CN104462820B - A kind of method of power network shaft tower coordinate error detection - Google Patents
A kind of method of power network shaft tower coordinate error detection Download PDFInfo
- Publication number
- CN104462820B CN104462820B CN201410752980.0A CN201410752980A CN104462820B CN 104462820 B CN104462820 B CN 104462820B CN 201410752980 A CN201410752980 A CN 201410752980A CN 104462820 B CN104462820 B CN 104462820B
- Authority
- CN
- China
- Prior art keywords
- point
- error
- mrow
- shaft tower
- msub
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Analysis (AREA)
- Navigation (AREA)
Abstract
The present invention provides a kind of overhead power transmission line pole tower coordinate classification error-detecting method, comprises the following steps:1) overhead line structures error situation is divided into clearly error type and indefinite error type, altogether 7 group;2) fill a vacancy entering row interpolation by error detection circuit missing shaft tower coordinate;3) clearly error type error detection rule is set up;4) using error detection rule, draw by the clearly error shaft tower of error detection circuit, record shaft tower error type is gone forward side by side row interpolation correction;5) error rate that indefinite error type is carried out to the circuit coordinate after correction is examined, and is drawn by the error probability of all kinds of indefinite error types of error detection shaft tower;6) expert's given threshold is more than by the indefinite wrong shaft tower type error probability of error detection circuit, is judged as the shaft tower coordinate mistake, misregistration type;7) comprehensive descision, completes all shaft tower error detection work of circuit.The inventive method, an error probability can be drawn to each shaft tower, indefinite gray area is preferably handled.Other this method can summarize new error type, be finally reached the wrong shaft tower coordinate for detecting the overwhelming majority.
Description
Technical field
The present invention relates to power network shaft tower coordinate measurement and correction technique field, and in particular to a kind of power network shaft tower coordinate measurement
Method.
Background technology
In order to timely search line fault point, the workload of line walking is reduced, it is ensured that the reliability service of power network, it is necessary to carry
For more accurately shaft tower coordinate.But it is due to various mistakes and error, causes tower spotting inaccurate.This paper main purposes are
The shaft tower coordinate of possible mistake is verified, line fault tower spotting is solved the problems, such as.
Power network shaft tower can produce mistake, by analysis, mainly have following two situations to cause when longitude and latitude is recorded
Error in data:
1st, gps data Wrong localization
The gps coordinate of same shaft tower, has " degree every minute and second ", " degree point " and " decimal point " three kinds of forms to show in GPS measuring machines.
Because the gps data of " degree every minute and second ", " degree point " records inconvenience in computer, part staff, will " degree in logging data
Every minute and second ", " degree point " form are changed without formula and are recorded as " decimal point " form with regard to mistake, may will result in difference tens
Kilometer.
2nd, survey crew's measurement error
Need to measure each base shaft tower using conventional method measurement shaft tower gps coordinate, and require staff
Must stand to the center of shaft tower, but be due to transmission line of electricity shaft tower radix it is many, and most of shaft tower is located at high mountain, mound
The traffic difficulties such as mound area, it is difficult to ensure that every survey crew can measure shaft tower GPS coordinate data as requested.In addition, bar
The gps coordinate of tower reaches after decimal point 5, and survey crew can " least bit of difference, loses it once imprudence records wrong one digit number, also
A thousand li ", the gps data for causing part measurement to be returned is inaccurate.
The content of the invention
It is an object of the invention to provide a kind of method of power network shaft tower coordinate error detection, the present invention draws shaft tower by probing into
Clearly error type and indefinite error type, using classification error-detecting method, design program (rule) detects clearly to malfunction
Power network shaft tower coordinate detects indefinite error power network shaft tower coordinate with similarity, and that realizes power network shaft tower coordinate goes out false retrieval
Survey.
The purpose of the present invention can be realized by following technical measures:
A kind of overhead power transmission line pole tower coordinate classification error-detecting method, comprises the following steps:
The error situation of overhead line structures is divided into clearly error type and indefinite error type by S101;
S102 carries out average interpolation to the shaft tower coordinate lacked in power network shaft tower to be detected and filled a vacancy;
S103 sets up clearly error type error detection rule;
S104 is drawn by the clearly error shaft tower of error detection circuit, recording bar using the clearly error type error detection rule
Tower, which malfunctions, type and to carry out the average interpolation of the step S102 and fills a vacancy operation;
S105 average interpolation is filled a vacancy after circuit coordinate carry out it is indefinite error type detection, draw by error detection shaft tower
It is all kinds of it is indefinite error types error probability;
The indefinite wrong shaft tower type error probability that S106 is detected power network shaft tower is more than expert's given threshold, is judged as
The shaft tower coordinate mistake, record error type;
S107 completes all bars of circuit according to clearly error type and indefinite error type detection result, comprehensive descision
Tower error detection works.
Regard each overhead line structures as a point in the step S101, then the clearly error type includes four
Kind:Repeat point, isolated point, shift point, fallibility flex point;Indefinite error type includes three kinds:Similar isolated point error, angle go out
It is wrong, apart from long error.
The deletion condition of described step S102 missing shaft tower coordinates includes four kinds:Whole piece power network line missing, grid line
Road intercalary delection, power network line start missing, power network line and terminate missing, and average interpolation is carried out respectively and is filled a vacancy.Assuming that shaft tower
Actual number is N, and its corresponding average interpolation, which is filled a vacancy, operates calculating process as follows:
S102-1 whole pieces circuit all lack or whole piece circuit only one of which point situation, the benefit value meaning of such case is not
Greatly, it is inherent shortcoming, does not process.
S102-2 middle lines have lacked n (n=1,2 ... N-2) point, if the previous point coordinates of missing section is (x1,y1), lack section
Latter point coordinates is (x2,y2), then the coordinate of middle i-th of missing point is:
S102-3 circuits start to have lacked n (n=1,2 ... N-2) point, if first point for having data is (x after missing section1,
y1), it is (x to lack second point after section2,y2), then i-th point of the coordinate lacked is
xi=x1-(x2-x1)*i,yi=y1-(y2-y1) * i, i=1,2 ..., n.
S102-4 circuit latter ends have lacked n (n=1,2 ... N-2) point, if first point for having data is (x before missing section1,
y1), it is (x to lack second point before section2,y2), then i-th point of the coordinate lacked is
xi=x1+(x2-x1)*i,yi=y1+(y2-y1) * i, i=1,2 ..., n.
The described step S103 type error detections that clearly malfunction include repeating point (i.e. excessively similitude) debugging, isolated point looking into
Wrong, drift point debugging and fallibility flex point debugging, particular content are as follows:
S103-1 repeats point (i.e. excessively similitude) debugging, including:
S103-1-1 is (x provided with two point (shaft tower) coordinates1,y1), (x2,y2), first to whole piece line scanning one time, such as
The absolute value of the distance of fruit current point and next point is less than 0.000001, then marks next point to attach most importance to complex point;
S103-1-2 will be marked as repeat point point, carry out as step S102 average interpolation fill a vacancy operation-.
S103-2:Isolated point debugging, including:
S103-2-1 extracts 5 attributes of every circuit interlude point:
Every circuit, which is removed, starts 2 points of each points with terminating interlude after 2 points, there is 5 attributes:L1 is itself and the left side
The distance of consecutive points, L2 is its distance with the right consecutive points, the distance for the left adjoint point that L3 is it with being separated by a bit, L4 for its with
It is separated by the distance of the right adjoint point of a bit, angle θ is its angle with the line of left and right consecutive points.Due to 2 points of the beginning of every circuit
With terminating at 2 points, without attribute, it is believed that they do not isolate.
S103-2-2 takes maximum in L1, L2 to be L5, and it is L6 to take maximum in L3, L4.
S103-2-3 is less than 60 degree (being threshold value, self controllable system) as angle θ, and L5 is more than 2000 meters (being threshold value, self controllable system),
When L6 is more than 2000 meters (being threshold value, self controllable system), it is believed that the point is clear and definite isolated point.
S103-2-4 is filled a vacancy using the average difference in the step S102 operates amendment isolated point.
S103-3:Drift point debugging, including:The feature of S103-3-1 shift points clear and definite first, shift point is one section (i.e. whole
Section drift) occur, in power network shaft tower, there may be many drift sections, and may be comprising 2 or multiple points in each drift section
Shift point.Extract the attribute L1, L2 of each measuring point to be checked in every power network line;Due to every circuit beginning a little with end
The a little not attribute, it is not shift point to assert it.
S103-3-2 is as some point B to be detected, and its left adjoint point is A, and right adjoint point is C, if the attribute L1 of B points>3000
(being arrange parameter, self controllable system), L2<When 2000 (being arrange parameter, self controllable system), it is believed that point B is the starting of shift point
Point, the coordinate numbering for recording it is mark.
Next S103-3-3 sentences the subsequent point C of breakpoint B, judges that whether C points arrive subsequent point D distance less than 2000 again
(being arrange parameter, self controllable system);Continued executing with if it has not, returning to S103-3-2 to C points.
If it is, judging this section of shift point number, it is specifically:When CD distances are less than 2000, and AD distances subtract AC apart from small
When the 30% of AC distances, C points are shift point, and n is used for recording shift point number, whenever a point is judged for shift point more, n
Plus 1.Continue next point D detection, the detection of upper two conditions is repeated, when the two conditions are met when different, to next
Point returns to S103-3-1 and continued executing with.
The data that S103-3-4 is finally recorded are a two-dimensional array, and array has 2 values per a line, respectively with starting
The subscript of the point of point mark marks, the point number n of drift.Average is carried out with the formula of the step S102-2 according to this record
Interpolation, which fills a vacancy to operate, corrects the shift point.Because the point number of drift is n, and shift point is only possible in section, therefore is used
S102-2 method amendment.
S103-4:Fallibility flex point debugging, including:
S103-4-1 fallibility flex points only appear in interlude, and Origin And Destination will not be fallibility flex point, and fallibility record is every
3 attributes that bar circuit interlude is each put, respectively angle θ, k1、k2.Remember r1For its line segment slope with left adjoint point line, r2
For its line segment slope with right adjoint point line, and angle k is calculated by slope1And k2, k1=arctan (r1),k2=arctan
(r2)。
S103-4-2 works as angle θ<60 ° (being threshold value, self controllable system), angle k1With k2Poor absolute value be less than 30 ° (be threshold
Value, self controllable system) when, the point is judged for fallibility flex point, must be point error.
S103-4-3 is filled a vacancy operation using the average interpolation in the step S102.
Described S105 average interpolation is filled a vacancy after circuit coordinate carry out it is indefinite error type detection, draw tested
The error probability of all kinds of indefinite error types of wrong shaft tower, the detection of indefinite error type includes:Similar isolated point error
Error rate detection, angle error point error rate examine, apart from it is long error point error rate detection comprise the following steps that:
S105-1:The error rate detection of similar isolated point error, including:
S105-1-1 determines attribute, extracts every circuit and removes 5 category of 2 points of the beginning with terminating interlude point after 2 points
Property, it is respectively:L1, L2, L3, L4, angle θ;Derivative attribute:L5、L6.
The isolated point sample of 10 errors of artificial selection, extracts isolated point corresponding attribute angle θ, L5, L6.
S105-1-2, in order to avoid the attribute θ, L5, L6 that extract have extremum interference, by attribute discretization, i.e., to category
Property value classification represent, such as attribute L5, with 500 meters for step-length, it is different it is interval in a different value represent this
The size of value, more than ten thousand metres, is uniformly given as ten thousand metres in interval.Different attribute can have different specific discretization sides
Method.
The property value of measuring point to be checked also carries out discretization, obtains attribute θ, L5, L6 of each measuring point to be checked
It is expressed as a 3-dimensional vector (X1,X2,X3), attribute θ, L5, L6 of each sample in 10 error isolated point samples
One (X of correspondence1i,X2i,X3i) (i=1,2 ... 10 are 10 error samples).
S105-1-3 defines similarity formula:
, wherein min (), which is represented, takes less value in both, and max () represents to take value larger in both.
Wherein a, b, c are respectively the significance level of attribute θ, L5, L6 in similar isolated point error check, are visited by experiment
Beg for, find a=0.5, b=0.25 in similarity formula, during c=0.25, that is, think important journey of the angle in isolated point judgement
When degree accounts for 1/2, L5 and L6 and respectively accounts for 1/4, error detection effect is best.Due to there are 10 isolated point samples, then 10 r are obtainedi(i=1,
2 ..., 10) r=max (r, are takeni) for the similarity size of measuring point to be checked and isolated point.
S105-2:The error rate of angle error point is examined, including:
Angles of the S105-2-1 first to each point is handled, the angle more than 100 °, it is believed that it will not be necessarily angle
Degree error.In order to obtain the error probability of angle error point, the angle that angle is more than 100 ° is defined as 100 °, so it
Error probability is set to 0%.
S105-2-2 angles error point error rate formula be:
R=(100- corner dimensions)/100, wherein r ∈ [0,1]
S105-3:Apart from the error rate detection of long error point, including:
S105-3-1 measuring points to be checked and adjacent 2 points of distance are L1, L2, L=max (L1, L2) are taken, to L>=2000
Point, it is believed that it has an error probability apart from long error point, if L is more than 7000, it is believed that the point must be apart from long
Error point.
S105-3-2 is defined as follows apart from the error rate of long error point:
Wherein r ∈ [0,1].
The method of the comprehensive descision of the step S107 is:Specify type error detection and in indefinite type error-detecting method, 7 is small
In class error-detecting method, as long as there is a class method to be judged as mistake, then comprehensive descision result is just mistake;All do not misdeem, then the shaft tower
Coordinate does not have mistake.
Present invention contrast prior art, has the following advantages:
The classification similarity detection power network shaft tower that patent of the present invention is proposed sits calibration method, compared to traditional direct basis
The method that program (rule) carrys out error detection, can draw an error probability, rather than immediately arrive at some shaft tower to each shaft tower
Certain mistake, have ignored indefinite gray area.Other this method can continue to optimize model, and test mould is being carried out to sample
When type effect is fine or not, if it find that some wrong shaft tower coordinate is not detected, then this kind of shaft tower can be summarized
The feature of coordinate, is summarized as new error type, so as to carry out similarity examination again, is finally reached the mistake for detecting the overwhelming majority
Miss shaft tower coordinate.
Brief description of the drawings
Fig. 1 is a kind of schematic flow sheet of overhead power transmission line pole tower coordinate classification error-detecting method in the embodiment of the present invention;
Fig. 2 be the embodiment of the present invention in shaft tower clearly malfunction type with it is indefinite error type schematic diagram;
Fig. 3 is the schematic diagram of the isolated vertex type of shaft tower in the embodiment of the present invention;
Fig. 4 is the schematic diagram of drift vertex type in the embodiment of the present invention;
Fig. 5 is the schematic diagram of fallibility Kneetype in the embodiment of the present invention;
Embodiment
Method comprises the following steps:To the power network to be detected, the figure of some circuits of picture is first passed through, comparison is summed up clear and definite
The type that malfunctions and indefinite error type;Then the type that clearly malfunctioned to each class designs a program (rule) to sentence
It is disconnected, and correct;To indefinite error vertex type, a similarity is obtained using the method for similarity examination.
Above-mentioned technical problem is solved, the present invention is adopted the following technical scheme that:
S101:Overhead line structures error situation is divided into clearly error type and indefinite error type;
It is artificial to participate in judging to determine clearly error type and indefinite error type, one by drawing shaft tower line map
As have following several basic error types.Clearly malfunction type:Repeat point, isolated point, shift point, fallibility flex point;It is not clear
True error type:Isolated point error, angle malfunction, apart from long error, and each error type cases diagram is shown in Fig. 2,3,4,5.
S102:Fill a vacancy entering row interpolation by error detection circuit missing shaft tower coordinate, the detailed process that interpolation is filled a vacancy is as follows;
Show that the situation of missing shaft tower coordinate is divided into following 4 kinds through probing into:In the middle of whole piece power network line missing, power network line
Missing, power network line start missing, power network line and terminate missing, it is assumed that the actual number of shaft tower is N, its corresponding interpolation operation
It is as follows:
S102-1 whole pieces circuit all lack or whole piece circuit only one of which point situation, the benefit value meaning of such case is not
Greatly, it is inherent shortcoming, does not process.
S102-2 middle lines have lacked n (n=1,2 ... N-2) point, if the previous point coordinates of missing section is (x1,y1), lack section
Latter point coordinates is (x2,y2), then the coordinate of middle i-th of missing point is:
S102-3 circuits start to have lacked n (n=1,2 ... N-2) point, if first point for having data is (x after missing section1,
y1), it is (x to lack second point after section2,y2), then i-th point of the coordinate lacked is
xi=x1-(x2-x1)*i,yi=y1-(y2-y1) * i, i=1,2 ..., n.
S102-4 circuit latter ends have lacked n (n=1,2 ... N-2) point, if first point for having data is (x before missing section1,
y1), it is (x to lack second point before section2,y2), then i-th point of the coordinate lacked is
xi=x1+(x2-x1)*i,yi=y1+(y2-y1) * i, i=1,2 ..., n.
S103:Set up clearly error type error detection rule.
S103-1 repeats point (excessively similitude) debugging step.
S103-1-1 is (x provided with two shaft tower coordinates1,y1), (x2,y2), first to whole piece line scanning one time, if worked as
The absolute value of the distance of preceding point and next point is less than 0.000001, then marks next point complex point of attaching most importance to (to be with a flag
Array) it is designated as 1;
S103-1-2 by indicate 1 point, carry out interpolation processing (average), interpolation method handled by S101 method.
S103-2:Isolated point debugging step, isolated point schematic diagram is shown in Fig. 3.
S103-2-1 extract every circuit interlude point 5 attributes (remove each two points of head and the tail, without attribute, it is believed that
They do not isolate).
Each point of interlude, there is 5 attributes, and L1 is its distance with left side consecutive points, and L2 is itself and the right consecutive points
Distance, L3 is the distance of its left adjoint point with being separated by any, the distance for the right adjoint point that L4 is it with being separated by a bit, and angle 1 is it
With the angle of the line of left and right consecutive points.
S103-2-2 takes maximum in L1, L2 to be L5, and it is L6 to take maximum in L3, L4.
S103-2-3 is less than 60 degree (being threshold value, self controllable system) when angle 1, and L5 is more than 2000 meters (being threshold value, self controllable system),
When L6 is more than 2000 meters (being threshold value, self controllable system), it is believed that the point is clear and definite isolated point.
S103-2-4 amendment isolated points are average interpolation, use the method in S102.
S103-3:Drift point debugging step, shift point schematic diagram is shown in Fig. 4.
S103-3-1, which is first extracted, each treats cautious attribute L1, L2.
S103-3-2 works as the attribute L1 of some point to be detected>3000 (being arrange parameter, self controllable system), L2<2000 (are
Arrange parameter, self controllable system) when think that the point is likely to be the starting point of shift point, record it coordinate numbering be
mark.State, illustrated with reference to accompanying drawing 4 for convenience.Such as Fig. 4, A, B, C point are the point from left to right adjoined each other.Detect A
During point attribute, L1 is unsatisfactory for>3000, L2<2000;B points be the right adjoint point of A points, i.e., next point, detection B point attributes when, AB>
3000, BC<2000, it is believed that B points are probably shift point starting point, then the coordinate numbering of B points is designated as mark.
Next S103-3-3 judges the subsequent point C points of B points, judges whether the distance of C points to subsequent point D points is less than 2000
(being arrange parameter, self controllable system),
Continued executing with if it has not, carrying out return S103-3-1 to C points.
If it is, judging this section of shift point number, n is used for recording shift point number, judges a point for drift whenever more
When moving point, n adds 1.When CD distance be less than 2000, and AD distances subtract AC distance be less than AC distances 30% when, it is believed that C points be drift
Point;Continue the detection of next point, repeat the detection of both the above condition, when meeting upper two conditions when different, to next
Individual point returns to S103-3-1 and continued executing with.
The data that S103-3-4 is finally recorded are that a two-dimensional array, array has 2 values per a line, respectively drifts about
Starting point subscript m ark, the point number n of drift.According to this record above S102-2 method interpolation amendment shift point.
S103-4:Fallibility flex point debugging step, fallibility flex point schematic diagram is shown in Fig. 5.
S103-4-1 fallibility flex points, only appear in interlude, it is believed that during Origin And Destination will not be fallibility flex point, record
Between 3 attributes each putting of section, respectively angle 1, k1、k2.Remember r1For its line segment slope with left adjoint point line, r2For itself and the right side
The line segment slope of adjoint point line, angle k is obtained by slope1=arctan (r1),k2=arctan (r2)。
S103-4-2 works as angle 1<60 ° (being threshold value, self controllable system), angle k1With k2Poor absolute value be less than 30 ° (be threshold
Value, self controllable system) when, the point is judged for fallibility flex point, must be point error.
S103-4-3 is modified to the average interpolation fallibility flex point in S102.
S104:Using error detection rule, show that, by the clearly error shaft tower of error detection circuit, record shaft tower error type is simultaneously carried out
Interpolation correction;
S105:The error rate that indefinite error type is carried out to the circuit coordinate after correction is examined, and draws tested shaft tower
The error probability of all kinds of indefinite error types;
Indefinite error type having substantially:Isolated point error, angle are malfunctioned, apart from long error, and error rate is examined
Step difference is as follows:
S105-1:The error rate of isolated point error is examined, and isolated point schematic diagram is shown in Fig. 3.
S105-1-1 determines attribute
Each point of 5 attributes (ignoring two points of head and the tail) interlude of every circuit interlude point is extracted, there are 5 category
Property, it is respectively:L1, L2, L3, L4, angle 1;Derivative attribute:L5、L6.
The isolated point sample of 10 errors of artificial selection, extracts the corresponding attribute angle 1, L5 of typical failure point, L6.
Isolated point sample is shown in Table 1:
The isolated point sample of table 1.
S105-1-2 data predictions, in order to avoid extremum must be disturbed, selection is by attribute discretization, attribute after discretization
It see the table below shown.
The attribute discretization of table 2.
The property value of measuring point to be checked is also by above-mentioned steps discretization is also pressed, and the attribute for obtaining each point represents it is one 3
Dimensional vector (X1,X2,X3), each error sample has (an X1i,X2i,X3i) (i=1,2 ... 10 are 10 error samples).
S105-1-3 defines similarity formula:
Wherein pass through experiment discussion, when finding a=2, that is, think that significance level of the angle in isolated point judgement accounts for 1/2,
When L5 and L6 respectively accounts for 1/4, error detection effect is best.Due to there are 10 isolated point samples, then 10 r are obtainedi(i=1,2 ..., 10),
Take r=max (ri) with regard to the similarity size for measuring point to be checked and isolated point.
S105-1-4 similarity error detections are related to optimization details:
A, for different types of error point, different attributes can be extracted so that these category this kind of points of performance difference and
The difference of normal point.
B, carry out similarity system design when, it is believed that some attributes are more important with respect to other attributes, as angle accounts for 2/
5, other 3 attributes respectively account for 1/5, and the analog result so made is more preferable.
C, after the erroneous point of certain quasi-representative is found, the attributes of some erroneous points may be excessively abnormal, as distance is more than
50000 meters, so in similarity system design, influence can be caused larger.So carrying out discretization, such as larger than 50000 to property value
It is considered that being just that 50000,0-500m can consider as class, etc..
D, certain type sample typically malfunction and have a little looked for 10, can compare the similar value of tested point and this 10 points, choosing
Maximum similar value is as the similar value with such, so as to ensure pardon of the typical fault type to this class error point of choosing.
E, the definition of likelihood can voluntarily consider how optimization defines, and be two vectorial similitudes.
F, model can be continued to optimize, and such as either with or without the error point detected, then can sum up a kind of new error
Type, then detected by similarity, so that sophisticated model, reaches the effect of anticipation.
S105-2:The error rate of angle error point is examined.Because the definition of flex point is that angle is less than 60 °, other attributes thousand
Become ten thousand to change, so when flex point similarity checks mistake, if angle is smaller, the probability of point error is bigger.First draw and belong to
The angle each put.
Angle first to each point is handled, during angle more than 100 °, it is believed that it will not be necessarily flex point error.
In order to obtain an error probability, the angle that angle is more than 100 ° is defined as 100 °, so its error probability is set to 0%.
Angle error point error rate be:
R=(100- corner dimensions)/100, wherein r ∈ [0,1]
S105-3:Error rate apart from long error point is examined.
The point and adjacent 2 points of distance are L1, L2, L=max (L1, L2) are taken, to L>=2000 point, it is believed that it has one
Individual error probability, if L thinks that the point necessarily malfunctions more than 7000.
Apart from the error rate of long error point:
Wherein r ∈ [0,1].
S106:Expert's given threshold is more than by the indefinite wrong shaft tower type error probability of error detection circuit, this is judged as
Shaft tower coordinate mistake, misregistration type;
S107:According to clearly error type and indefinite error type detection result, comprehensive descision completes all bars of circuit
Tower error detection works.
The method of comprehensive descision is:Specify in type error detection and indefinite type error-detecting method, as long as there is a kind of method to sentence
Break as mistake, then comprehensive descision result is just mistake;All do not misdeem, then the shaft tower coordinate does not have mistake.
During comprehensive descision one is arranged in following table, 1 represents error, and 0 representative does not malfunction.
Illustrate below in conjunction with the accompanying drawings:
Below in conjunction with the accompanying drawings and specific embodiment, technical scheme is further described.
Shown in Figure 1, a kind of overhead power transmission line pole tower coordinate classification error-detecting method comprises the following steps:
The tens thousand of overhead line structures coordinates (error shaft tower to be detected) gathered according to certain power network, therefrom select 10 circuits to make
The classification error-detecting method is examined for sample, the correct coordinates of 10 overhead line structures have determined that.Shaft tower number to be detected
According to shown in following sheet form.
The shaft tower data to be detected of table 3.
Shaft tower title | Shaft tower longitude | Shaft tower latitude |
1 shaft tower | 113.665694 | 24.687697 |
2 shaft towers | 113.665395 | 24.688 |
3 shaft towers | 113.665084 | 24.688264 |
4 shaft towers | 113.664756 | 24.688539 |
5 shaft towers | 113.664111 | 24.68899 |
6 shaft towers | 113.664061 | 24.689214 |
7 shaft towers | 113.663834 | 24.689485 |
…… | …… | …… |
48 shaft towers | 113.64051 | 24.789503 |
1 shaft tower | 113.65388 | 24.690671 |
2 shaft towers | 113.666855 | 24.686709 |
3 shaft towers | 113.653652 | 24.690597 |
…… | …… | …… |
54 shaft towers | 113.5825 | 24.750833 |
S101:Overhead line structures error situation is divided into clearly error type and indefinite error type, altogether 7 group;
Probing into the clear and definite error type drawn has:Repeat point, isolated point, shift point, fallibility flex point;Indefinite error class
Type has:Isolated point error, angle malfunction, apart from long error, and details are shown in Fig. 2.
S102:Fill a vacancy entering row interpolation by error detection circuit missing shaft tower coordinate;
S103:Clearly error type error detection rule is set up, detailed rules sees above step S103 in specification;
S104:Using error detection rule, show that, by the clearly error shaft tower of error detection circuit, record shaft tower error type is simultaneously carried out
Interpolation correction;
Programmable detection result is as shown in the table, to it is each it is to be detected point carried out whether be clearly malfunction type judgement,
If the type that clearly malfunctions, then to the slotting intermediate value amendment of error point.After following table is longitude and latitude, judged result and modification before modification
Shaft tower warp, in latitude table, wherein complex point of whether attaching most importance to, whether be isolated point, whether be shift point, whether be that fallibility flex point 4 is arranged
In, 1 representative is the type, and 0 is represented as not being the type.
The Programmable detection result of table 4.
S105:The error rate that indefinite error type is carried out to the circuit coordinate after correction is examined, and draws tested shaft tower
The error probability of all kinds of indefinite error types.
The error probability of indefinite error type is calculated, each shaft tower to be detected is for indefinite type of error:It is lonely
Vertical point error, angle malfunction, respectively obtain an error probability apart from long error, as a result as shown in the table.
The error probability of the indefinite error type of table 5.
S106:Expert's given threshold is more than by the indefinite wrong shaft tower type error probability of error detection circuit, this is judged as
Shaft tower coordinate mistake, misregistration type.
Probe into obtain the power network line it is each it is indefinite error type threshold value be set to 0.7, error detection effect is good.
S107:According to clearly error type and indefinite error type detection result, comprehensive descision completes all bars of circuit
Tower error detection works;
The method of comprehensive descision is:Specify in type error detection and indefinite type error-detecting method, in 7 group error-detecting methods,
As long as there is a class method to be judged as mistake, then comprehensive descision result is just mistake;All do not misdeem, then the shaft tower coordinate does not have mistake.
Comprehensive descision result see the table below, during comprehensive descision one is arranged in table, and 1 represents error, and 0 representative does not malfunction.
The comprehensive descision result of table 6.
Analysis result finds the method that the classification similarity detects shaft tower coordinate, more flexible in error detection, and can
Preferably detect the shaft tower that makes mistake.
In summary analyze, a kind of method of power network shaft tower coordinate error detection of the present invention be based on classification debugging to it is similar
Degree processing, finds out the power network shaft tower of error.This method can draw an error probability to each shaft tower to be detected, rather than
The certain mistake of some shaft tower is immediately arrived at, indefinite gray area is have ignored.It can control in the quantity of error shaft tower, processing
There is certain flexibility, and effect is better than program (rule) directly judgement.
The implementation of the present invention is not limited to this, under the premise of above-mentioned basic fundamental thought of the invention, according to this area
Ordinary technical knowledge and customary means make the modification of other diversified forms to present invention, replace or change, and all fall within
Within rights protection scope of the present invention.
Claims (2)
- The error-detecting method 1. a kind of overhead power transmission line pole tower coordinate is classified, comprises the following steps:S101, is divided into clearly error type and indefinite error type by the error situation of overhead line structures;S102, carries out average interpolation to the shaft tower coordinate lacked in the power network shaft tower by error detection and fills a vacancy;S103, sets up clearly error type error detection rule;S104, using the clearly error type error detection rule, show that, by the clearly error shaft tower of error detection circuit, record shaft tower goes out Wrong type simultaneously carries out the average interpolation described in step S102 and filled a vacancy operation;S105, the circuit coordinate after filling a vacancy average interpolation carries out the detection of indefinite error type, draws by error detection shaft tower The error probability of all kinds of indefinite error types;S106, is more than expert's given threshold by the indefinite wrong shaft tower type error probability of error detection circuit, is judged as the shaft tower Coordinate mistake, misregistration type;In S107, the step S101-S106, as long as there is a class method to be judged as mistake, then comprehensive descision result is just mistake;All do not have Judge error, then the shaft tower coordinate does not have mistake;In the step S101, described clearly error type has four kinds:Repeat point, isolated point, shift point, fallibility flex point; Described indefinite error type has three kinds:Isolated point error, angle malfunction, apart from long error;In the step S102, the deletion condition has 4 kinds:Whole piece power network line missing, power network line intercalary delection, power network Circuit starts missing, power network line and terminates missing;The clearly error type error detection rule of the step S103 is as follows:S103-1 repeats point debugging stepS103-1-1 is (x provided with two shaft tower coordinates1,y1), (x2,y2), first to whole piece line scanning one time, if current point with The poor absolute value of next point is less than 0.000001, then marks next point complex point of attaching most importance to be designated as 1 with a flag;S103-1-2 by indicate 1 point, carry out interpolation processing, interpolation method handled by S102 method;S103-2:Isolated point debugging stepS103-2-1 extracts 5 attributes of every circuit interlude point:Each point of interlude, there is 5 attributes, and L1 is itself and a left side The distance of side consecutive points, L2 is its distance with the right consecutive points, the distance for the left adjoint point that L3 is it with being separated by a bit, and L4 is it With the distance for the right adjoint point for being separated by a bit, angle 1 is its angle with the line of left and right consecutive points;S103-2-2 takes maximum in L1, L2 to be L5, and it is L6 to take maximum in L3, L4, and L1, L2, L3 and L4 unit are rice;S103-2-3 is less than 60 degree when angle 1, and L5 is more than 2000 meters, when L6 is more than 2000 meters, it is believed that the point is clear and definite isolated point;S103-2-4 amendment isolated points are average interpolation, use the method in S102;S103-3:Drift point debugging stepS103-3-1, which is first extracted, each treats cautious attribute L1, L2;L1 is the distance of the measuring point to be checked and left adjoint point, and L2 treats for this The distance of test point and right adjoint point;S103-3-2 works as the attribute L1 of some point to be detected>3000 meters, L2<Think that the point is likely to be shift point at 2000 meters Starting point, it is mark to record the measuring point coordinate to be checked numbering;When detecting B point attributes, AB>3000, BC<2000, it is believed that B Point is probably shift point starting point, then the coordinate numbering of B points is designated as into mark;Next S103-3-3 judges the subsequent point C points of B points, judges whether the distance of C points to subsequent point is less than 2000;Continued executing with if it has not, carrying out return S103-3-1 to C points;If it is, judge this section of shift point number, n is used for recording shift point number, whenever judging a point for shift point more When, n adds 1;When CD distances are less than 2000;And AD distances subtract AC distance be less than AC distances 30% when, it is believed that C points are shift point, Continue the detection of next point, repeat the detection of both the above condition, when meeting upper two conditions when different, to next point S103-3-1 is returned to continue executing with;The data that S103-3-4 is finally recorded are that a two-dimensional array, array has 2 values, the starting respectively drifted about per a line Point subscript m ark, the point number n of drift;According to this record above S102 method interpolation amendment shift point;S103-4:Fallibility flex point debugging stepS103-4-1 fallibility flex points, only appear in interlude, it is believed that Origin And Destination will not be fallibility flex point, record interlude 3 attributes each put, respectively angle 1, K1, K2, left side bearing slope over 10 r1, right side bearing slope over 10 r2;Angle k is obtained by slope1 =arctan (r1),k2=arctan (r2);S103-4-2 works as angle 1<60 °, angle k1With k2Poor absolute value be less than 30 ° when, judge the point for fallibility flex point, necessarily It is point error;S103-4-3 is modified to slotting intermediate value.
- The error-detecting method 2. overhead power transmission line pole tower coordinate according to claim 1 is classified, it is characterised in that:It is described not The error rate of clearly error type is examined, and is comprised the following steps that:S105-1:The error rate of isolated point error is examinedS105-1-1 determines attribute5 attributes of every circuit interlude point are extracted, each point of interlude there are 5 attributes:L1 is itself and left side consecutive points Distance, L2 is its distance with the right consecutive points, the distance for the left adjoint point that L3 is it with being separated by a bit, and L4 is it with being separated by one The distance of the right adjoint point of point, angle 1 is its angle with the line of left and right consecutive points;Remember that maximum is L5 in L1, L2, takes L3, L4 Middle maximum is L6;The isolated point sample of 10 errors of artificial selection, extracts the corresponding attribute angle 1, L5 of typical failure point, L6;S105-1-2 numbers Data preprocess, in order to avoid extremum must be disturbed, is selected attribute discretization;The property value of measuring point to be checked also by above-mentioned steps discretization is also pressed, obtain each point attribute represent be a 3-dimensional to Measure (X1,X2,X3), each error sample has (an X1i,X2i,X3i), i=1,2 ... 10, it is 10 error samples;S105-1-3 defines similarity formula:<mrow> <msub> <mi>r</mi> <mi>i</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>a</mi> <mo>*</mo> <mfrac> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>X</mi> <mrow> <mn>1</mn> <mi>i</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>X</mi> <mrow> <mn>1</mn> <mi>i</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>+</mo> <mfrac> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>2</mn> </msub> <mo>,</mo> <msub> <mi>X</mi> <mrow> <mn>2</mn> <mi>i</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>2</mn> </msub> <mo>,</mo> <msub> <mi>X</mi> <mrow> <mn>2</mn> <mi>i</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>+</mo> <mfrac> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>3</mn> </msub> <mo>,</mo> <msub> <mi>X</mi> <mrow> <mn>3</mn> <mi>i</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>3</mn> </msub> <mo>,</mo> <msub> <mi>X</mi> <mrow> <mn>3</mn> <mi>i</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> </mrow> <mn>4</mn> </mfrac> <mo>,</mo> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>...</mo> <mn>10</mn> </mrow>As a=2, that is, angle is thought when the significance level during isolated point judges accounts for 1/2, L5 and L6 and respectively accounts for 1/4, error detection effect It is best;Due to there are 10 isolated point samples, then 10 r are obtainedi(i=1,2...10), takes r=max (ri) just for measuring point to be checked and The similarity size of isolated point;S105-1-4 similarity error detectionsA, for different types of error point, extract different attributes so that these category this kind of points of performance difference and normal point Difference;B, carry out similarity system design when, it is believed that some attributes are more important with respect to other attributes:Angle accounts for 2/5, other 3 category Property takes L1, L2, L3 or L4 wherein three, respectively accounts for 1/5;C, after the erroneous point of certain quasi-representative is found, to property value carry out discretization, think 50000 meters, 0- more than 50000 meters 500 meters are thought a class;D, certain type sample typically malfunction and have a little looked for 10, compare the similar value of tested point and this 10 points, and choosing is maximum similar Value is used as such similar value;E, likelihood are defined as two vectorial similitudes;F, model are continued to optimize, such as either with or without the error point detected, then sum up a kind of new error type, then by similar Spend to detect;S105-2:The error rate of angle error point is examined;The definition of flex point is that angle is less than 60 °, and other attributes are ever-changing, When flex point similarity checks mistake, if angle is smaller, the probability of point error is bigger;First draw the angle for belonging to each point;Angle first to each point is handled:During angle more than 100 °, the angle that angle is more than 100 ° is defined as 100 °, Error probability is set to 0%;Angle error point error rate be:R=(100- corner dimensions)/100, wherein r ∈ [0,1]S105-3:Error rate apart from long error point is examinedIt is L1, L2 apart from long error point and adjacent 2 points of distance, L=max (L1, L2) is taken, to L>=2000 meters of point, recognizes There is an error probability for the point, if L thinks that the point necessarily malfunctions more than 7000 meters, L unit is rice;Apart from the error rate of long error point:Wherein r ∈ [0,1].
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410752980.0A CN104462820B (en) | 2014-12-10 | 2014-12-10 | A kind of method of power network shaft tower coordinate error detection |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410752980.0A CN104462820B (en) | 2014-12-10 | 2014-12-10 | A kind of method of power network shaft tower coordinate error detection |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104462820A CN104462820A (en) | 2015-03-25 |
CN104462820B true CN104462820B (en) | 2017-10-27 |
Family
ID=52908849
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410752980.0A Active CN104462820B (en) | 2014-12-10 | 2014-12-10 | A kind of method of power network shaft tower coordinate error detection |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104462820B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111256702B (en) * | 2020-04-27 | 2020-08-21 | 天津市普迅电力信息技术有限公司 | Unmanned aerial vehicle autonomous inspection method for inspection of power tower |
CN112163055B (en) * | 2020-09-09 | 2023-11-21 | 成都深瑞同华科技有限公司 | Tower labeling method |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102855649A (en) * | 2012-08-23 | 2013-01-02 | 山东电力集团公司电力科学研究院 | Method for splicing high-definition image panorama of high-pressure rod tower on basis of ORB (Object Request Broker) feature point |
CN103246936A (en) * | 2013-04-24 | 2013-08-14 | 广东电网公司中山供电局 | System and method for pre-warning of typhoon risks of overhead transmission lines of grid |
-
2014
- 2014-12-10 CN CN201410752980.0A patent/CN104462820B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102855649A (en) * | 2012-08-23 | 2013-01-02 | 山东电力集团公司电力科学研究院 | Method for splicing high-definition image panorama of high-pressure rod tower on basis of ORB (Object Request Broker) feature point |
CN103246936A (en) * | 2013-04-24 | 2013-08-14 | 广东电网公司中山供电局 | System and method for pre-warning of typhoon risks of overhead transmission lines of grid |
Also Published As
Publication number | Publication date |
---|---|
CN104462820A (en) | 2015-03-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104330089B (en) | A kind of method that map match is carried out using history gps data | |
CN106970945A (en) | A kind of track preprocess method of taxi data set | |
CN102750413B (en) | Data processing and mapping method of topographic surveying of electric transmission line tower positions | |
CN107704992A (en) | The method and device of transmission line lightning stroke risk assessment | |
CN104951997A (en) | Multidimensional graphic and model quality analysis method facing power distribution network | |
CN111950627A (en) | Multi-source information fusion method and application thereof | |
CN108830933A (en) | A kind of electric force pole tower tower body method for reconstructing, system, medium and equipment | |
CN103247062A (en) | Method for surveying and mapping map by collecting farmland key points | |
CN103758017A (en) | Detection method and detection system for three-dimensional pavement elevation grid numerical value | |
CN106195646A (en) | Leak and be distributed estimating device, deduction system and presumption method | |
CN104777401A (en) | Power grid fault positioning method | |
CN102236600A (en) | Method and device for obtaining code coverage rate | |
CN107391515A (en) | Power system index analysis method based on Association Rule Analysis | |
CN106208050B (en) | A kind of grid branch static parameter detection and identification method based on PMU | |
CN104462820B (en) | A kind of method of power network shaft tower coordinate error detection | |
CN116308958A (en) | Carbon emission online detection and early warning system and method based on mobile terminal | |
CN104731970B (en) | The quality testing and control method of the multi-source heterogeneous data of highway | |
CN105468517A (en) | Statistical error positioning method based on reduction of black-box test case | |
CN102506812A (en) | VT checking method for stability judgment of reference points in deformation monitoring | |
CN111311173A (en) | National county level unit economic arrangement and spatialization method | |
CN104537581A (en) | Method for positioning temporary voltage drop source on line by adopting fuzzy similarity match | |
CN109324264A (en) | A kind of discrimination method and device of distribution network line impedance data exceptional value | |
CN107561470A (en) | A kind of fault detector evaluation of running status system | |
CN104590325A (en) | Railway station line and section insulating point GPS coordinate automatic surveying and mapping system and method | |
CN105717517B (en) | A kind of vehicle-mounted big dipper multi-mode GNSS high accuracy roads basis collecting method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |