,(lk,tk), the important sampled point serial number collection Q of reference audio；
Wherein, m is that digital audio radian time series passes through compressed radian number, and k is the reference audio important radian time
Sequence passes through compressed radian number；
6) minmal sequence number set Φ=P ∪ Q is calculated；
7) calculating the important radian time series of digital audio needs increased sampled point serial number set Φ-P；And reference audio is important
Radian time series needs increased sampled point serial number set Φ-Q；
8) according to sampled point serial number set Φ-P and digital audio radian time series S ' need to be increased1, digital sound after being expanded
Frequently important radian time series Ω1={ (r1,t1),(r2,t2),...,(rK,tK) and digitized audio samples point time series
Ψ1={ (y1,t1),(y2,t2),...,(yK,tK), wherein K=| Φ |, yKFor k-th in digitized audio samples point time series
The audio amplitude of sampled point；
9) according to reference sample point serial number set Φ-Q and reference audio radian time series S ' need to be increased0, join after being expanded
Examine the important radian time series Ω of audio0={ (l1,t1),(l2,t2),...,(lK,tK) and the reference audio sampled point time
Sequence Ψ0={ (x1,t1),(x2,t2),...,(xK,tK), wherein K=| Φ |, xKFor in reference audio sampled point time series
The audio amplitude of k-th sampled point；
10) the important radian time series of digital audio and digitized audio samples point time series after the expansion obtained to step 8)
It is normalized；
11) the important radian time series of reference audio and reference audio sampled point time series after the expansion obtained to step 9)
It is normalized；
12) audio is calculated with sampling point sequence according to the radian time series after step 10) and step 11) normalized to adopt
Sample radian time series trend similarity, specific formula for calculation is:
Wherein Δ yi=yi+1-yi, Δ xi=xi+1-xi, Δ ti=ti+1-ti；
13) audio sample time series amplitude is calculated according to the sampling point sequence after step 10) and step 11) normalized
Deviation, specific formula for calculation are:
14) similarity is calculated, specific formula for calculation is:
δ=α × η+β × ε, wherein α and β meet: alpha+beta=1 and α, β are >=0；
15) breakdown judge；
Similarity threshold is set as δ0；0≤δ0≤1；
If δ >=δ0, then it is determined as dissmilarity；Export failure；Conversely, it is similar, then export fault-free.
2. the failure automatic detection method of voicefrequency circuit according to claim 1, which is characterized in that the tool of the step 2)
Body step is:
2.1) reference audio signal input audio Acquisition Circuit is exported using T as the digital audio and video signals in period, and is denoted as
Digital audio time series, expression are:
S1={ (y1,t1),(y2,t2),...,(yn,tn),
Wherein, yiFor the amplitude of sampling instant digital audio and video signals, tnFor the sampled point time information of digital audio and video signals；
2.2) reference audio signal is sampled by the period of T, obtains reference audio time series, expression is:
S0={ (x1,t1),(x2,t2),...,(xn,tn),
Wherein, xiFor the amplitude of sampling instant reference audio signal, tnFor the sampled point time information of reference audio signal.
3. the failure automatic detection method of voicefrequency circuit according to claim 1, which is characterized in that the tool of the step 3)
Body step is:
For the radian rad of digital audio time series ith sample pointyiIt indicates are as follows:
For the radian rad of reference audio time series ith sample pointxiIt indicates are as follows:
4. the failure automatic detection method of voicefrequency circuit according to claim 1, which is characterized in that the tool of the step 5)
Body step is:
The compression of digital audio radian time series:
A1: the trend deviation of digital audio radian time series is calculated；Specific formula for calculation is:
θy=| rj-ri|, meet 0 < j-i≤N, wherein i, j ∈ [1, n], computer capacity is [1, n-1]；
A2: the important radian time series set S " of the corresponding digital audio of digital audio radian time series is established1={ (r1,
t1),(r2,t2),...,(rm,tm) and the important sampled point serial number collection P of digital audio；
A radian is selected, according to the calculating of step A1) the radian trend deviation and judges whether to meet deviation threshold condition θy≥
Θ, if satisfied, then record the radian while recording the sequence number that the radian corresponds to sampled point；If not satisfied, then carrying out next
The calculating of radian, until cambered all calculate of institute is completed, the finally obtained important radian time series set S " of digital audio1
={ (r1,t1),(r2,t2),...,(rm,tm) and the important sampled point serial number collection P of digital audio；
The compression of reference audio radian time series:
B1: the trend deviation of reference audio radian time series is calculated；Specific formula for calculation is:
θx=| lj-li|, meet 0 < j-i≤N, wherein i, j ∈ [1, n], computer capacity is [1, n-1]；
B2: the important radian time series set S " of the corresponding reference audio of reference audio radian time series is established0={ (l1,
t1),(l2,t2),...,(lk,tk) and the important sampled point serial number collection Q of reference audio；
A radian is selected, according to the calculating of step B1) the radian trend deviation and judges whether to meet deviation threshold condition θx≥
Θ, if satisfied, then record the radian while recording the sequence number that the radian corresponds to sampled point；If not satisfied, then carrying out next
The calculating of radian, until cambered all calculate of institute is completed, the finally obtained important radian time series set S " of reference audio0
={ (l1,t1),(l2,t2),...,(lk,tk) and the important sampled point serial number collection Q of reference audio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711145812.5A CN107743292B (en) | 2017-11-17 | 2017-11-17 | A kind of failure automatic detection method of voicefrequency circuit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711145812.5A CN107743292B (en) | 2017-11-17 | 2017-11-17 | A kind of failure automatic detection method of voicefrequency circuit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107743292A CN107743292A (en) | 2018-02-27 |
| CN107743292B true CN107743292B (en) | 2019-09-10 |
Family
ID=61234766
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711145812.5A Active CN107743292B (en) | 2017-11-17 | 2017-11-17 | A kind of failure automatic detection method of voicefrequency circuit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107743292B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110636428A (en) * | 2018-06-21 | 2019-12-31 | 杭州海康威视数字技术股份有限公司 | Audio fault detection method and system |
| CN112689872B (en) * | 2018-11-21 | 2022-11-15 | 深圳市欢太科技有限公司 | Audio detection method, computer-readable storage medium and electronic device |
| EP3694230A1 (en) * | 2019-02-08 | 2020-08-12 | Ningbo Geely Automobile Research & Development Co. Ltd. | Audio diagnostics in a vehicle |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2116999A1 (en) * | 2007-09-11 | 2009-11-11 | Panasonic Corporation | Sound judging device, sound sensing device, and sound judging method |
| CN101577605A (en) * | 2008-05-08 | 2009-11-11 | 吴志军 | Speech LPC hiding and extraction algorithm based on filter similarity |
| CN102411301A (en) * | 2011-09-05 | 2012-04-11 | 广东电网公司电力科学研究院 | Controlled object frequency domain phase characteristic identification method and apparatus thereof |
| CN102779509A (en) * | 2011-05-11 | 2012-11-14 | 联想(北京)有限公司 | Voice processing equipment and voice processing method |
| CN104133851A (en) * | 2014-07-07 | 2014-11-05 | 小米科技有限责任公司 | Audio similarity detecting method, audio similarity detecting device and electronic equipment |
-
2017
- 2017-11-17 CN CN201711145812.5A patent/CN107743292B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2116999A1 (en) * | 2007-09-11 | 2009-11-11 | Panasonic Corporation | Sound judging device, sound sensing device, and sound judging method |
| CN101577605A (en) * | 2008-05-08 | 2009-11-11 | 吴志军 | Speech LPC hiding and extraction algorithm based on filter similarity |
| CN102779509A (en) * | 2011-05-11 | 2012-11-14 | 联想(北京)有限公司 | Voice processing equipment and voice processing method |
| CN102411301A (en) * | 2011-09-05 | 2012-04-11 | 广东电网公司电力科学研究院 | Controlled object frequency domain phase characteristic identification method and apparatus thereof |
| CN104133851A (en) * | 2014-07-07 | 2014-11-05 | 小米科技有限责任公司 | Audio similarity detecting method, audio similarity detecting device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107743292A (en) | 2018-02-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108073153B (en) | Sensor interface apparatus, metrical information communication system and method, storage medium | |
| Cochran et al. | Earthquake early warning ShakeAlert system: Testing and certification platform | |
| CN107743292B (en) | A kind of failure automatic detection method of voicefrequency circuit | |
| CN103308846B (en) | Method and device for detecting functional performance of integrated chip based on model identification | |
| CN104517606A (en) | Method and device for recognizing and testing speech | |
| CN109341848B (en) | Safety monitoring system in tunnel operation stage | |
| CN110261746B (en) | Cable defect detection method based on periodic attenuation characteristics of oscillating wave voltage | |
| CN102006198B (en) | Method and device for acquiring network failure association rule | |
| CN106197650A (en) | A kind of method judging that audio signal is noiseless | |
| CN111678992A (en) | Nondestructive testing method for identifying damage type of concrete structure | |
| CN112668612A (en) | Partial discharge signal clustering analysis method based on grids | |
| CN108470163B (en) | Rail turnout plate gap defect detection method and terminal equipment | |
| CN104483016B (en) | Complete acquisition method and system of vibrating signals | |
| CN112600659B (en) | Method and system for detecting side channel leakage of safety chip based on neural network | |
| CN102901611B (en) | A kind of quick arrange distinguish method of satellite antenna structural damage based on modal test | |
| CN102004177B (en) | Oscilloscope and method for identifying time sequence of universal serial bus signals by using same | |
| KR101646981B1 (en) | Data processing system for evaluating safety of structure, and method for the same | |
| CN111767546A (en) | Deep learning-based input structure inference method and device | |
| CN109100243B (en) | Detection method for reinforced direct shear test | |
| CN107271204B (en) | Non- constant amplitude thresholding data compression method and apparatus | |
| CN114001892B (en) | Time synchronization method between free jet test measurement systems | |
| CN108846106A (en) | A kind of method and apparatus for judging to whether there is identical audio in multiple audios | |
| KR940004332A (en) | Circuit test method and delay defect detector | |
| CN102692459A (en) | Method and apparatus for data compression | |
| CN111737096B (en) | Method and device for detecting integral grading of TPC-H pre-test system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |