CN106845337A - A kind of arc length difference sequence method of ship Power Disturbance timi requirement and type identification - Google Patents
A kind of arc length difference sequence method of ship Power Disturbance timi requirement and type identification Download PDFInfo
- Publication number
- CN106845337A CN106845337A CN201611130754.4A CN201611130754A CN106845337A CN 106845337 A CN106845337 A CN 106845337A CN 201611130754 A CN201611130754 A CN 201611130754A CN 106845337 A CN106845337 A CN 106845337A
- Authority
- CN
- China
- Prior art keywords
- disturbance
- arc length
- length difference
- signal
- difference sequence
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Length, Angles, Or The Like Using Electric Or Magnetic Means (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The present invention proposes a kind of arc length difference sequence method of ship electrical energy power quality disturbance timi requirement and type identification.The method is processed the filtered electric power signal of morphological filter based on arc length difference sequence, it is possible to achieve common 5 kinds single disturbances and 3 kinds of compound timi requirements for disturbing and type identification in Ship Electrical Power System.The method is first using a kind of improved alternately mixing morphological filter, treatment is filtered to ship electric power signal, then a kind of definition of arc length difference sequence is proposed, and for the precise time Location perturbation time, extract disturbance time and frequency domain characteristics amplitude variation characteristic and crest feature, characteristic quantity and disturbance time are finally input to the type identifier for defining disturbance type identification rule, disturbance type identification is realized.The disturbance that the method for the present invention can be used in the actual Ship Electrical Power System of quick type identification, it is adaptable to real-time detection and type identification.
Description
Technical field:
The present invention relates to ship Power Disturbance timi requirement and type identification, the more particularly to ship based on arc length difference sequence
The disturbance timi requirement and type identification of single disturbance and compound disturbance in oceangoing ship power system.
Background technology:
China's shipping industry is quickly grown, and with the huge innovation of ship energy device and power set technology, electric power is pushed away
Enter class type ship as the main flow ship type in market of ship.In the popularization of electric propulsion class type ship, it is ensured that marine vessel power system
System is safe and reliable and stably operation is its important technological problems, and rapidly and accurately timi requirement and type identification ship electricity
Disturbance in energy is the important foundation for solving this technical problem.Therefore, to ship Power Disturbance timi requirement and type identification
Carrying out research has deep meaning.
Morphological filter filtering method is a kind of filtering of the Time-Domain Nonlinear based on set theory for coming from image characteristics extraction
Method, with calculate it is simple, real-time the characteristics of, be the time-domain filtering method for attracting attention in recent years.Based on classical form
The denoising method research for learning principle launches around two main aspects:Choose adaptive structure element and design New Operators letter
Number.However, although these researchs improve filter effect to some extent, algorithm complexity and amount of calculation all significantly increase
Plus, extensive application power is worth weighing.
The common method of ship Power Disturbance timi requirement and type identification be Fourier analysis, Short Time Fourier Analysis,
And wavelet analysis, electric power signal is analyzed from frequency domain angle.But, it is flooded with the working environment of Ship Electrical Power System big
The electromagnetic noise of amount and high intensity, especially military boats and ships, are distributed with much noise frequency in frequency domain, are analyzed using frequency domain method
Electric power signal, timi requirement and type identification disturbance are a difficult tasks.Under strong noise background, the time domain waveform of signal
Information is not likely to produce distortion, and such as sinusoidal voltage waveform receives noise pollution, and waveform becomes more coarse, floods the letter of details area
Breath, but waveform now will have the feature of sinusoidal waveform, and the waveform of details area, institute careful will be portrayed by filtering
It is to solve Ship Electrical Power System to disturb to carry out disturbance timi requirement with type identification with the time domain waveform information for making full use of signal
The method a kind of more feasible compared with frequency domain method of timi requirement and type identification.
In view of analyzing above, determine to solve the Ship Electrical Power System Power Disturbance time in the case of by strong noise pollution
Position and type identification problem, from the angle of time-domain analysis, propose that one kind is based on improving alternately compound filter and arc length difference sequence
The ship Power Disturbance timi requirement and kind identification method of row.
The content of the invention:
The purpose of the present invention is directed under ship strong electromagnetic noise operation environment, electric power signal disturbance in Ship Electrical Power System
Timi requirement and identification problem, propose it is a kind of based on improve alternately mixing morphological filter and arc length difference sequence ship electricity
Timi requirement and kind identification method can be disturbed.The method is filtered by improving alternately mixing morphological filter to ship electric power signal
Ripple, with reference to the definition of arc length difference sequence, is accurately positioned disturbance time of origin, extract filtered signal waveforms amplitude feature and
Crest number feature, and the disturbance type identifier for defining type identification rule is input to, realize that final disturbance type is known
Not.
If in signal certain sampled point be a points, after a points the per quart signal period (or 1/2nd signal periods,
One signal period) point be designated as b points, c points, d points etc. successively, and a points are called the starting point of arc length sequence, b points, c points, d points
Deng being periodic point that arc length sequence is chosen by a quarter (or 1/2nd) the electricity cycle, these points half interval contour two-by-two
Arc length can constitute arc length sequenceThen (single order) arc length difference sequence can be defined as follows formula (1):
Nonzero term in Δ s is referred to as the first nonzero element, the second nonzero element, the 3rd nonzero element etc. successively.
Illustrated for ease of method and step, if the amplitude of normal voltage signal is A, the frequency of signal is 50Hz, each electricity week
Phase samples k point, then the sampling number of the electric cycle (0.005s) of a quarter isIf X0={ x1,x2,…,xnIt is sampling letter
Number, and x1It is zero crossing.The step of the inventive method, is as follows:
Step 1) number of winning the confidence amplitude A on the basis of be worth, standardization is taken to original sampled signal.(2) are believed sampling as the following formula
Number X0It is standardized, obtains normalized signal X1:
Step 2) using it is improved alternately mixing morphological filter to step 1) in treatment after signal be filtered, improve
Alternating mixing morphological filter such as following formula (3) it is shown:
Almix (n)=((f) OC (g)+(f) CO (g)) (n)/2 (3)
Wherein:
[(f) OC (g)] (n)=(f ο gg) (n) (4)
[(f) CO (g)] (n)=(fg ο g) (n) (5)
F is sampled signal, and g (m) is structural element, Df1={ p ..., q }, p=[N/1000], q=N- [N/1000];Dg
={ 0,1,2 ..., M }, M are the number of element in structural element g (m);N and M are integer, N >=M;
Step 3) the disturbance time is accurately positioned, it is divided into three steps:A. by 1/2nd electric computation of Period steps 2) in
Filtering signal arc length difference sequence, Location perturbation time interval;B. by a quarter electricity computation of Period step 2) in filter
The arc length difference sequence of ripple signal, the precise moments in Location perturbation interval;C. in comprehensive disturbance time interval and disturbance interval
Precise moments, complete disturbance time precise positioning, detailed process is as follows:
Step 3.1) by 1/2nd electric computation of Period steps 2) in filtering signal arc length difference sequence, Location perturbation
Time interval:
Step 3.1.1) first, with zero crossing x1It is starting point, periodic point is taken by a quarter electricity cycle.Calculated for two cycles
Arc length between point, obtains arc length sequence for X3, finally it is calculated arc length difference sequence X shown in formula (1) for another example4;
Step 3.1.2) the first nonzero element in extraction step arc length difference sequence, search non-with first in arc length sequence
The related arc of neutral element, the time interval [t where this arc1,t1+ 0.005] just there is the moment comprising disturbance;
Step 3.2) by a quarter electricity computation of Period step 2) in filtering signal arc length difference sequence, Location perturbation
Precise moments in interval:
Step 3.2.1) to step 2) in filtered signal first a quarter cycle in point it is selected according to user
Sample frequency carry out double sampling;
Step 3.2.2) using the sampled point in first a quarter electricity cycle successively as starting point, by 1/2nd electricity
Cycle takes periodic point, is calculated arc length difference sequence group;
Step 3.2.3) extraction step 3.2.2) in arc length difference sequence group in the first nonzero element, composition is first non-
Neutral element set;
Step 3.2.4) to step 3.2.3) in the first nonzero element set in element take absolute value, and be designated as X5,
The maximum in these elements is searched, serial number j of this maximum in set is remembered;
Step 3.3) precise moments of the comprehensive disturbance time interval with disturbance in interval, disturbance time precise positioning is completed,
(10) can be calculated disturbance moment T as the following formula, and unit is the second:
Step 4) according to step 3) in disturbance timi requirement result and step 2) in filtered signal, extract disturbance hair
The arc length difference sequence X in raw stage4' and disturbance stage of development filtered signal X2', disturbance type identification is carried out, it is divided into two
Step:A. to X4' characteristic processing is carried out, temporarily liter, temporarily drop, transient oscillation and diagnosis disturbance can be recognized;B. to X2' carry out at feature
Reason, can recognize interruption, transient oscillation, temporarily rise/temporarily drop harmonic, note filtered signal X2' in have Z sampled point, specifically
Disturbance type identification step is as follows;
Step 4.1) to the arc length difference sequence X of disturbance stage of development4' carry out characteristic processing:
Step 4.1.1) to the arc length difference sequence X of disturbance stage of development4' carry out medium filtering, the template length of filtering
ForAnd even number is taken as, obtain waveform amplitude Characteristics amount S after disturbance occurs;
Step 4.1.2) whether waveforms amplitude characteristic quantity S is judged more than 0, if so, then going to step 4.3);If it is not, then turning step
It is rapid 4.4);
Step 4.1.3) judge whether first element of S and last element are more than 0.1, i.e. Sstart-Send> 0.1,
If so, then judging that disturbance is disturbed as transient oscillation;If it is not, being then judged to temporarily rise disturbance;
Step 4.1.4) whether S is judged more than -0.9 and less than -0.1, if so, being then judged to temporarily drop disturbance;If it is not, then sentencing
It is set to interruption;
Step 4.2) takeFiltered signal X to disturbing stage of development2' carry out at feature
Reason:
Step 4.2.1) to the filtered signal X of disturbance stage of development2' crest number detection is carried out, as a result it is designated as fluctuation
Frequecy characteristic P;
Step 4.2.2) judge whether P is equal to zero, if so, being then judged to interrupt;Otherwise go to step 4.7);
Step 4.2.3) judge whether M≤P≤M+1 sets up, if being then judged to temporarily rise or temporarily drop disturbance, otherwise turn step
It is rapid 4.8);
Step 4.2.4) judge whether P >=N sets up, if being then judged to transient oscillation, otherwise it is judged as harmonic disturbance, most
Higher hamonic wave is [P/ [z/k]] subharmonic.
Beneficial effect:
The ship Power Disturbance time for alternately mixing morphological filter and arc length difference sequence based on improvement in the present invention
Positioning and kind identification method, having two the characteristics of its is maximum is respectively:Pole in morphological filter is replaced using mean value function
Value computing, makes that filtered signal waveform is more smooth, and wave character is more significantly;A kind of definition of arc length difference sequence is proposed,
And the extraction of perturbation features is realized based on this sequence.The method, from the angle of time-domain waveform analysis, is the ship Power Disturbance time
Positioning provide one with disturbance identification method research in new Research Thinking, the characteristics of with theory innovation.The method can be real
The precise time positioning of spot oceangoing ship electric power signal disturbance, effectively 5 kinds of single type disturbances of identification and 3 kinds of complicated type disturbances.Should
Method can be used for the quick positioning of Power Disturbance and identification in actual ship.
Brief description of the drawings:
The present invention is further illustrated below in conjunction with the drawings and specific embodiments.
Fig. 1 ship Power Disturbance timi requirements and type identification flow chart;
Fig. 2 ship Power Disturbance time precise positioning flow charts;
Fig. 3 ship electric energy type identification flow charts;
Fig. 4 voltage swells disturb timi requirement and extract result with perturbation features;
Fig. 5 voltage Sag Disturbances timi requirement extracts result with perturbation features;
Fig. 6 voltage interruptions disturb timi requirement and extract result with perturbation features;
Fig. 7 voltage harmonics disturb timi requirement and extract result with perturbation features;
Fig. 8 voltage transients duration of oscillation is positioned and extracts result with perturbation features;
Fig. 9 voltage dips add the compound disturbance timi requirement of harmonic wave to extract result with perturbation features;
Figure 10 voltage swells add the compound disturbance timi requirement of harmonic wave to extract result with perturbation features;
Voltage Sag Disturbance timi requirements of the Figure 11 containing phase hit extracts result with perturbation features;
Specific embodiment:
In order that technological means, creation characteristic, reached purpose and effect that the present invention is realized are easy to understand, tie below
Conjunction is specifically illustrating, and embodiments of the present invention are expanded on further.
First, if the amplitude of normal voltage signal is A, the frequency of signal is 50Hz, k point of each electric periodic sampling, then
A quarter electricity the cycle (0.005s) sampling number beIf X0={ x1,x2,…,xnIt is sampled signal, and x1It is zero passage
Point.
The basic procedure of holistic approach is filtered as shown in figure 1, being passed in and out to standardized ship electric power signal, then will filtering
Signal is separately input to disturb timi requirement process and disturbed depth process afterwards, then by timi requirement result and filtered signal one
Rise and be input to type recognition procedure, finally realize timi requirement and type identification, the specific implementation step of method is as follows:
Step 1) number of winning the confidence amplitude A on the basis of be worth, standardization is taken to original sampled signal.(1) is believed sampling as the following formula
Number X0It is standardized, obtains normalized signal X1:
Step 2) using it is improved alternately mixing morphological filter to step 1) in treatment after signal be filtered, improve
Alternating mixing morphological filter such as following formula (2) it is shown:
Almix (n)=((f) OC (g)+(f) CO (g)) (n)/2 (2)
Wherein:
[(f) OC (g)] (n)=(f ο gg) (n) (3)
[(f) CO (g)] (n)=(fg ο g) (n) (4)
F is sampled signal, and g (m) is structural element, Df1={ p ..., q }, p=[N/1000], q=N- [N/1000];Dg
={ 0,1,2 ..., M }, M are the number of element in structural element g (m);N and M are integer, N >=M;
Step 3) the disturbance time is accurately positioned, as shown in Fig. 2 being divided into three steps:A. based on 1/2nd electric cycles
Calculate step 2) in filtering signal arc length difference sequence, Location perturbation time interval;B. by a quarter electricity computation of Period step
It is rapid 2) in filtering signal arc length difference sequence, Location perturbation interval in precise moments;C. the disturbance time is completed accurately fixed
Position, detailed process is as follows:
Step 3.1) by 1/2nd electric computation of Period steps 2) in filtering signal arc length difference sequence, Location perturbation
Time interval:
Step 3.1.1) first, with zero crossing x1It is starting point, periodic point is taken by a quarter electricity cycle.Calculated for two cycles
Arc length between point, obtains arc length sequence for X3, finally it is calculated arc length difference sequence X shown in formula (1) for another example4;
Step 3.1.2) the first nonzero element in extraction step arc length difference sequence, search non-with first in arc length sequence
The related arc of neutral element, the time interval [t where this arc1,t1+ 0.005] just there is the moment comprising disturbance;
Step 3.2) by a quarter electricity computation of Period step 2) in filtering signal arc length difference sequence, Location perturbation
Precise moments in interval:
Step 3.2.1) to step 2) in filtered signal first a quarter cycle in point it is selected according to user
Sample frequency carry out double sampling;
Step 3.2.2) using the sampled point in first a quarter electricity cycle successively as starting point, by 1/2nd electricity
Cycle takes periodic point, is calculated arc length difference sequence group;
Step 3.2.3) extraction step 3.2.2) in arc length difference sequence group in the first nonzero element, composition is first non-
Neutral element set;
Step 3.2.4) to step 3.2.3) in the first nonzero element set in element take absolute value, and be designated as X5,
The maximum in these elements is searched, serial number j of this maximum in set is remembered;
Step 3.3) precise moments of the comprehensive disturbance time interval with disturbance in interval, disturbance time precise positioning is completed,
(9) can be calculated disturbance moment T as the following formula, and unit is the second:
Step 4) according to step 3) in disturbance timi requirement result and step 2) in filtered signal, extract disturbance hair
The arc length difference sequence X in raw stage4' and disturbance stage of development filtered signal X2', disturbance type identification is carried out, such as Fig. 3 institutes
Show, be divided into two steps:A. to X4' characteristic processing is carried out, temporarily liter, temporarily drop, transient oscillation and diagnosis disturbance can be recognized;B. to X2' enter
Row characteristic processing, can recognize interruption, transient oscillation, temporarily rise/temporarily drop harmonic, note filtered signal X2' in have Z sampling
Point, disturbance type identification step is as follows;
Step 4.1) to the arc length difference sequence X of disturbance stage of development4' carry out characteristic processing:
Step 4.1.1) to the arc length difference sequence X of disturbance stage of development4' carry out medium filtering, the template length of filtering
ForAnd even number is taken as, obtain waveform amplitude Characteristics amount S after disturbance occurs;
Step 4.1.2) whether waveforms amplitude characteristic quantity S is judged more than 0, if so, then going to step 4.3);If it is not, then turning step
It is rapid 4.4);
Step 4.1.3) judge whether first element of S and last element are more than 0.1, i.e. Sstart-Send> 0.1,
If so, then judging that disturbance is disturbed as transient oscillation;If it is not, being then judged to temporarily rise disturbance;
Step 4.1.4) whether S is judged more than -0.9 and less than -0.1, if so, being then judged to temporarily drop disturbance;If it is not, then sentencing
It is set to interruption;
Step 4.2) takeFiltered signal X to disturbing stage of development2' carry out characteristic processing:
Step 4.2.1) to the filtered signal X of disturbance stage of development2' crest number detection is carried out, as a result it is designated as fluctuation
Frequecy characteristic P;
Step 4.2.2) judge whether P is equal to zero, if so, being then judged to interrupt;Otherwise go to step 4.7);
Step 4.2.3) judge whether M≤P≤M+1 sets up, if being then judged to temporarily rise or temporarily drop disturbance, otherwise turn step
It is rapid 4.8);
Step 4.2.4) judge whether P >=N sets up, if being then judged to transient oscillation, otherwise it is judged as harmonic disturbance, most
Higher hamonic wave is [P/ [z/k]] subharmonic.
The inventive method is verified and is described as follows by l-G simulation test:Common disturbance voltage is temporary in structure microgrid
5 kinds of single type disturbances such as liter, voltage dip, interruption, harmonic wave, transient oscillation, temporarily rise and add harmonic wave, temporarily drop plus harmonic wave and contain phase
3 kinds of compound type disturbances such as temporary drop of position saltus step, the signal model that totally eight types are disturbed, as shown in table 1.
Parameter setting in the Power Disturbance signal simulation model of table 1 and emulation
Wherein u (t) is defined as follows formula (10):
The emulation signal of all types of disturbances is generated in MATLAB, and adds white noise, make signal to noise ratio for 40dB, each
200 points of electric periodic sampling, sample frequency is 10kHz, is sampled 20 electric cycles, totally 4001 sampled points, and unification is arranged on adopts
During disturbing signal, i.e. 0.082s~0.203s are produced between sampling point 820~2030.The different ginsengs of signal model in simulation process
The column of parameter one is set during number presses table 1.
The change curve of part key feature amount is as shown in Fig. 4~Figure 11 in simulation process, wherein X2It is by improving shape
Signal after state filter filtering, X3It is for initial with j-th sampled point found within first a quarter electricity cycle
Point, periodic point, the arc length sequence being ultimately formed, X are chosen with 1/2nd electric cycles4It is according to arc length sequence X3Calculate
The arc length difference sequence for arriving, X5For step 3.6) in take absolute value after the first nonzero element set, S be step 4.1) in waveform
The amplitude amount of levying.
Disturbing signal is processed using the method for the present invention, double sampling problem (skipping step 3.3) is not considered,
Positioning result is as shown in table 2 with specific perturbation features.
The type timi requirement of table 28 and perturbation features table
As shown in Table 2, disturbance positioning time T and true time of origin error are 0.1~0.2ms, with sample frequency
Improve, error will be smaller, it is sufficient to meet in engineering the need for quick Location perturbation.Meanwhile, the recognition result for disturbing type is complete
Correctly.When there is single disturbance and compound disturbance, the increase and decrease of the positive and negative sign waveforms amplitude of waveforms amplitude characteristic quantity S symbol,
Its numerical values recited characterizes waveforms amplitude intensity of variation.Vibration frequency feature P is the statistics to crest number, to it is single disturbance or
Load disturbance all has very strong recognition capability, with certain antijamming capability.
When voltage swell is disturbed, as shown in Figure 4.Arc length sequence X3Numerical value becomes big during disturbing;Arc length difference sequence exists
Disturbance occurs and finish time numerical nonzero, pulse occurs, is zero at other moment;First nonzero element set X5Middle maximum
It is the 22nd element, is 0.0822s by formula (4) the calculation perturbation moment, is 0.2ms with actual disturbance moment error;Waveforms amplitude
The amount of levying S rises to 0.5 during disturbing, and is determined as temporarily rising disturbance according to disturbed depth algorithm.
During voltage Sag Disturbance, as shown in Figure 5.Arc length sequence X3Numerical value reduces during disturbing;Arc length difference sequence exists
Disturbance occurs and finish time numerical nonzero, pulse occurs, is zero at other moment;First nonzero element set X5Middle maximum
It is the 22nd element, is 0.0822s by formula (4) the calculation perturbation moment, is 0.2ms with actual disturbance moment error;Waveforms amplitude
The amount of levying S drops to -0.3 during disturbing, and temporarily drop disturbance is determined as according to disturbed depth algorithm.
When voltage interruption is disturbed, as shown in Figure 6.Arc length sequence X3Numerical value reduces during disturbing;Arc length difference sequence exists
Disturbance occurs and finish time numerical nonzero, pulse occurs, is zero at other moment;First nonzero element set X5Middle maximum
It is the 21st element, is 0.0821s by formula (4) the calculation perturbation moment, is 0.1ms with the error at actual disturbance moment;Waveform width
The value amount of levying S drops to -0.9 during disturbing, and is determined as interrupting disturbance according to disturbed depth algorithm.
During voltage transient oscillation disturbances, as shown in Figure 7.Arc length sequence X3Numerical value becomes big during disturbing;Arc length difference sequence
It is listed in disturbance to occur and finish time numerical nonzero, pulse occurs, is zero at other moment;First nonzero element set X5In most
Big value is the 20th element, is 0.082s by formula (4) the calculation perturbation moment, identical with the actual disturbance moment;The waveforms amplitude amount of levying S
Show as first rising during disturbing declining afterwards, first and last element differs by more than 0.1, meanwhile, crest testing result is P=119, root
Transient oscillation is determined as according to disturbed depth algorithm to disturb.
When voltage harmonic is disturbed, as shown in Figure 8.Arc length sequence X3Numerical value becomes big during disturbing;Arc length difference sequence exists
Disturbance occurs and finish time numerical nonzero, pulse occurs, is zero at other moment;First nonzero element set X5Middle maximum
It is the 22nd element, is 0.0822s by formula (4) the calculation perturbation moment, is 0.2ms with the error at actual disturbance moment;Waveform width
The value amount of levying S rises to 0.5 during disturbing, meanwhile, crest testing result is P=42, is determined as according to disturbed depth algorithm humorous
Wave disturbance.
When voltage dip adds harmonic disturbance, as shown in Figure 9.Arc length sequence X3Numerical value first becomes big rear reduction during disturbing;
Arc length difference sequence occurs and finish time numerical nonzero in disturbance, pulse occurs, is zero at other moment;First nonzero element
Set X5Middle maximum is the 20th element, is 0.082s by formula (4) the calculation perturbation moment, identical with the actual disturbance moment;Ripple
The shape amplitude amount of levying S drops to -0.11 during disturbing, meanwhile, crest testing result is P=42, is sentenced according to disturbed depth algorithm
Wei not drop temporarily plus harmonic disturbance.
When voltage swell adds harmonic disturbance, as shown in Figure 10.Arc length sequence X3Numerical value becomes big during disturbing;Arc length difference
Sequence occurs and finish time numerical nonzero in disturbance, pulse occurs, is zero at other moment;First nonzero element set X5In
Maximum is the 20th element, is 0.0820s by formula (4) the calculation perturbation moment, identical with the actual disturbance moment;Waveforms amplitude is levied
Amount S rises to 0.9 during disturbing, meanwhile, crest testing result is P=42, and being determined as temporarily liter according to disturbed depth algorithm adds
Harmonic disturbance.
When temporary drop containing phase hit is disturbed, as shown in figure 11.Arc length sequence X3Numerical value reduces during disturbing;Arc length
Difference sequence occurs and finish time numerical nonzero in disturbance, pulse occurs, is zero at other moment;First nonzero element set
X5Middle maximum is the 22nd element, is 0.0822s by formula (4) the calculation perturbation moment, and the error with the actual disturbance moment is
0.2ms;The waveforms amplitude amount of levying S drops to -0.5 during disturbing, meanwhile, crest testing result is P=7, according to disturbed depth
Algorithm is determined as the temporary drop disturbance containing phase hit.
Claims (2)
1. a kind of ship electrical energy power quality disturbance timi requirement and the arc length difference sequence method of type identification, define arc length difference sequence
Row are as follows:
If certain sampled point is a points in signal, per quart signal period after a points, or 1/2nd signal periods, or one
The point of individual signal period is designated as b points, c points, d points etc. successively, and a points are called the starting point of arc length sequence, and b points, c points, d points etc. are
Arc length sequence presses a quarter, or 1/2nd, or a periodic point for electric cycle selection, and these put the arc length of half interval contour two-by-two
Arc length sequence can be constituted isThen arc length difference sequence can be defined as follows formula (1), by the non-zero in Δ s
Item is referred to as the first nonzero element, the second nonzero element, the 3rd nonzero element etc. successively:
If being sampled to electric power signal according to each k point of electric cycle (0.02s) sampling, then a quarter electricity cycle
The sampling number of (0.005s) isX0={ x1,x2,…,xnIt is sampled signal, and x1It is zero crossing;It is characterized in that ship
Electrical energy power quality disturbance timi requirement is comprised the following steps with the arc length difference sequence method of type identification:
Step 1) number of winning the confidence amplitude A on the basis of be worth, standardization is taken to original sampled signal.(2) are to sampled signal X as the following formula0Enter
Row standardization, obtains normalized signal X1:
Step 2) using it is improved alternately mixing morphological filter to step 1) in treatment after signal be filtered, improved friendship
For mixed style wave filter such as following formula (3) Suo Shi:
Almix (n)=((f) OC (g)+(f) CO (g)) (n)/2 (3)
Wherein:
F is sampled signal, and g (m) is structural element, Df1={ p ..., q }, p=[N/1000], q=N- [N/1000];Dg=0,
1,2 ..., M }, M is the number of element in structural element g (m);N and M are integer, N >=M;
Step 3) the disturbance time is accurately positioned, it is divided into three steps:A. by 1/2nd electric computation of Period steps 2) in filter
The arc length difference sequence of ripple signal, Location perturbation time interval;B. by a quarter electricity computation of Period step 2) in filtering letter
Number arc length difference sequence, Location perturbation interval in precise moments;C. the essence during comprehensive disturbance time interval and disturbance are interval
At the true moment, disturbance time precise positioning is completed, detailed process is as follows:
Step 3.1) by 1/2nd electric computation of Period steps 2) in filtering signal arc length difference sequence, the Location perturbation time
It is interval:
Step 3.1.1) first, with zero crossing x1 as starting point, periodic point is taken by a quarter electricity cycle.Between calculating two periodic points
Arc length, obtain arc length sequence for X3, finally it is calculated arc length difference sequence X shown in formula (1) for another example4;
Step 3.1.2) the first nonzero element in extraction step arc length difference sequence, search in arc length sequence with the first non-zero entry
Plain related arc, the time interval [t where this arc1,t1+ 0.005] just there is the moment comprising disturbance;
Step 3.2) by a quarter electricity computation of Period step 2) in filtering signal arc length difference sequence, Location perturbation is interval
In precise moments:
Step 3.2.1) to step 2) in filtered signal first a quarter cycle in point according to adopting that user selectes
Sample frequency carries out double sampling;
Step 3.2.2) using the sampled point in first a quarter electricity cycle successively as starting point, by 1/2nd electric cycles
Periodic point is taken, arc length difference sequence group is calculated;
Step 3.2.3) extraction step 3.2.2) in arc length difference sequence group in the first nonzero element, constitute the first non-zero entry
Element set;
Step 3.2.4) to step 3.2.3) in the first nonzero element set in element take absolute value, and be designated as X5, search this
Maximum in a little elements, remembers serial number j of this maximum in set;
Step 3.3) precise moments of the comprehensive disturbance time interval with disturbance in interval, disturbance time precise positioning is completed, press
Formula (10) can be calculated disturbance moment T, and unit is the second:
Step 4) according to step 3) in disturbance timi requirement result and step 2) in filtered signal, extract disturbance occur rank
The arc length difference sequence X of section4' and disturbance stage of development filtered signal X2', disturbance type identification is carried out, it is divided into two steps:a.
To X4' characteristic processing is carried out, temporarily liter, temporarily drop, transient oscillation and diagnosis disturbance can be recognized;B. to X2' characteristic processing is carried out, can
To recognize interruption, transient oscillation, temporarily rise/temporarily drop harmonic, note filtered signal X2' in have Z sampled point, specifically disturb class
Type identification step is as follows;
Step 4.1) to the arc length difference sequence X of disturbance stage of development4' carry out characteristic processing:
Step 4.1.1) to the arc length difference sequence X of disturbance stage of development4' medium filtering is carried out, the template length of filtering isAnd even number is taken as, obtain waveform amplitude Characteristics amount S after disturbance occurs;
Step 4.1.2) whether waveforms amplitude characteristic quantity S is judged more than 0, if so, then going to step 4.3);If it is not, then going to step
4.4);
Step 4.1.3) judge whether first element of S and last element are more than 0.1, i.e. Sstart-Send> 0.1, if
It is then to judge that disturbance is disturbed as transient oscillation;If it is not, being then judged to temporarily rise disturbance;
Step 4.1.4) whether S is judged more than -0.9 and less than -0.1, if so, being then judged to temporarily drop disturbance;If it is not, being then judged to
Interrupt;
Step 4.2) takeFiltered signal X to disturbing stage of development2' carry out characteristic processing:
Step 4.2.1) to the filtered signal X of disturbance stage of development2' crest number detection is carried out, as a result it is designated as vibration frequency special
Levy P;
Step 4.2.2) judge whether P is equal to zero, if so, being then judged to interrupt;Otherwise go to step 4.7);
Step 4.2.3) judge whether M≤P≤M+1 sets up, if being then judged to temporarily rise or temporarily drop disturbance, otherwise go to step
4.8);
Step 4.2.4) judge whether P >=N sets up, if being then judged to transient oscillation, otherwise it is judged as harmonic disturbance, most high order
Harmonic wave is [P/ [z/k]] subharmonic.
2. crest number detection according to claim 1, it is characterized in that the signal waveform in certain time of detection regulation
Amplitude be positive crest value number, if a period of time in amplitude it is equal, only remember this section of initial time of time be peak
Value, P=1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611130754.4A CN106845337B (en) | 2016-12-09 | 2016-12-09 | Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611130754.4A CN106845337B (en) | 2016-12-09 | 2016-12-09 | Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106845337A true CN106845337A (en) | 2017-06-13 |
CN106845337B CN106845337B (en) | 2020-05-29 |
Family
ID=59140671
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611130754.4A Active CN106845337B (en) | 2016-12-09 | 2016-12-09 | Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106845337B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117173477A (en) * | 2023-09-05 | 2023-12-05 | 北京交通大学 | Domain generalization pedestrian re-identification method based on multi-layer data disturbance strategy |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102565574A (en) * | 2011-12-06 | 2012-07-11 | 中国人民解放军海军工程大学 | Ship electric energy quality monitoring system |
CN106250904A (en) * | 2016-05-18 | 2016-12-21 | 国网新疆电力公司电力科学研究院 | Based on Power Disturbance analyser and the sorting technique of improving S-transformation |
-
2016
- 2016-12-09 CN CN201611130754.4A patent/CN106845337B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102565574A (en) * | 2011-12-06 | 2012-07-11 | 中国人民解放军海军工程大学 | Ship electric energy quality monitoring system |
CN106250904A (en) * | 2016-05-18 | 2016-12-21 | 国网新疆电力公司电力科学研究院 | Based on Power Disturbance analyser and the sorting technique of improving S-transformation |
Non-Patent Citations (2)
Title |
---|
李庚银等: "基于数学形态学和网格分形的电能质量扰动检测及定位", 《中国电机工程学报》 * |
赵静等: "利用广义形态滤波与差分熵的电能质量扰动检测", 《中国电机工程学报》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117173477A (en) * | 2023-09-05 | 2023-12-05 | 北京交通大学 | Domain generalization pedestrian re-identification method based on multi-layer data disturbance strategy |
Also Published As
Publication number | Publication date |
---|---|
CN106845337B (en) | 2020-05-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106250904B (en) | Electric energy disturbance analyzer based on improved S transformation and classification method | |
CN109946512B (en) | A kind of dynamic power analysis method for improving frequency domain interpolation | |
CN110068759A (en) | A kind of fault type preparation method and device | |
CN105675956B (en) | A kind of Voltage flicker detection based on windowed interpolation Short Time Fourier Transform | |
CN102841251A (en) | Electric car charging station harmonic wave detection method in microgrid | |
CN107255772B (en) | Semi-supervised voltage sag accident source identification method | |
CN110441654A (en) | Based on the distribution network electric energy quality disturbance detecting method for improving EWT and CMPE | |
CN111291452B (en) | Ship electric propulsion system fault mode risk determination method and system | |
CN108957175A (en) | Electrical energy power quality disturbance recognition methods based on improved HHT algorithm | |
CN106407659A (en) | Air quality index (AQI) predicting method and device | |
CN106845337A (en) | A kind of arc length difference sequence method of ship Power Disturbance timi requirement and type identification | |
Tang et al. | Design of power quality monitoring system based on LabVIEW | |
CN105606892A (en) | Power grid harmonic and inter-harmonic analysis method based on SST transformation | |
Chu et al. | TDV-LSTM: A new methodology for series arc fault detection in low power AC systems | |
CN104808060B (en) | A kind of digital measuring method of electrical signal phase difference | |
CN116070384A (en) | Transient stability evaluation method and system based on power grid feature arrangement importance | |
CN106598910A (en) | EMD (Empirical Mode Decomposition) end effect inhibiting method and system | |
CN109902961A (en) | The multi-scale filtering method and system filtered for power signal in Energy Decomposition | |
Liu et al. | Microscopic image enhancement of Chinese Herbal Medicine based on fuzzy set | |
CN113176431A (en) | Voltage sag initial time positioning analysis method and device | |
CN109581054A (en) | A kind of real-time harmonic rapid detection method of bank base conversion power supply system peculiar to vessel | |
Subtirelu | Harmonic distortions analyzer for power rectifiers | |
Huang et al. | Power quality disturbances recognition based on HS-transform | |
Lan et al. | A Novel Motor Fault Diagnosis Method Based on Generative Adversarial Learning with Distribution Fusion of Discrete Working Conditions. | |
CN105044457B (en) | A kind of antimierophonic trend of harmonic detection method of power |
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 |