CN114113891B - Single-phase grounding fault judging method of pulse neural membrane system based on distributed population - Google Patents

Single-phase grounding fault judging method of pulse neural membrane system based on distributed population Download PDF

Info

Publication number
CN114113891B
CN114113891B CN202111403455.4A CN202111403455A CN114113891B CN 114113891 B CN114113891 B CN 114113891B CN 202111403455 A CN202111403455 A CN 202111403455A CN 114113891 B CN114113891 B CN 114113891B
Authority
CN
China
Prior art keywords
fault
population
phase
membrane system
value
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
CN202111403455.4A
Other languages
Chinese (zh)
Other versions
CN114113891A (en
Inventor
范松海
刘益岑
吴天宝
马小敏
刘小江
罗磊
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.)
Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd
Original Assignee
Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd
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 Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd filed Critical Electric Power Research Institute of State Grid Sichuan Electric Power Co Ltd
Priority to CN202111403455.4A priority Critical patent/CN114113891B/en
Publication of CN114113891A publication Critical patent/CN114113891A/en
Application granted granted Critical
Publication of CN114113891B publication Critical patent/CN114113891B/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
    • G01R31/08—Locating faults in cables, transmission lines, or networks
    • G01R31/081—Locating faults in cables, transmission lines, or networks according to type of conductors
    • G01R31/085—Locating faults in cables, transmission lines, or networks according to type of conductors in power transmission or distribution lines, e.g. overhead
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
    • Y04S10/52—Outage or fault management, e.g. fault detection or location

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention discloses a single-phase grounding fault judging method of a pulse neural membrane system based on a distributed population, which aims at a power distribution network neutral point ungrounded system to establish a PSCAD/EMTDC simulation model and acquire fault data of a fault line and a non-fault line under different conditions; establishing a fault line selection model according to the fault data and the non-fault data to reflect the line fault condition; solving an objective function of the fault line selection model by using an optimized pulse neural membrane system; outputting the final optimized parameters (determining z value) to obtain P i ＝z·α i +(1‑z)·β i And z. Calculating P from test data i Calculated P i When the voltage is greater than the threshold value, the single-phase earth fault of the line is judged. The method has the advantages that the single-phase grounding fault of the system with the neutral point not grounded can be judged by using the pulse neural membrane system of the distributed population, the parallel calculation of multiple servers can be adopted, the judging time is shortened, and the judging accuracy is improved.

Description

Single-phase grounding fault judging method of pulse neural membrane system based on distributed population
Technical Field
The invention relates to the technical field of power distribution network systems, in particular to a single-phase grounding fault judging method of a pulse neural membrane system based on a distributed population.
Background
In the power distribution network faults, single-phase earth faults account for about 80% of the total faults, and when single-phase earth faults occur, faults can be timely and accurately removed, so that fault equipment is prevented from being damaged due to long-time fault operation, and power failure accidents with larger area are avoided. There is a need for a method of quickly monitoring and identifying faulty lines that can accurately identify and ablate faulty lines. However, as the single-phase grounding fault information features are weak, the operation mode of the power distribution network is complex, the fault accuracy of single-phase grounding is difficult to be effectively improved, and a manual trial-pull method is still widely adopted for line selection in a plurality of substations, so that the fault processing efficiency is low. Therefore, the demand of effectively selecting lines for single-phase ground faults is continuously increased through the dispatching SCADA system, however, a large amount of data exists in the dispatching system, a related mathematical model is required to be established for processing a large amount of steady-state information, and fault line selection is performed through the mathematical model.
Disclosure of Invention
The invention aims to provide a single-phase grounding fault judging method of a pulse neural membrane system based on a distributed population.
The technical scheme for realizing the purpose of the invention is as follows:
a single-phase grounding fault judging method of a pulse neural membrane system based on a distributed population comprises the following steps:
s10, collecting phase current and three-phase reactive power of each phase of an ith line of a neutral point ungrounded system of the power distribution network before and after a fault; calculating the absolute value of the difference value of the phase current of each phase in the ith line before and after the fault, and selecting the maximum absolute value as the phase current variation delta I of the ith line i The method comprises the steps of carrying out a first treatment on the surface of the Calculating the absolute value of the difference value of the three-phase reactive power of the ith line before and after the fault as the reactive power variation delta Q of the ith line i ；i＝1,2,...,n；
S11, phase current variation DeltaI i And reactive power variation Δq i And (3) respectively normalizing:
Figure BDA0003365566560000011
wherein alpha is i And beta i Respectively normalizing the phase current variation and the reactive power variation;
s12, calculating the fault probability P of the ith line i ＝z·α i +(1-z)·β i Z is a weight parameter;
s13, e.g. P i If the signal is larger than the threshold value, judging that the ith line has single-phase grounding fault; the weight parameter z is obtained by training a pulse neural membrane system of a distributed population by using training set data; wherein,
the pulsed neural membrane system of the distributed population, comprising: neuron number, i, initialization probability
Figure BDA0003365566560000021
Probability of variation
Figure BDA0003365566560000022
Number of sub-population M, number of individuals n in sub-population, total number of individuals H, migration interval M 1 Migration quantity M 2 The ith sub-population probability matrix->
Figure BDA0003365566560000023
And binary matrix->
Figure BDA0003365566560000024
Training a pulsed neural membrane system of a distributed population using training set data as follows:
s20, let iteration number gen=0, maximum iteration number maxgen;
s21, if the gen is larger than or equal to maxgen, training is completed;
S22，gen＝gen+1；
s23, performing cross processing;
s24, if gen is M 1 If not, continuing, otherwise, jumping to S26;
s25, information exchange processing is carried out:
s251, let i=1;
s252, if i > m, executing the following steps; otherwise, jumping to S26;
i. construction of candidate removal Individual set S g,em ：
Figure BDA0003365566560000025
wherein,
Figure BDA0003365566560000026
indicating removal of population P i g,em The ith individual of the g generation in ∈,>
Figure BDA0003365566560000027
representation of
Figure BDA0003365566560000028
Is adapted to the value of->
Figure BDA0003365566560000029
Representing population P i g,em Average fitness value of (a);
from S g,em Selecting individual immigration
Figure BDA00033655665600000210
Figure BDA00033655665600000211
wherein, the euclidean distance is indicated, ESNPS g,em Is from S g,em Is a group of the individuals in the group,
Figure BDA00033655665600000212
is an immigrating population P i g,im Average distance of (2);
constructing a matrix Dg ,im ：
Figure BDA00033655665600000213
wherein, the euclidean distance is indicated,
Figure BDA00033655665600000214
indicating removal of population P i g,im The ith individual of the g generation, D g,im Representing population P i g,im A matrix formed by Euclidean distances between every two of the medium individuals;
selecting a replaced individual
Figure BDA0003365566560000031
Figure BDA0003365566560000032
wherein,
Figure BDA0003365566560000033
and->
Figure BDA0003365566560000034
The representation being from D g,im Two individuals with shortest Euclidean distance;
Figure BDA0003365566560000035
and->
Figure BDA0003365566560000036
Are respectively->
Figure BDA0003365566560000037
And->
Figure BDA0003365566560000038
Is a fitness value of (a);
s253, let i=i+1, return to S252;
s26, performing mutation treatment;
s27, returning to S21.
The method has the advantages that the single-phase grounding fault of the system with the neutral point not grounded can be judged by using the pulse neural membrane system of the distributed population, the parallel calculation of multiple servers can be adopted, the judging time is shortened, and the judging accuracy is improved.
Drawings
Fig. 1 is a 10kV small-current grounding system distribution network simulation model built by PSCAD/EMTDC.
Fig. 2 is a schematic diagram of a system architecture of a pulsed neural membrane system of a distributed population.
FIG. 3 is a graph of the trend of z values after 20 independent optimizations.
Detailed Description
The invention provides a steady-state information single-phase grounding fault line selection method of a pulse neural membrane system based on a distributed population, which aims at a power distribution network neutral point ungrounded system to establish a PSCAD/EMTDC simulation model and acquire fault data of fault lines and non-fault lines under different conditions; establishing a fault line selection model according to the fault data and the non-fault data to reflect the line fault condition; solving an objective function of the fault line selection model by using an optimized pulse neural membrane system; outputting the final optimized parameters (determining z value) to obtain P i ＝z·α i +(1-z)·β i And z. Calculating P from test data i Calculated P i When the voltage is greater than the threshold value, the single-phase earth fault of the line is judged. The method comprises the following steps:
a steady state information single-phase-to-earth fault line selection method of a pulse neural membrane system based on a distributed population comprises the following steps:
s10: and establishing a PSCAD/EMTDC simulation model aiming at a neutral point ungrounded system of the power distribution network, and acquiring fault data of a fault line and a non-fault line under different conditions.
S11: and establishing a fault line selection model according to the fault data and the non-fault data to reflect the line fault condition.
S111: collecting current data and reactive power data of three phases before and after a fault in a line to obtain an effective value of the component in the line, and simultaneously calculating a phase current variation delta I i Reactive power variation delta Q i I is the line number. Wherein the phase current variation ΔI i The current data of three phases in the ith line are respectively calculated absolute values of difference values of each phase before and after the fault, and the largest absolute value is taken as the phase current variation of the line.Reactive power variation delta Q i Is the absolute value of the difference between the reactive power data of the ith line before and after the fault.
S112: in a non-grounding system, the phase current and reactive power change amount change is obvious when single-phase grounding fault occurs, and the phase current and reactive power change amount can be changed
Taking the result as a line selection criterion. The normalization processing of the phase current and reactive power variation is as follows:
Figure BDA0003365566560000041
and->
Figure BDA0003365566560000042
Wherein alpha is i And beta i The values after normalization of the phase current and reactive power variation are respectively given, and n is the total number of lines.
S113: and in order to perform fusion criteria, the line selection weights in the two variable quantities are optimized. Namely, calculating the corresponding fault probability of all feeder lines at the bus, namely P i ＝z·α i +(1-z)·β i Wherein P is i For the failure probability of the feeder line, z is a weight parameter that needs to be optimized. In neutral point ungrounded systems, when calculated P i If the voltage is more than 0.5, the single-phase earth fault of the line is judged.
S12: and solving an objective function of the fault line selection model by using a pulse neural membrane system of the distributed population.
Model P built at S11 i ＝z·α i +(1-z)·β i In, alpha i And beta i From the three-phase current data and the reactive power data, z and P are known from the calculation in S11 i Unknown, given or optimized by z-value, the corresponding P can be determined i Values. Thus, based on the three-phase current data, reactive power data and tag data f obtained from the training data set i An objective function of the impulse membrane system is established. The method is characterized in that the thinking is that the label value f of each group of data in the training set is set i And P calculated after determination of z i The sum of squares of the differences is minimal as follows:
Figure BDA0003365566560000043
where O_z is the final optimized z value and N is the training set data number. The training input data of the objective function is single-phase current normalized value alpha i Reactive power normalized value beta i And a failure tag value f i 。
The specific steps of optimizing parameters by using the pulse neural membrane system of the distributed population are as follows:
s121: setting the number of neurons involved in evolution as l and establishing a network frame: initializing probabilities
Figure BDA0003365566560000044
Probability of variation
Figure BDA0003365566560000045
Number of individuals H, number of sub-populations M, number of individuals n in sub-populations, migration interval M 1 Migration quantity M 2 Iteration number gen=0, maximum iteration number max gen, ith sub-population probability matrix +.>
Figure BDA0003365566560000046
(/>
Figure BDA0003365566560000047
Element->
Figure BDA0003365566560000048
Elements representing the jth row and the kth column in the ith sub-population) and a binary matrix +.>
Figure BDA0003365566560000051
(/>
Figure BDA0003365566560000052
Element->
Figure BDA0003365566560000053
The table represents the elements of the j-th row and the k-th column in the i-th sub-population).
S122: the initialization population is divided into m sub-populations, each sub-population containing n individuals.
S123: if the iteration number gen is equal to or greater than maxgen, the algorithm returns to S1236.
S124：gen＝gen+1。
The following steps S125-S1215 cross-process the data.
S125：i＝1。
S126: if i > m, the algorithm returns to S1216.
S127：j＝1。
S128: if j > n, the algorithm returns to S1215.
S129：k＝1。
S1210: if k > l, the algorithm returns to S1215.
S1211: if random numbers
Figure BDA0003365566560000054
The algorithm performs the following two steps, otherwise the algorithm returns to S1212.
i. Of the m×n individuals, two individuals C different from the current individual j are selected k1 And C k2 C, i.e k1 ≠C k2 Not equal to j, if individual C k1 And C k2 The fitness function value f (C) k1 ) And f (C) k2 ) There is a relation f (C k1 )＞f(C k2 ) Then the current individual j is directed to C k1 Learning, i.e. b k ＝b k1 Otherwise, the current individual j is directed to C k2 Learning, i.e. b k ＝b k2 The method comprises the steps of carrying out a first treatment on the surface of the Wherein b k 、b k1 And b k2 Respectively the current individual, C k1 And C k2 The kth bit binary of the individual.
if b k =1, then the current rule probability value is
Figure BDA0003365566560000055
Otherwise, the current rule probability value is +.>
Figure BDA0003365566560000056
Wherein->
Figure BDA0003365566560000057
And->
Figure BDA0003365566560000058
The weight coefficients are respectively corresponding weight coefficients, and the value is a random number between 0.4 and 0.6.
S1212: if the best individual in the sub-population i corresponds to a binary bit
Figure BDA0003365566560000059
Then
Figure BDA00033655665600000510
Otherwise->
Figure BDA00033655665600000511
S1213: k=k+1 and returns to S1210.
S1214: j=j+1 and returns to S128.
S1215: i=i+1 and returns to S126.
The data is cross-processed in steps S125-S1215, and the method of this embodiment may be used, or other cross-processing methods of the film system may be used. The goal of the crossover process is to maintain the diversity requirements of the entire population to enhance the ability of the algorithm to exploit the optimal values.
S1216: if gen is M 1 The algorithm returns to S1217, otherwise it returns to S1220.
Steps S1217 to S1219 below perform information exchange processing on the data.
S1217：i＝1。
S1218: if i > m, the algorithm performs the following steps; otherwise, the algorithm returns to S1219;
i. according to the following formula (1), a candidate removal individual set S is constructed g,en ：
Figure BDA0003365566560000061
wherein,
Figure BDA0003365566560000062
indicating removal of population P i g,em The ith individual of the g generation in ∈,>
Figure BDA0003365566560000063
representation of
Figure BDA0003365566560000064
Is adapted to the value of->
Figure BDA0003365566560000065
Representing population P i g,em Is used for the average fitness value of (a).
From S according to the following formula (2) g,em Selecting individual immigration
Figure BDA0003365566560000066
Figure BDA0003365566560000067
wherein, the euclidean distance is indicated, ESNPS g,em Is from S g,em Is a group of the individuals in the group,
Figure BDA0003365566560000068
is an immigrating population P i g,im Is a function of the average distance of (a).
Constructing a matrix D according to the following formula (3) g,im ：
Figure BDA0003365566560000069
wherein, the euclidean distance is indicated,
Figure BDA00033655665600000610
indicating removal of population P i g,im The ith individual of the g generation, D g,im Representing population P i g,im The euclidean distance between every two individuals.
Selecting a replaced individual according to formula (4)
Figure BDA00033655665600000611
Figure BDA00033655665600000612
wherein,
Figure BDA00033655665600000613
and->
Figure BDA00033655665600000614
The representation being from D g,im The two individuals with the shortest euclidean distance,
Figure BDA00033655665600000615
and->
Figure BDA00033655665600000616
Are respectively->
Figure BDA00033655665600000617
And->
Figure BDA00033655665600000618
Is used for the adaptation value of the (c).
S1219 i=i+1 and returns to S1218.
Steps S1217 to S1219 perform information exchange processing on the data. The information exchange process provided in this embodiment has two advantages:
(1) So that the individual learning object in the algorithm is not a single optimal value in a single direction, but a plurality of optimal values in a plurality of directions, which enhances the searching efficiency of the algorithm.
(2) The exchange of information from multiple sub-populations facilitates parallel operations for multiple servers, which will greatly reduce the time penalty of the optimization algorithm.
The following steps S1220-S1234 perform mutation processing on the data.
S1220: calculating the current global optimum G bestfit (gen), current sample diversity value DP average (gen) column labels R corresponding to the best individuals of the current generation bestfit 。
S1221: if G bestfit (gen)＞G bestfit (gen-1), then let P cm1 =0; if G bestfit (gen)＝G bestfit (gen-1), then cause
Figure BDA0003365566560000071
Wherein P is cm1 And G bestfit (gen-1) represents a local optimum trigger condition and a previous generation global optimum, respectively; n (N) nimaxgen > 1 and N nimaxgen E, N, i.e. taking a natural number greater than 1.
S1222: calculating sample diversity ratio
Figure BDA0003365566560000072
Wherein DP average (0) Is the initial sample diversity value.
S1223: if the trigger condition rand is satisfied at the same time 1 (.)＜P cm1 And rand 2 (.)＞P cm2 Continuing, otherwise, turning to S1225; wherein rand is 1 (-) and rand 2 (.) is a random number taken from between 0 and 1.
S1224：i＝1。
S1225: if i > m, the algorithm returns to S1235.
S1226：j＝1。
S1227: if j > n, the algorithm returns to S1225.
S1228: if j +.R bestfit ，R bestfit If the number is the optimal individual number, the process continues, otherwise, the process goes to S1324 to continue.
S1229：k＝1。
S1230: if k > l, the algorithm returns to S1227.
S1331: if random free quantity
Figure BDA0003365566560000073
Then the current rule probability value is mutated P ij (gen)＝rand 4 (-) otherwise, not performing a mutation operation; the mutation operation is as follows: cross mutation is carried out on 20% of n-1 individuals except the optimal individuals in n individuals with the probability of 5%, so that a new individual population is obtained; wherein rand is 3 (-) and rand 4 (.) is a random number taken from between 0 and 1.
S1232: k=k+1 and returns to S1230.
S1233: j=j+1 and returns to S1227.
S1234: i=i+1 and returns to S1225.
S1220-S1234 performs mutation processing on the data, and the method of this embodiment can be adopted, or other mutation processing methods of the membrane system can be adopted. The mutation treatment is beneficial to strengthening the exploration capacity of the algorithm and preventing the algorithm from being in premature or locally optimal in the optimization process.
S1235: the algorithm returns to S123.
S1236: outputs f (z) and z.
S13: outputting the final optimized parameters to obtain P i ＝z·α i +(1-z)·β i And z. Calculating P from test data i Calculated P i When the voltage is greater than the threshold value (which can be set to 0.5), the single-phase earth fault of the line is judged.
The following describes the embodiments of the present invention with reference to the drawings.
The 10kV small current grounding system distribution network simulation model built by using PSCAD/EMTDC is shown in figure 1, and comprises 5 lines (Line 1-Line 5), wherein the types of the lines are overhead lines, cable lines and overhead cable mixed lines. The simulation parameters associated with the model are shown in table 1.
TABLE 1 simulation model parameter table
Figure BDA0003365566560000081
FIG. 2 is a schematic diagram of a distributed population membrane system.
The invention adopts a distributed population structure, the total population quantity is 50, the total population is divided into 5 small populations, each sub population is independently evolved, and each sub population is mutually communicated.
FIG. 3 is a graph of the trend of z values after 20 independent optimizations.
FIG. 3 shows that after 20 independent evolutions, the average value of z is 0.4969, and the objective function P is obtained i ＝0.4969α i +0.5031β i . And verifying 40 groups of test data to obtain the accuracy rate of 98.79%. Therefore, the established route selection model and the parameter optimization method are feasible and effective for steady-state information single-phase grounding fault route selection.
The steady-state information single-phase earth fault line selection method based on the optimized pulse neural membrane system provided by the invention can accurately determine the optimal parameters and select the fault line on the premise of only knowing the steady-state information.

Claims (1)

1. The single-phase grounding fault judging method of the pulse neural membrane system based on the distributed population is characterized by comprising the following steps of:
s10, collecting phase current and three-phase reactive power of each phase of an ith line of a neutral point ungrounded system of the power distribution network before and after a fault; calculating the absolute value of the difference value of the phase current of each phase in the ith line before and after the fault, and selecting the maximum absolute value as the phase current variation delta I of the ith line i The method comprises the steps of carrying out a first treatment on the surface of the Calculating the absolute value of the difference value of the three-phase reactive power of the ith line before and after the fault as the reactive power variation delta Q of the ith line i ；i＝1，2，...，n；
S11, phase current variation DeltaI i And reactive power variation Δq i And (3) respectively normalizing:
Figure FDA0003365566550000011
wherein alpha is i And beta i Respectively normalizing the phase current variation and the reactive power variation;
s12, calculating the fault probability P of the ith line i ＝z·α i +(1-z)·β i Z is a weight parameter;
s13, e.g. P i If the signal is larger than the threshold value, judging that the ith line has single-phase grounding fault;
the weight parameter z is obtained by training a pulse neural membrane system of a distributed population by using training set data; wherein,
the pulsed neural membrane system of the distributed population, comprising: neuron number, i, initialization probability
Figure FDA0003365566550000019
Mutation probability->
Figure FDA0003365566550000013
Number of sub-population M, number of individuals n in sub-population, total number of individuals H, migration interval M 1 Migration quantity M 2 The ith sub-population probability matrix->
Figure FDA0003365566550000014
And binary matrix->
Figure FDA0003365566550000015
Training a pulsed neural membrane system of a distributed population using training set data as follows:
s20, let iteration number gen=0, maximum iteration number maxgen;
s21, if the gen is larger than or equal to maxgen, training is completed;
S22，gen＝gen+1；
s23, performing cross processing;
s24, if gen is M 1 And then on,otherwise, jumping to S26;
s25, information exchange processing is carried out:
s251, let i=1;
s252, if i > m, executing the following steps; otherwise, jumping to S26;
i. construction of candidate removal Individual set S g，em ：
Figure FDA0003365566550000016
wherein,
Figure FDA0003365566550000017
indicating removal of the population->
Figure FDA0003365566550000018
The ith individual of the g generation in ∈,>
Figure FDA0003365566550000021
representation of
Figure FDA0003365566550000022
Is adapted to the value of->
Figure FDA0003365566550000023
Representing population->
Figure FDA0003365566550000025
Average fitness value of (a);
from S g，em Selecting individual immigration
Figure FDA0003365566550000026
Figure FDA0003365566550000027
wherein, table of IIndicating Euclidean distance, ESNPS g，em Is from S g，em Is a group of the individuals in the group,
Figure FDA0003365566550000028
is an immigrating population
Figure FDA0003365566550000029
Average distance of (2);
constructing matrix D g，im ：
Figure FDA00033655665500000210
wherein, the euclidean distance is indicated,
Figure FDA00033655665500000211
indicating removal of the population->
Figure FDA00033655665500000212
The ith individual of the g generation, D g，im Representing population->
Figure FDA00033655665500000213
A matrix formed by Euclidean distances between every two of the medium individuals;
selecting a replaced individual
Figure FDA00033655665500000214
Figure FDA00033655665500000215
wherein,
Figure FDA00033655665500000216
and->
Figure FDA00033655665500000217
The representation being from D g，im Two individuals with shortest Euclidean distance; />
Figure FDA00033655665500000218
And->
Figure FDA00033655665500000219
Are respectively->
Figure FDA00033655665500000220
And->
Figure FDA00033655665500000221
Is a fitness value of (a);
s253, let i=i+1, return to S252;
s26, performing mutation treatment;
s27, returning to S21.
CN202111403455.4A 2021-11-20 2021-11-20 Single-phase grounding fault judging method of pulse neural membrane system based on distributed population Active CN114113891B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111403455.4A CN114113891B (en) 2021-11-20 2021-11-20 Single-phase grounding fault judging method of pulse neural membrane system based on distributed population

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111403455.4A CN114113891B (en) 2021-11-20 2021-11-20 Single-phase grounding fault judging method of pulse neural membrane system based on distributed population

Publications (2)

Publication Number Publication Date
CN114113891A CN114113891A (en) 2022-03-01
CN114113891B true CN114113891B (en) 2023-04-25

Family

ID=80371892

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111403455.4A Active CN114113891B (en) 2021-11-20 2021-11-20 Single-phase grounding fault judging method of pulse neural membrane system based on distributed population

Country Status (1)

Country Link
CN (1) CN114113891B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109490702A (en) * 2018-10-08 2019-03-19 西南交通大学 A kind of method for diagnosing faults based on adaptive optimization pulse nerve membranous system
CN110598831A (en) * 2019-08-14 2019-12-20 西安理工大学 Improved backtracking search optimization algorithm based on multiple strategies
CN112505532A (en) * 2020-12-14 2021-03-16 电子科技大学 Analog circuit single fault diagnosis method based on improved particle swarm optimization
CN113448319A (en) * 2021-07-20 2021-09-28 国网四川省电力公司电力科学研究院 Fault diagnosis method based on rapid self-adaptive optimization pulse neurolemma system
CN113625118A (en) * 2021-08-17 2021-11-09 广西电网有限责任公司 Single-phase earth fault line selection method based on optimized pulse neurolemma system

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170286828A1 (en) * 2016-03-29 2017-10-05 James Edward Smith Cognitive Neural Architecture and Associated Neural Network Implementations
US20220358346A1 (en) * 2021-04-30 2022-11-10 Wisconsin Alumni Research Foundation Systems, methods, and media for generating and using spiking neural networks with improved efficiency

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109490702A (en) * 2018-10-08 2019-03-19 西南交通大学 A kind of method for diagnosing faults based on adaptive optimization pulse nerve membranous system
CN110598831A (en) * 2019-08-14 2019-12-20 西安理工大学 Improved backtracking search optimization algorithm based on multiple strategies
CN112505532A (en) * 2020-12-14 2021-03-16 电子科技大学 Analog circuit single fault diagnosis method based on improved particle swarm optimization
CN113448319A (en) * 2021-07-20 2021-09-28 国网四川省电力公司电力科学研究院 Fault diagnosis method based on rapid self-adaptive optimization pulse neurolemma system
CN113625118A (en) * 2021-08-17 2021-11-09 广西电网有限责任公司 Single-phase earth fault line selection method based on optimized pulse neurolemma system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Jianping Dong.A distribute adaptive optimization spiking neural P system for approximately solving combinatorial optimization problems.《ELSEVIER》.2022,第596卷全文. *
王琳.基于扩展P系统的粒子群算法及其在聚类分析中的应用.《中国博士学位论文全文数据库信息科技辑》.2020,全文. *

Also Published As

Publication number Publication date
CN114113891A (en) 2022-03-01

Similar Documents

Publication Publication Date Title
CN110082640B (en) Distribution network single-phase earth fault identification method based on long-time memory network
CN109459669B (en) 10kV power distribution network single-phase earth fault section positioning method
CN110609200B (en) Power distribution network earth fault protection method based on fuzzy metric fusion criterion
CN108920812A (en) A kind of machining surface roughness prediction technique
CN105512808A (en) Power system transient stability assessment method based on big data
CN104008424B (en) A kind of Optimal PMU Allocation in Power System
Saroha et al. Multi step ahead forecasting of wind power by genetic algorithm based neural networks
CN110417005B (en) Transient stability serious fault screening method combining deep learning and simulation calculation
CN113536509B (en) Micro-grid topology identification method based on graph convolution network
CN111798095A (en) Power cable state evaluation method based on neural network
CN111650472A (en) Method for positioning voltage sag source
CN113985194A (en) Power distribution network fault positioning method based on stack self-encoder
Zheng et al. Real-time transient stability assessment based on deep recurrent neural network
CN114113891B (en) Single-phase grounding fault judging method of pulse neural membrane system based on distributed population
CN112215279A (en) Power grid fault diagnosis method based on immune RBF neural network
CN106208043A (en) A kind of low and medium voltage distribution network key power consumption link discrimination method
CN114266396A (en) Transient stability discrimination method based on intelligent screening of power grid characteristics
CN113886183A (en) Method for measuring and calculating voltage sag event occurrence time
CN113625118B (en) Single-phase earth fault line selection method based on optimized pulse neural membrane system
CN111814394B (en) Power system safety assessment method based on correlation and redundancy detection
Yan et al. A new evolutionary neural network algorithm based on improved genetic algorithm and its application in power transformer fault diagnosis
CN111597640A (en) Method for predicting demand load of hybrid power ship under condition classification
Khudayarov et al. Power system steady state calculations using artificial neural networks
CN111313408A (en) Power grid fragile line identification method considering transient energy correlation
Bai et al. Contingency-based voltage stability monitoring via neural network with multi-level feature fusion

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