CN106764468B - A kind of leakage early warning system and adaptive spectrum noise-eliminating method - Google Patents
A kind of leakage early warning system and adaptive spectrum noise-eliminating method Download PDFInfo
- Publication number
- CN106764468B CN106764468B CN201710006221.3A CN201710006221A CN106764468B CN 106764468 B CN106764468 B CN 106764468B CN 201710006221 A CN201710006221 A CN 201710006221A CN 106764468 B CN106764468 B CN 106764468B
- Authority
- CN
- China
- Prior art keywords
- noise
- spectrum
- unit
- acquisition
- subelement
- 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
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
- F17D—PIPE-LINE SYSTEMS; PIPE-LINES
- F17D5/00—Protection or supervision of installations
- F17D5/02—Preventing, monitoring, or locating loss
- F17D5/06—Preventing, monitoring, or locating loss using electric or acoustic means
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Examining Or Testing Airtightness (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a kind of leakage early warning system and adaptive spectrum noise-eliminating methods, it include: power supply unit, acquisition unit, plus and blowup unit, processor unit and GPRS transmission unit, processor unit includes: to obtain the ADC subelement of discrete sampling data to noise information sample conversion；FFT (fast Fourier) transformation is carried out to each acquisition noise sample, obtains noise frequency domain information, and calculate the FFT transform subelement of amplitude spectrum；Environmental Noise Influence is eliminated by adaptive spectrum denoising algorithm in frequency domain, retains the self-adapted noise elimination subelement of metastable leakage noise information；Frequency spectrum correlation analysis subelement；Early warning analysis subelement and GPRS transmission unit.When alerting generation, transmitting warning information to server or designated mobile phone.It is poor that the present invention can solve current leakage prior-warning device interference free performance, can only cannot monitor for 24 hours in night use, and there are the universal technical problems such as higher of rate of false alarm and rate of failing to report.
Description
Technical field
The present invention relates to Signal and Information Processing technical fields, and in particular, to a kind of leakage method for early warning and equipment.
Background technique
Leaking source of early warning is to be mainly used in water supply network, is typically placed on well interior conduit, passes through seismoelectric sensor
Duct noise is picked up, by analyzing the intensity and spectrum distribution feature of noise, judges whether leakage occurs, and in time believe early warning
Breath is sent to monitoring center or chief leading cadre's mobile phone.
When pipe leakage, fluid media (medium) high speed passes through leakage gap, due to vibration, friction, deceleration, expansion, shock etc.,
Fluid generates eddy stress or shearing force, forms leakage noise.Sound wave is leaked with the high-order sound of the plane wave of non-frequency dispersion and frequency dispersion
Mode form is propagated in pipeline fluid, is influenced by fluid and pipeline attenuation characteristic, and as propagation distance increases, noise intensity is fast
Speed weakens.Using this feature of noise is generated when pipe leakage, have now been developed that a variety of pipeline leakage detection/monitorings are set both at home and abroad
Standby (such as: leakage measuring instrument by sonic, correlator, leakage early warning etc.), noise detecting method is also water supply network currently main application method.
Since leakage noise is extremely faint, vulnerable to ambient noise interference, how correctly to identify leakage noise, need to leakage
The feature of noise and ambient noise respectively is analyzed.Theoretical and actual test proves: pipeline environment noise belongs to Gauss white noise
Sound, spectral power distribution meet Gaussian Profile；Versus environmental noise, any point in leakage noise transmission approach, leaks noise
With more stable noise intensity and spectrum distribution feature.Wherein noise intensity feature decision method is also that current external mainstream is seeped
The method that leakage early warning product (such as: Britain is bold and generous) generallys use.
Using noise intensity feature decision method, acquisition noise data during night 2:00~4:00 are typically chosen in, it is main
Syllabus is to reduce wrong report and miss probability to reduce ambient noise interference.When acquisition, according to fixed intervals T seconds (general 5
~10 seconds, different manufacturers were slightly different) acquisition noise data and noise intensity is calculated, acquisition total degree n times (General N >=1000
It is secondary), for the ease of analysis, the collected noise intensity discrete series of n times are defined as F (t), while defining strength range threshold value
(general setting 20db or so, different manufacturers omit by Vth (general setting 5db or so, different manufacturers are slightly different) and excision value Vcut
It is different), leak determination method:
1, when (F (t) >=Vcut and when MAX (F (t))-MIN (F (t))≤Vth has leakage to MIN；
2, when (F (t)<Vcut or when MAX (F (t))-MIN (F (t))>Vth, ne-leakage occurs MIN；
Wherein MAX () takes sequence maximum value, and MIN () takes sequence minimum value.This method is mainly according to leakage noise intensity
It is relatively stable, and the stronger feature of ambient noise randomness determines whether leakage occurs.
Although leakage strength characteristic distinguished number has the advantages that algorithm is simple, hardware cost is low etc., interference free performance is poor,
Can only be in night use, the deficiencies of there are rate of false alarm and generally higher rates of failing to report.
Summary of the invention
The object of the present invention is to provide a kind of leakage early warning system and adaptive spectrum noise-eliminating methods, to solve current leakage
Prior-warning device interference free performance is poor, can only cannot monitor within 24 hours, there are rate of false alarm and rate of failing to report are generally higher in night use
Etc. technical problems.
For achieving the above object, the technical solution adopted in the present invention is as follows:
A kind of leakage early warning system, comprising:
1) power supply unit provides electric power safeguard for equipment；
2) acquisition unit, for picking up the leakage noise sample of pipeline weak；
3) plus and blowup unit is used for collected faint duct noise plus and blowup, and send to processor
It sets ADC and carries out data acquisition；
4) processor unit: it is responsible for Data Management Analysis and early warning output；The unit includes following several subelements:
A) to noise information sample conversion, discrete sampling data ADC subelement: are obtained；
B) FFT transform subelement: FFT transform is carried out to acquisition noise sample, obtains noise frequency domain information, and calculate amplitude
Spectrum；
C) self-adapted noise elimination subelement: primary complete collection process, i.e. acquisition interval+acquisition duration are defined, is adopted for one
Collect the period, continuous m collection period is defined as a de-noising period, defines a Variable Learning for each frequency point, study becomes
The position for indicating current frequency point regulation coefficient is measured, FFT transform is executed to each acquisition data, and obtain input range spectrum,
Definition study amplitude spectrum, in each collection period, to input range spectrum, frequency point is compared one by one with study amplitude spectrum, to retain
Metastable leakage noise information；Include:
Comparing unit compares for composing to input range and learning amplitude spectrum, obtains spectral change state；
Unit tracks spectral change state for learning to the Variable Learning；
Adjustment unit obtains de-noising amplitude spectrum for being adjusted to the study amplitude spectrum；
D) frequency spectrum correlation analysis subelement: data carry out frequency spectrum related operation by frequency spectrum correlation analysis algorithm after de-noising,
Obtain noise correlation coefficients value；Include:
Receiving unit, for receiving the amplitude spectrum；
Arithmetic element: for carrying out related operation to the amplitude spectrum received and previous de-noising periodic amplitude spectrum, phase is obtained
Coefficient values；
Storage unit: for storing the amplitude spectrum received；
E) early warning analysis subelement: comprehensive statistics analysis is carried out to correlated results, whether confirmation leakage occurs；Include:
Receiving unit, for receiving relevance degree；
Statistic unit: for for statistical analysis to the relevance degree received；
Judging unit: for judging whether leakage occurs according to statistical result；
5) GPRS transmission unit: when alerting generation, transmitting warning information to server or designated mobile phone.
A kind of adaptive spectrum noise-eliminating method of novel leakage early warning system, the system include: power supply unit, acquisition list
Member, plus and blowup unit, processor unit and GPRS transmission unit；The processor unit includes: to sample to noise information
Conversion obtains the ADC subelement of discrete sampling data；FFT transform is carried out to acquisition noise sample, obtains noise frequency domain information,
And calculate the FFT transform subelement of amplitude spectrum；Environmental Noise Influence is eliminated by adaptive spectrum Denoising Method in frequency domain, retains phase
To the self-adapted noise elimination subelement of stable leakage noise information；Data carry out frequency spectrum by frequency spectrum correlation analysis algorithm after de-noising
Related operation obtains the frequency spectrum correlation analysis subelement of noise correlation coefficients value；Comprehensive statistics analysis is carried out to correlated results, really
Recognize the early warning analysis subelement whether leakage occurs；Wherein self-adapted noise elimination is suddenly as follows:
Acquisition interval is TS twice in succession for definition；Define a length of TL when single acquisition；Primary complete collection process is defined, i.e.,
Acquisition interval+acquisition duration is a collection period TP, then TP=TS+TL；
Continuous m collection period is defined as a de-noising cycle T F；It defines regulation coefficient array G [n], wherein n=2*m
+ 1, enable G [n]=- 1/2, -1/4 ... ..-1/2m, 0,1/2m ... ..1/4,1/2；
A Variable Learning g is defined for each frequency point, and 0≤g < n, Variable Learning g are for indicating current frequency point adjustment system
Several positions corresponds to the subscript of G [n] array, and Variable Learning g is initialized as m；
FFT transform is executed to each acquisition data, and obtaining input range spectrum is Vi；Definition study amplitude spectrum Vb；Every
To input range spectrum Vi, frequency point is compared a collection period one by one with study amplitude spectrum Vb, and is adjusted as follows:
First by frequency point regularized learning algorithm variable:
Then, by frequency point regularized learning algorithm amplitude spectrum: Vb (n)=Vb (n) * (1+G [g (n)])；
The present invention eliminates the influence of ambient noise using adaptive spectrum noise cancellation apparatus；It is comprehensive using frequency spectrum related algorithm device
Assessment noise intensity and spectrum distribution feature are closed, so the present invention has preferable interference free performance, extremely low wrong report and fails to report
The features such as rate is suitble to continue to monitor for 24 hours, meets leakage early warning requirement of real-time.
Detailed description of the invention
Fig. 1 is inventive algorithm flow diagram；
Fig. 2 is the time diagrams such as collection period of the invention, de-noising period；
Fig. 3 is frequency spectrum correlation timing schematic diagram of the invention；
Fig. 4 is adaptive spectrum denoising algorithm schematic diagram of the present invention；
Fig. 5 is the example adaptive spectrum denoising algorithm flow chart of embodiments of the present invention；
Fig. 6 is the example frequency spectrum related algorithm flow chart of embodiments of the present invention；
Fig. 7 is the example early warning analysis algorithm flow chart of embodiments of the present invention.
Fig. 8 is the circuit theory of constitution block diagram of invention.
Fig. 9 is algorithm process flow diagram of the invention；
Figure 10 is interrupt processing flow chart of the invention；
Figure 11 is the example adaptive spectrum denoising algorithm flow chart of embodiments of the present invention；
Figure 12 is the example frequency spectrum related algorithm flow chart of embodiments of the present invention；
Figure 13 is the example early warning analysis algorithm flow chart of embodiments of the present invention.
Specific embodiment
Below in conjunction with attached drawing, detailed description of the preferred embodiments.It should be understood that this place is retouched
The specific embodiment stated is merely to illustrate and explain the present invention, and is not intended to restrict the invention.
It is influenced by aqueous fluid and pipeline attenuation characteristic, leakage noise will be much higher than low in communication process high frequency attenuation speed
The frequency rate of decay, therefore leakage noise frequency can be acquired and be mainly distributed within the scope of 50~2Khz, according to sampling thheorem, sampling frequency
Rate should be higher than that 4Khz, generally take 8Khz sample frequency in practical application.
According to leakage noise behavior: having metastable noise intensity and spectrum distribution feature.This algorithm synthesis foundation
Noise intensity and spectrum distribution feature, are handled and are analyzed from frequency domain to noise data, and algorithm process process is as shown in Figure 1；
Each acquisition noise data are converted through FFT (fast Fourier) first, obtain noise frequency domain information；By " adaptive
Answer frequency spectrum de-noising " algorithm elimination Environmental Noise Influence, retain metastable leakage noise information；Data are by " frequency spectrum after de-noising
Correlation analysis " algorithm is carrying out frequency spectrum correlation operation according to the de-noising period, obtains degree of relevancy on noise temporal；Correlation knot
Fruit is for statistical analysis by " early warning analysis " module, is confirmed whether leakage.
The algorithm course of work for ease of description, acquisition interval is TS twice in succession for definition, is defined a length of when single acquisition
TL, defining primary complete collection process (acquisition interval+acquisition duration) is a collection period TP, then TP=TS+TL.It will be continuous
M collection period is defined as a de-noising cycle T F, such as Fig. 2 (collection period of the invention, de-noising period time diagram,
Note: m=5 in figure) shown in:
Each collection period TP acquires a noise data, acquires duration TL, and acquisition data through FFT operation, obtain first
Input range spectrum, then by " adaptive spectrum de-noising " algorithm process to eliminate environmental disturbances noise；It is every to pass through a de-noising week
Current de-noising amplitude spectrum and previous de-noising periodic amplitude spectrum are executed " frequency spectrum correlation analysis " operation, obtain noise phase by phase TF
Relationship number r (i).As shown in Fig. 3 (frequency spectrum correlation timing schematic diagram of the invention, note: m=5 in figure).
Calculated correlation coefficient r (i) carries out statistical disposition by " early warning analysis " module, judges whether leakage occurs.
(note: actual test shows: when TS >=5 second；TL >=1 second；When m >=5, performance is best.)
In order to better illustrate the present invention, various pieces are described in detail below:
Adaptive spectrum de-noising part:
Theory and practice proves: pipeline environment noise belongs to white Gaussian noise, and spectral power distribution meets Gaussian Profile；And
Leaking noise has more stable noise intensity and more stable spectrum distribution.For the spy of above-mentioned ambient noise and leakage noise
Point designs the self-adapted noise elimination model as shown in Fig. 4 (adaptive spectrum denoising algorithm schematic diagram).
In order to illustrate the working principle of self-adapted noise elimination, regulation coefficient array G [n] is defined first, wherein n=2*m+1, enable
G [n]=- 1/2, -1/4 ... ..-1/2m, 0,1/2m ... ..1/4,1/2 }；If setting m=5, i.e. 5 collection period definition
For a filtering cycle, then regulation coefficient array G (n)={ -1/2, -1/4, -1/8, -1/16, -1/32,0,1/32,1/16,1/
8,1/4,1/2 }.A Variable Learning g is defined for each frequency point, and 0≤g < n, Variable Learning g are for indicating current frequency point adjustment
The position of coefficient corresponds to the subscript of G [n] array, and when initial acquisition, Variable Learning g is set as m.Each acquisition data are held
Row FFT transform, and obtaining input range spectrum is Vi；Definition study amplitude spectrum Vb, study amplitude spectrum Vb are for tracking study input
Noise variation eliminates Environmental Noise Influence, the noise after finally obtaining de-noising according to the strong feature of ambient noise time randomness
Amplitude spectrum.
When work, acquisition data, which are converted and calculated through FFT (fast Fourier), obtains input range spectrum Vi, by input range
Frequency point is compared spectrum Vi one by one with study amplitude spectrum Vb, and operation in the following manner:
First by frequency point regularized learning algorithm variable:
Then, by frequency point regularized learning algorithm amplitude spectrum: Vb (n)=Vb (n) * (1+G [g (n)])；
Fig. 5 is the example adaptive spectrum denoising algorithm flow chart of embodiment according to the present invention, first to acquisition number
According to execution FFT transform, and calculates and obtain input range spectrum Vi；Then compare input range spectrum Vi and study amplitude spectrum by frequency point
Vb obtains spectral change state；According to the spectral change state of acquisition, regularized learning algorithm variable g and study amplitude spectrum Vb；Work as completion
Study amplitude spectrum Vb is passed to next stage and handled by one de-noising period.
Frequency spectrum correlation analysis part:
We select simple correlation coefficient algorithm, also known as Pearson correlation coefficient or " Pearson came product in frequency spectrum correlation analysis
Square related coefficient ", it describes the tightness degree contacted between two spacing variables, is a kind of linearly dependent coefficient.Pearson's phase
Relationship number is the statistic for reflecting two linear variable displacement degrees of correlation, and related coefficient is indicated with r.
After a complete de-noising period, amplitude modal data by adaptive spectrum de-noising, with previous de-noising week
The amplitude modal data of phase obtains the correlation degree of different time leakage noise according to Pearson correlation coefficient formula deployment analysis.
Operational formula:
Such method, by amplitude modal data directly as the input data deployment analysis of correlation analysis, correlated results is comprehensive
Embody leakage noise intensity and spectrum distribution feature.Noise intensity method is relied on relative to simple, there is better accuracy
And objectivity, advantageously reduce leakage misinformation probability.
The frequency spectrum correlation coefficient r of acquisition, value range are [- 1,1], and r>0 indicates to be positively correlated, and r<0 indicates negatively correlated, | r |
Illustrate the height of degree of correlation between variable.Distinguishingly, r=1 is known as perfect positive correlation, and r=-1 is known as perfect negative correlation, r
=0 is referred to as uncorrelated.Theory and practice confirmation, when | r | when being more than or equal to 0.8, both thought that two variables had very strong linear phase
Guan Xing.
Fig. 6 is the example frequency spectrum related algorithm flow chart of embodiment according to the present invention, first to current study amplitude
The spectrum Vb and amplitude spectrum VPREV in previous de-noising period does correlation analysis, obtains correlation coefficient r value；Vb is saved in VPREV later
In；The correlation coefficient r value of calculating is finally passed to early warning analysis to handle.
Early warning analysis principle:
Through frequency spectrum correlation analysis, the correlation coefficient r value of acquisition, the signal for completing certain time by early warning analysis algorithm is related
Degree statistics is handled with early warning, is shown according to the practical application of current leakage early warning and actual test: when continuous time >=60 minute
When the r value of all acquisitions >=0.8, pipe leakage occurs；When the conditions set forth above are not met, pipeline ne-leakage.Using this early warning
Judgment mode relative to tradition only night acquisition and alarm mode, can accomplish 24 hours monitor, possess better real-time and
Practicability.
Fig. 7 is the example early warning analysis algorithm flow chart of embodiment according to the present invention, the phase obtained to correlation analysis
Relationship number r value is counted, if r >=0.8, serial correlation degree variables TC=TC+1；Otherwise TC=0；Judge continuous phase
Close whether degree variables TC is more than early warning number (such as 60 times, early warning number=pre-warning time/TF), if it exceeds early warning time
Number, then early warning exports.
Hardware components
With the practical application of leakage source of early warning and universal, under the premise of assurance function and perfect performance, more to setting
Standby power consumption and cost propose that high requirement, each restraining factors of choosing comprehensively equipment source of early warning propose design as shown in Figure 8
Structure.
Power supply unit: comprehensive assessment equipment power dissipation and equipment volume, using two section ER14505 battery parallel power supplies.
Acquisition unit: using the most common piezoelectric ceramic process sensor of water supply industry, and piezoelectric ceramics has extremely sensitive
Characteristic, extremely weak mechanical oscillation can be converted into electric signal, can be used for sonar system, meteorological detection, telemetering environment protect
Shield, household electrical appliance etc..Piezoelectric ceramics, which makes it even to the sensitivity of external force, can sense that more than ten meters of outer winged insects pat wing to sky
The disturbance of gas makes piezoelectricity seismic detector of it, can accurately measure earthquake intensity, indicates the azimuth-range of earthquake.
Plus and blowup unit: using three-level low noise, consumption high gain amplifier, and amplifier selects LT1492CS8, this amplification
Circuit can provide > gain amplifier of 80db.
MSP430 processor: when internal resource, this equipment use MSP430F5418A model to comprehensive assessment sexual valence.
MSP430 series monolithic is that Texas Instruments (TI) starts 16 super low-power consumptions of one kind introduced to the market for 1996, has essence
The mixed-signal processor (Mixed Signal Processor) of simple instruction set (RISC).It has the advantage that
Processing capacity is strong: MSP430 series monolithic is one 16 single-chip microcontrollers, is used reduced instruction set computer (RISC)
Structure, with addressing system abundant (7 kinds of source operand addressing, 4 kinds of destination operand addressing), 27 succinct core instructions
And a large amount of dummy instruction；Data storage can all participate in a variety of operations in a large amount of register and piece；There are also efficient
It tables look-up process instruction.These features, which ensure that, can compile efficient source program.
Arithmetic speed is fast: MSP430 series monolithic can realize the instruction cycle of 40ns under the driving of 25MHz crystal.
16 data widths, the instruction cycle of 40ns and multi-functional hardware multiplier (being able to achieve multiply-add operation) match, energy
Realize certain algorithms (such as FFT) of Digital Signal Processing.
Super low-power consumption: the super low-power consumption performance of MSP430 single-chip microcontroller has been industry undisputable fact.MSP430 series monolithic
The supply voltage of machine uses 1.8-3.6V.There are five types of low-power consumption mode (LPM0~LPM4) altogether.It is reachable under real-time clock mode
2.5 μ A, under RAM holding mode, minimum reachable 0.1 μ A.
Resourceful in piece: each series of MSP430 series monolithic is all integrated with more rich interior peripheral hardware.They divide
It is not house dog (WDT), analog comparator, timer, UART, SPI, I2C, hardware multiplier, liquid crystal driver, 10/12
The various combination of several peripheral modules such as ADC, 16 sigma-delta ADC, DMA, real-time clock (RTC) and USB controller.10/12
Hardware A/D converter has higher conversion rate, reaches as high as 200kbps, can satisfy most of data acquisition applications.At this
In design, the ADC of 12 high throw-over rate fully meets design requirement, is also beneficial to power consumption and cost control.
GPRS radio-cell (General Packet Radio Service): being the letter of general packet radio service technology
Claim, it is a kind of available mobile data services of gsm mobile telephone user, belongs to the data transmission skill in Generation Mobile Telecommunication System
Art.Since GPRS promotes the advantages such as, signal stabilization most wide compared with early, coverage rate, use cost be low, while considering to leak source of early warning
Amount of communication data is minimum, is lower than 5Mbyte within one month, therefore GPRS communication mode is undoubtedly best selection.
Software section:
Software carries out analysis and assessment, new algorithm compositive index noise intensity to pipe leakage using novel leakage warning algorithm
It with spectrum distribution feature, is handled and is analyzed from frequency domain to noise data, (example is of the invention by algorithm process process such as Fig. 9
Algorithm process flow diagram) signal:
Acquisition noise data are converted through FFT (fast Fourier) first, obtain noise frequency domain information；By " adaptive frequency
Compose de-noising " algorithm elimination Environmental Noise Influence, retain metastable leakage noise information；Data are by " frequency spectrum is related after de-noising
Analysis " algorithm carries out frequency spectrum correlation operation according to the de-noising period, obtains degree of relevancy on noise temporal；Correlated results by
" early warning analysis " module carries out Statistic analysis, is confirmed whether leakage.
Because acquisition leakage noise frequency is mainly distributed within the scope of 50~2Khz, according to sampling thheorem, sample frequency is answered
Higher than 4Khz, 8Khz sample frequency is chosen in the design.Consider equipment actual power loss and operand, acquire 100ms data every time,
That is 800 valid data samples obtain 400 frequency spectrums (symmetry of FFT), spectral resolution 10Hz after FFT transform.In order to
Compaction algorithms amount, and consider that leaking noise is mainly distributed on 50~2Khz range, all take preceding 200 frequency spectrums to participate in rear class fortune
It calculates.
Comprehensive assessment actual application environment and equipment power dissipation and equipment performance, taking 60 seconds is a collection period, i.e. TP=
60 seconds；5 collection period are a de-noising period, i.e. TF=TP*m=300 seconds, m=5；Continuous 20 correlation analyses result r value
>=0.8, confirm leakage, i.e., pre-warning time=20*TF=6000 seconds=100 minutes.
In order to simplify realization, shorten processing time, raising treatment effeciency, reduction equipment power dissipation, all acquisitions and operation work
Work will be completed in RTC Interruption, and RTC fixed time interval 1 minute, i.e. configures RTC Interruption, fixed time interval 1 within TP=60 seconds
Minute, equipment only wakes up CPU when RTC is interrupted, and when interruption is exited, reenters low-power consumption mode.Interrupt total process flow
As shown in Figure 10 (example interrupt processing flow chart of the invention): illustrating algorithm process process in order to clearer, to key
Algorithm and processing step, will deployment analysis one by one.
Adaptive spectrum de-noising is realized:
Adaptive spectrum de-noising realizes that starting ADC12 first simultaneously configures DMA channel, 8K acquisition rate in RTC Interruption
Continuous acquisition 100ms data；After the completion of to be collected, FFT transform is carried out to acquisition data, and calculate amplitude spectrum；By frequency point ratio
Amplitude spectrum is composed and learnt compared with input range, obtains spectral change state；According to spectral change state, regularized learning algorithm variable g；By frequency
Point adjustment bin magnitudes；Into junior " frequency spectrum correlation analysis " processing routine.Flow chart such as Figure 11 (embodiment party according to the present invention
The example adaptive spectrum denoising algorithm flow chart of formula) shown in.
Frequency spectrum correlation analysis is realized:
Correlation analysis is done to the current study amplitude spectrum Vb and amplitude spectrum VPREV in previous de-noising period first, is obtained related
Coefficient r value；Vb to VPREV is saved later；Correlation coefficient r value passes to early warning analysis processing for statistical analysis；Correlation analysis
It completes；Shown in flow chart such as Figure 12 (the example frequency spectrum related algorithm flow chart of embodiment according to the present invention).Such method,
Correlation analysis is unfolded directly as the input data of correlation analysis in amplitude modal data, correlated results synthesis embodies leakage
Noise intensity and spectrum distribution feature.Noise intensity method is relied on relative to simple, there is better accuracy and objectivity, have
Misinformation probability is leaked conducive to reducing.
Early warning analysis is realized:
The correlation coefficient r value that statistics " correlation analysis " obtains, if r >=0.8, serial correlation degree variables TC=TC+
1；Otherwise TC=0；Judge whether serial correlation degree variables TC is more than associated numbers of times 20 times (100 minutes), if it exceeds related
Number, then early warning exports；Warning information passes through GPRS teletransmission to server or designated mobile phone；So far early warning analysis is completed；Process
Figure is as shown in Figure 13 (the example early warning analysis algorithm flow chart of embodiment according to the present invention).
It is described the prefered embodiments of the present invention in detail above in conjunction with attached drawing, still, the present invention is not limited to above-mentioned realities
The detail in mode is applied, within the scope of the technical concept of the present invention, a variety of letters can be carried out to technical solution of the present invention
Monotropic type, these simple variants all belong to the scope of protection of the present invention.
It is further to note that specific technical features described in the above specific embodiments, in not lance
In the case where shield, it can be combined in any appropriate way.
In addition, various embodiments of the present invention can be combined randomly, as long as it is without prejudice to originally
The thought of invention, it should also be regarded as the disclosure of the present invention.
Table 1:
Claims (2)
1. a kind of leakage early warning system characterized by comprising
1) power supply unit provides electric power safeguard for equipment；
2) acquisition unit, for picking up the leakage noise sample of pipeline weak；
3) plus and blowup unit is used for collected faint duct noise plus and blowup, and send to ADC built in processor
Carry out data acquisition；
4) processor unit: it is responsible for Data Management Analysis and early warning output；The unit includes following several subelements:
A) to noise information sample conversion, discrete sampling data ADC subelement: are obtained；
B) FFT transform subelement: FFT transform is carried out to acquisition noise sample, obtains noise frequency domain information, and calculate amplitude spectrum；
C) self-adapted noise elimination subelement: defining primary complete collection process, i.e. acquisition interval+acquisition duration, is an acquisition week
Continuous m collection period is defined as a de-noising period by the phase, defines a Variable Learning for each frequency point, Variable Learning is used
In the position for indicating current frequency point regulation coefficient, FFT transform is executed to each acquisition data, and obtain input range spectrum, definition
Learn amplitude spectrum, frequency point is compared one by one with study amplitude spectrum to input range spectrum in each collection period, opposite to retain
Stable leakage noise information；Include:
Comparing unit compares for composing to input range and learning amplitude spectrum, obtains spectral change state；
Unit tracks spectral change state for learning to the Variable Learning；
Adjustment unit obtains de-noising amplitude spectrum for being adjusted to the study amplitude spectrum；
D) frequency spectrum correlation analysis subelement: data carry out frequency spectrum related operation by frequency spectrum correlation analysis algorithm after de-noising, obtain
Noise correlation coefficients value；Include:
Receiving unit, for receiving the amplitude spectrum；
Arithmetic element: for carrying out related operation to the amplitude spectrum received and previous de-noising periodic amplitude spectrum, phase relation is obtained
Numerical value；
Storage unit: for storing the amplitude spectrum received；
E) early warning analysis subelement: comprehensive statistics analysis is carried out to correlated results, whether confirmation leakage occurs；Include:
Receiving unit, for receiving relevance degree；
Statistic unit: for for statistical analysis to the relevance degree received；
Judging unit: for judging whether leakage occurs according to statistical result；
5) GPRS transmission unit: when alerting generation, transmitting warning information to server or designated mobile phone.
2. a kind of adaptive spectrum noise-eliminating method of novel leakage early warning system, the system include: power supply unit, acquisition unit,
Plus and blowup unit, processor unit and GPRS transmission unit；The processor unit includes: to turn to noise information sampling
It changes, obtains the ADC subelement of discrete sampling data；FFT transform is carried out to acquisition noise sample, obtains noise frequency domain information, and
Calculate the FFT transform subelement of amplitude spectrum；Environmental Noise Influence is eliminated by adaptive spectrum Denoising Method in frequency domain, is retained opposite
The self-adapted noise elimination subelement of stable leakage noise information；Data carry out frequency spectrum phase by frequency spectrum correlation analysis algorithm after de-noising
Operation is closed, the frequency spectrum correlation analysis subelement of noise correlation coefficients value is obtained；Comprehensive statistics analysis, confirmation are carried out to correlated results
The early warning analysis subelement whether leakage occurs；Wherein self-adapted noise elimination is suddenly as follows:
Acquisition interval is TS twice in succession for definition；Define a length of TL when single acquisition；Primary complete collection process is defined, that is, is acquired
Interval+acquisition duration is a collection period TP, then TP=TS+TL；
Continuous m collection period is defined as a de-noising cycle T F；It defines regulation coefficient array G [n], wherein n=2*m+1,
Enable G [n]=- 1/2, -1/4 ... ..-1/2m, 0,1/2m ... ..1/4,1/2；
A Variable Learning g is defined for each frequency point, and 0≤g < n, Variable Learning g are for indicating current frequency point regulation coefficient
Position corresponds to the subscript of G [n] array, and Variable Learning g is initialized as m；
FFT transform is executed to each acquisition data, and obtaining input range spectrum is Vi；Definition study amplitude spectrum Vb；It is adopted each
The collection period, frequency point was compared one by one with study amplitude spectrum Vb to input range spectrum Vi, and was adjusted as follows:
First by frequency point regularized learning algorithm variable:
Then, by frequency point regularized learning algorithm amplitude spectrum: Vb (n)=Vb (n) * (1+G [g (n)]).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710006221.3A CN106764468B (en) | 2017-01-05 | 2017-01-05 | A kind of leakage early warning system and adaptive spectrum noise-eliminating method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710006221.3A CN106764468B (en) | 2017-01-05 | 2017-01-05 | A kind of leakage early warning system and adaptive spectrum noise-eliminating method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106764468A CN106764468A (en) | 2017-05-31 |
| CN106764468B true CN106764468B (en) | 2019-01-15 |
Family
ID=58949539
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710006221.3A Active CN106764468B (en) | 2017-01-05 | 2017-01-05 | A kind of leakage early warning system and adaptive spectrum noise-eliminating method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106764468B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108916660B (en) * | 2018-03-27 | 2020-03-31 | 南京施迈艾库智能科技有限公司 | Water leakage point detection method based on Hall flowmeter |
| CN108679455A (en) * | 2018-05-17 | 2018-10-19 | 中山市顺康塑料制品有限公司 | Pipeline leakage monitor and line leakage method |
| CN112996012A (en) * | 2019-12-12 | 2021-06-18 | 北京普源精电科技有限公司 | Spectrum data processing method, device, equipment and storage medium |
| CN111256753A (en) * | 2020-01-09 | 2020-06-09 | 成都理工大学 | Intelligent collapse monitoring device and monitoring method thereof |
| CN111476137B (en) * | 2020-04-01 | 2023-08-01 | 北京埃德尔黛威新技术有限公司 | Novel pipeline leakage early warning online relevant positioning data compression method and device |
| CN112651623A (en) * | 2020-12-23 | 2021-04-13 | 贵州树精英教育科技有限责任公司 | Academic ability level testing system and algorithm |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11271168A (en) * | 1998-03-25 | 1999-10-05 | Mitsui Eng & Shipbuild Co Ltd | Leakage detection method |
| CN101319955A (en) * | 2007-06-07 | 2008-12-10 | 北京昊科航科技有限责任公司 | Method for extracting leakage of pipe monitored by infrasonic wave |
| CN102269333B (en) * | 2011-07-20 | 2013-11-06 | 中国海洋石油总公司 | Method for eliminating pipe blockage acoustic signal strong interference by utilizing frequency domain self-adaptive filtering |
| CN103278816B (en) * | 2013-05-14 | 2015-05-20 | 陕西延长石油(集团)有限责任公司研究院 | Petroleum leakage radar detecting system based on linear frequency modulation signal system |
| CN103644460A (en) * | 2013-10-09 | 2014-03-19 | 中国石油大学(华东) | Filtering optimal selection method for leakage sound wave signals of gas transmission line |
| CN206514082U (en) * | 2017-01-05 | 2017-09-22 | 北京埃德尔黛威新技术有限公司 | A kind of new seepage early warning system |
-
2017
- 2017-01-05 CN CN201710006221.3A patent/CN106764468B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106764468A (en) | 2017-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106764468B (en) | A kind of leakage early warning system and adaptive spectrum noise-eliminating method | |
| Lucarini et al. | A new mathematical framework for atmospheric blocking events | |
| Lin | Power harmonics and interharmonics measurement using recursive group-harmonic power minimizing algorithm | |
| Huang et al. | Enhancement of digital equivalent voltage flicker measurement via continuous wavelet transform | |
| CN106523928B (en) | Pipeline leakage detection method based on the screening of sound wave real time data two level | |
| CN107328528A (en) | A kind of ultrasonic wave air tightness detection system and method based on multisensor | |
| CN112014810B (en) | High-precision electronic reconnaissance signal parameter measurement method based on FPGA | |
| CN206514082U (en) | A kind of new seepage early warning system | |
| Zhou et al. | Pipeline signal feature extraction with improved VMD and multi-feature fusion | |
| CN104253659B (en) | Spectrum sensing method and device | |
| CN201378190Y (en) | Digital thunder and lightning detection device | |
| CN106678552B (en) | A kind of novel leakage method for early warning | |
| Vedros et al. | On the limits of em based detection of control logic injection attacks in noisy environments | |
| Bi et al. | Knock Detection in Spark Ignition Engines Base on Complementary Ensemble Empirical Mode Decomposition‐Hilbert Transform | |
| Shao et al. | Segmented Kalman filter based antistrong transient impact method for vortex flowmeter | |
| Xu et al. | Bubble detection in sodium flow using EVFM and correlation coefficient calculation | |
| CN107180627A (en) | The method and apparatus for removing noise | |
| Yang et al. | Design of sensor nodes in underwater sensor networks | |
| CN202126334U (en) | Valve leakage detecting device | |
| Shao et al. | Adaptive filter based anti-strong transient impact method for vortex flowmeter | |
| CN110796036B (en) | Method for improving identification precision of structural modal parameters | |
| CN105786671B (en) | A kind of detection generates the method, apparatus and calculating equipment of power consumption application | |
| CN106685633A (en) | Full-digital phase shift weak signal phase lock detection method and equipment | |
| CN103581923B (en) | A kind of energy detection method based on weighting matrix filtering | |
| Guri et al. | Trusted detection of sensitive activities on mobile phones using power consumption measurements |
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 |