CN104373820A - Method for lowering false alarm rate of pipeline leakage monitoring - Google Patents

Abstract

The invention discloses a method for lowering the false alarm rate of pipeline leakage monitoring. The method includes the following steps that a first sound wave monitor arranged at a pipeline origin station and a second sound wave monitor arranged at a pipeline tail station are used for continuously monitoring an origin station sound wave signal and a tail station sound wave signal in a pipeline in real time respectively; when it is monitored that abnormal signals exist at the origin station and the tail station at the same time, the abnormal signal peak value in the origin station sound wave signal and the abnormal signal peak value in the tail station sound wave signal are calculated; whether the peak value ratio between the origin station abnormal signal peak value and the tail station abnormal signal peak value is within a preset range or not is judged, if yes, an origin station abnormal signal and a tail station abnormal signal are same source signals, and a leakage alarm is given out; if not, the origin station abnormal signal and the tail station abnormal signal are non-same source signals, and the leakage alarm cannot be given out. By means of the method for lowering the false alarm rate of pipeline leakage monitoring, the system can be prevented from giving out a false alarm when it is judged that the signals are not same source signals, the false alarm rate of pipeline leakage monitoring at present can be effectively lowered, leakage monitoring reliability is improved, and the method has good robustness.

Description

Reduce the method for line leakage rate of false alarm

Technical field

The present invention relates to line leakage field, particularly relate to a kind of method reducing line leakage rate of false alarm.

Background technique

In the Cemented filling such as oil gas, hazardous chemical, pipe laying distance, complex circuit, the pipe leakage that normal generation causes due to corrosion and ageing, engineering construction, artificial destruction etc., causes fire good to leak the heavy losses such as blast, environmental pollution, casualties.

At present, the primary metering method for line leakage has: sonic method, Fiber Optic Sensor, negative pressure wave method etc.Line leakage method cost wherein based on sound wave is lower, easily realizes, and has higher leakage monitoring sensitivity and positioning precision, receives a lot of concerns in recent years.The sensor type that sonic method uses comprises: sonic sensor, acceleration transducer, pickup device and piezoelectric pressure transducer etc.But, no matter select which kind of sensor to implement line leakage, inevitably cause false alarm because sonic sensor sensitivity is higher, if especially when upstream and downstream monitoring station monitors different undesired signals respectively, then unavoidably cause wrong report.

Therefore, explore a kind of reliable abnormal signal homology identifying method according to acoustic signals propagation attenuation model, the reliability of reduction rate of false alarm, raising leakage monitoring system is of great importance.

Summary of the invention

Based on this, for realizing a kind of method reducing line leakage rate of false alarm that the object of the invention provides, comprise the following steps:

Initial station acoustic signals and the terminal acoustic signals of pipe interior is monitored respectively in real time, continuously by the second sound wave monitor of the first sound wave monitor and pipeline terminal that are arranged on pipeline initial station;

When monitoring initial station and terminal there is abnormal signal simultaneously, calculate the initial station abnormal signal peak value in the acoustic signals of initial station and the terminal abnormal signal peak value in terminal acoustic signals respectively;

Judge peakedness ratio between described initial station abnormal signal peak value and terminal abnormal signal peak value whether in preset range, if so, then described pipeline leaks; If not, then described pipeline leaks.

As a kind of embodiment reducing the method for line leakage rate of false alarm, second sound wave monitor of described the first sound wave monitor and pipeline terminal by being arranged on pipeline initial station monitors initial station acoustic signals and the terminal acoustic signals of pipe interior respectively in real time, continuously, comprises the following steps:

With sampling period T sample respectively described first sound wave monitor and described second sound wave monitor export described initial station acoustic signals and described terminal acoustic signals two paths of signals, whole minute moment to gather two paths of signals stamp time tag respectively;

Extract continuous two minutes described initial station acoustic signals and described terminal acoustic signals respectively, wherein last minute signal is historical data, within latter one minute, signal is real time data, and described historical data and described real time data form the complete pending initial station acoustic signals of a frame and terminal acoustic signals;

Wherein, T be greater than 0 positive number.

As a kind of embodiment reducing the method for line leakage rate of false alarm, also comprise the step of the preset range of the ratio determined between described initial station abnormal signal peak value and terminal abnormal signal peak value, this step specifically comprises the following steps:

Obtain many group initial station acoustic signals and terminal acoustic signals and carry out filtering, denoising and remove average value processing, obtaining the signal after denoising;

Signal after denoising is obtained to the peak of the correlation coefficient of initial station abnormal signal and terminal abnormal signal by correlation computations, be denoted as CorrPos, and according to formula:

$1=\frac{V_{\mathrm{up}}(L+\mathrm{CorrPos}{V}_{\mathrm{dn}})}{V_{\mathrm{up}}+V_{\mathrm{dn}}},$

Determine the distance of abnormal signal occurrence positions apart from initial station, wherein, l is the distance of abnormal signal occurrence positions apart from initial station, and L is pipeline total length, V _dnwith V _upbe respectively from initial station to terminal with from terminal to the acoustic signals velocity of propagation of initial station;

After the monitoring of Preset Time, obtaining position location is the initial station abnormal signal of initial station and the terminal abnormal signal of homology, and position location is the initial station abnormal signal of terminal and the terminal abnormal signal of homology, and obtain initial station normal signal and the terminal normal signal of the first predetermined number;

Discrete Fourier transform is carried out to the signal after described denoising, obtains the frequency spectrum of the signal after denoising;

The frequency spectrum of comparison normal signal and abnormal signal, intercepts the frequency spectrum of described abnormal signal;

Fourier inversion is carried out to the frequency spectrum of the described abnormal signal after intercepting, obtains the signal after the time domain reconstruction of described abnormal signal;

The peakedness ratio of the peak value of the initial station peak value of abnormal signal and the terminal abnormal signal of homology is determined according to the amplitude of the signal after time domain reconstruction;

According to the propagation attenuation formula of sound wave along pipeline:

$\mathrm{Peak}={\mathrm{Peak}}_0{e}^{-\mathrm{\α}\frac{1}{L}},$

Wherein, Peak ₀for the initial magnitude of abnormal sound wave origination point acoustic signals, Peak is the signal amplitude after propagation attenuation, l is the distance of abnormal signal occurrence positions apart from initial station, L is pipeline total length, α is Acoustic Wave Propagation damping coefficient, the acoustic signals downstream propagation that initial station produces is to terminal, and when terminal generation acoustic signals adverse current propagates into initial station, it meets respectively:

${\mathrm{Peak}}_{\mathrm{dn}}={\mathrm{Peak}}_{\mathrm{up}}{e}^{-{\mathrm{\α}}_s\frac{L}{L}}$

${\mathrm{Peak}}_{\mathrm{up}}={\mathrm{Peak}}_{\mathrm{dn}}{e}^{-{\mathrm{\α}}_n\frac{L}{L}}$

Peak _upfor the abnormal signal amplitude that initial station collects, Peak _dnfor the abnormal signal amplitude of the homology that terminal collects, α _swith α _nbe respectively propagation attenuation coefficient when following current and adverse current are propagated;

Can be calculated acoustic signals downstream propagation damping coefficient α further _swith adverse current propagation attenuation factor alpha _nfor:

${\mathrm{\α}}_s=-1n\left(\frac{{\mathrm{Peak}}_{\mathrm{dn}}}{{\mathrm{Peak}}_{\mathrm{up}}}\right)$

${\mathrm{\α}}_n=-1n\left(\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}\right)$

Peakedness ratio according to the abnormal signal of described initial station abnormal signal and terminal homology obtains α _sand α _nvalue;

In conjunction with the peakedness ratio determination acoustic signals downstream propagation damping coefficient α of the paired initial station abnormal signal of the second predetermined number and the terminal abnormal signal of homology _swith adverse current propagation attenuation factor alpha _nscope be:

α _smin≤α _s≤α _smax

α _nmin≤α _n≤α _nmax；

For the leakage acoustic signals occurred in apart from initial station l distance, the peakedness ratio r of the terminal abnormal signal of its initial station abnormal signal and homology meets following relation:

$r=\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}=\frac{e^{-{\mathrm{\α}}_n\frac{l}{L}}}{e^{-{\mathrm{\α}}_s\frac{(L-l)}{L}}};$

Acoustic Wave Propagation damping coefficient is formed four groups of [α _smin, α _nmin], [α _smin, α _nmax], [α _smax, α _nmin], [α _smax, α _nmax], obtain four r values, and maximum value is designated as r _max, minimum value is designated as r _min, then the span of the peakedness ratio of the terminal abnormal signal of initial station abnormal signal and homology is:

$r_{\min }\≤\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}\≤r_{\max }.$

As a kind of embodiment reducing the method for line leakage rate of false alarm, judge peakedness ratio between described initial station abnormal signal peak value and terminal abnormal signal peak value whether in preset range, if so, then described pipeline leaks; If not, then described pipeline leaks, and comprises the following steps:

Calculate the peakedness ratio between described initial station abnormal signal peak value and described terminal abnormal signal peak value;

According to formula and α _swith α _nvalue calculate the preset range of described peakedness ratio;

Judge described peakedness ratio whether in calculated described preset range;

If so, then first and last station abnormal signal is homologous signal, sends leakage alarm;

If not, then first and last station abnormal signal is non-homogeneous signal, does not send leakage alarm.

As a kind of embodiment reducing the method for line leakage rate of false alarm, the described amplitude according to the signal after time domain reconstruction determines the peakedness ratio of the initial station peak value of abnormal signal and the peak value of terminal abnormal signal, comprises the following steps:

Divide positive and negative interval according to the polarity of described amplitude, and the peak value obtaining each interval is designated as Peak [i], maximum value is got in positive interval, gets minimum value between minus zone, wherein, and 1≤i≤N _peak, N _peakit is the interval sum in the pending initial station acoustic signals or terminal acoustic signals that a frame is complete;

Calculate the average of positive peak and negative peak, be designated as meanVP and meanVN respectively, and the peak value of meanVP will be greater than in positive signal peak value according to following formulae discovery normalization peak value projecting degree index L _p, and when positive peak is less than meanVP L in season _pbe 0,

$L_p=\frac{\mathrm{Peak}\left[i\right]-\mathrm{meanVP}}{\mathrm{Peak}\left[i\right]},1\≤i\≤N_{\mathrm{Peak}},\mathrm{Peak}\left[i\right]>0;$

Amplitude in negative peak is greater than the peak value of meanVN according to following formulae discovery normalization peak value projecting degree index L _n, and when negative peak is greater than meanVN L in season _nbe 0,

$L_N=\frac{\mathrm{Peak}\left[i\right]-\mathrm{meanVN}}{\mathrm{Peak}\left[i\right]},1\&le;i\&le;N_{\mathrm{Peak}},\mathrm{Peak}\left[i\right]<0;$

Be 0 by described normalization peak value projecting degree index clearly lower than the signal of the signal spacing of predetermined threshold value, determine described normalization peak value projecting degree index higher than predetermined threshold value peak value location between and there is moment LeakPos;

When having multiple abnormal signal, calculate the difference of the LeakPos of initial station one abnormal signal and each abnormal signal of terminal successively:

DT＝LeakPos _up(i)-LeakPos _dn(j)

When difference DT and described CorrPos closest to time, determine that this group signal is respectively initial station abnormal signal and corresponding terminal abnormal signal, now interval sequence number is designated as Range respectively _upwith Range _dn;

And then obtain the peak value Peak (Range of initial station abnormal signal _up) and the peak value Peak (Range of terminal abnormal signal _dn), and obtain the peak value of described initial station abnormal signal and the peakedness ratio of described terminal abnormal signal further.

As a kind of embodiment reducing the method for line leakage rate of false alarm, when using initial station acoustic signals and the terminal acoustic signals of the first sound wave monitor of pipeline initial station and the second sound wave monitor monitoring pipe interior of pipeline terminal, global positioning system is utilized to carry out precision time service.

Beneficial effect of the present invention comprises:

A kind of method reducing line leakage rate of false alarm provided by the invention, it is by setting up Acoustic Wave Propagation attenuation model in pipeline, judges whether the peakedness ratio of the abnormal signal that the sound wave monitoring instrument at first and last station receives judges in preset range that received abnormal signal is homologous signal on earth.And stop system to send false alarm when judging not to be homologous signal.Effectively can reduce the rate of false alarm of existing line leakage, improve the reliability of leakage monitoring, and the method have good robustness.

Accompanying drawing explanation

Fig. 1 is a kind of flow chart reducing a specific embodiment of the method for line leakage rate of false alarm of the present invention;

Fig. 2 is a kind of flow chart reducing another specific embodiment of the method for line leakage rate of false alarm of the present invention;

To be that the present invention is a kind of reduce the sample 1 first and last station signal denoising of an instantiation of line leakage rate of false alarm method and go average value processing result schematic diagram Fig. 3;

To be that the present invention is a kind of reduce the sample 2 first and last station signal denoising of an instantiation of line leakage rate of false alarm method and go average value processing result schematic diagram Fig. 4;

Fig. 5 is a kind of sample 1 first and last station signal Fourier inversion result schematic diagram after frequency spectrum intercepts reducing an instantiation of line leakage rate of false alarm method of the present invention;

Fig. 6 is a kind of sample 2 first and last station signal Fourier inversion result schematic diagram after frequency spectrum intercepts reducing an instantiation of line leakage rate of false alarm method of the present invention.

Embodiment

In order to make object of the present invention, technological scheme and advantage clearly understand, be described below in conjunction with the embodiment of accompanying drawing to the method for reduction line leakage rate of false alarm of the present invention.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.

The method of the reduction line leakage rate of false alarm of one embodiment of the invention, as shown in Figure 1, comprises the following steps:

S100, monitors initial station acoustic signals and the terminal acoustic signals of pipe interior respectively in real time, continuously by the second sound wave monitor of the first sound wave monitor and pipeline terminal that are arranged on pipeline initial station.

S200, when monitoring abnormal signal, by calculating the position location of signal correction coefficient determination abnormal signal, calculates the abnormal signal peak value in the acoustic signals of initial station and the abnormal signal peak value in terminal acoustic signals respectively.

It should be noted that, sound wave monitoring instrument (comprising the first sound wave monitor and the second sound wave monitor) continues to monitor acoustic signals herein, general when pipeline does not leak, and the signal received is comparatively mild, does not significantly fluctuate.But, if the pipeline of monitoring have leak or stronger undesired signal (as adjusted pump, adjusting valve etc.) time, the signal that sound wave monitoring instrument receives has obvious fluctuation, then now say ripple monitor and monitored abnormal signal.And the judgement of abnormal signal in conjunction with existing techniques in realizing, can be described in detail herein no longer one by one.

S300, judge peakedness ratio between described initial station abnormal signal peak value and terminal abnormal signal peak value whether in preset range, if so, then first and last station abnormal signal is homologous signal, sends leakage alarm; If not, then first and last station abnormal signal is non-homogeneous signal, does not send leakage alarm.

For the conveyance conduit that any one is determined, when there is homology abnormal signal, peakedness ratio between the peak value of the abnormal signal that it is arranged at first stop and the sound wave monitoring instrument being arranged on terminal monitors is all within the specific limits, therefore, by judging whether the peakedness ratio of the abnormal signal that the sound wave monitoring instrument at first and last station receives judges in preset range that received abnormal signal is homologous signal on earth.And stop system to send false alarm when judging not to be homologous signal.Effectively can reduce the rate of false alarm of existing line leakage, improve the reliability of leakage monitoring.

Also it should be noted that, initial station herein refers to the initial position that the logistics of conveyance conduit delivered inside is dynamic and is provided with monitoring point herein.Terminal is that the final position that the logistics of conveyance conduit delivered inside is moved also is provided with monitoring point herein.Distance between initial station and terminal is the length of delivery line substantially.Certainly, two also can be selected on pipeline close to first and last station with reference to method of the present invention and monitor at a distance of the point of fixed length.

Wherein in an embodiment, step S100, is monitored initial station acoustic signals and the terminal acoustic signals of pipe interior respectively in real time, continuously, comprises the following steps by the second sound wave monitor of the first sound wave monitor and pipeline terminal that are arranged on pipeline initial station:

S110, with sampling period T sample respectively described first sound wave monitor and described second sound wave monitor export described initial station acoustic signals and described terminal acoustic signals two paths of signals, whole minute moment to gather two paths of signals stamp time tag respectively.

Wherein, for the monitoring time of the sound wave monitoring instrument of initial station and terminal by with GPS (GlobalPositioning System, global positioning system) carry out under precision time service, thus the synchronous of initial station acoustic signals and terminal Acoustic Signal Acquisition can be guaranteed.Improve the precision of monitoring.

S120, if collection signal length per minute is N/2 point, extract continuous two minutes described initial station acoustic signals and described terminal acoustic signals respectively, wherein last minute signal is historical data, within latter one minute, signal is real time data, and described historical data and described real time data form the complete pending initial station acoustic signals of a frame and terminal acoustic signals.Wherein, T be greater than 0 positive number, N is positive integer.

Concrete, realize by building Acoustic Wave Propagation attenuation model in pipeline in preset range the present invention of the ratio between initial station abnormal signal peak value and terminal abnormal signal peak value.It carries out abnormal signal to the signal collected and diagnoses and location, and screening obtains wherein at first stop saequential transmission to abnormal signal and the inverse abnormal signal passing to initial station of terminal of terminal.Utilize above-mentioned saequential transmission with the inverse abnormal signal that passes and choose some normal signal, denoising carries out discrete Fourier transform after going average, obtain the frequency spectrum of each signal, the frequency spectrum of abnormal signal and normal signal is contrasted, determine the abnormal signal frequency band range being different from normal signal.The frequency spectrum of above-mentioned abnormal signal is carried out frequency domain filtering by the frequency band range determined, the time domain reconstruction signal after utilizing Fourier inversion to obtain frequency domain filtering.Signal after reconstruct is carried out positive and negative interval division, in outstanding signal location, finds peak value, calculate the ratio of first and last station peak value.Calculate damping coefficient according to Acoustic Wave Propagation decay formula, set up the Acoustic Wave Propagation attenuation model in corresponding pipeline.

Concrete, S010, determines that the step of the preset range of the ratio between described initial station abnormal signal peak value and terminal abnormal signal peak value comprises the following steps:

S011, obtains many group initial station acoustic signals and terminal acoustic signals and carries out filtering, denoising and remove average value processing, obtaining the signal after denoising.

S012, is obtained the peak of the correlation coefficient of initial station abnormal signal and terminal abnormal signal, is denoted as CorrPos by correlation computations to the signal after denoising.According to formula:

$1=\frac{V_{\mathrm{up}}(L+\mathrm{CorrPos}{V}_{\mathrm{dn}})}{V_{\mathrm{up}}+V_{\mathrm{dn}}}.---\left(1\right)$

Calculate abnormal signal occurrence positions, wherein, L is pipeline total length, and Vdn and Vup is respectively from initial station to terminal with from terminal to the acoustic signals velocity of propagation of initial station.Obtaining position location is the initial station abnormal signal of initial station and the terminal abnormal signal of homology, and position location is the initial station abnormal signal of terminal and the terminal abnormal signal of homology, and obtain initial station normal signal and the terminal normal signal of some (the first predetermined number).

It should be noted that, described Preset Time is determined according to the actual requirements herein, and the present invention needs the terminal abnormal signal of many group initial station abnormal signals and homology in the process determining preset range.Therefore, Preset Time herein can be determined according to the complexity of the actual precision associative operation wanting to reach.Equally, normal signal mainly plays reference role, and its concrete quantity also can be chosen according to actual conditions.

S013, carries out discrete Fourier transform to the signal after described denoising, obtains the frequency spectrum of the signal after denoising.Frequency spectrum X (k) of signal is obtained after discrete Fourier transform:

X(k)＝DFT[x(n)]0≤k≤N-1 (2)

In formula, x (n) is time-domain signal, and N is that discrete Fourier transform is counted, and also always counts for signal.

S014, the frequency spectrum of comparison normal signal and abnormal signal, intercepts the frequency spectrum of described abnormal signal.Be specially: the spectrogram drawing abnormal signal and normal signal, observe each abnormal signal of statistics and the frequency band range belonging to normal signal, initial and the end sequence number of the frequency band being different from a section narrower of normal signal at abnormal signal place in frequency spectrum is designated as fst and fend respectively, the frequency spectrum of this frequency band is retained remaining and do clearing process.Thus the narrower frequency band that can obtain accurately corresponding to abnormal signal.Frequency spectrum wherein after adjustment is:

Said normal signal comprises initial station normal signal and terminal normal signal herein, and abnormal signal comprises initial station abnormal signal and terminal abnormal signal.

S015, carries out Fourier inversion to the frequency spectrum of the abnormal signal after described intercepting, obtains the signal after the time domain reconstruction of described abnormal signal.

S016, determines the peakedness ratio of the initial station peak value of abnormal signal and the peak value of terminal abnormal signal according to the amplitude of the signal after time domain reconstruction.

S017, according to the propagation attenuation formula of sound wave along pipeline:

$\mathrm{Peak}={\mathrm{Peak}}_0{e}^{-\mathrm{\&alpha;}\frac{1}{L}}---\left(4\right)$

Wherein, Peak ₀for leaking the initial magnitude of origination point acoustic signals, Peak is the signal amplitude after propagation attenuation, and l is the distance of signal occurrence positions apart from initial station, and L is pipeline total length, and α is Acoustic Wave Propagation damping coefficient.Obtain the acoustic signals downstream propagation of initial station generation to terminal, or when terminal generation acoustic signals adverse current propagates into initial station, it meets respectively:

${\mathrm{Peak}}_{\mathrm{dn}}={\mathrm{Peak}}_{\mathrm{up}}{e}^{-{\mathrm{\&alpha;}}_s\frac{L}{L}}---\left(5\right)$

${\mathrm{Peak}}_{\mathrm{up}}={\mathrm{Peak}}_{\mathrm{dn}}{e}^{-{\mathrm{\&alpha;}}_n\frac{L}{L}}---\left(6\right)$

In formula, the propagation length of acoustic signals is L, Peak _upfor the abnormal signal amplitude that initial station collects, Peak _dnfor terminal collects the abnormal signal amplitude of homology, α _swith α _nbe respectively propagation attenuation coefficient when following current and adverse current are propagated.

S018, can be calculated acoustic signals downstream propagation damping coefficient α further _swith adverse current propagation attenuation factor alpha _nfor:

${\mathrm{\&alpha;}}_s=-1n\left(\frac{{\mathrm{Peak}}_{\mathrm{dn}}}{{\mathrm{Peak}}_{\mathrm{up}}}\right)---\left(7\right)$

${\mathrm{\&alpha;}}_n=-1n\left(\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}\right)---\left(8\right)$

S019, the peakedness ratio of the initial station abnormal signal determined according to step S016 and the terminal abnormal signal of homology obtains α _sand α _noccurrence.

S020, in conjunction with the peakedness ratio determination acoustic signals downstream propagation damping coefficient α of the paired initial station abnormal signal of the second predetermined number and the terminal abnormal signal of homology _swith adverse current propagation attenuation factor alpha _nscope be:

$\begin{array}{c}{\mathrm{\&alpha;}}_{s\min }\&le;{\mathrm{\&alpha;}}_s\&le;{\mathrm{\&alpha;}}_{s\max }\\ {\mathrm{\&alpha;}}_{n\min }\&le;{\mathrm{\&alpha;}}_n\&le;{\mathrm{\&alpha;}}_{n\max }\end{array}---\left(9\right)$

As a kind of embodiment, described second predetermined number can choose 10 or 12.Certainly, according to the requirement of available accuracy, also more quantity can be selected.

After determining the damping coefficient of co-current flow and counter-current flow, continue to perform next step, S021, for the leakage acoustic signals occurred in apart from initial station l distance, the peakedness ratio r of the terminal abnormal signal of its initial station abnormal signal and homology meets following relation:

$r=\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}=\frac{e^{-{\mathrm{\&alpha;}}_n\frac{l}{L}}}{e^{-{\mathrm{\&alpha;}}_s\frac{(L-l)}{L}}}---\left(10\right)$

Acoustic Wave Propagation damping coefficient is formed four groups of [α _smin, α _nmin], [α _smin, α _nmax], [α _smax, α _nmin], [α _smax, α _nmax], obtain four r values, and maximum value is designated as r _max, minimum value is designated as r _min, then the span of the peakedness ratio of initial station abnormal signal and terminal abnormal signal is:

$r_{\min }\&le;\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}\&le;r_{\max }---\left(11\right)$

Finally determine initial station abnormal signal peak value and terminal abnormal signal peak value between ratio at r _minand r _maxbetween, acoustic signals propagation attenuation model in the pipeline that above formula is also structure.

As above-mentioned known, when the ratio of the described abnormal signal peak value that the abnormal signal peak value and described terminal of determining described initial station are determined is whether in preset range, also need the occurrence of the occurrence positions determination preset range of the abnormal signal according to the supposition tentatively determined.Accordingly, step S300, judge peakedness ratio between described initial station abnormal signal peak value and terminal abnormal signal peak value whether in preset range, if so, then first and last station abnormal signal is homologous signal, sends leakage alarm; If not, then first and last station abnormal signal is non-homogeneous signal, does not send leakage alarm, comprises the following steps:

S310, calculates the peakedness ratio between initial station abnormal signal peak value and terminal abnormal signal peak value.

S320, according to formula (10) and l, α _swith α _nvalue calculate the preset range of described peakedness ratio.

S330, judges described peakedness ratio whether in calculated described preset range; If so, then first and last station abnormal signal is homologous signal, sends leakage alarm; If not, then first and last station abnormal signal is non-homogeneous signal, does not send leakage alarm.

Concrete, step S310, calculates the peakedness ratio between initial station abnormal signal peak value and terminal abnormal signal peak value and step S016, determines the peakedness ratio of the initial station peak value of abnormal signal and the peak value of terminal abnormal signal according to the amplitude of the signal after time domain reconstruction, method is identical, comprises the following steps:

S311, divides positive and negative interval according to the polarity of described amplitude, and the peak value obtaining each interval is designated as Peak [i], and maximum value is got in positive interval, gets minimum value between minus zone, wherein, and 1≤i≤N _peak, N _peakit is the interval sum in the pending initial station acoustic signals or terminal acoustic signals that a frame is complete;

S312, calculates the average of positive peak and negative peak, is designated as meanVP and meanVN respectively, and will be greater than the peak value of meanVP in positive signal peak value according to following formulae discovery normalization peak value projecting degree index L _p, be 0 when positive peak is less than meanVP LP in season,

$L_p=\frac{\mathrm{Peak}\left[i\right]-\mathrm{meanVP}}{\mathrm{Peak}\left[i\right]},1\&le;i\&le;N_{\mathrm{Peak}},\mathrm{Peak}\left[i\right]>0---\left(12\right)$

Amplitude in negative peak is greater than the peak value of meanVN according to following formulae discovery normalization peak value projecting degree index L _n, be 0 when negative peak is greater than meanVN LN in season,

$L_N=\frac{\mathrm{Peak}\left[i\right]-\mathrm{meanVN}}{\mathrm{Peak}\left[i\right]},1\&le;i\&le;N_{\mathrm{Peak}},\mathrm{Peak}\left[i\right]<0---\left(13\right)$

Be 0 by described normalization peak value projecting degree index clearly lower than the signal of the signal spacing of predetermined threshold value, determine between the peak value location of described normalization peak value projecting degree index higher than predetermined threshold value, and determine that moment LeakPos occurs its signal.

It should be noted that, when peak value projecting degree index is more more outstanding close to signal when 1 or-1 herein; When projecting degree index close to 0 time represent that signal is not given prominence to.To determine between outstanding signal peak location according to peak value projecting degree index and predetermined threshold value and determine that the moment occurs outstanding signal.The signal differential in each interval is calculated, the step-length of the difference time-domain signal that to be w, x be after denoising according to formula d [i]=x [i]-x [i+w].In the signal differential d of outstanding signal, maximizing position is designated as the generation moment LeakPos of current interval signal, and the initial station abnormal signal generation moment is designated as LeakPos _upi (), the terminal abnormal signal generation moment is designated as LeakPos _dnj (), i, j are positive integer, and are all less than or equal to the interval sum of affiliated signal.In the signal of first and last station, all projecting degree indexs are lower than being all clearly 0 in the signal spacing of predetermined threshold value.

S313, when having multiple abnormal signal, calculates the difference of the LeakPos of initial station one abnormal signal and each abnormal signal of terminal successively:

DT＝LeakPos _up(i)-LeakPos _dn(j) (14)

When the CorrPos in difference DT and aforementioned S012 closest to time, determine that this group signal is respectively initial station abnormal signal and corresponding terminal abnormal signal;

S314, obtains the peak value Peak (Range of initial station abnormal signal _up) and the peak value Peak (Range of terminal abnormal signal _dn), and obtain the peak value of described initial station abnormal signal and the peakedness ratio of described terminal abnormal signal further.

It should be noted that, after determining the preset range of initial station abnormal signal peak value and terminal abnormal signal peakedness ratio, just reality is monitored the homology of abnormal signal herein.In observation process, as shown in Figure 2, after monitoring abnormal signal, also to perform step S011 ~ S015 and signal is processed and calculates abnormal signal position location for distance initial station distance l, correlation coefficient peak CorrPos.And continue to perform step S311 ~ S314.As shown in Figure 2, after obtaining time domain reconstruction signal, continue to divide positive and negative interval (S311) by the polarity of signal amplitude, find out abnormal signal according to positioning result in conjunction with Difference Calculation and the moment occurs, obtain abnormal signal peak value, according between normalization peak value projecting degree determination abnormal signal location (S312 ~ S313), perform step S314 afterwards and calculate the initial station peak value of abnormal signal and the peakedness ratio of described terminal abnormal signal, finally carry out judging whether pipeline leaks according to preset range, if then send leakage alarm, if not, then be judged to be non-homogeneous signal, do not report to the police.

The application of example to method of the present invention that act one is concrete is below described.

Choose two abnormal signal samples to identify, sample 1 is undesired signal, sample 2 is leakage signal, two samples are the complete signal of a frame.As shown in Figure 2, and method of the present invention can realize with any Programming with Pascal Language algorithm flow, and runs on computers.

Step 1. is by the abnormal signal of a certain amount of pipe under test with the spectrogram of normal signal is added up, comparison, and abnormal signal frequency band range is the 5 to the 45 spectral line and the 5967 to 5997 spectral line (symmetry properties of Fourier transformation frequency spectrum) in frequency spectrum.According to α in the attenuation model that historical data is set up _smax=-0.2785, α _smin=-0.4884, α _nmax=1.9640, α _nmin=1.5746.By abnormal signal diagnostic method, diagnose out in this two frame signal and all there is abnormal signal, signal length is N=6000 point, and sample rate is 50Hz, pipeline total length L=12.409km.

Step 2. adopts moving average filter denoising, and yardstick is 100; Through moving average filter denoising with go average, obtain ambipolar first and last station signal.Wherein Fig. 3 is the sample 1 first and last station signal after process, and Fig. 4 is sample 2 first and last station signal.The figure of two figure middle and upper parts is initial station signal schematic representation, and the figure of bottom is the signal schematic representation of terminal.

Step 3. calculates the correlation coefficient peak value of sample 1 and sample 2, and the abnormal signal position location calculating sample 1 is l ₁=8.6161km, the abnormal signal position location of sample 2 is l ₂=0.6006km

Spectral line in the frequency spectrum of first and last station signal except the frequency band range of the place of abnormal signal described in step 1 is all clearly 0 by step 4..

Frequency spectrum after above-mentioned frequency domain filtering is obtained time domain reconstruction waveform by Fourier inversion by step 5., and as shown in Figure 5, the time domain reconstruction waveform of sample 2 as shown in Figure 6 for the time domain reconstruction waveform of sample 1.

The signal at first and last station, according to the polarity of signal amplitude, is divided into some positive and negative intervals by step 6..In this example, signal spacing, sample 1 initial station number is 46, and terminal signal spacing number is 49; Signal spacing, sample 2 initial station number is 37, and terminal signal spacing number is 55.

Step 7. sample 1 initial station abnormal signal is positioned at the 37th interval, and terminal abnormal signal is positioned at the 34th interval, and interval interior signal is done the Difference Calculation that time delay yardstick is 50, and after the signal differential of initial station, extreme point position is x ₁(4248), terminal is y ₁(3908), initial station abnormal signal peak value Peak _up1=227.3970, terminal abnormal signal peak value Peak _dn1=581.4869, Amplitude Ration is r ₁=Peak _up1/ Peak _dn1=0.3911.

Sample 2 initial station abnormal signal is positioned at the 9th interval, and terminal abnormal signal is positioned at the 17th interval, and interval interior signal is done the Difference Calculation that time delay yardstick is 50, and after the abnormal signal difference of initial station, extreme point position is x ₂(957), terminal is y ₂(1775), initial station abnormal signal peak value Peak _up2=181.4235, terminal abnormal signal peak value Peak _dn2=294.1498, Amplitude Ration is r ₂=Peak _up2/ Peak _dn2=0.6168.

Sample 1 and the orientation distance of sample 2 are substituted into model and show that sample 1 Amplitude Ration interval is for [0.2203,0.3087] by step 8., and the Amplitude Ration interval of sample 2 is [0.5713,0.7109].The Amplitude Ration r of sample 2 ₂in interval, be judged to be the leakage signal of homology; The Amplitude Ration r of sample 1 ₁outside interval, be judged to be non-homogeneous undesired signal.

The method of reduction line leakage rate of false alarm of the present invention, based on Acoustic Wave Propagation attenuation model, according to the modeling of pipeline actual signal, reduces the impact of transducer sensitivity and transmitter magnification factor greatly.Due to the time-frequency characteristic of undesired signal and leakage signal time-frequency characteristic more close, adopt mode identification method to be in the past difficult to distinguish, easily cause wrong report.And this method can realize the identification of pipe leakage undesired signal effectively and quickly, improve the accuracy that line leakage is reported to the police, and there is good robustness.

The above embodiment only have expressed several mode of execution of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.

Claims (6)

1. reduce a method for line leakage rate of false alarm, it is characterized in that, comprise the following steps:

Initial station acoustic signals and the terminal acoustic signals of pipe interior is monitored respectively in real time, continuously by the second sound wave monitor of the first sound wave monitor and pipeline terminal that are arranged on pipeline initial station;

When monitoring initial station and terminal there is abnormal signal simultaneously, calculate the initial station abnormal signal peak value in the acoustic signals of initial station and the terminal abnormal signal peak value in terminal acoustic signals respectively;

Judge peakedness ratio between described initial station abnormal signal peak value and terminal abnormal signal peak value whether in preset range, if so, then described pipeline leaks; If not, then described pipeline leaks.

2. the method for reduction line leakage rate of false alarm according to claim 1, it is characterized in that, second sound wave monitor of described the first sound wave monitor and pipeline terminal by being arranged on pipeline initial station monitors initial station acoustic signals and the terminal acoustic signals of pipe interior respectively in real time, continuously, comprises the following steps:

With sampling period T sample respectively described first sound wave monitor and described second sound wave monitor export described initial station acoustic signals and described terminal acoustic signals two paths of signals, whole minute moment to gather two paths of signals stamp time tag respectively;

Extract continuous two minutes described initial station acoustic signals and described terminal acoustic signals respectively, wherein last minute signal is historical data, within latter one minute, signal is real time data, and described historical data and described real time data form the complete pending initial station acoustic signals of a frame and terminal acoustic signals;

Wherein, T be greater than 0 positive number.

3. the method for reduction line leakage rate of false alarm according to claim 2, it is characterized in that, also comprise the step of the preset range of the ratio determined between described initial station abnormal signal peak value and terminal abnormal signal peak value, this step specifically comprises the following steps:

Obtain many group initial station acoustic signals and terminal acoustic signals and carry out filtering, denoising and remove average value processing, obtaining the signal after denoising;

Signal after denoising is obtained to the peak of the correlation coefficient of initial station abnormal signal and terminal abnormal signal by correlation computations, be denoted as CorrPos, and according to formula:

$1=\frac{V_{\mathrm{up}}(L+\mathrm{CorrPos}{V}_{\mathrm{dn}})}{V_{\mathrm{up}}+V_{\mathrm{dn}}},$

Determine the distance of abnormal signal occurrence positions apart from initial station, wherein, l is the distance of abnormal signal occurrence positions apart from initial station, and L is pipeline total length, V _dnwith V _upbe respectively from initial station to terminal with from terminal to the acoustic signals velocity of propagation of initial station;

After the monitoring of Preset Time, obtaining position location is the initial station abnormal signal of initial station and the terminal abnormal signal of homology, and position location is the initial station abnormal signal of terminal and the terminal abnormal signal of homology, and obtain initial station normal signal and the terminal normal signal of the first predetermined number;

Discrete Fourier transform is carried out to the signal after described denoising, obtains the frequency spectrum of the signal after denoising;

The frequency spectrum of comparison normal signal and abnormal signal, intercepts the frequency spectrum of described abnormal signal;

Fourier inversion is carried out to the frequency spectrum of the described abnormal signal after intercepting, obtains the signal after the time domain reconstruction of described abnormal signal;

The peakedness ratio of the peak value of the initial station peak value of abnormal signal and the terminal abnormal signal of homology is determined according to the amplitude of the signal after time domain reconstruction;

According to the propagation attenuation formula of sound wave along pipeline:

$\mathrm{Peak}={\mathrm{Peak}}_0{e}^{-\mathrm{\&alpha;}\frac{1}{L}},$

Wherein, Peak ₀for the initial magnitude of abnormal sound wave origination point acoustic signals, Peak is the signal amplitude after propagation attenuation, l is the distance of abnormal signal occurrence positions apart from initial station, L is pipeline total length, α is Acoustic Wave Propagation damping coefficient, the acoustic signals downstream propagation that initial station produces is to terminal, and when terminal generation acoustic signals adverse current propagates into initial station, it meets respectively:

${\mathrm{Peak}}_{\mathrm{dn}}={\mathrm{Peak}}_{\mathrm{up}}{e}^{-{\mathrm{\&alpha;}}_s\frac{L}{L}}$

${\mathrm{Peak}}_{\mathrm{up}}={\mathrm{Peak}}_{\mathrm{dn}}{e}^{-{\mathrm{\&alpha;}}_n\frac{L}{L}}$

Peak _upfor the abnormal signal amplitude that initial station collects, Peak _dnfor the abnormal signal amplitude of the homology that terminal collects, α _swith α _nbe respectively propagation attenuation coefficient when following current and adverse current are propagated;

Can be calculated acoustic signals downstream propagation damping coefficient α further _swith adverse current propagation attenuation factor alpha _nfor:

${\mathrm{\&alpha;}}_s=-1n\left(\frac{{\mathrm{Peak}}_{\mathrm{dn}}}{{\mathrm{Peak}}_{\mathrm{up}}}\right)$

${\mathrm{\&alpha;}}_n=-1n\left(\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}\right)$

Peakedness ratio according to the abnormal signal of described initial station abnormal signal and terminal homology obtains α _sand α _nvalue;

In conjunction with the peakedness ratio determination acoustic signals downstream propagation damping coefficient α of the paired initial station abnormal signal of the second predetermined number and the terminal abnormal signal of homology _swith adverse current propagation attenuation factor alpha _nscope be:

α _smin≤α _s≤α _smax

α _nmin≤α _n≤α _nmax；

For the leakage acoustic signals occurred in apart from initial station l distance, the peakedness ratio r of the terminal abnormal signal of its initial station abnormal signal and homology meets following relation:

$r=\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}=\frac{e^{-{\mathrm{\&alpha;}}_n\frac{l}{L}}}{e^{-{\mathrm{\&alpha;}}_s\frac{(L-l)}{L}}};$

Acoustic Wave Propagation damping coefficient is formed four groups of [α _smin, α _nmin], [α _smin, α _nmax], [α _smax, α _nmin], [α _smax, α _nmax], obtain four r values, and maximum value is designated as r _max, minimum value is designated as r _min, then the span of the peakedness ratio of the terminal abnormal signal of initial station abnormal signal and homology is:

$r_{\min }\&le;\frac{{\mathrm{Peak}}_{\mathrm{up}}}{{\mathrm{Peak}}_{\mathrm{dn}}}\&le;r_{\max }.$

4. the method for reduction line leakage rate of false alarm according to claim 3, is characterized in that, judge peakedness ratio between described initial station abnormal signal peak value and terminal abnormal signal peak value whether in preset range, if so, then described pipeline leaks; If not, then described pipeline leaks, and comprises the following steps:

Calculate the peakedness ratio between described initial station abnormal signal peak value and described terminal abnormal signal peak value;

According to formula and α _swith α _nvalue calculate the preset range of described peakedness ratio;

Judge described peakedness ratio whether in calculated described preset range;

If so, then first and last station abnormal signal is homologous signal, sends leakage alarm;

If not, then first and last station abnormal signal is non-homogeneous signal, does not send leakage alarm.

5. the method for reduction line leakage rate of false alarm according to claim 3, is characterized in that, the described amplitude according to the signal after time domain reconstruction determines the peakedness ratio of the initial station peak value of abnormal signal and the peak value of terminal abnormal signal, comprises the following steps:

Divide positive and negative interval according to the polarity of described amplitude, and the peak value obtaining each interval is designated as Peak [i], maximum value is got in positive interval, gets minimum value between minus zone, wherein, and 1≤i≤N _peak, N _peakit is the interval sum in the pending initial station acoustic signals or terminal acoustic signals that a frame is complete;

Calculate the average of positive peak and negative peak, be designated as meanVP and meanVN respectively, and the peak value of meanVP will be greater than in positive signal peak value according to following formulae discovery normalization peak value projecting degree index L _p, and when positive peak is less than meanVP L in season _pbe 0,

$L_p=\frac{\mathrm{Peak}\left[i\right]-\mathrm{meanVP}}{\mathrm{Peak}\left[i\right]},1\&le;i\&le;N_{\mathrm{Peak}},\mathrm{Peak}\left[i\right]>0;$

Amplitude in negative peak is greater than the peak value of meanVN according to following formulae discovery normalization peak value projecting degree index L _n, and when negative peak is greater than meanVN L in season _nbe 0,

$L_N=\frac{\mathrm{Peak}\left[i\right]-\mathrm{meanVN}}{\mathrm{Peak}\left[i\right]},1\&le;i\&le;N_{\mathrm{Peak}},\mathrm{Peak}\left[i\right]<0;$

Be 0 by described normalization peak value projecting degree index clearly lower than the signal of the signal spacing of predetermined threshold value, determine described normalization peak value projecting degree index higher than predetermined threshold value peak value location between and there is moment LeakPos;

When having multiple abnormal signal, calculate the difference of the LeakPos of initial station one abnormal signal and each abnormal signal of terminal successively:

DT＝LeakPos _up(i)-LeakPos _dn(j)

When difference DT and described CorrPos closest to time, determine that this group signal is respectively initial station abnormal signal and corresponding terminal abnormal signal, now interval sequence number is designated as Range respectively _upwith Range _dn;

And then obtain the peak value Peak (Range of initial station abnormal signal _up) and the peak value Peak (Range of terminal abnormal signal _dn), and obtain the peak value of described initial station abnormal signal and the peakedness ratio of described terminal abnormal signal further.

6. the method for reduction line leakage rate of false alarm according to claim 2, it is characterized in that, when using initial station acoustic signals and the terminal acoustic signals of the first sound wave monitor of pipeline initial station and the second sound wave monitor monitoring pipe interior of pipeline terminal, global positioning system is utilized to carry out precision time service.