Nothing Special   »   [go: up one dir, main page]

CN109375037B - Design method of single-phase earth fault alarm of medium-voltage ship power system - Google Patents

Design method of single-phase earth fault alarm of medium-voltage ship power system Download PDF

Info

Publication number
CN109375037B
CN109375037B CN201811365359.3A CN201811365359A CN109375037B CN 109375037 B CN109375037 B CN 109375037B CN 201811365359 A CN201811365359 A CN 201811365359A CN 109375037 B CN109375037 B CN 109375037B
Authority
CN
China
Prior art keywords
alarm
confidence
alarm confidence
dynamic
static
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
CN201811365359.3A
Other languages
Chinese (zh)
Other versions
CN109375037A (en
Inventor
徐晓滨
翁旭
胡燕祝
高海波
高迪驹
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.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi 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 Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN201811365359.3A priority Critical patent/CN109375037B/en
Publication of CN109375037A publication Critical patent/CN109375037A/en
Application granted granted Critical
Publication of CN109375037B publication Critical patent/CN109375037B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Locating Faults (AREA)
  • Monitoring And Testing Of Nuclear Reactors (AREA)

Abstract

The invention discloses a design method of a single-phase earth fault alarm of a medium-voltage ship power system, and belongs to the field of safe operation and maintenance of ships. The invention firstly carries out static fusion on two alarm confidence coefficients containing uncertainty at each moment by utilizing an alarm confidence coefficient static fusion rule to obtain the static alarm confidence coefficient at each moment, then carries out dynamic fusion on the static alarm confidence coefficient at the current moment and the dynamic alarm confidence coefficient at the previous moment by utilizing an alarm confidence coefficient dynamic fusion rule to obtain the dynamic alarm confidence coefficient at the current moment, and judges whether to send out an alarm or not under a relevant judgment criterion.

Description

Design method of single-phase earth fault alarm of medium-voltage ship power system
Technical Field
The invention relates to a design method of a single-phase earth fault alarm of a medium-voltage ship power system, and belongs to the field of safe operation and maintenance of ships.
Background
In the power system for the medium-voltage ship, the number of cables is large, the total length is long, the equivalent ground capacitance is obviously larger than that of the power system for the low-voltage ship, and the damage of single-phase ground fault cannot be ignored. The single-phase earth fault is a frequent fault, which can cause the non-earth voltage to rise relative to the earth voltage, and the long-term fault operation will affect the insulation level of the cable, resulting in the accelerated aging of the cable. It also causes arc at the ground point (discontinuity), resulting in a serious fault, which develops into a two-phase and three-phase (short circuit) ground fault. In general, effective protection against such faults can be achieved by setting appropriate directional zero-sequence overcurrent protection.
The single-phase earth fault in the medium-voltage marine power system is to quantitatively analyze the phase and amplitude of the generated zero-sequence voltage and current and then design an alarm to effectively monitor the single-phase earth state of the medium-voltage power system in real time. However, whether an alarm is given or not can be determined only according to whether a single input signal variable triggers a set alarm threshold, and the mechanism for triggering alarm generation by the threshold under the single variable often causes the situations of false alarm, false alarm failure and the like in the practical application of a medium-voltage power system. The basic indicators that typically measure the performance of an alarm are false alarm rate, and average delay time. Digital filtering, time delay and dead zone setting are several types of common alarm design methods based on absolute threshold discrimination modes, and the traditional alarm methods adopt a mechanism that the process variable is immediately alarmed when exceeding the threshold and is immediately relieved when being lower than the threshold. However, in consideration of the complex situation of the medium-voltage power system, these conventional methods cannot well eliminate various uncertain interferences in the operation of the medium-voltage power system and the information acquisition of the sensor, which results in an excessive false alarm rate and a too high false negative rate, and cannot achieve the purpose of accurately determining the operation state of the medium-voltage power system.
Disclosure of Invention
The invention aims to provide a design method of a single-phase earth fault alarm of a medium-voltage ship power system, which is different from an absolute threshold alarm judgment mode adopted in a traditional alarm design method.
The invention comprises the following steps:
(1) for a medium-voltage ship power system, the topology of a medium-voltage part is mostly a busbar segmented structure, each segment of busbar is supplied with power by at least one generator, load devices such as a daily transformer, a propulsion frequency converter and the like are separately connected to different busbar segments, according to the Steel sea vessel entry standard, when a single-phase ground fault occurs in the system, a zero-sequence voltage U with the inequality of 3.3 KV-10.8 KV is formed between a fault point and the ground, zero-sequence current I and a phase D are generated in a pair mode through the ground capacitance of a cable and equipment and the neutral resistance of a generator, a zero-sequence current signal is acquired by a zero-sequence current sensor, and the peak value is obtained through Fourier decomposition and is marked as x1Zero is acquired by a zero-sequence current sensorSequence current phase, noted as x2Zero sequence current peak value x1Starting from 0 to increase, zero sequence current phase x2The deviation is started from 180 degrees to-90 degrees, meanwhile, the medium-voltage power system is changed from a normal working state to a fault state, and the sensor collects zero-sequence current signals and zero-sequence current phases once every 1 second.
(2) The identification frame of the single-phase earth fault alarm is set as Θ ═ NA, a }, where NA ═ 0 represents that the power system is in a normal operation state, and a ═ 1 represents that the power system is in a single-phase earth fault state.
(3) From step (1), x1And x2Alarm variables of zero sequence current peak value and zero sequence current phase which are needed to be monitored by the single-phase earth fault alarm respectively, let x1(k) And x2(k) K is 1,2,3, …, K, which is the online measurement sequence of the sampling value of the zero sequence current peak value and the sampling value of the zero sequence current phase by the sensor, K is the sampling time, the sampling number K is more than 3600, and the alarm variable x of the zero sequence current peak value is constructed1Normal distribution X for normal operating condition and single-phase earth fault condition1,NA～N(μ1,σ1 2) And X1,A～N(μ2,σ2 2) The corresponding probability density function is f1,NA(x) And f1,A(x) Constructing alarm variable x of zero sequence current phase2Normal distribution X for normal operating condition and single-phase earth fault condition2,NA～N(μ3,σ3 2) And X2,A～N(μ4,σ4 2) The corresponding probability density function is f2,NA(x) And f2,A(x) Wherein 0A is not more than mu1<μ2<3A,σ1>0,σ2>0,180°≤μ3<360°,-90°<μ4<0°,σ3>0,σ4>0。
(4) Setting two input fault characteristics of single-phase earth fault alarm as x1(k) And x2(k) The end point set of the reference interval is T ═ Tn’1,2 …, N, and Qm’1,2 …, M, which form the zero sequence current respectivelyPeak value x of the sampled value1(k) N-1 reference intervals of { [ T ]1,T2],[T2,T3],…,[TN-2,TN-1],[TN-1,TN]Sample value x of zero sequence current phase2(k) M-1 reference intervals of { [ Q ]1,Q2],[Q2,Q3],…,[QM-2,QM-1],[QM-1,QM]And have a value of (mu)1-3*σ1)<T1<T2<…<TN<(μ2+3*σ2)，(μ4-3*σ4)＝Q1<Q2<…<QM＝(μ3+3*σ3) The output of the alarm is the single-phase grounding state of the medium-voltage power system, which is recorded as y (k), and the reference value set is W ═ Ke1,2}, where W1＝NA＝0，W2＝A＝1。
(5) Constructing an input x by step (3)1(k) And x2(k) Sequence of measured values S ═ x for output y (k)1(k)，x2(k) Y (K), K is 1,2,3, …, K, K is not less than 3600}, and x is equal to or less than 36001(k)、x2(k) And y (k) is expressed as a sample set Z ═ x1(k),x2(k),y(k)]Form (b), ascertaining x1(k) And x2(k) Therein are respectively β1Each measured value satisfies N (mu)1,σ1 2) And N (mu)3,σ3 2) β for outputs y (k) equal to 02Each measured value satisfies N (mu)2,σ2 2) And N (mu)4,σ4 2) Corresponding to output y (k) equal to 1 and having β1+β2X is equal to K1(k) And y (k) is expressed as a sample set Z' ═ x1(k),y(k)]X is to be2(k) And y (k) is expressed as a sample set Z ═ x2(k),y(k)]Using sets of samples [ x ] respectively1(k),y(k)]And [ x ]2(k),y(k)]Constructing input fault characteristics x by probability distribution on corresponding intervals1(k) And x2(k) Table of probability distribution with medium voltage power system single phase grounding state y (k) as shown in table 1 and table 2 below, respectively, where N-1, 2 …, N-1, M-1, 2, …, M-1, e-12, there are
Figure BDA0001868376040000031
Figure BDA0001868376040000032
Obviously, there are
Figure BDA0001868376040000033
αnTo fall at a reference point TnSum of probability distributions of above, and having an＝γ1,n+γ2,n，ηmTo fall at a reference point QmSum of probability distributions of (1) and has ηm＝ε1,m+ε2,m。
TABLE 1 sample [ x1(k),y(k)]Table of probability distribution
Figure BDA0001868376040000034
TABLE 2 sample [ x2(k),y(k)]Table of probability distribution
Figure BDA0001868376040000035
(6) According to the sample [ x ] obtained in the step (5)1(k),y(k)]Can obtain the current input fault feature x1(k) Fall at a reference point TnWhen above, WeConfidence of occurrence is
Figure BDA0001868376040000036
And is provided with
Figure BDA0001868376040000041
A reference point T can be defined which corresponds tonAlarm confidence of
Figure BDA0001868376040000042
In the same way, pairIn the sample [ x2(k),y(k)]The probability distribution table of (2) is defined corresponding to the reference point QmAlarm confidence of
Figure BDA0001868376040000043
Wherein
Figure BDA0001868376040000044
For the new measurement value x1(t) and x2(T), T ═ 1,2,3, …, which necessarily falls within reference point TnOr QmAt this time, the alarm confidence corresponding to the reference point is activated, and they are given by equation (2) and equation (3), respectively.
(7) Respectively obtaining the process variable x at the t moment according to the step (6)1And x2Alarm confidence of
Figure BDA0001868376040000045
And
Figure BDA0001868376040000046
then, the alarm confidence coefficient of each moment is fused by using an alarm confidence coefficient static fusion rule
Figure BDA0001868376040000047
And
Figure BDA0001868376040000048
performing static fusion to obtain static alarm confidence B at t momenttThe specific calculation process is as follows:
Bt＝Ro,b(2)，o＝{We|e＝1,2} (5)
Figure BDA0001868376040000049
wherein at the setting wi＝riWhen i is 1,2, there are
Figure BDA00018683760400000410
(8) Obtaining the static alarm confidence B at the time t according to the step (7)tThen, the static alarm confidence at the current time t is fused with the dynamic alarm confidence at the past time by using an alarm confidence dynamic fusion rule to obtain the dynamic alarm confidence at the current time t, which is marked as Et＝[Et(NA),Et(A)]The method comprises the following specific steps:
(8-1) when t is 1, there is E1＝B1＝[E1(NA),E1(A)]I.e., the dynamic alarm confidence is the static alarm confidence obtained at that time.
(8-2) setting an alarm confidence weighting wi1, i 1,2, reliability r of the static alarm confidence at the current timetCalculated by the following formula:
Figure BDA00018683760400000411
wherein r is00.5 is the initial value of reliability, τ is the reward and punishment coefficient, and is calculated by the following formula:
Figure BDA0001868376040000051
setting eNA＝(1,0,0)，eA＝(0,1,0)，B′t＝[Bt(NA),Bt(A),0]，E′t＝[Et(NA),Et(A),0]φ is a reliability enhancement factor, calculated by the following equation:
Figure BDA0001868376040000052
Figure BDA0001868376040000053
Figure BDA0001868376040000054
Figure BDA0001868376040000055
Figure BDA0001868376040000056
is the average value of the reliability of the dynamic alarm confidence at the first moment
Figure BDA0001868376040000057
(8-2-1) when t is 2, obtaining static alarm confidence degrees B at t-1 and t-2 by using equations (5), (6) and (7)1And B2，R11, get
Figure BDA00018683760400000510
R2＝rtAnd fusing the alarm confidence coefficient dynamic fusion rules by using the alarm confidence coefficient dynamic fusion rules to obtain a fusion result:
Figure BDA0001868376040000058
Figure BDA0001868376040000059
namely, the static alarm confidence coefficients at the time t-1 and the time t-2 are fused to obtain the dynamic alarm confidence coefficient at the time t-2.
(8-2-2) when t is>When t is less than or equal to l, the reliability of the dynamic alarm confidence coefficient at the t-1 moment is taken
Figure BDA0001868376040000061
Calculating and obtaining the reliability r of the confidence coefficient of the static alarm at the current time t by using the formulas (8) to (13)tHaving R2＝rtDynamic alarm confidence E for time t-1 using equations (15) and (16)t-1And the static alarm confidence B of the current time ttAre fused to obtainDynamic alarm confidence E to current time tt。
(8-2-3) when t is>When l is needed, the reliability of the dynamic alarm confidence coefficient at the t-1 moment is calculated by using the formula (14)
Figure BDA0001868376040000062
Calculating and obtaining the reliability r of the confidence coefficient of the static alarm at the current time t by using the formulas (8) to (14)tHaving R2＝rtDynamic alarm confidence E for time t-1 using equations (15) and (16)t-1And the static alarm confidence B of the current time ttFusing to obtain the dynamic alarm confidence E of the current time tt。
(9) Obtaining the dynamic alarm confidence E at the time t according to the step (8)t＝(Et(NA),Et(A) Giving an alarm criterion: if Et(NA)≥Et(A) If the output y (t) is 0, no alarm is given, namely the medium-voltage power system is in a normal operation state, and if the output E is not in the normal operation statet(NA)<Et(A) And outputting y (t) equal to 1, and alarming, namely, explaining that the medium-voltage power system is in a single-phase grounding state at the moment.
The invention provides a design method of a single-phase earth fault alarm of a medium-voltage ship power system, which comprises the steps of firstly, carrying out static fusion on two alarm confidence coefficients containing uncertainty at each moment by utilizing an alarm confidence coefficient static fusion rule to obtain the static alarm confidence coefficient at each moment, then, carrying out dynamic fusion on the static alarm confidence coefficient at the current moment and the dynamic alarm confidence coefficient at the previous moment by utilizing an alarm confidence coefficient dynamic fusion rule to obtain the dynamic alarm confidence coefficient at the current moment, and judging whether to give an alarm or not under a relevant judgment criterion. The program (compiling environment LabVIEW, C + + and the like) compiled by the method can run on a monitoring alarm computer, and is combined with hardware such as a sensor, a data collector, a data memory and the like to form an online alarm system, so that the real-time alarm function of the running state of the medium-voltage power system is realized.
Drawings
FIG. 1 is a block flow diagram of the method of the present invention;
FIGS. 2(a) and 2(b) are diagrams each showing example x of the method of the present invention1And x2The sample data sequence of (1);
FIGS. 3(a) and 3(b) are diagrams of example x of the method of the present invention1And x2The test sample sequence of (1).
Detailed Description
In order to be better suitable for a medium-pressure ship power system and reduce the false alarm rate and the false missing alarm rate of a single-phase earth fault alarm, the invention adopts a confidence fusion technology combining static fusion and dynamic fusion which is different from the traditional alarm, namely, two alarm confidences containing uncertainty at each moment are firstly statically fused by using an alarm confidence static fusion rule to obtain the static alarm confidence at each moment, and then the alarm confidence dynamic fusion rule is used to dynamically fuse the static alarm confidence at the current moment and the dynamic alarm confidence at the previous moment, so that the more accurate judgment on the running state of the medium-pressure ship power system is obtained, and the purposes of real-time monitoring and accurate alarm are achieved.
As shown in fig. 1, the present invention comprises the following steps:
(1) for a medium-voltage ship power system, the topology of a medium-voltage part is mostly a busbar segmented structure, each segment of busbar is supplied with power by at least one generator, load devices such as a daily transformer, a propulsion frequency converter and the like are separately connected to different busbar segments, according to the Steel sea vessel entry standard, when a single-phase ground fault occurs in the system, a zero-sequence voltage U with the inequality of 3.3 KV-10.8 KV is formed between a fault point and the ground, zero-sequence current I and a phase D are generated in a pair mode through the ground capacitance of a cable and equipment and the neutral resistance of a generator, a zero-sequence current signal is acquired by a zero-sequence current sensor, and the peak value is obtained through Fourier decomposition and is marked as x1The zero sequence current phase is acquired by the zero sequence current sensor and is marked as x2Zero sequence current peak value x1Starting from 0 to increase, zero sequence current phase x2The deviation is from 180 degrees to-90 degrees, and the medium-voltage power system is normally operatedWhen the working state is changed into a fault state, the sensor collects zero-sequence current signals and zero-sequence current phases once every 1 second.
(2) The identification frame of the single-phase earth fault alarm is set as Θ ═ NA, a }, where NA ═ 0 represents that the power system is in a normal operation state, and a ═ 1 represents that the power system is in a single-phase earth fault state.
(3) From step (1), x1And x2Alarm variables of zero sequence current peak value and zero sequence current phase which are needed to be monitored by the single-phase earth fault alarm respectively, let x1(k) And x2(k) K is 1,2,3, …, K, which is the online measurement sequence of the sampling value of the zero sequence current peak value and the sampling value of the zero sequence current phase by the sensor, K is the sampling time, the sampling number K is more than 3600, and the alarm variable x of the zero sequence current peak value is constructed1Normal distribution X for normal operating condition and single-phase earth fault condition1,NA～N(μ1,σ1 2) And X1,A～N(μ2,σ2 2) The corresponding probability density function is f1,NA(x) And f1,A(x) Constructing alarm variable x of zero sequence current phase2Normal distribution X for normal operating condition and single-phase earth fault condition2,NA～N(μ3,σ3 2) And X2,A～N(μ4,σ4 2) The corresponding probability density function is f2,NA(x) And f2,A(x) Wherein 0A is not more than mu1<μ2<3A,σ1>0,σ2>0,180°≤μ3<360°,-90°<μ4<0°,σ3>0,σ4>0。
This is exemplified here for ease of understanding. Construction of a Normal distribution X1,NA～N(0.5A,0.52)、X1,A～N(2A,0.52)、X2,NA～N(180°,902)、X2,A～N(-45°,302) It is obvious that they correspond to probability density functions, respectively
Figure BDA0001868376040000081
Figure BDA0001868376040000082
Alarm variable x for respectively representing zero sequence current peak value1Alarm variable x of zero sequence current phase2Normal distribution with respect to normal operating conditions and single-phase ground fault conditions.
(4) Setting two input fault characteristics of single-phase earth fault alarm as x1(k) And x2(k) The end point set of the reference interval is T ═ Tn’1,2 …, N, and Qm’1,2 …, M, which respectively form the sampling value x of the zero-sequence current peak value1(k) N-1 reference intervals of { [ T ]1,T2],[T2,T3],…,[TN-2,TN-1],[TN-1,TN]Sample value x of zero sequence current phase2(k) M-1 reference intervals of { [ Q ]1,Q2],[Q2,Q3],…,[QM-2,QM-1],[QM-1,QM]And have a value of (mu)1-3*σ1)<T1<T2<…<TN<(μ2+3*σ2)，(μ4-3*σ4)＝Q1<Q2<…<QM＝(μ3+3*σ3) The output of the alarm is the single-phase grounding state of the medium-voltage power system, which is recorded as y (k), and the reference value set is W ═ Ke1,2}, where W1＝NA＝0，W2＝A＝1。
To facilitate understanding of the set of two input fault signature reference interval endpoints and the reference value of the output single-phase-to-ground state, this is exemplified herein. Obtaining the sampling value x of the zero-sequence current peak value through the normal distribution of the sampling values of the zero-sequence current peak value and the zero-sequence current phase constructed in the step (3)1(k) Has a distribution interval of [0A,3A ]]Sampling value x of zero sequence current phase2(k) The distribution interval is [ -90 DEG, 360 DEG ]]The corresponding output operating state results are discrete sequences 0 and 1. Therefore, the set of reference values W ═ 0,1} for the output single-phase grounding state and e ═ can be set1,2, input x1(k) Is {0,0.1,0.2,0.3,0.5,0.7,0.9,1.1,1.3,1.5,1.8,2.1,2.4,2.7,3.0}, corresponding to the reference interval { [0,0.1, 2.7 }],[0.1,0.2],[0.2,0.3],[0.3,0.5],[0.5,0.7],[0.7,0.9],[0.9,1.1],[1.1,1.3],[1.3,1.5],[1.5,1.8],[1.8,2.1],[2.1,2.4],[2.4,2.7],[2.7,3.0]In (unit: A), input x2(k) Is equal to-90, -30,0,45,90,135,160,165,170,175,180,360, corresponding to the reference interval-90, -30],[-30,0],[0,45],[45,90],[90,135],[135,160],[160,165],[165,170],[170,175],[175,180],[180,360]} (unit:deg.C).
(5) Constructing an input x by step (3)1(k) And x2(k) Sequence of measured values S ═ x for output y (k)1(k)，x2(k) Y (K), K is 1,2,3, …, K, K is not less than 3600}, and x is equal to or less than 36001(k)、x2(k) And y (k) is expressed as a sample set Z ═ x1(k),x2(k),y(k)]Form (b), ascertaining x1(k) And x2(k) Therein are respectively β1Each measured value satisfies N (mu)1,σ1 2) And N (mu)3,σ3 2) β for outputs y (k) equal to 02Each measured value satisfies N (mu)2,σ2 2) And N (mu)4,σ4 2) Corresponding to output y (k) equal to 1 and having β1+β2X is equal to K1(k) And y (k) is expressed as a sample set Z' ═ x1(k),y(k)]X is to be2(k) And y (k) is expressed as a sample set Z ═ x2(k),y(k)]。
For the sake of understanding, it is exemplified that data K-4800 set is constructed as a training sample by the step (3), and arranged in a sequence S, wherein β are confirmed13600 x1(k) And x2(k) Measured values of (a) are measured when the medium voltage power system is in a normal operation state, corresponding to outputs y (k) of 0, β respectively21200 x1(k) And x2(k) The measured value of (a) is that the medium voltage power system is in the single phase earth fault state, corresponding to the output y (k) being 1, there is β1+β2K4800, which respectively form 4800 sets of samples Z' ═ x1(k),y(k)]And Z ═ x2(k),y(k)]。
Using sample sets x respectively1(k),y(k)]And [ x ]2(k),y(k)]Constructing input fault characteristics x by probability distribution on corresponding intervals1(k) And x2(k) The probability distribution table of the single-phase grounding state y (k) of the medium-voltage power system is shown in the following table 1 and table 2, respectively, wherein N is 1,2 …, N-1, M is 1,2, …, M-1, e is 1,2, and
Figure BDA0001868376040000091
Figure BDA0001868376040000092
obviously, there are
Figure BDA0001868376040000093
αnTo fall at a reference point TnSum of probability distributions of above, and having an＝γ1,n+γ2,n，ηmTo fall at a reference point QmSum of probability distributions of (1) and has ηm＝ε1,m+ε2,m。
TABLE 1 sample [ x1(k),y(k)]Table of probability distribution
Figure BDA0001868376040000094
TABLE 2 sample [ x2(k),y(k)]Table of probability distribution
Figure BDA0001868376040000095
To facilitate understanding of the probability distribution table, this is exemplified here. Respectively obtaining all sets of samples [ x ] 4800 by combining corresponding probability density functions along with the reference interval in the step (4)1(k),y(k)]And [ x ]2(k),y(k)]The probability distribution on each corresponding interval can be used to respectively construct the sample [ x1(k),y(k)]And [ x ]2(k),y(k)]The probability distribution tables of (2) are shown in tables 3 and 4 below.
TABLE 3 sample [ x1(k),y(k)]Table of probability distribution
Figure BDA0001868376040000101
TABLE 4 sample [ x2(k),y(k)]Table of probability distribution
Figure BDA0001868376040000102
(6) According to the sample [ x ] obtained in the step (5)1(k),y(k)]Can obtain the current input fault feature x1(k) Fall at a reference point TnWhen above, WeConfidence of occurrence is
Figure BDA0001868376040000103
And is provided with
Figure BDA0001868376040000104
A reference point T can be defined which corresponds tonAlarm confidence of
Figure BDA0001868376040000105
Similarly, for sample [ x2(k),y(k)]The probability distribution table of (2) is defined corresponding to the reference point QmAlarm confidence of
Figure BDA0001868376040000106
Wherein
Figure BDA0001868376040000107
For the new measurement value x1(t) and x2(T), T ═ 1,2,3, …, which necessarily falls within reference point TnOr QmAt this time, the alarm confidence corresponding to the reference point is activated, and the alarm confidence is respectively activatedThe formula (2) and the formula (3).
This is exemplified herein for ease of understanding the activation process for alarm confidence. For time t, there is a new measurement value x1When (t) is 1.31, it can be seen that it falls within the interval [1.3,1.5 ] corresponding to table 3]Then there is
Figure BDA0001868376040000111
Figure BDA0001868376040000112
Then the alarm confidence can be obtained as
Figure BDA0001868376040000113
At the same time t, there is a new measurement value x2(t) — 15, for the same reason
Figure BDA0001868376040000114
(7) Respectively obtaining the process variable x at the t moment according to the step (6)1And x2Alarm confidence of
Figure BDA0001868376040000115
And
Figure BDA0001868376040000116
then, the alarm confidence coefficient of each moment is fused by using an alarm confidence coefficient static fusion rule
Figure BDA0001868376040000117
And
Figure BDA0001868376040000118
performing static fusion to obtain static alarm confidence B at t momenttThe specific calculation process is as follows:
Bt＝Ro,b(2)，o＝{We|e＝1,2} (5)
Figure BDA0001868376040000119
wherein,at setting wi＝riWhen i is 1,2, there are
Figure BDA00018683760400001110
This is exemplified here for ease of understanding. Obtaining the process variable x related to the same time t according to the step (6)1And x2Alarm confidence of
Figure BDA00018683760400001111
And
Figure BDA00018683760400001112
get wi＝riAnd (5) statically fusing 0.9 and 1 and 2 by using formulas (5) to (7) to obtain a static alarm confidence B at the time tt＝(0.116,0.884)。
(8) Obtaining the static alarm confidence B at the time t according to the step (7)tThen, the static alarm confidence at the current time t is fused with the dynamic alarm confidence at the past time by using an alarm confidence dynamic fusion rule to obtain the dynamic alarm confidence at the current time t, which is marked as Et＝[Et(NA),Et(A)]The method comprises the following specific steps:
(8-1) when t is 1, there is E1＝B1＝[E1(NA),E1(A)]That is, the dynamic alarm confidence is the static alarm confidence obtained at the moment;
(8-2) setting an alarm confidence weighting wi1, i 1,2, reliability r of the static alarm confidence at the current timetCalculated by the following formula:
Figure BDA00018683760400001113
wherein r is00.5 is the initial value of reliability, τ is the reward and punishment coefficient, and is calculated by the following formula:
Figure BDA00018683760400001114
setting eNA＝(1,0,0)，eA＝(0,1,0)，B′t＝[Bt(NA),Bt(A),0]，E′t＝[Et(NA),Et(A),0]φ is a reliability enhancement factor, calculated by the following equation:
Figure BDA0001868376040000121
Figure BDA0001868376040000122
Figure BDA0001868376040000123
Figure BDA0001868376040000124
Figure BDA0001868376040000125
is the average value of the reliability of the dynamic alarm confidence at the first moment
Figure BDA0001868376040000126
(8-2-1) when t is 2, obtaining static alarm confidence degrees B at t-1 and t-2 by using equations (5), (6) and (7)1And B2，R11, get
Figure BDA0001868376040000127
R2＝rtThe alarm confidence coefficient dynamic fusion rules are utilized to fuse the alarm confidence coefficient dynamic fusion rules to obtain a fusion result
Figure BDA0001868376040000128
Figure BDA0001868376040000129
Fusing the static alarm confidence coefficients at the time t-1 and the time t-2 to obtain a dynamic alarm confidence coefficient at the time t-2;
(8-2-2) when t is>When t is less than or equal to l, the reliability of the dynamic alarm confidence coefficient at the t-1 moment is taken
Figure BDA00018683760400001210
Calculating and obtaining the reliability r of the confidence coefficient of the static alarm at the current time t by using the formulas (8) to (13)tHaving R2＝rtDynamic alarm confidence E for time t-1 using equations (15) and (16)t-1And the static alarm confidence B of the current time ttFusing to obtain the dynamic alarm confidence E of the current time tt；
(8-2-3) when t is>When l is needed, the reliability of the dynamic alarm confidence coefficient at the t-1 moment is calculated by using the formula (14)
Figure BDA0001868376040000131
Calculating and obtaining the reliability r of the confidence coefficient of the static alarm at the current time t by using the formulas (8) to (14)tHaving R2＝rtDynamic alarm confidence E for time t-1 using equations (15) and (16)t-1And the static alarm confidence B of the current time ttFusing to obtain the dynamic alarm confidence E of the current time tt。
For enhancing the understanding of step (8), this is exemplified here. First, assume that the new measured value x at 4 times of 1,2,3,4 is knowni(t), i is 1,2, and the static alarm confidence at these 4 times is calculated sequentially according to equations (5) to (7), as shown in table 5:
TABLE 5 static alarm confidence table
Figure BDA0001868376040000132
The dynamic alarm confidence levels at these 4 moments can be given according to step (8) as follows:
when t is 1, as obtained according to step (8-1), E1＝B1＝(0.9,0.1)；
When t is 2, according to step (8-2-1), R is taken1＝1,
Figure BDA0001868376040000135
Calculating the reliability r of the static alarm confidence at time 22: τ is calculated to 1 by equation (9), and d is calculated by equation (11)1D is calculated by equation (12) when the value is 0.42Further, when the value is equal to 0.1, phi is calculated to 0.6 by the formula (10). Initial value r of reliability0Substituting the reward and punishment coefficient tau and the reliability enhancement factor phi into a formula (8) to obtain r20.8, i.e. R2E is set to 0.8 by equations (15) and (16)1(0.9,0.1) and B2Dynamic fusion is carried out when the t is equal to 2, and the dynamic alarm confidence E is obtained2＝(0.92,0.08)；
(for convenience of explanation, take l as 3.)
When t is 3, according to step (8-2-2),
Figure BDA0001868376040000133
r is calculated by equations (8) to (13)3When R is 0.65, then2＝r3E is set to 0.65 using equations (15) and (16)2(0.92,0.08) and B3Dynamic fusion is carried out when the t is equal to 3, and the dynamic alarm confidence E is obtained3＝(0.86,0.14)；
When t is 4, the calculation is performed by using formula (14) according to step (8-2-3)
Figure BDA0001868376040000134
Then R is1R is calculated from equations (8) to (14) at 0.824When R is 0.75, then R2＝r4E is set to 0.75 by equations (15) and (16)3(0.86,0.14) and B4Dynamic fusion is carried out when the t is equal to 4, and the dynamic alarm confidence E is obtained4＝(0.67,0.33)。
(9) Obtaining the dynamic alarm confidence E at the time t according to the step (8)t＝(Et(NA),Et(A) Giving an alarm criterion: if Et(NA)≥Et(A) If the output y (t) is 0, no alarm is given, namely the medium-voltage power system is in a normal operation state, and if the output E is not in the normal operation statet(NA)<Et(A) And outputting y (t) equal to 1, and alarming, namely, explaining that the medium-voltage power system is in a single-phase grounding state at the moment.
In the above example, according to the dynamic alarm confidence output at 4 moments, an alarm result is made according to step (9), as shown in table 6:
TABLE 6 alarm results
Time t Dynamic alarm confidence Et Alarm result
t＝1 E1＝(0.9,0.1) Normal operation (NA)
t＝2 E2＝(0.92,0.08) Normal operation (NA)
t＝3 E3＝(0.86,0.14) Normal operation (NA)
t＝4 E4＝(0.67,0.33) Normal operation (NA)
Embodiments of the method of the present invention are described in detail below with reference to the accompanying drawings:
the flow chart of the method of the invention is shown in figure 1, and the core part is as follows: after determining alarm variables of zero sequence current peak values and zero sequence current phases of a medium-voltage power system to be monitored and sample data sequences thereof, firstly, performing static fusion on two alarm confidence coefficients containing uncertainty at each moment by using an alarm confidence coefficient static fusion rule to obtain a static alarm confidence coefficient at each moment, then performing dynamic fusion on the static alarm confidence coefficient at the current moment and a dynamic alarm confidence coefficient at the previous moment by using an alarm confidence coefficient dynamic fusion rule to obtain a dynamic alarm confidence coefficient at the current moment, and judging whether to give an alarm or not under a relevant judgment criterion.
The following is a combination of x shown in FIGS. 2(a) and 2(b)1And x2The best embodiment is given, and the steps of the method of the present invention are described in detail.
1. Sample data sequence x for respectively giving sampling value of zero sequence current peak value and sampling value of zero sequence current phase of medium voltage power system1(k) And x2(k)。
Zero sequence current peak value x1Sample data sequence x of (2)1(k) As shown in FIG. 2(a), K is 4800, and x is statistically determined1The range of variation is [0A,3A ]](ii) a Zero sequence current phase x2Sample data sequence x of (2)2(k) As shown in FIG. 2(b), K is 4800, and x is statistically determined2The range of variation is [ -90 °,180 °]。
2. And selecting a set of two input fault characteristic reference interval endpoints and a reference value of an output single-phase grounding state.
Obtaining the sampling value x of the zero-sequence current peak value through the normal distribution of the sampling values of the zero-sequence current peak value and the zero-sequence current phase constructed in the step (3)1(k) Has a distribution interval of [0A,3A ]]Zero sequence electricitySampled value x of the flow phase2(k) The distribution interval is [ -90 DEG, 360 DEG ]]The corresponding output operating state results are discrete sequences 0 and 1. Therefore, the output single-phase grounding state reference value set W ═ 0,1, and e ═ 1,2, and the input x can be set1(k) Is {0,0.1,0.2,0.3,0.5,0.7,0.9,1.1,1.3,1.5,1.8,2.1,2.4,2.7,3.0}, corresponding to the reference interval { [0,0.1, 2.7 }],[0.1,0.2],[0.2,0.3],[0.3,0.5],[0.5,0.7],[0.7,0.9],[0.9,1.1],[1.1,1.3],[1.3,1.5],[1.5,1.8],[1.8,2.1],[2.1,2.4],[2.4,2.7],[2.7,3.0]In (unit: A), input x2(k) Is equal to-90, -30,0,45,90,135,160,165,170,175,180,360, corresponding to the reference interval-90, -30],[-30,0],[0,45],[45,90],[90,135],[135,160],[160,165],[165,170],[170,175],[175,180],[180,360]} (unit:deg.C).
3. Sample of construction [ x ]1(k),y(k)]And [ x ]2(k),y(k)]Table of probability distribution.
Constructing data of K-4800 group as training sample, arranging sequence S, and confirming β13600 x1(k) And x2(k) Measured values of (a) are measured when the medium voltage power system is in a normal operation state, corresponding to outputs y (k) of 0, β respectively21200 x1(k) And x2(k) The measured value of (a) is that the medium voltage power system is in the single phase earth fault state, corresponding to the output y (k) being 1, there is β1+β2K4800, which respectively form 4800 sets of samples Z' ═ x1(k),y(k)]And Z ═ x2(k),y(k)]. Respectively obtaining all sets of samples [ x ] 4800 by combining corresponding probability density functions along with the reference interval in the step (4)1(k),y(k)]And [ x ]2(k),y(k)]The probability distribution on each corresponding interval can be used to respectively construct the sample [ x1(k),y(k)]And [ x ]2(k),y(k)]The probability distribution tables (2) are shown in tables 7 and 8 below:
TABLE 7 samples [ x ]1(k),y(k)]Table of probability distribution
Figure BDA0001868376040000151
TABLE 8 sample [ x2(k),y(k)]Table of probability distribution
Figure BDA0001868376040000152
4. New measured value xi(t) activating to obtain a corresponding alarm confidence level.
Obtaining a sample [ x ] according to step (5) of the invention1(k),y(k)]And [ x ]2(k),y(k)]After the probability distribution table, step (6) of the method according to the invention makes it possible to carry out the updating of the measured value xi(t) activation results in two corresponding alarm confidences. For time t, there is a new measurement value x1When (t) is 1.31, it can be seen that it falls within the interval [1.3,1.5 ] corresponding to table 3]Then there is
Figure BDA0001868376040000161
Then the alarm confidence can be obtained as
Figure BDA0001868376040000162
At the same time t, there is a new measurement value x2(t) — 15, for the same reason
Figure BDA0001868376040000163
5. And performing static fusion on the two alarm confidence coefficients by using an alarm confidence coefficient static fusion rule to obtain the static alarm confidence coefficient at each moment.
Obtaining the process variable x related to the same time t according to the step (6)1And x2Alarm confidence of
Figure BDA0001868376040000164
Get wi＝riAnd (5) statically fusing 0.9 and 1 and 2 by using formulas (5) to (7) to obtain a static alarm confidence B at the time tt＝(0.116,0.884)。
6. And dynamically fusing the static alarm confidence coefficient at the current moment with the dynamic alarm confidence coefficient at the previous moment by using an alarm confidence coefficient dynamic fusion rule to obtain the dynamic alarm confidence coefficient at the current moment.
The 4 time instants of known t ═ 1,2,3,4 are the measured values xi(t), i is 1,2, and the static alarm confidence at these 4 times is calculated sequentially according to equations (5) to (7), as shown in table 9:
TABLE 9 static alarm confidence table
Figure BDA0001868376040000165
The dynamic alarm confidence levels at these 4 moments can be given according to step (8) as follows:
when t is 1, as obtained according to step (8-1), E1＝B1＝(0.9,0.1)；
When t is 2, according to step (8-2-1), R is taken1＝1,
Figure BDA0001868376040000166
Calculating the reliability r of the static alarm confidence at time 22: τ is calculated to 1 by equation (9), and d is calculated by equation (11)1D is calculated by equation (12) when the value is 0.42Further, when the value is equal to 0.1, phi is calculated to 0.6 by the formula (10). Initial value r of reliability0Substituting the reward and punishment coefficient tau and the reliability enhancement factor phi into a formula (8) to obtain r20.8, i.e. R2E is set to 0.8 by equations (15) and (16)1(0.9,0.1) and B2Dynamic fusion is carried out when the t is equal to 2, and the dynamic alarm confidence E is obtained2＝(0.92,0.08)；
(for convenience of explanation, take l as 3.)
When t is 3, according to step (8-2-2),
Figure BDA0001868376040000171
r is calculated by equations (8) to (13)3When R is 0.65, then2＝r3E is set to 0.65 using equations (15) and (16)2(0.92,0.08) and B3Dynamic fusion is carried out when the t is equal to 3, and the dynamic alarm confidence E is obtained3＝(0.86,0.14)；
When t is 4, the calculation is performed by using formula (14) according to step (8-2-3)
Figure BDA0001868376040000172
Then R is1R is calculated from equations (8) to (14) at 0.824When R is 0.75, then R2＝r4E is set to 0.75 by equations (15) and (16)3(0.86,0.14) and B4Dynamic fusion is carried out when the t is equal to 4, and the dynamic alarm confidence E is obtained4＝(0.67,0.33)。
7. And making an alarm decision according to the alarm criterion.
An alarm result can be made according to step (9), as shown in table 10:
TABLE 10 alarm results
Time t Dynamic alarm confidence Et Alarm result
t＝1 E1＝(0.9,0.1) Normal operation (NA)
t＝2 E2＝(0.92,0.08) Normal operation (NA)
t＝3 E3＝(0.86,0.14) Normal operation (NA)
t＝4 E4＝(0.67,0.33) Normal operation (NA)

Claims (1)

1. A design method of a single-phase earth fault alarm of a medium-voltage ship power system is characterized by comprising the following steps:
(1) for a medium-voltage ship power system, the topology of a medium-voltage part is mostly a busbar segmented structure, each segment of busbar is powered by at least one generator, when a single-phase earth fault occurs in the system, a zero-sequence voltage U with the voltage of 3.3-10.8 KV is formed between a fault point and the ground, zero-sequence current I and a phase D thereof are generated in a pair of ground capacitors of cables and equipment and a neutral point resistor of the generator, a zero-sequence current signal is acquired by a zero-sequence current sensor, and a peak value is obtained through Fourier decomposition and is marked as x1The zero sequence current phase is acquired by the zero sequence current sensor and is marked as x2Zero sequence current peak value x1Starting from 0 to increase, zero sequence current phase x2The direction of the deviation is changed from 180 degrees to-90 degrees, meanwhile, the medium-voltage power system is changed from a normal working state to a fault state, and the sensor collects zero-sequence current signals and zero-sequence current phases once every 1 second;
(2) setting an identification frame of the single-phase earth fault alarm as theta ═ NA, a }, wherein NA ═ 0 represents that the power system is in a normal operation state, and a ═ 1 represents that the power system is in a single-phase earth fault state;
(3) from step (1), x1And x2Zero sequence current signal peak value and zero sequence current phase position which are needed to be monitored by the single-phase earth fault alarm respectively, and let x1(k) And x2(k) K is 1,2,3, …, K, which is the online measurement sequence of the sampling value of the zero sequence current peak value and the sampling value of the zero sequence current phase of the sensor, respectively, K is the sampling time, the sampling number K is more than 3600, and the zero sequence current signal peak value x is constructed1Normal distribution for normal operating condition and single-phase earth fault conditionX1,NA～N(μ1,σ1 2) And X1,A～N(μ2,σ2 2) The corresponding probability density function is f1,NA(x) And f1,A(x) Constructing the zero sequence current phase x2Normal distribution X for normal operating condition and single-phase earth fault condition2,NA～N(μ3,σ3 2) And X2,A～N(μ4,σ4 2) The corresponding probability density function is f2,NA(x) And f2,A(x) Wherein 0A is not more than mu1<μ2<3A,σ1>0,σ2>0,180°≤μ3<360°,-90°<μ4<0°,σ3>0,σ4>0；
(4) Setting two input fault characteristics of single-phase earth fault alarm as x1(k) And x2(k) The end point set of the reference interval is T ═ Tn’1,2 …, N, and Qm’1,2 …, M, which respectively form the sampling value x of the zero-sequence current peak value1(k) N-1 reference intervals of { [ T ]1,T2],[T2,T3],…,[TN-2,TN-1],[TN-1,TN]Sample value x of zero sequence current phase2(k) M-1 reference intervals of { [ Q ]1,Q2],[Q2,Q3],…,[QM-2,QM-1],[QM-1,QM]And have a value of (mu)1-3*σ1)<T1<T2<…<TN<(μ2+3*σ2)，(μ4-3*σ4)＝Q1<Q2<…<QM＝(μ3+3*σ3) The output of the alarm is the single-phase grounding state of the medium-voltage power system, which is recorded as y (k), and the reference value set is W ═ Ke1,2}, where W1＝NA＝0，W2＝A＝1；
(5) Constructing an input x by step (3)1(k) And x2(k) Sequence of measured values S ═ x for output y (k)1(k)，x2(k) Y (K), K is 1,2,3, …, K, K is not less than 3600}, and x is equal to or less than 36001(k)、x2(k) And y (k) is expressed as a sample set Z ═ x1(k),x2(k),y(k)]Form (b), ascertaining x1(k) And x2(k) Therein are respectively β1Each measured value satisfies N (mu)1,σ1 2) And N (mu)3,σ3 2) β for outputs y (k) equal to 02Each measured value satisfies N (mu)2,σ2 2) And N (mu)4,σ4 2) Corresponding to output y (k) equal to 1 and having β1+β2X is equal to K1(k) And y (k) is expressed as a sample set Z' ═ x1(k),y(k)]X is to be2(k) And y (k) is expressed as a sample set Z ═ x2(k),y(k)]Using sets of samples [ x ] respectively1(k),y(k)]And [ x ]2(k),y(k)]Constructing input fault characteristics x by probability distribution on corresponding intervals1(k) And x2(k) Probability distribution table of single-phase grounding state y (k) of the medium-voltage power system;
(6) according to the sample [ x ] obtained in the step (5)1(k),y(k)]Can obtain the current input fault feature x1(k) Fall at a reference point TnWhen above, WeConfidence of occurrence is
Figure FDA0002445560790000021
And is provided with
Figure FDA0002445560790000022
A reference point T can be defined which corresponds tonAlarm confidence of
Figure FDA0002445560790000023
Similarly, for sample [ x2(k),y(k)]The probability distribution table of (2) is defined corresponding to the reference point QmAlarm confidence of
Figure FDA0002445560790000024
Wherein
Figure FDA0002445560790000025
For newly-advanced zero-sequence current signal peak value x1(t) and zero sequence current phase x2(T), T ═ 1,2,3, …, which necessarily falls within reference point TnOr QmAt this time, the alarm confidence corresponding to the reference point is activated, and the alarm confidence is given by the formula (2) and the formula (3);
(7) respectively obtaining the process variable x at the t moment according to the step (6)1And x2Alarm confidence of
Figure FDA0002445560790000027
And
Figure FDA0002445560790000026
then, the alarm confidence coefficient of each moment is fused by using an alarm confidence coefficient static fusion rule
Figure FDA0002445560790000028
And
Figure FDA0002445560790000029
performing static fusion to obtain static alarm confidence B at t momenttThe specific calculation process is as follows:
Bt＝Ro,b(2)，o＝{We|e＝1,2} (5)
Figure FDA0002445560790000031
wherein at the setting wi＝riWhen i is 1,2, there are
Figure FDA0002445560790000032
(8) Obtaining the static alarm confidence B at the time t according to the step (7)tThen, the static alarm confidence at the current time t is fused with the dynamic alarm confidence at the past time by using an alarm confidence dynamic fusion rule to obtain the dynamic alarm confidence at the current time t, which is marked as Et＝[Et(NA),Et(A)]The method comprises the following specific steps:
(8-1) when t is 1, there is E1＝B1＝[E1(NA),E1(A)]That is, the dynamic alarm confidence is the static alarm confidence obtained at the moment;
(8-2) setting an alarm confidence weighting wi1, i 1,2, reliability r of the static alarm confidence at the current timetCalculated by the following formula:
Figure FDA0002445560790000033
wherein r is00.5 is the initial value of reliability, τ is the reward and punishment coefficient, and is calculated by the following formula:
Figure FDA0002445560790000034
setting eNA＝(1,0,0)，eA＝(0,1,0)，B′t＝[Bt(NA),Bt(A),0]，E′t＝[Et(NA),Et(A),0]φ is a reliability enhancement factor, calculated by the following equation:
Figure FDA0002445560790000035
Figure FDA0002445560790000036
Figure FDA0002445560790000037
Figure FDA0002445560790000041
Figure FDA0002445560790000042
is the average value of the reliability of the dynamic alarm confidence at the first moment
Figure FDA0002445560790000043
(8-2-1) when t is 2, obtaining static alarm confidence degrees B at t-1 and t-2 by using equations (5), (6) and (7)1And B2，R11, get
Figure FDA0002445560790000044
R2＝rtThe alarm confidence coefficient dynamic fusion rules are utilized to fuse the alarm confidence coefficient dynamic fusion rules to obtain a fusion result
Figure FDA0002445560790000045
Figure FDA0002445560790000046
Fusing the static alarm confidence coefficients at the time t-1 and the time t-2 to obtain a dynamic alarm confidence coefficient at the time t-2;
(8-2-2) when t is>When t is less than or equal to l, the reliability of the dynamic alarm confidence coefficient at the t-1 moment is taken
Figure FDA0002445560790000047
Calculating and obtaining the reliability r of the confidence coefficient of the static alarm at the current time t by using the formulas (8) to (13)tHaving R2＝rtDynamic alarm confidence E for time t-1 using equations (15) and (16)t-1And the static alarm confidence B of the current time ttCarrying out fusion to obtainDynamic alarm confidence E at time tt；
(8-2-3) when t is>When l is needed, the reliability of the dynamic alarm confidence coefficient at the t-1 moment is calculated by using the formula (14)
Figure FDA0002445560790000048
Calculating and obtaining the reliability r of the confidence coefficient of the static alarm at the current time t by using the formulas (8) to (14)tHaving R2＝rtDynamic alarm confidence E for time t-1 using equations (15) and (16)t-1And the static alarm confidence B of the current time ttFusing to obtain the dynamic alarm confidence E of the current time tt；
(9) Obtaining the dynamic alarm confidence E at the time t according to the step (8)t＝(Et(NA),Et(A) Giving an alarm criterion: if Et(NA)≥Et(A) If the output y (t) is 0, no alarm is given, namely the medium-voltage power system is in a normal operation state, and if the output E is not in the normal operation statet(NA)<Et(A) And outputting y (t) equal to 1, and alarming, namely, explaining that the medium-voltage power system is in a single-phase grounding state at the moment.
CN201811365359.3A 2018-11-16 2018-11-16 Design method of single-phase earth fault alarm of medium-voltage ship power system Active CN109375037B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811365359.3A CN109375037B (en) 2018-11-16 2018-11-16 Design method of single-phase earth fault alarm of medium-voltage ship power system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811365359.3A CN109375037B (en) 2018-11-16 2018-11-16 Design method of single-phase earth fault alarm of medium-voltage ship power system

Publications (2)

Publication Number Publication Date
CN109375037A CN109375037A (en) 2019-02-22
CN109375037B true CN109375037B (en) 2020-06-30

Family

ID=65389169

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811365359.3A Active CN109375037B (en) 2018-11-16 2018-11-16 Design method of single-phase earth fault alarm of medium-voltage ship power system

Country Status (1)

Country Link
CN (1) CN109375037B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112986754B (en) * 2021-05-11 2021-10-19 国网江西省电力有限公司电力科学研究院 Small current grounding system fault identification method and device based on data driving
CN114005256B (en) * 2021-10-25 2023-06-20 山东科技大学 Alarm delay design method and system based on alarm duration probability function
CN115112360B (en) * 2022-06-22 2024-04-16 南京智慧水运科技有限公司 Rudder fault diagnosis method based on reliability updating and fusion

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5600524A (en) * 1995-05-04 1997-02-04 Leviton Manufacturing Co., Inc. Intelligent ground fault circuit interrupter
US20030001589A1 (en) * 1999-12-20 2003-01-02 Stmicroelectronics, Inc. Redundant electric fuses
US7113988B2 (en) * 2000-06-29 2006-09-26 International Business Machines Corporation Proactive on-line diagnostics in a manageable network
CA2535058A1 (en) * 2005-12-19 2007-06-19 A.U.G. Signals Ltd. An adaptive constant false alarm rate detection system
CN100462729C (en) * 2006-06-15 2009-02-18 西安交通大学 On-line monitoring method for winding turn-to-turn short-circuit of distribution generator stator based on multi-criterion mixing
CN101718634B (en) * 2009-11-20 2012-05-09 西安交通大学 Equipment state comprehensive dynamic alarming method based on multivariate probability model
CN101986358A (en) * 2010-08-31 2011-03-16 彭浩明 Neural network and fuzzy control fused electrical fire intelligent alarm method
CN102789676B (en) * 2012-08-10 2014-04-02 杭州电子科技大学 Method for designing industrial alarm on basis of alarm evidence fusion
CN203759824U (en) * 2013-12-25 2014-08-06 北京东方巨融科技开发有限公司 Intelligent forest fire prevention monitoring alarm system
CN103941158B (en) * 2014-04-17 2017-01-25 国家电网公司 Power distribution network fault diagnosis method based on multi-source information
CN104600680A (en) * 2015-02-16 2015-05-06 国家电网公司 Intelligent alarming method based on data fusion
CN104764979A (en) * 2015-03-30 2015-07-08 国网河南省电力公司经济技术研究院 Virtual information fusion power grid alarming method based on probabilistic reasoning
CN105929301B (en) * 2016-04-12 2019-01-04 扬州盈信电气工程有限公司 A kind of grounding net of transformer substation corrosion failure point intelligent monitoring device
CN106771883B (en) * 2017-01-22 2022-07-26 中国电力科学研究院 Multi-source information power distribution fault positioning method and system based on cloud
CN107767639A (en) * 2017-08-25 2018-03-06 武汉米风通信技术有限公司 A kind of warning system and method based on the enhancing of Big Dipper ground and Internet of Things fusion

Also Published As

Publication number Publication date
CN109375037A (en) 2019-02-22

Similar Documents

Publication Publication Date Title
CN109375037B (en) Design method of single-phase earth fault alarm of medium-voltage ship power system
Yu et al. An adaptive high and low impedance fault detection method
US11656263B2 (en) Effective feature set-based high impedance fault detection
CN111812451B (en) Phase current transient fault component-based distributed line selection method for power distribution network
SE459059B (en) PROTECT FOR HIGH-RESISTANT EARTH ERRORS
Luo et al. Stacked auto-encoder-based fault location in distribution network
CN110618314B (en) Harmonic wave responsibility division method for resisting short-circuit fault interference of power distribution system
CN109975653B (en) 10kV distribution line fault location method
CN102565629B (en) A kind of transmission line of alternation current Fault Phase Selection test simulation method based on lumped parameter Π model
CN110045232B (en) Method for identifying ground fault phase of neutral point non-effective grounding system
CN112034387B (en) Power transmission line short-circuit fault diagnosis method and device based on prediction sequence
CN108548999B (en) Cable insulation state evaluation method
KR102260550B1 (en) Facility health monitoring method by measuring the electric circuit constant inside the power facility in operation
Chouidira et al. Fuzzy logic based broken bar fault diagnosis and behavior study of induction machine
CN117849691A (en) Multi-dimensional collaborative operation monitoring and early warning system and method for capacitive voltage transformer
CN110705031B (en) Excitation inrush current identification method based on second-order Taylor coefficient
CN102035173A (en) Single phase earthed relay protection device
CN114910744B (en) High-resistance ground fault detection method based on S transformation and self-adaptive average singular entropy
KR102419753B1 (en) Facility health monitoring method by measuring the electric circuit constant inside the power facility in operation
Jung et al. Return voltage-a reliable diagnosis for water treed cables?
CN110161363B (en) Transformer running state vibration and sound detection method and system based on main frequency characteristic quantity
Zhang et al. A fast fault detection method for controlled fault interruption based on wavelet singularities and BP neural network
CN115616332B (en) AC power transmission line lightning stroke interference identification method and system based on extension fusion
CN113189446B (en) Low-current ground fault line selection method
EP3422514A1 (en) Method and assembly for detecting a winding defect in a transformer

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