CN107563436A - A kind of method and system of the automatic identification low-level jet stream on barb figure - Google Patents
A kind of method and system of the automatic identification low-level jet stream on barb figure Download PDFInfo
- Publication number
- CN107563436A CN107563436A CN201710765185.9A CN201710765185A CN107563436A CN 107563436 A CN107563436 A CN 107563436A CN 201710765185 A CN201710765185 A CN 201710765185A CN 107563436 A CN107563436 A CN 107563436A
- Authority
- CN
- China
- Prior art keywords
- low
- jet stream
- level jet
- wind
- history
- 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.)
- Granted
Links
Landscapes
- Air Conditioning Control Device (AREA)
Abstract
The present invention relates to a kind of method and system of the automatic identification low-level jet stream on barb figure, this method includes：Determine that history normalizes Wind outline；Determine the characteristic parameter scope of low-level jet stream；History normalization Wind outline is chronological, obtain the normalized anemobiagraph of history；Member function corresponding with characteristic quantity is constructed according to the characteristic quantity of each point in figure, according to characteristic quantity and member function corresponding with characteristic quantity, determines low-level jet stream point；Low-level jet stream point is filtered and clustering processing, obtains low-level jet stream region；Member function corresponding with characteristic quantity is constructed according to the characteristic quantity in each low-level jet stream region, according to characteristic quantity and member function corresponding with characteristic quantity, it is determined that final low-level jet stream region.A kind of system is further related to, the system includes：Receiver, processor and output equipment.It can judge to whether there is low-level jet stream from barb figure in real time by the present invention, low-level jet stream region can be positioned exactly, and export the characteristic parameter of low-level jet stream.
Description
Technical field
The invention belongs to weather radar field, more particularly to a kind of method of automatic identification low-level jet stream on barb figure and
System.
Background technology
At present, when adjudicating low-level jet stream intensity on barb figure, the barb for generally first extracting each height of any instant out is made
Wind outline, then the characteristic quantity such as calculation of wind speed maximum and place height, wind speed highly compare, more than this height Wind Speed Shear.Though
Right this way can preferably characterize some information of low-level jet stream, but relatively simple if for Intelligent Recognition, as certainly
Dynamicization flow is it is possible that higher rate of failing to report, it is difficult to adjudicates whether low-level jet stream is deposited from newest barb figure in real time
The parameters such as the duration on low-level jet stream, altitude range, maximum wind velocity, torrent intensity can not accurately exported；And for
Different purposes, each characteristic quantity is typically used alone, integrated level is relatively low, is not easy to the overall cognitive of torrent.
The content of the invention
The technical problems to be solved by the invention are：The method of existing identification low-level jet stream is manual identified mostly or showed
Real time implementation is also not implemented in some automatic identifying methods, higher rate of failing to report occurs when especially identifying, it is difficult in real time from most
On new barb figure adjudicate low-level jet stream whether there is, can not accurately export the duration on low-level jet stream, altitude range,
The parameters such as maximum wind velocity, so as to judge the intensity of low-level jet stream.
To solve technical problem above, the invention provides a kind of side of the automatic identification low-level jet stream on barb figure
Method, this method include：
S1, according to the history Wind outline of low-level jet stream in the Wind outline of current time low-level jet stream and preset time period, really
Determine history normalization Wind outline；
S2, Wind outline is normalized according to the historical data of low-level jet stream and the history, determines the spy of the low-level jet stream
Levy parameter area；
S3, history normalization Wind outline is chronological, the normalized anemobiagraph of history is obtained, wherein described go through
The normalized anemobiagraph of history be using the time as abscissa, highly be ordinate, speed for ordinate image；
S4, the characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while according to the spy of the low-level jet stream
Levy parameter area and construct corresponding with characteristic quantity member function, and according to the characteristic quantity and corresponding with the characteristic quantity
Member function, determine low-level jet stream point；
S5, the low-level jet stream point is filtered and clustering processing, obtains multiple low-level jet stream regions；
S6, the characteristic quantity in each low-level jet stream region is calculated, while according to the characteristic parameter model of the low-level jet stream
Construction member function corresponding with the characteristic quantity is enclosed, and according to the characteristic quantity and member's letter corresponding with the characteristic quantity
Number, it is determined that final low-level jet stream region.
Beneficial effects of the present invention：By the above method, can judge to whether there is low latitude from barb figure in real time
Torrent, low-level jet stream region can be more accurately positioned, duration, the height of low-level jet stream can be checked on barb figure
The parameters such as scope, maximum wind velocity, torrent intensity, characteristic quantity can also be increased and decreased according to the demand of different user to obtain different meanings
On low-level jet stream intensity.
Further, the characteristic parameter scope includes：
The lower boundary altitude range of low-level jet stream band, the coboundary altitude range of low-level jet stream band, low-level jet stream band are most
Big wind speed range, low-level jet stream band maximum wind velocity where altitude range, low-level jet stream band wind speed dropping section gradient mould
Scope and low-level jet stream band maximum wind velocity at Gaussian curvature scope.
Above-mentioned further beneficial effect：Adjusted on the basis of relevant low-level jet stream priori according to local historical data
Whole parameter, add the regional adaptability of this method.
Further, it is determined that scope and the low-level jet stream of the mould of the gradient of the wind speed dropping section of the low-level jet stream band
The process of the scope of Gaussian curvature at the maximum wind velocity of band is：
It is maximum that the history is normalized into the mould of the gradient of the wind speed dropping section of low-level jet stream band and rapid zone on Wind outline
Gaussian curvature at wind speed forms statistic histogram；
The statistic histogram is analyzed, determines the scope of the mould of the gradient and the scope of the Gaussian curvature.
Above-mentioned further beneficial effect：Gaussian curvature is the product of two principal curvatures, to reduce amount of calculation, specifies two
Principal curvatures is curved surface cuts curvature of a curve along horizontal, ordinate direction, and the scope of the mould by determining gradient and Gaussian curvature
Scope, crest can be searched out, so can more accurately detect low-level jet stream.
Further, the S4 includes：
S41, the characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while according to the spy of the low-level jet stream
Levy parameter area and construct corresponding with characteristic quantity member function, further according to the characteristic quantity and corresponding with the characteristic quantity
Member function determines member's item of the characteristic quantity, finally according to member's item of the characteristic quantity, obtains corresponding with each putting
Fraction；
S42, judges whether fraction corresponding to each point is more than pre-determined threshold thsIf being more than, the point is that the low latitude is anxious
Flow point.
Above-mentioned further beneficial effect：By above-mentioned method can with the altitude range of Primary Location low-level jet stream and when
Between scope.
Further, the S41 includes：
S411, calculate the mould and Gaussian curvature of the gradient each put in the normalized anemobiagraph of the history；
S412, the member function of the mould of the gradient corresponding with the mould of the gradient and the Gaussian curvature is constructed respectively
With the member function of the Gaussian curvature；
S413, according to the mould of the gradient, the Gaussian curvature, the gradient mould member function and the Gauss it is bent
The member function of rate, it is determined that member's item of the mould for the gradient each put and member's item of Gaussian curvature；
S414, according to member's item of the mould of the gradient and member's item of the Gaussian curvature, obtain corresponding with each putting
Fraction.
Above-mentioned further beneficial effect：By member's item of the mould of above-mentioned calculating gradient and member's item of Gaussian curvature,
The point for possessing low-level jet stream feature in image can be covered as much as possible.
Further, clustering processing includes in the S5：
Travel through each row point in the normalized anemobiagraph of the history from left to right, if it is latter row exist low-level jet stream point and
Overlapping with the low-level jet stream point of previous column in height, then the two is combined into a low-level jet stream region；Will if not overlapping
Preceding next two columns are divided into two different low-level jet stream regions, and the low-level jet stream region wherein where previous column is terminated, latter row are opened
The low-level jet stream region for beginning new.
Further, the S6 includes：
S61, according to the coboundary height average, lower boundary height average, maximum wind velocity in each low-level jet stream region
Member's item of the average of place height, the average of maximum wind velocity, obtain fraction corresponding with each region；
S62, judges whether fraction corresponding to each region is more than pre-determined threshold thoIf being more than, the region be it is described most
Whole low-level jet stream region.
Above-mentioned further beneficial effect：Low latitude can be positioned on the normalized anemobiagraph of history by above-mentioned method
Torrent region.
Further, the S61 includes：
S611, calculate the coboundary height average, lower boundary height average, maximum wind velocity in each low-level jet stream region
The average of place height, the average of maximum wind velocity；
S612, construct respectively and the coboundary height average, the lower boundary height average, maximum wind velocity place
The member function of the coboundary height average, the lower boundary are high corresponding to the average of height, the average of the maximum wind velocity
The member function of the average of height where spending the member function of average, the maximum wind velocity, the maximum wind velocity average into
Member's function；
S613, according to height where the coboundary height average, the lower boundary height average, the maximum wind velocity
Average, the average of the maximum wind velocity, the member function of the coboundary height average, the member of the lower boundary height average
Function, the member function of average of height where the maximum wind velocity, the maximum wind velocity average member function, it is determined that often
Member's item of the coboundary height average in individual region, member's item of the lower boundary height average, maximum wind velocity institute
Member's item of the average of member's item, the maximum wind velocity in the average of height；
S614, according to member's item of the coboundary height average, the lower boundary height average member's item, it is described most
Member's item of member's item of the average of big wind speed place height, the average of the maximum wind velocity, is obtained and each low-level jet stream area
Fraction corresponding to domain.
Above-mentioned further beneficial effect：Member's item, lower boundary by above-mentioned calculating coboundary height average is highly equal
Member's item of the average of height where member's item of value, maximum wind velocity, maximum wind velocity average member's item, can be as more as possible
Reject the region for not possessing low-level jet stream key feature in image in ground.
Further, this method also includes：
S7, calculate the height on the upper and lower border at each moment and the wind in the final low-level jet stream region
Fast maximum, and the height on the upper and lower border at each moment and the wind speed maximum are subjected to line respectively, obtain whole
The altitude curve and wind speed maximum curve on the upper and lower border in the individual final low-level jet stream region；
S8, construct the member function and maximum wind velocity of the width of the final low-level jet stream region low-to-medium altitude rapid zone
Member function；
S9, according to the difference of the average of the coboundary altitude curve and the average of the lower boundary altitude curve, the wind
The member function of the average of fast maximum curve, the member function of the width and the maximum wind velocity, it is anxious to obtain the low latitude
The intensity of stream.
Further, the S1 includes：
S11, barb figure is constructed, the Wind outline at each moment is obtained on barb figure；
S12, the first history Wind outline of the first preset time period is obtained on the Wind outline, according to the first preset time
Seek the first average Wind outline of the first history Wind outline in interval；
S13, the second history Wind outline of the second preset time period is obtained on the Wind outline, according to the second preset time
Seek the second average Wind outline of the second history Wind outline in interval；
S14, according to the described first average Wind outline and the second average Wind outline, determine the 3rd history Wind outline；
S15, according to the Wind outline at current time and the 3rd history Wind outline, determine that history normalizes Wind outline.
The invention further relates to a kind of system of the method for the automatic identification low-level jet stream on barb figure, the system includes：Should
System includes：Receiver, processor and output equipment；
The receiver, for receiving the Wind outline at current time；
The processor, gone through for low-level jet stream in the Wind outline according to current time low-level jet stream and preset time period
History Wind outline, determine that history normalizes Wind outline；Wind outline is normalized according to the historical data of low-level jet stream and the history, really
The characteristic parameter scope of the fixed low-level jet stream；History normalization Wind outline is chronological, obtain history normalization
Anemobiagraph, wherein the normalized anemobiagraph of the history be using the time as abscissa, highly be ordinate, speed is ordinate
Image；The characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while is joined according to the feature of the low-level jet stream
Number scope constructs member function corresponding with the characteristic quantity, and according to the characteristic quantity and member corresponding with the characteristic quantity
Function, determine low-level jet stream point；The low-level jet stream point is filtered and clustering processing, obtains multiple low-level jet stream regions；
Calculate the characteristic quantity in each low-level jet stream region, at the same according to the characteristic parameter scope of low-level jet stream construction with it is described
Member function corresponding to characteristic quantity, and according to the characteristic quantity and member function corresponding with the characteristic quantity, it is determined that final
Low-level jet stream region；
The output equipment, for exporting the characteristic parameter in the final low-level jet stream region.
Beneficial effects of the present invention：, can be in real time automatically from wind when air-ground torrent occurs by the system of the present invention
Low-level jet stream region is identified on plumage figure, and greatly reduces rate of failing to report, can also in real time export and continue on low-level jet stream
The parameters such as time, altitude range, maximum wind velocity, and then export the low-level jet stream intensity in user's meaning.
Brief description of the drawings
Fig. 1 is a kind of flow chart for the method that low-level jet stream is identified on barb figure of the embodiment of the present invention 1；
Fig. 2 is a kind of flow chart for the method that low-level jet stream is identified on barb figure of the embodiment of the present invention 4；
Fig. 3 is a kind of flow chart for the method that low-level jet stream is identified on barb figure of the embodiment of the present invention 5；
Fig. 4 is a kind of flow chart for the method that low-level jet stream is identified on barb figure of the embodiment of the present invention 8；
Fig. 5 is a kind of schematic diagram for the system that low-level jet stream is identified on barb figure of the embodiment of the present invention 10.
Embodiment
The principle and feature of the present invention are described below in conjunction with accompanying drawing, the given examples are served only to explain the present invention, and
It is non-to be used to limit the scope of the present invention.
As shown in figure 1, the embodiment of the present invention 1 is a kind of method that low-level jet stream is identified on barb figure, this method includes：
S1, according to the history Wind outline of low-level jet stream in the Wind outline of current time low-level jet stream and preset time period, really
Determine history normalization Wind outline；
S2, Wind outline is normalized according to the historical data of low-level jet stream and the history, determines the spy of the low-level jet stream
Levy parameter area；
S3, history normalization Wind outline is chronological, the normalized anemobiagraph of history is obtained, wherein described go through
The normalized anemobiagraph of history be using the time as abscissa, highly be ordinate, speed for ordinate image；
S4, the characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while according to the spy of the low-level jet stream
Levy parameter area and construct corresponding with characteristic quantity member function, and according to the characteristic quantity and corresponding with the characteristic quantity
Member function, determine low-level jet stream point；
S5, the low-level jet stream point is filtered and clustering processing, obtains multiple low-level jet stream regions；
S6, the characteristic quantity in each low-level jet stream region is calculated, while according to the characteristic parameter model of the low-level jet stream
Construction member function corresponding with the characteristic quantity is enclosed, and according to the characteristic quantity and member's letter corresponding with the characteristic quantity
Number, it is determined that final low-level jet stream region.
It should be noted that by above-mentioned method, can judge in real time from barb figure anxious with the presence or absence of low latitude
Stream, can more accurately position low-level jet stream region, duration, the height model of low-level jet stream can be checked on barb figure
Enclose, the parameter such as maximum wind velocity, torrent intensity, characteristic quantity can also be increased and decreased according to the demand of different user and obtained in different meanings
Low-level jet stream intensity.
Alternatively, characteristic parameter scope includes described in another embodiment 2：
The lower boundary altitude range of low-level jet stream band, the coboundary altitude range of low-level jet stream band, low-level jet stream band are most
Big wind speed range, low-level jet stream band maximum wind velocity where altitude range, low-level jet stream band wind speed dropping section gradient mould
Scope and low-level jet stream band maximum wind velocity at Gaussian curvature scope.
It should be noted that first, obtained according to the priori of low-level jet stream below with reference to information：Rapid zone lower boundary
Altitude range [hdr1,hdr2], rapid zone coboundary altitude range [hur1,hur2], altitude range [h where torrent maximum wind velocitymr1,
hmr2], torrent maximum wind velocity scope [vdr,vur].And wherein signified priori includes：Definition, document, meteorological personnel are to locality
Experience of low-level jet stream etc..
Secondly, respectively to the lower boundary height of rapid zone, coboundary height, most strong wind in local history low-level jet stream data
Height, maximum wind velocity do statistic histogram where fast, obtain corresponding to above-mentioned 4 parameter areas：[hdl1,hdl2]、[hul1,
hul2]、[hml1,hml2]、[vdl,vul]。
Thirdly, the parameter according to obtained from the statistic histogram that parameter and local history that priori obtains are done, really
Surely the common factor in the section of parameter is corresponded to, and it is respectively h to make 4 borders occured simultaneouslydi、hui、hmi、vdi(i=1,2).
Finally, local history low-level jet stream data are asked " the normalized Wind outline of history ", according to the normalized wind of history
Profile obtains the Gauss at the scope of the mould of the gradient of the wind speed dropping section of low-level jet stream band and the maximum wind velocity of low-level jet stream band
The scope of curvature.
Alternatively, learnt by the above embodiments 1 and embodiment 2, the low-level jet stream is determined in another embodiment 3
The scope of Gaussian curvature at the maximum wind velocity of the scope of the mould of the gradient of the wind speed dropping section of band and the low-level jet stream band
Process is：
It is maximum that the history is normalized into the mould of the gradient of the wind speed dropping section of low-level jet stream band and rapid zone on Wind outline
Gaussian curvature at wind speed forms statistic histogram；
The statistic histogram is analyzed, determines the scope of the mould of the gradient and the scope of the Gaussian curvature.
It should be noted that it is further anxious to local history low latitude on the basis of embodiment 2 in the present embodiment 3
Flow data seeks the normalized Wind outline of history, then the wind speed dropping section of rapid zone on the normalized Wind outline of history respectively
Gaussian curvature at the mould and rapid zone maximum wind velocity of gradient does statistic histogram, obtains the scope [g of the mould of gradient1,g2], it is bent
Rate scope [c1,c2]。
Alternatively, as shown in Fig. 2 on the basis of above-described embodiment 1 or embodiment 2 further in the present embodiment 4
The S4 includes：
S41, the characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while according to the spy of the low-level jet stream
Levy parameter area and construct corresponding with characteristic quantity member function, further according to the characteristic quantity and corresponding with the characteristic quantity
Member function determines member's item of the characteristic quantity, finally according to member's item of the characteristic quantity, obtains corresponding with each putting
Fraction；
S42, judges whether fraction corresponding to each point is more than pre-determined threshold thsIf being more than, the point is that the low latitude is anxious
Flow point.
It should be noted that in the present embodiment 4 step S41 be " the normalized Wind outline of history " is chronological,
Obtain using the time as abscissa, be highly ordinate, speed for ordinate image.Three coordinates are discrete coordinates, then
On the normalized Wind outline figure of this history, the mould and Gaussian curvature of the gradient of each point in image are calculated.Gaussian curvature is two
The product of principal curvatures, to reduce amount of calculation, it is that curved surface cuts curvature of a curve along horizontal, ordinate direction to specify two principal curvatures.This
It is usually what is set up at crest summit, is advantageous to torrent identification.
Step S42, when using fuzzy logic identification low-level jet stream, first by the mould and curvature of the gradient of each point in image
The close number of dimensionless, value is converted into by " member function ", then by its weighted average, finally sets threshold judgement.Due to " going through
The part curvature of non-low-level jet stream is small on the normalized Wind outline of history ", gradient mould is big, therefore by curvature when setting member function
The small value of the mould of big value and gradient is as low-level jet stream.
Alternatively, as shown in figure 3, further being wrapped on the basis of above-described embodiment 4 described in the present embodiment 5 in S41
Include：
S411, calculate the mould and Gaussian curvature of the gradient each put in the normalized anemobiagraph of the history；
S412, the member function of the mould of the gradient corresponding with the mould of the gradient and the Gaussian curvature is constructed respectively
With the member function of the Gaussian curvature；
S413, according to the mould of the gradient, the Gaussian curvature, the gradient mould member function and the Gauss it is bent
The member function of rate, it is determined that member's item of the mould for the gradient each put and member's item of Gaussian curvature；
S414, according to member's item of the mould of the gradient and member's item of the Gaussian curvature, obtain corresponding with each putting
Fraction.
It should be noted that being step S41 in the present embodiment 5, first calculate every in the normalized Wind outline model of history
The mould and Gaussian curvature of the gradient of individual low-level jet stream point；
Step S412, make the member function M (g) of the mould of gradient and the member function M (c) of curvature：
In formula, kgAnd kcIt is set by the user, g2And c1Definition refer to the mould and curvature of gradient.The purpose of this step is to the greatest extent may be used
The point for possessing low-level jet stream feature in image can be covered more.Therefore, first, weighted average uses arithmetic mean；Second, member
The value of function from 1 drop to 1/2 drop to again 0 expression from low-level jet stream judgement set up to judgement whether sets up can again to can not adjudicate
It is vertical；3rd, M (g) is in g>G2, M (c) are in c<Just start to decline from 1 during c1.
Step S413, for each point (x, y) in image, merge the mould and curvature of gradient：
In formula, g (x, y), c (x, y), V (x, y) are the mould of the gradient of each point, curvature, speed, a in imagegAnd acBy user
Setting, is just set to 0.5.Because the calculating of " history Wind outline " has combined, without two kinds of situations of low-level jet stream, if therefore certain wind exterior feature
Without low-level jet stream on line, then the wind speed corresponding to it on " the normalized Wind outline of history " is most is less than 0；Conversely, then it is corresponding
" the normalized Wind outline of history " on low-level jet stream region wind speed necessarily be more than 0.So directly by the point of V (x, y)≤0
The S (x, y) of (x, y), which is set to 0, can reduce follow-up amount of calculation, reduce feature and low latitude of the non-low-level jet stream area due to gradient and curvature
Mistake is identified caused by torrent is close.
Step S414, gating limit ths, S (x, y)>thsPoint (x, y) judgement be low-level jet stream region, wherein S (x, y) is
Refer to fraction corresponding to each low-level jet stream point.
Alternatively, obtain described in the present embodiment 6 that clustering processing includes in S5 by above-described embodiment：
The low-level jet stream point judged by accident in the low-level jet stream point is rejected using Density Weighted method；
Wherein, the Density Weighted method refers to using rectangular window as center neighborhood of a point, and thresholding th is exceeded in rectangular windows's
The total amount of all low-level jet stream points, which is done, in the quantity and window of low-level jet stream point compares, and output center is used as multiplied by using the amplitude of central point
The amplitude of point, rejects the low-level jet stream point judged by accident in the low-level jet stream point.
It should be noted that " Density Weighted " method is used to reject the erroneous judgement region being sporadicly distributed in the present embodiment." density
Method is for weighting "：Using rectangular window as center neighborhood of a point, thresholding th is exceeded in windowsPoint quantity and the total amount at window midpoint do
Than multiplied by the amplitude of the point using the amplitude of central point as output center.So, the output amplitude of the point inside low-level jet stream area is not
Become, output amplitude reduction half of boundary point or so, the output amplitude of the point in fragmentary erroneous judgement region declines to a great extent, you can sets
Thresholding thdFilter out low-level jet stream region.
And cluster and refer to cluster low-level jet stream point, its method is：The normalized wind speed of the history is traveled through from left to right
Each row point in figure, if low-level jet stream point and overlapping with the low-level jet stream point of previous column in height be present in latter row,
The two is combined into a low-level jet stream region；Preceding next two columns are divided into two different low-level jet stream regions if not overlapping, wherein
Low-level jet stream region where previous column terminates, latter row start new low-level jet stream region；It is anxious that if low latitude is not present in latter row
The then low-level jet stream region where previous column be present and terminate in flow point and previous column；If there is low-level jet stream point and previous column in latter row
In the absence of then starting new region.
Secondly, for every piece of low-level jet stream region in image, the coboundary height average of each moment torrent in region is sought
hd, lower boundary height average hu, height where maximum wind velocity average hm, maximum wind velocity average vm。
Then construct it is above-mentioned in each parameter member function：
Three height use M (h) member functions, and wind speed uses M (vm) member function.In formula, h=hd(hu、hm) when, h1
=hd1(hu1、hm1)、h2=hd2(hu2、hm2)、kh1=kd1(ku1、km1)、kh2=kd2(ku2、km2).M (h) kh1(i.e. kd1、ku1、
km1) and kh2(i.e. kd2、ku2、km2) be set by the user.This step purpose is to reject in image not possess low-level jet stream as much as possible
The point of key feature.Therefore, first, weighted average uses geometric average；Second, the value of member function drops to 1/2 from 1 and dropped to again
Whether 0 expression sets up from low-level jet stream judgement establishment to judgement can be invalid to adjudicating again.
For every piece of low-level jet stream region in image, by hu、hd、hm、vmBring following formula fusion into：
In formula, i represents i-th piece of low-level jet stream region, βhd、βhu、βhm、βvmIt is set by the user, is just set to βhd=βhu=βvm
=0.3, βhm=0.1.As long as it can be seen that there is one to export 0 after its member function in three height, a speed, then O (i)=
0, i.e. i-th piece of region is not adjudicated as low-level jet stream.
Finally, gating limit tho, O (i)>thoRegion unit i judgement be low-level jet stream.)
Alternatively, obtaining the S6 described in the present embodiment 7 by the above embodiments includes：
S61, according to the coboundary height average, lower boundary height average, maximum wind velocity in each low-level jet stream region
Member's item of the average of place height, the average of maximum wind velocity, obtain fraction corresponding with each region；
S62, judges whether fraction corresponding to each region is more than pre-determined threshold thoIf being more than, the region be it is described most
Whole low-level jet stream region.
Alternatively, as shown in figure 4, further this method is also wrapped in the present embodiment 8 on the basis of above-described embodiment 2
Include：
S7, calculate the height on the upper and lower border at each moment and the wind in the final low-level jet stream region
Fast maximum, and the height on the upper and lower border at each moment and the wind speed maximum are subjected to line respectively, obtain whole
The altitude curve and wind speed maximum curve on the upper and lower border in the individual final low-level jet stream region；
S8, construct the member function and maximum wind velocity of the width of the final low-level jet stream region low-to-medium altitude rapid zone
Member function；
S9, according to the difference of the average of the coboundary altitude curve and the average of the lower boundary altitude curve, the wind
The member function of the average of fast maximum curve, the member function of the width and the maximum wind velocity, it is anxious to obtain the low latitude
The intensity of stream.
It should be noted that in the step S7 of the present embodiment 8, for every piece of low-level jet stream region in image, by region
In most forward time start output time as torrent；Torrent is used as if the time in region most rearward is not current time
Terminate output time, otherwise represent that current time is just influenceed by low-level jet stream, the end time wouldn't export.Output area it is upper,
Altitude curve where lower boundary altitude curve and maximum wind velocity, export the average of every curve.Each moment in output area
Line, the line of wind speed maximum of wind speed average.Export the wind speed average of bulk portion and the average of wind speed maximum line.
In step S8, the member function M (h of the width of low-level jet stream band are constructedu-hd) and maximum wind velocity member function M
(vm)：
This step purpose is that the low-level jet stream that will be had confirmed that provides intensity by user's request.Therefore, first, weighted average uses
Arithmetic mean；Second, the value of member function drops to 0 expression torrent intensity from most by force to most weak from 1；3rd, by M (hu-hd)=M
(h) in hu1-hd21 is got, is successively decreased to both sides, M2(vm) in vm>vuWhen just reach 1.
S9, for every piece of low-level jet stream region in image, make the average of upper and lower boundary height curve in step S8
Difference represents torrent bandwidth, brings the average of the curve of height where itself and maximum wind velocity into following formula, obtains torrent intensity：
I (i)=αhM(h(i))+αvM2(vm(i))
In formula, i represents i-th piece of low-level jet stream region, ahAnd avIt is set by the user, is just set to 0.5.
If identification low-level jet stream has other purposes (such as predicting precipitation and monitoring wind shear), can be by increased feature
Amount (such as wind speed highly ratio) can also optionally remove some weighted term to eliminate corresponding spy by incorporating I (i) from I (i)
The influence that sign amount is judged torrent intensity.
Alternatively, further S1 includes described in the present embodiment 9 on the basis of above-described embodiment 1 or embodiment 2：
S11, barb figure is constructed, the Wind outline at each moment is obtained on barb figure；
S12, the first history Wind outline of the first preset time period is obtained on the Wind outline, according to the first preset time
Seek the first average Wind outline of the first history Wind outline in interval；
S13, the second history Wind outline of the second preset time period is obtained on the Wind outline, according to the second preset time
Seek the second average Wind outline of the second history Wind outline in interval；
S14, according to the described first average Wind outline and the second average Wind outline, determine the 3rd history Wind outline；
S15, according to the Wind outline at current time and the 3rd history Wind outline, determine that history normalizes Wind outline.
It should be noted that it is that earth's surface frictional force reduces with height in the present embodiment 9, old place turns wind speed with altitude increasing
Add, therefore the coboundary wind speed of low-level jet stream band is commonly greater than lower boundary wind speed.This not only reduces cutting for low-level jet stream coboundary
Become, add the difficulty that rapid zone is extracted from Wind outline, also increase the probability that high-altitude strong wind is mistaken for torrent.Therefore must be first
The influence of earth's surface friction is eliminated from the Wind outline at current time, torrent identification could be started.Removing method is as follows：
S11, construct using the time as abscissa, be highly the barb figure of ordinate, the wind at each moment is extracted from barb figure
Profile.
S12, for all Wind outlines of last year, seek average profile of the every month per half an hour：
In formula, Pi,j,kFor the i-th moment, jth day, the Wind outline of the kth moon,For i-th0Wind outline before and a half hours
Bar number, makes M0=0, NkFor the number of days of the kth moon.
S13, with mode is solved in S12, the previous day at current time is asked to the first 30 days average profiles per half an hour
S14, willWithWeighting obtains " history Wind outline "：
In formula, weight λfAnd λn0.5 just is set to, can be manually adjusted.If climatic year change in recent years is obvious, or last year works as
Preceding middle of the month low-level jet stream accumulation duration exceedes two weeks, then reduces λf；If now just when seasonal variations, or the previous day at current time
Duration is accumulated more than 15 days to preceding 30 days low-to-medium altitude torrent, then reduces λn。
S15, make the difference to obtain " the normalized Wind outline of history " with the Wind outline and " history Wind outline " at current time：
Adjust λfAnd λnAfterwards, " history Wind outline "Low-to-medium altitude torrent does not account for leading role, turns with can preferably embodying
Wind is with highly increased characteristic.Then cut from the Wind outline at current time " history Wind outline ", you can more significantly eliminate
The influence of earth's surface friction.
Correspondingly, as shown in figure 5, further relating to one kind automatic identification low-level jet stream on barb figure in the embodiment of the present invention 10
Method system, the system includes：Receiver, processor and output equipment；
The receiver, for receiving the Wind outline at current time；
The processor, gone through for low-level jet stream in the Wind outline according to current time low-level jet stream and preset time period
History Wind outline, determine that history normalizes Wind outline；Wind outline is normalized according to the historical data of low-level jet stream and the history, really
The characteristic parameter scope of the fixed low-level jet stream；History normalization Wind outline is chronological, obtain history normalization
Anemobiagraph, wherein the normalized anemobiagraph of the history be using the time as abscissa, highly be ordinate, speed is ordinate
Image；The characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while is joined according to the feature of the low-level jet stream
Number scope constructs member function corresponding with the characteristic quantity, and according to the characteristic quantity and member corresponding with the characteristic quantity
Function, determine low-level jet stream point；The low-level jet stream point is filtered and clustering processing, obtains multiple low-level jet stream regions；
Calculate the characteristic quantity in each low-level jet stream region, at the same according to the characteristic parameter scope of low-level jet stream construction with it is described
Member function corresponding to characteristic quantity, and according to the characteristic quantity and member function corresponding with the characteristic quantity, it is determined that final
Low-level jet stream region；
The output equipment, for exporting the characteristic parameter in the final low-level jet stream region.
It should be noted that the method in the above embodiments 1-9 can be applied to the system in the present embodiment 10,
All it is one-to-one in its technical scheme.By the system of the present invention, low latitude can be identified from barb figure automatically in real time
Torrent region, and greatly reduce rate of failing to report, can also export in real time the duration on low-level jet stream, altitude range,
The parameters such as maximum wind velocity, and then export the low-level jet stream intensity in user's meaning.
In this manual, identical embodiment or example are necessarily directed to the schematic representation of above-mentioned term.Moreover,
Specific features, structure, material or the feature of description can be in any one or more embodiments or example in an appropriate manner
With reference to.In addition, in the case of not conflicting, those skilled in the art can implement the difference described in this specification
Example or example and the feature of different embodiments or example are combined and combined.
The foregoing is only presently preferred embodiments of the present invention, be not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent substitution and improvements made etc., it should be included in the scope of the protection.
Claims (10)
- A kind of 1. method of the automatic identification low-level jet stream on barb figure, it is characterised in that this method includes：S1, according to the history Wind outline of low-level jet stream in the Wind outline of current time low-level jet stream and preset time period, it is determined that going through History normalizes Wind outline；S2, Wind outline is normalized according to the historical data of low-level jet stream and the history, determine the feature ginseng of the low-level jet stream Number scope；S3, history normalization Wind outline is chronological, the normalized anemobiagraph of history is obtained, wherein the history is returned One change anemobiagraph be using the time as abscissa, highly be ordinate, speed for ordinate image；S4, the characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while joined according to the feature of the low-level jet stream Number scope constructs member function corresponding with the characteristic quantity, and according to the characteristic quantity and member corresponding with the characteristic quantity Function, determine low-level jet stream point；S5, the low-level jet stream point is filtered and clustering processing, obtains multiple low-level jet stream regions；S6, the characteristic quantity in each low-level jet stream region is calculated, while according to the characteristic parameter scope structure of the low-level jet stream Make member function corresponding with the characteristic quantity, and according to the characteristic quantity and member function corresponding with the characteristic quantity, really Fixed final low-level jet stream region.
- 2. according to the method for claim 1, it is characterised in that the characteristic parameter scope includes：The lower boundary altitude range of low-level jet stream band, the coboundary altitude range of low-level jet stream band, the most strong wind of low-level jet stream band Fast scope, low-level jet stream band maximum wind velocity where altitude range, low-level jet stream band wind speed dropping section gradient mould model Enclose the scope with the Gaussian curvature at the maximum wind velocity of low-level jet stream band.
- 3. according to the method for claim 2, it is characterised in that determine the gradient of the wind speed dropping section of the low-level jet stream band The scope of mould and the maximum wind velocity of the low-level jet stream band at the process of scope of Gaussian curvature be：The history is normalized to the mould and rapid zone maximum wind velocity of the gradient of the wind speed dropping section of low-level jet stream band on Wind outline The Gaussian curvature at place forms statistic histogram；The statistic histogram is analyzed, determines the scope of the mould of the gradient and the scope of the Gaussian curvature.
- 4. method according to claim 1 or 2, it is characterised in that the S4 includes：S41, the characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while joined according to the feature of the low-level jet stream Number scope constructs member function corresponding with the characteristic quantity, further according to the characteristic quantity and member corresponding with the characteristic quantity Function determines member's item of the characteristic quantity, finally according to member's item of the characteristic quantity, obtains with each putting corresponding fraction；S42, judges whether fraction corresponding to each point is more than pre-determined threshold thsIf being more than, the point is the low-level jet stream point.
- 5. according to the method for claim 4, it is characterised in that the S41 includes：S411, calculate the mould and Gaussian curvature of the gradient each put in the normalized anemobiagraph of the history；S412, member function and the institute of the mould of the gradient corresponding with the mould of the gradient and the Gaussian curvature are constructed respectively State the member function of Gaussian curvature；S413, according to the mould of the gradient, the Gaussian curvature, the gradient mould member function and the Gaussian curvature Member function, it is determined that member's item of the mould for the gradient each put and member's item of Gaussian curvature；S414, according to member's item of the mould of the gradient and member's item of the Gaussian curvature, obtain with each putting corresponding minute Number.
- 6. according to any described methods of claim 1-3, it is characterised in that clustering processing includes in the S5：Each row point in the normalized anemobiagraph of the history is traveled through from left to right, if latter row have low-level jet stream point and in height Overlapping with the low-level jet stream point of previous column on degree, then the two is combined into a low-level jet stream region；If not overlapping by before and after Two row are divided into two different low-level jet stream regions, and the low-level jet stream region wherein where previous column terminates, latter row start newly Low-level jet stream region.
- 7. according to the method for claim 2, it is characterised in that the S6 includes：S61, according to where the coboundary height average, lower boundary height average, maximum wind velocity in each low-level jet stream region Member's item of the average of height, the average of maximum wind velocity, obtain fraction corresponding with each region；S62, judges whether fraction corresponding to each region is more than pre-determined threshold th。If being more than, the region is described final Low-level jet stream region.
- 8. according to the method for claim 2, it is characterised in that this method also includes：S7, the height on the upper and lower border at each moment and the wind speed maximum in the final low-level jet stream region are calculated, And the height on the upper and lower border at each moment and the wind speed maximum are subjected to line respectively, obtain it is whole it is described most The altitude curve and wind speed maximum curve on the upper and lower border in whole low-level jet stream region；S8, construct the member function of width and the member of maximum wind velocity of the final low-level jet stream region low-to-medium altitude rapid zone Function；S9, according to the difference of the average of the coboundary altitude curve and the average of the lower boundary altitude curve, the wind speed most Average, the member function of the width and the member function of the maximum wind velocity of big value curve, obtain the low-level jet stream Intensity.
- 9. method according to claim 1 or 2, it is characterised in that the S1 includes：S11, barb figure is constructed, the Wind outline at each moment is obtained on barb figure；S12, the first history Wind outline of the first preset time period is obtained on the Wind outline, according to the first prefixed time interval Seek the first average Wind outline of the first history Wind outline；S13, the second history Wind outline of the second preset time period is obtained on the Wind outline, according to the second prefixed time interval Seek the second average Wind outline of the second history Wind outline；S14, according to the described first average Wind outline and the second average Wind outline, determine the 3rd history Wind outline；S15, according to the Wind outline at current time and the 3rd history Wind outline, determine that history normalizes Wind outline.
- 10. a kind of system of the automatic identification low-level jet stream on barb figure, it is characterised in that utilize any institutes of claim 1-9 The method stated, the system include：Receiver, processor and output equipment；The receiver, for receiving the Wind outline at current time；The processor, the history wind for low-level jet stream in the Wind outline according to current time low-level jet stream and preset time period Profile, determine that history normalizes Wind outline；According to the historical data of low-level jet stream and history normalization Wind outline, institute is determined State the characteristic parameter scope of low-level jet stream；History normalization Wind outline is chronological, obtain the normalized wind of history Speed figure, wherein the normalized anemobiagraph of the history be using the time as abscissa, highly be ordinate, speed for ordinate figure Picture；The characteristic quantity of each point in the normalized anemobiagraph of the history is calculated, while according to the characteristic parameter model of the low-level jet stream Construction member function corresponding with the characteristic quantity is enclosed, and according to the characteristic quantity and member's letter corresponding with the characteristic quantity Number, determines low-level jet stream point；The low-level jet stream point is filtered and clustering processing, obtains multiple low-level jet stream regions；Meter The characteristic quantity in each low-level jet stream region is calculated, while according to the characteristic parameter scope of low-level jet stream construction and the spy Member function corresponding to sign amount, and according to the characteristic quantity and member function corresponding with the characteristic quantity, it is determined that final is low Empty torrent region；The output equipment, for exporting the characteristic parameter in the final low-level jet stream region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710765185.9A CN107563436B (en) | 2017-08-30 | 2017-08-30 | A kind of method and system of the automatic identification low-level jet stream on barb figure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710765185.9A CN107563436B (en) | 2017-08-30 | 2017-08-30 | A kind of method and system of the automatic identification low-level jet stream on barb figure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107563436A true CN107563436A (en) | 2018-01-09 |
| CN107563436B CN107563436B (en) | 2019-11-15 |
Family
ID=60978384
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710765185.9A Active CN107563436B (en) | 2017-08-30 | 2017-08-30 | A kind of method and system of the automatic identification low-level jet stream on barb figure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107563436B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109597079A (en) * | 2018-12-11 | 2019-04-09 | 北京无线电测量研究所 | Low-level jet stream recognition methods and device based on wind profile radar |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130325213A1 (en) * | 2012-05-31 | 2013-12-05 | Optical Air Data Systems, Llc | LDV for Airdrops |
| EP2884300A1 (en) * | 2012-08-09 | 2015-06-17 | Mitsubishi Electric Corporation | Signal processing device and signal processing method in wind profiler |
| CN105068075A (en) * | 2015-06-30 | 2015-11-18 | 江苏省气象科学研究所 | Near-surface gale calculation method |
| CN105607063A (en) * | 2016-01-05 | 2016-05-25 | 北京无线电测量研究所 | Detection method and system for low-altitude wind shear at airport |
| CN106779201A (en) * | 2016-12-08 | 2017-05-31 | 贵州电网有限责任公司电力科学研究院 | A kind of wind speed modeling method of consideration plateau mountain area blower fan position |
-
2017
- 2017-08-30 CN CN201710765185.9A patent/CN107563436B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130325213A1 (en) * | 2012-05-31 | 2013-12-05 | Optical Air Data Systems, Llc | LDV for Airdrops |
| EP2884300A1 (en) * | 2012-08-09 | 2015-06-17 | Mitsubishi Electric Corporation | Signal processing device and signal processing method in wind profiler |
| CN105068075A (en) * | 2015-06-30 | 2015-11-18 | 江苏省气象科学研究所 | Near-surface gale calculation method |
| CN105607063A (en) * | 2016-01-05 | 2016-05-25 | 北京无线电测量研究所 | Detection method and system for low-altitude wind shear at airport |
| CN106779201A (en) * | 2016-12-08 | 2017-05-31 | 贵州电网有限责任公司电力科学研究院 | A kind of wind speed modeling method of consideration plateau mountain area blower fan position |
Non-Patent Citations (1)
| Title |
|---|
| 吴君: "《利用风廓线雷达资料分析气旋暴雨与低空急流的关系》", 《安徽农业科学》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109597079A (en) * | 2018-12-11 | 2019-04-09 | 北京无线电测量研究所 | Low-level jet stream recognition methods and device based on wind profile radar |
| CN109597079B (en) * | 2018-12-11 | 2023-03-21 | 北京无线电测量研究所 | Low-altitude torrent identification method and device based on wind profile radar |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107563436B (en) | 2019-11-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106910185B (en) | A kind of DBCC disaggregated model construction method based on CNN deep learning | |
| Wu et al. | A novel color image segmentation method and its application to white blood cell image analysis | |
| CN102496013B (en) | Chinese character segmentation method for off-line handwritten Chinese character recognition | |
| CN107844785B (en) | A kind of method for detecting human face based on size estimation | |
| CN106096561A (en) | Infrared pedestrian detection method based on image block degree of depth learning characteristic | |
| CN102902967A (en) | Method for positioning iris and pupil based on eye structure classification | |
| CN106778664A (en) | The dividing method and its device of iris region in a kind of iris image | |
| CN105069477A (en) | Method for AdaBoost cascade classifier to detect image object | |
| CN102645679A (en) | Mesocyclone identification method based on Doppler radar echo images | |
| CN103559508B (en) | A kind of based on continuous Adaboost video vehicle detection method | |
| CN106295591A (en) | Gender identification method based on facial image and device | |
| CN101751679A (en) | Sorting method, detecting method and device of moving object | |
| CN110610142A (en) | Airport flight waveform similarity measurement method based on trend distance and spectral clustering | |
| CN107832730A (en) | Improve the method and face identification system of face recognition accuracy rate | |
| CN110781873A (en) | Driver fatigue grade identification method based on bimodal feature fusion | |
| CN109101108A (en) | Method and system based on three decision optimization intelligence cockpit human-computer interaction interfaces | |
| CN110264241A (en) | A kind of advertisement dispensing evaluation optimal sysytem based on big data | |
| CN101853397A (en) | Bionic human face detection method based on human visual characteristics | |
| CN104156945A (en) | Method for segmenting gray scale image based on multi-objective particle swarm optimization algorithm | |
| CN103310227A (en) | Automatic window width and window level extraction method based on neural network | |
| CN103810500A (en) | Place image recognition method based on supervised learning probability topic model | |
| CN109978168A (en) | The origin cause of formation automated reasoning method and system of landing again based on timing QAR parameter curve cluster | |
| CN110660070A (en) | Rice vein image extraction method and device | |
| CN103020953A (en) | Segmenting method of fingerprint image | |
| CN107563436B (en) | A kind of method and system of the automatic identification low-level jet stream on barb figure |
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 |