CN106845337B - Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type - Google Patents

Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type Download PDF

Info

Publication number
CN106845337B
CN106845337B CN201611130754.4A CN201611130754A CN106845337B CN 106845337 B CN106845337 B CN 106845337B CN 201611130754 A CN201611130754 A CN 201611130754A CN 106845337 B CN106845337 B CN 106845337B
Authority
CN
China
Prior art keywords
disturbance
arc length
time
length difference
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.)
Active
Application number
CN201611130754.4A
Other languages
Chinese (zh)
Other versions
CN106845337A (en
Inventor
卓金宝
施伟锋
卢捷
毋恒先
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Maritime University
Original Assignee
Shanghai Maritime University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Maritime University filed Critical Shanghai Maritime University
Priority to CN201611130754.4A priority Critical patent/CN106845337B/en
Publication of CN106845337A publication Critical patent/CN106845337A/en
Application granted granted Critical
Publication of CN106845337B publication Critical patent/CN106845337B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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 And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
  • Measurement Of Length, Angles, Or The Like Using Electric Or Magnetic Means (AREA)

Abstract

The invention provides an arc length difference sequence method for ship power quality disturbance time positioning and type identification. The method processes the electric energy signal filtered by the morphological filter based on the arc length difference sequence, and can realize time positioning and type identification of 5 common single disturbances and 3 common composite disturbances in a ship power system. The method comprises the steps of firstly adopting an improved alternating mixed form filter to filter ship electric energy signals, then providing definition of an arc length difference sequence, using the definition to accurately time and position disturbance time, extracting disturbance time-frequency domain characteristics, amplitude variation characteristics and wave crest characteristics, and finally inputting characteristic quantity and disturbance time into a type identifier with a disturbance type identification rule defined to realize disturbance type identification. The method can be used for quickly identifying the disturbance in the actual ship power system in a type mode, and is suitable for real-time detection and type identification.

Description

Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type
The technical field is as follows:
the invention relates to ship electric energy disturbance time positioning and type identification, in particular to disturbance time positioning and type identification of single disturbance and composite disturbance in a ship power system based on an arc length difference sequence.
Background art:
the ship industry in China develops rapidly, and with the great innovation of ship energy devices and power device technologies, electric propulsion type ships become mainstream ship types in the ship market. In the popularization of electric propulsion type ships, the important technical problem is to ensure the safe, reliable and stable operation of a ship electric power system, and the rapid and accurate time positioning and the disturbance in the type identification of ship electric energy are important foundations for solving the technical problem. Therefore, the method has profound significance in researching ship electric energy disturbance time positioning and type identification.
The morphological filter filtering method is a time domain nonlinear filtering method based on set theory and derived from image feature extraction, has the characteristics of simple calculation and strong real-time performance, and is a time domain filtering method which is concerned in recent years. The research of the denoising method based on the classical morphological principle is spread around two main aspects: and selecting a self-adaptive structural element and designing a novel operator function. However, although these studies improve the filtering effect to different degrees, the algorithm complexity and the calculation amount are greatly increased, and the generalization application capability is worth to be deduced.
The common methods for positioning the ship electric energy disturbance time and identifying the type are Fourier analysis, short-time Fourier analysis and wavelet analysis, and electric energy signals are analyzed from the aspect of frequency domain. However, the working environment of the ship power system is filled with a large amount of high-intensity electromagnetic noise, especially for military ships, a large amount of noise frequencies are distributed in the frequency domain, and the time positioning and the type identification disturbance are a difficult task by analyzing the electric energy signal by adopting a frequency domain method. Under the background of strong noise, time domain waveform information of the signal is not easy to generate distortion, for example, a sinusoidal voltage waveform is polluted by noise, the waveform becomes rougher and submerges information of a detail area, but the waveform still has characteristics of the sinusoidal waveform at the moment, and the waveform of the detail area can be finely described through filtering, so that disturbance time positioning and type identification are carried out by fully utilizing the time domain waveform information of the signal, and the method is a more feasible method compared with a frequency domain method for solving the problem of disturbance time positioning and type identification of a ship power system.
In view of the above analysis, in order to solve the problems of positioning and type identification of the electric energy disturbance time of the ship power system under the condition of strong noise pollution, a ship electric energy disturbance time positioning and type identification method based on an improved alternating hybrid filter and an arc length difference sequence is provided from the perspective of time domain analysis.
The invention content is as follows:
the invention aims to provide a ship electric energy disturbance time positioning and type identification method based on an improved alternative mixed form filter and an arc length difference sequence, aiming at the problems of time positioning and identification of electric energy signal disturbance in a ship electric power system in a ship strong electromagnetic noise working environment. The method comprises the steps of filtering ship electric energy signals by improving an alternative mixed form filter, accurately positioning disturbance occurrence time by combining the definition of an arc length difference sequence, extracting waveform amplitude characteristics and wave crest number characteristics of the filtered signals, inputting the waveform amplitude characteristics and the wave crest number characteristics into a disturbance type identifier with a type discrimination rule defined, and realizing final disturbance type identification.
A certain sampling point in the signal is set as a point a, the points of each quarter signal period (or one half signal period and one signal period) after the point a are sequentially marked as a point b, a point c, a point d and the like, the point a is called as the starting point of an arc length sequence, the point b, the point c, the point d and the like are period points selected by the arc length sequence according to one quarter (or one half and one) electric period, and the arc lengths of curves between every two points can form the arc length sequence
Figure GDA0002416208590000028
Then the (first order) arc length difference sequence can be defined as the following formula (1):
Figure GDA0002416208590000021
the non-zero term in Δ s is referred to as a first non-zero element, a second non-zero element, a third non-zero element, and the like in order.
For the convenience of the method steps, let the amplitude of the normal voltage signal be A, the frequency of the signal be 50Hz, and k points are sampled in each electrical cycle, so that the number of sampling points in a quarter electrical cycle (0.005s) is set as
Figure GDA0002416208590000022
Let X0＝{x1,x2,…,xnIs the sampled signal, and x1Is the zero crossing. The method comprises the following steps:
step 1) taking the amplitude A of the signal as a reference value, and standardizing the original sampling signal. Sample signal X is sampled according to the following formula (2)0Carrying out standardization to obtain a standardized signal X1：
Figure GDA0002416208590000023
Step 2) filtering the signal processed in step 1) by using an improved alternative mixed-mode filter, wherein the improved alternative mixed-mode filter is shown as the following formula (3):
almix(n)＝((f)OC(g)+(f)CO(g))(n)/2 (3)
wherein:
Figure GDA0002416208590000024
Figure GDA0002416208590000025
Figure GDA0002416208590000026
Figure GDA0002416208590000027
Figure GDA0002416208590000031
Figure GDA0002416208590000032
f is the sampling signal, g (m) is the structural element, Df1＝{p,…,q}，p＝[N/1000]，q＝N-[N/1000]；Dg(iii) {0,1,2, …, M }, where M is the number of elements in the structural element g (M); n and M are integers, and N is more than or equal to M;
step 3) accurately positioning the disturbance time, which comprises the following three steps: a. calculating an arc length difference sequence of the filtering signals in the step 2) according to a half electric cycle, and positioning a disturbance time interval; b. calculating an arc length difference sequence of the filtering signals in the step 2) according to a quarter of the electric cycle, and positioning the accurate time in the disturbance interval; c. and (3) integrating the disturbance time interval and the precise moment in the disturbance interval to finish precise positioning of the disturbance time, wherein the specific process is as follows:
step 3.1) calculating the arc length difference sequence of the filtering signals in the step 2) according to one half of the electric cycle, and positioning a disturbance time interval:
step 3.1.1) First, with zero crossing x1As a starting point, a cycle point is taken in a quarter of an electrical cycle. Calculating the arc length between two periodic points to obtain an arc length sequence X3Finally, calculating to obtain an arc length difference sequence X as shown in formula (1)4；
Step 3.1.2) extracting the first non-zero element in the arc length difference sequence in the step, searching an arc related to the first non-zero element in the arc length sequence, wherein the arc is positioned in a time interval [ t [ [ t ]1,t1+0.005]The disturbance occurrence time is included;
step 3.2) calculating the arc length difference sequence of the filtering signals in the step 2) according to a quarter of the electric cycle, and positioning the accurate time in the disturbance interval:
step 3.2.1) carrying out secondary sampling on the points in the first quarter period of the filtered signal in the step 2) according to the sampling frequency selected by the user;
step 3.2.2) taking sampling points in the first quarter of the electrical cycle as starting points in sequence, taking periodic points according to a half of the electrical cycle, and calculating to obtain an arc length difference sequence group;
step 3.2.3) extracting first non-zero elements in the arc length difference sequence group in the step 3.2.2) to form a first non-zero element set;
step 3.2.4) taking the absolute value of the elements in the first non-zero element set in step 3.2.3), and recording as X5Searching the maximum value in the elements, and recording the serial number of the maximum value in the set as j;
step 3.3) integrating the disturbance time interval and the precise time in the disturbance interval to finish precise positioning of the disturbance time, and calculating the disturbance time T according to the following formula (10), wherein the unit is second:
Figure GDA0002416208590000041
step 4) extracting an arc length difference sequence X at the disturbance generation stage according to the disturbance time positioning result in the step 3) and the filtered signal in the step 2)4' sum disturbance generation stage filtered signal X2' identifying the disturbance type, which comprises two steps: a. to X4' feature processing, which can identify the transient rise, the transient fall, the transient oscillation and the diagnostic disturbance; b. to X2' feature processing, recognizing interrupts, transient oscillations, ramp-up/ramp-down, and harmonics, recording the filtered signal X2The method comprises the following steps of firstly, acquiring a plurality of sampling points, wherein the sampling points are Z, and the specific disturbance type identification step is as follows;
step 4.1) arc length difference sequence X for disturbance generation stage4' performing characteristic processing:
step 4.1.1) arc length difference sequence X for disturbance generation stage4' median filtering is performed with a filter template length of
Figure GDA0002416208590000042
And taking the number as an even number to obtain a waveform amplitude characteristic quantity S after the disturbance occurs;
step 4.1.2) judging whether the waveform amplitude characteristic quantity S is larger than 0, if so, turning to step 4.3); if not, turning to step 4.4);
step 4.1.3) determining whether the first element and the last element of S are greater than 0.1, i.e. Sstart-SendIf the transient oscillation disturbance is greater than 0.1, judging the disturbance as transient oscillation disturbance; if not, determining to be transient-rise disturbance;
step 4.1.4) judging whether S is larger than-0.9 and smaller than-0.1, if so, judging that the disturbance is temporarily reduced; if not, determining to be interrupted;
step 4.2) taking
Figure GDA0002416208590000043
Filtered signal X to disturbance generation stage2' performing characteristic processing:
step 4.2.1) post-filter signal X for disturbance occurrence stage2Detecting the number of wave crests, and recording the result as a fluctuation frequency characteristic P;
step 4.2.2) judging whether P is equal to zero, if so, judging to be interrupted; otherwise, turning to the step 4.7);
step 4.2.3) judging whether P is more than or equal to M and less than or equal to M +1, if so, judging that the disturbance is temporarily increased or temporarily decreased, and otherwise, turning to the step 4.8);
and 4.2.4) judging whether P is more than or equal to N, if so, judging the transient oscillation, otherwise, judging the harmonic disturbance, and the highest harmonic is P/[ z/k ] ] subharmonic.
Has the advantages that:
the ship electric energy disturbance time positioning and type identification method based on the improved alternating mixed form filter and the arc length difference sequence has the following two biggest characteristics: the mean value function is adopted to replace extreme value operation in the morphological filter, so that the waveform of the filtered signal is smoother, and the waveform characteristic is more obvious; and (3) providing definition of an arc length difference sequence, and realizing extraction of disturbance characteristics based on the sequence. The method provides a new research idea for the research of the ship electric energy disturbance time positioning and disturbance identification method from the aspect of time domain waveform analysis, and has the characteristic of theoretical innovation. The method can realize accurate time positioning of the ship power signal disturbance, and effectively identify 5 single type disturbances and 3 complex type disturbances. The method can be used for quickly positioning and identifying the electric energy disturbance in the actual ship.
Description of the drawings:
the invention is further described below in conjunction with the appended drawings and the detailed description.
FIG. 1 is a flow chart of ship electric energy disturbance time positioning and type identification;
FIG. 2 is a flow chart of accurate positioning of the disturbance time of the ship electric energy;
FIG. 3 is a flow chart of a ship power type identification;
FIG. 4 shows the positioning of the voltage sag disturbance time and the extraction of disturbance characteristics;
FIG. 5 shows the voltage sag disturbance time positioning and disturbance feature extraction results;
FIG. 6 illustrates the positioning of the disturbance time of voltage interruption and the extraction of disturbance characteristics;
FIG. 7 shows the voltage harmonic disturbance time positioning and disturbance feature extraction results;
FIG. 8 shows the voltage transient oscillation time positioning and disturbance feature extraction results;
FIG. 9 shows the voltage sag plus harmonic composite disturbance time localization and disturbance feature extraction results;
FIG. 10 shows the voltage ramp plus harmonic composite disturbance time positioning and disturbance feature extraction results;
FIG. 11 shows the results of the positioning of the voltage sag disturbance time and the extraction of disturbance characteristics, including phase jump;
the specific implementation mode is as follows:
in order to make the technical means, the creation features, the achievement purposes and the effects of the invention easy to understand, the embodiments of the invention are further described below with reference to the specific drawings.
Firstly, assuming that the amplitude of a normal voltage signal is A, the frequency of the signal is 50Hz, k points are sampled in each electrical cycle, and the number of sampling points in a quarter electrical cycle (0.005s) is
Figure GDA0002416208590000051
Let X0＝{x1,x2,…,xnIs the sampled signal, and x1Is the zero crossing.
The basic flow of the overall method is shown in fig. 1, filtering the standardized ship electric energy signal in and out, respectively inputting the filtered signal into a disturbance time positioning process and a disturbance identification process, and inputting the time positioning result and the filtered signal into a type identification process, so as to finally realize time positioning and type identification, wherein the method comprises the following specific implementation steps:
step 1) taking the amplitude A of the signal as a reference value, and standardizing the original sampling signal. Sampling signal X according to the following formula (1)0Carrying out standardization to obtain a standardized signal X1：
Figure GDA0002416208590000061
Step 2) filtering the signal processed in step 1) by using an improved alternative mixed-mode filter, wherein the improved alternative mixed-mode filter is shown as the following formula (2):
almix(n)＝((f)OC(g)+(f)CO(g))(n)/2 (2)
wherein:
Figure GDA0002416208590000062
Figure GDA0002416208590000063
Figure GDA0002416208590000064
Figure GDA0002416208590000065
Figure GDA0002416208590000066
Figure GDA0002416208590000067
f is the sampling signal, N is the sample length, g (m) is the structural element, Df1＝{p,…,q}，p＝[N/1000]，q＝N-[N/1000]；Dg(iii) {0,1,2, …, M }, where M is the number of elements in the structural element g (M); n and M are integers, and N is more than or equal to M;
step 3) accurately positioning the disturbance time, as shown in fig. 2, the method is divided into three steps: a. calculating an arc length difference sequence of the filtering signals in the step 2) according to a half electric cycle, and positioning a disturbance time interval; b. calculating an arc length difference sequence of the filtering signals in the step 2) according to a quarter of the electric cycle, and positioning the accurate time in the disturbance interval; c. and finishing the accurate positioning of the disturbance time, wherein the specific process is as follows:
step 3.1) calculating the arc length difference sequence of the filtering signals in the step 2) according to one half of the electric cycle, and positioning a disturbance time interval:
step 3.1.1) first, with a zero crossing x1As a starting point, a cycle point is taken in a quarter of an electrical cycle. Calculating the arc length between two periodic points to obtain an arc length sequence X3Finally, calculating to obtain an arc length difference sequence X as shown in formula (1)4；
Step 3.1.2) extracting step arc length difference sequenceFind the arc in the arc length sequence related to the first non-zero element, and the time interval [ t ] of the arc1,t1+0.005]The disturbance occurrence time is included;
step 3.2) calculating the arc length difference sequence of the filtering signals in the step 2) according to a quarter of the electric cycle, and positioning the accurate time in the disturbance interval:
step 3.2.1) carrying out secondary sampling on the points in the first quarter period of the filtered signal in the step 2) according to the sampling frequency selected by the user;
step 3.2.2) taking sampling points in the first quarter of the electrical cycle as starting points in sequence, taking periodic points according to a half of the electrical cycle, and calculating to obtain an arc length difference sequence group;
step 3.2.3) extracting first non-zero elements in the arc length difference sequence group in the step 3.2.2) to form a first non-zero element set;
step 3.2.4) taking the absolute value of the elements in the first non-zero element set in step 3.2.3), and recording as X5Searching the maximum value in the elements, and recording the serial number of the maximum value in the set as j;
step 3.3) integrating the disturbance time interval and the precise time in the disturbance interval to finish precise positioning of the disturbance time, and calculating the disturbance time T according to the following formula (9), wherein the unit is second:
Figure GDA0002416208590000071
step 4) extracting an arc length difference sequence X at the disturbance generation stage according to the disturbance time positioning result in the step 3) and the filtered signal in the step 2)4' sum disturbance generation stage filtered signal X2', the disturbance type identification is performed, as shown in FIG. 3, in two steps: a. to X4' feature processing, which can identify the transient rise, the transient fall, the transient oscillation and the diagnostic disturbance; b. to X2' feature processing, recognizing interrupts, transient oscillations, ramp-up/ramp-down, and harmonics, recording the filtered signal X2Z sampling points are arranged in the method, and the disturbance type identification step is as follows;
step 4.1) arc length difference sequence X for disturbance generation stage4' performing characteristic processing:
step 4.1.1) arc length difference sequence X for disturbance generation stage4' median filtering is performed with a filter template length of
Figure GDA0002416208590000072
And taking the number as an even number to obtain a waveform amplitude characteristic quantity S after the disturbance occurs;
step 4.1.2) judging whether the waveform amplitude characteristic quantity S is larger than 0, if so, turning to step 4.3); if not, turning to step 4.4);
step 4.1.3) determining whether the first element and the last element of S are greater than 0.1, i.e. Sstart-SendIf the transient oscillation disturbance is greater than 0.1, judging the disturbance as transient oscillation disturbance; if not, determining to be transient-rise disturbance;
step 4.1.4) judging whether S is larger than-0.9 and smaller than-0.1, if so, judging that the disturbance is temporarily reduced; if not, determining to be interrupted;
step 4.2) taking
Figure GDA0002416208590000081
Filtered signal X to disturbance generation stage2' performing characteristic processing:
step 4.2.1) post-filter signal X for disturbance occurrence stage2Detecting the number of wave crests, and recording the result as a fluctuation frequency characteristic P;
step 4.2.2) judging whether P is equal to zero, if so, judging to be interrupted; otherwise, turning to the step 4.7);
step 4.2.3) judging whether P is more than or equal to M and less than or equal to M +1, if so, judging that the disturbance is temporarily increased or temporarily decreased, and otherwise, turning to the step 4.8);
and 4.2.4) judging whether P is more than or equal to N, if so, judging the transient oscillation, otherwise, judging the harmonic disturbance, and the highest harmonic is P/[ z/k ] ] subharmonic.
The method of the invention is verified and explained by simulation tests as follows: the method comprises the steps of constructing 5 single-type disturbances such as disturbance voltage temporary rise, voltage temporary drop, interruption, harmonic waves and transient oscillation, 3 composite-type disturbances such as temporary rise plus harmonic waves, temporary drop plus harmonic waves and temporary drop containing phase jump, and constructing signal models of eight types of disturbances, wherein the signal models are shown in table 1.
Meter 1 electric energy disturbance signal simulation model and parameter setting in simulation
Figure GDA0002416208590000082
Figure GDA0002416208590000091
Wherein u (t) is defined as the following formula (10):
Figure GDA0002416208590000092
generating simulation signals of various types of disturbances in MATLAB, adding white noise to enable the signal-to-noise ratio to be 40dB, sampling 200 points in each electrical period, sampling 10kHz, sampling 20 electrical periods, and setting 4001 sampling points uniformly between the sampling points 820-2030 to generate disturbance signals, namely, within the period of 0.082-0.203 s. Different parameters of the signal model in the simulation process are set according to the parameter column in table 1.
The variation curves of some key feature quantities in the simulation process are shown in fig. 4 to 11, where X2For signals filtered by the improved morphological filter, X3Taking the jth sampling point found in the first quarter of the electric cycle as the starting point, selecting the cycle point in the half of the electric cycle, and finally forming an arc length sequence, X4According to the arc length sequence X3Calculated arc length difference sequence, X5The first non-zero element set after the absolute value is obtained in the step 3.6), and S is the waveform amplitude characteristic quantity in the step 4.1).
The method of the invention is adopted to process the disturbance signal, the secondary sampling problem is not considered (namely step 3.3 is skipped), and the positioning result and the specific disturbance characteristics are shown in table 2.
TABLE 28 TYPE TIME LOCATION AND DISTURBING CHARACTERISTICS
Figure GDA0002416208590000093
As can be seen from Table 2, the error between the disturbance positioning time T and the real occurrence time is 0.1-0.2 ms, and the error is smaller along with the increase of the sampling frequency, so that the requirement of quickly positioning disturbance in engineering can be met. Meanwhile, the identification result of the disturbance type is completely correct. When single disturbance and composite disturbance occur, the plus and minus of the S sign of the waveform amplitude characteristic quantity represents the increase and decrease of the waveform amplitude, and the value represents the change degree of the waveform amplitude. The fluctuation frequency characteristic P is the statistics of the number of wave crests, has strong identification capability on single disturbance or load disturbance, and has certain anti-interference capability.
When the voltage is temporarily increased and disturbed, it is shown in fig. 4. Arc length sequence X3The value becomes large during the perturbation; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The maximum value of the medium value is the 22 nd element, the disturbance time is calculated according to the formula (4) and is 0.0822s, and the error with the actual disturbance time is 0.2 ms; the waveform amplitude characteristic quantity S rises to 0.5 during the disturbance period, and is judged as transient-rising disturbance according to a disturbance identification algorithm.
The voltage sag is disturbed as shown in fig. 5. Arc length sequence X3The value decreases during the perturbation; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The maximum value of the medium value is the 22 nd element, the disturbance time is calculated according to the formula (4) and is 0.0822s, and the error with the actual disturbance time is 0.2 ms; and the waveform amplitude characteristic quantity S is reduced to-0.3 in the disturbance period, and is judged to be sag disturbance according to a disturbance identification algorithm.
When the voltage is interrupted from the disturbance, as shown in fig. 6. Arc length sequence X3The value decreases during the perturbation; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The maximum value of the interference is the 21 st element, the disturbance moment is calculated as 0.0821s according to the formula (4), and the actual disturbance moment is calculatedThe error of the moving time is 0.1 ms; the waveform amplitude characteristic quantity S is reduced to-0.9 in the disturbance period, and the interruption disturbance is judged according to a disturbance identification algorithm.
When the voltage transient oscillation is disturbed, as shown in fig. 7. Arc length sequence X3The value becomes large during the perturbation; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The medium maximum value is the 20 th element, and the disturbance time is calculated to be 0.082s according to the formula (4) and is the same as the actual disturbance time; the waveform amplitude characteristic quantity S is shown as ascending first and then descending in the disturbance period, the phase difference of the first element and the last element is larger than 0.1, meanwhile, the peak detection result is P-119, and transient oscillation disturbance is judged according to a disturbance identification algorithm.
Voltage harmonics are disturbed as shown in fig. 8. Arc length sequence X3The value becomes large during the perturbation; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The maximum value of the medium value is the 22 nd element, the disturbance time is calculated according to the formula (4) and is 0.0822s, and the error between the disturbance time and the actual disturbance time is 0.2 ms; the waveform amplitude characteristic quantity S is increased to 0.5 in the disturbance period, meanwhile, the peak detection result is that P is 42, and harmonic disturbance is judged according to a disturbance identification algorithm.
Voltage sag plus harmonic disturbance is shown in fig. 9. Arc length sequence X3The value is increased and then decreased during the disturbance; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The medium maximum value is the 20 th element, and the disturbance time is calculated to be 0.082s according to the formula (4) and is the same as the actual disturbance time; the waveform amplitude characteristic quantity S is reduced to-0.11 in the disturbance period, meanwhile, the peak detection result is P-42, and the temporary reduction plus harmonic disturbance is judged according to the disturbance identification algorithm.
When the voltage is ramped plus harmonic disturbances, as shown in fig. 10. Arc length sequence X3The value becomes large during the perturbation; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The medium maximum value isThe 20 elements are calculated according to the formula (4) that the disturbance time is 0.0820s, which is the same as the actual disturbance time; the waveform amplitude characteristic quantity S is increased to 0.9 in the disturbance period, meanwhile, the peak detection result is that P is 42, and the transient-increase and harmonic disturbance is judged according to the disturbance identification algorithm.
When the dip disturbance including the phase jump is present, it is shown in fig. 11. Arc length sequence X3The value decreases during the perturbation; the arc length difference sequence has non-zero values at the time of disturbance occurrence and ending, generates pulses and is zero at other times; first set of non-zero elements X5The maximum value of the medium value is the 22 nd element, the disturbance time is calculated according to the formula (4) and is 0.0822s, and the error between the disturbance time and the actual disturbance time is 0.2 ms; and the waveform amplitude characteristic quantity S is reduced to-0.5 in the disturbance period, meanwhile, the peak detection result is P-7, and the transient-reduction disturbance containing phase jump is judged according to a disturbance identification algorithm.

Claims (2)

1. An arc length difference sequence method for ship power quality disturbance time positioning and type identification defines an arc length difference sequence as follows:
setting a certain sampling point in the signal as a point a, each quarter of signal period after the point a, or one half of signal period, or the points of one signal period as b point, c point, d point and e point in turn, and calling the point a as the starting point of arc length sequence, the point b, c point, d point and e point as the arc length sequence according to one quarter, one half or one cycle point selected by electric period, the arc lengths of curves between every two points can form the arc length sequence as
Figure FDA0002416208580000011
Then the arc length difference sequence can be defined as the following formula (1), and the non-zero term in Δ s is sequentially called as a first non-zero element, a second non-zero element, and a third non-zero element:
Figure FDA0002416208580000012
if sampling k points according to 0.02s of each electric cycle is adopted to sample the electric energy signal, the number of sampling points in a quarter of the electric cycle of 0.005s is equal to
Figure FDA0002416208580000013
X0＝{x1,x2,…,xnIs the sampled signal, and x1Is a zero crossing point; the method is characterized in that the arc length difference sequence method for the disturbance time positioning and the type identification of the ship power quality comprises the following steps:
step 1) taking the amplitude A of the signal as a reference value, standardizing the original sampling signal, and carrying out the following formula (2) on the sampling signal X0Carrying out standardization to obtain a standardized signal X1：
Figure FDA0002416208580000014
Step 2) filtering the signal processed in step 1) by using an improved alternative mixed-mode filter, wherein the improved alternative mixed-mode filter is shown as the following formula (3):
almix(n)＝((f)OC(g)+(f)CO(g))(n)/2 (3)
wherein:
Figure FDA0002416208580000015
Figure FDA0002416208580000016
Figure FDA0002416208580000017
Figure FDA0002416208580000018
Figure FDA0002416208580000019
Figure FDA00024162085800000110
f is the sampling signal, g (m) is the structural element, Df1＝{p,…,q}，p＝[N/1000]，q＝N-[N/1000]；Dg(iii) {0,1,2, …, M }, where M is the number of elements in the structural element g (M); n and M are integers, and N is more than or equal to M;
step 3) accurately positioning the disturbance time, which comprises the following three steps: a. calculating an arc length difference sequence of the filtering signals in the step 2) according to a half electric cycle, and positioning a disturbance time interval; b. calculating an arc length difference sequence of the filtering signals in the step 2) according to a quarter of the electric cycle, and positioning the accurate time in the disturbance interval; c. and (3) integrating the disturbance time interval and the precise moment in the disturbance interval to finish precise positioning of the disturbance time, wherein the specific process is as follows:
step 3.1) calculating the arc length difference sequence of the filtering signals in the step 2) according to one half of the electric cycle, and positioning a disturbance time interval:
step 3.1.1) first, with a zero crossing x1Taking periodic points according to a quarter of an electric cycle as a starting point, calculating the arc length between two periodic points to obtain an arc length sequence X3Finally, calculating to obtain an arc length difference sequence X as shown in formula (1)4；
Step 3.1.2) extracting the first non-zero element in the arc length difference sequence in the step, searching an arc related to the first non-zero element in the arc length sequence, wherein the arc is positioned in a time interval [ t [ [ t ]1,t1+0.005]The disturbance occurrence time is included;
step 3.2) calculating the arc length difference sequence of the filtering signals in the step 2) according to a quarter of the electric cycle, and positioning the accurate time in the disturbance interval:
step 3.2.1) carrying out secondary sampling on the points in the first quarter period of the filtered signal in the step 2) according to the sampling frequency selected by the user;
step 3.2.2) taking sampling points in the first quarter of the electrical cycle as starting points in sequence, taking periodic points according to a half of the electrical cycle, and calculating to obtain an arc length difference sequence group;
step 3.2.3) extracting first non-zero elements in the arc length difference sequence group in the step 3.2.2) to form a first non-zero element set;
step 3.2.4) taking the absolute value of the elements in the first non-zero element set in step 3.2.3), and recording as X5Searching the maximum value in the elements, and recording the serial number of the maximum value in the set as j;
step 3.3) integrating the disturbance time interval and the precise time in the disturbance interval to finish precise positioning of the disturbance time, and calculating the disturbance time T according to the following formula (10), wherein the unit is second:
Figure FDA0002416208580000031
step 4) extracting an arc length difference sequence X at the disturbance generation stage according to the disturbance time positioning result in the step 3) and the filtered signal in the step 2)4' sum disturbance generation stage filtered signal X2' identifying the disturbance type, which comprises two steps: a. to X4' feature processing, capable of identifying transient rises, dips, transient oscillations and diagnostic perturbations; b. to X2' feature processing, capable of identifying interrupts, transient oscillations, ramps/dips and harmonics, recording the filtered signal X2The method comprises the following steps of firstly, acquiring a plurality of sampling points, wherein the sampling points are Z, and the specific disturbance type identification step is as follows;
step 4.1) arc length difference sequence X for disturbance generation stage4' performing characteristic processing:
step 4.1.1) arc length difference sequence X for disturbance generation stage4' median filtering is performed with a filter template length of
Figure FDA0002416208580000032
And taking the number as an even number to obtain a waveform amplitude characteristic quantity S after the disturbance occurs;
step 4.1.2) judging whether the waveform amplitude characteristic quantity S is larger than 0, if so, turning to step 4.3); if not, turning to step 4.4);
step 4.1.3) determining whether the first element and the last element of S are greater than 0.1, i.e. Sstart-SendIf the value is more than 0.1, judging the disturbance as transient vibration if the value is more than 0.1Oscillating and disturbing; if not, determining to be transient-rise disturbance;
step 4.1.4) judging whether S is larger than-0.9 and smaller than-0.1, if so, judging that the disturbance is temporarily reduced; if not, determining to be interrupted;
step 4.2) taking
Figure FDA0002416208580000033
Filtered signal X to disturbance generation stage2' performing characteristic processing:
step 4.2.1) post-filter signal X for disturbance occurrence stage2Detecting the number of wave crests, and recording the result as a fluctuation frequency characteristic P;
step 4.2.2) judging whether P is equal to zero, if so, judging to be interrupted; otherwise, turning to the step 4.7);
step 4.2.3) judging whether P is more than or equal to M and less than or equal to M +1, if so, judging that the disturbance is temporarily increased or temporarily decreased, and otherwise, turning to the step 4.8);
and 4.2.4) judging whether P is more than or equal to N, if so, judging the transient oscillation, otherwise, judging the harmonic disturbance, and the highest harmonic is P/[ z/k ] ] subharmonic.
2. The arc length difference sequence method for ship power quality disturbance time positioning and type identification as claimed in claim 1, characterized in that the number of wave peak values whose amplitudes of signal waveforms in a certain specified period of time are positive is detected, and if the amplitudes in a certain period of time are equal, only the starting time of the certain period of time is recorded as the peak value, where P is 1.
CN201611130754.4A 2016-12-09 2016-12-09 Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type Active CN106845337B (en)

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 CN106845337A (en) 2017-06-13
CN106845337B true 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)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117173477B (en) * 2023-09-05 2024-06-21 北京交通大学 Domain generalization pedestrian re-identification method based on multi-layer data disturbance strategy

Citations (2)

* Cited by examiner, † Cited by third party
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

Patent Citations (2)

* Cited by examiner, † Cited by third party
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)

* Cited by examiner, † Cited by third party
Title
利用广义形态滤波与差分熵的电能质量扰动检测;赵静等;《中国电机工程学报》;20090305;第29卷(第7期);第121-127页 *
基于数学形态学和网格分形的电能质量扰动检测及定位;李庚银等;《中国电机工程学报》;20060228;第26卷(第8期);第25-30页 *

Also Published As

Publication number Publication date
CN106845337A (en) 2017-06-13

Similar Documents

Publication Publication Date Title
Laila et al. A refined Hilbert–Huang transform with applications to interarea oscillation monitoring
Lavopa et al. Real-time estimation of fundamental frequency and harmonics for active shunt power filters in aircraft electrical systems
Gargoom et al. Automatic classification and characterization of power quality events
Cupertino et al. Running DFT-based PLL algorithm for frequency, phase, and amplitude tracking in aircraft electrical systems
CN101964050B (en) Method for identifying modal parameter based on model order determination and signal noise elimination
CN103941072B (en) A kind of electric power signal mutation parameter measuring method based on real number Strong tracking filter
CN103424183B (en) Method for eliminating abnormal interference on detection for mechanical vibration signals
CN108535613B (en) Voltage flicker parameter detection method based on combined window function
CN105425011B (en) A kind of non-linear width phase detection method suitable for single-phase network deformation
CN106845334A (en) A kind of innovative noise extracting method based on mathematical morphology
CN103944174A (en) Low frequency oscillation online identification method based on cross-correlation function denoising algorithm
CN103163372A (en) Method for analyzing harmonics of power system by adopting Hilbert-Huang transform (HHT)
Chen et al. Nonstationary signal denoising using an envelope-tracking filter
Fu et al. An improved discrete Fourier transform-based algorithm for electric network frequency extraction
CN105930818A (en) Data processing method for increasing EMD denoising capability
CN106845337B (en) Arc length difference sequence method for positioning ship electric energy disturbance time and identifying type
CN111190049B (en) Method for detecting nano-volt level weak sinusoidal signal by chaotic system of principal component analysis
Shao et al. Feature patterns extraction-based amplitude/frequency modulation model for vortex flow sensor output signal
CN110826498A (en) Transient power quality detection method based on HHT
Lv et al. High-order synchroextracting chirplet transform for accurate instantaneous frequency estimation and its application in fault diagnosis of rotary machinery
Djurovic Estimation of the sinusoidal signal frequency based on the marginal median DFT
CN108959739A (en) A kind of analysis method and device of the pressure fluctuation of hydroenergy storage station transient process
CN117289022A (en) Power grid harmonic detection method and system based on Fourier algorithm
CN102680080B (en) Unsteady-state signal detection method based on improved self-adaptive morphological filtering
CN109030897B (en) Fourier transform-based electricity stealing detection method

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