CN104348173B - It is a kind of based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth - Google Patents
It is a kind of based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth Download PDFInfo
- Publication number
- CN104348173B CN104348173B CN201410469609.3A CN201410469609A CN104348173B CN 104348173 B CN104348173 B CN 104348173B CN 201410469609 A CN201410469609 A CN 201410469609A CN 104348173 B CN104348173 B CN 104348173B
- Authority
- CN
- China
- Prior art keywords
- population
- power
- lateral cross
- reactive power
- algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/30—Reactive power compensation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Theoretical Computer Science (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Genetics & Genomics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Physiology (AREA)
- Power Engineering (AREA)
Abstract
The invention discloses a kind of Method for Reactive Power Optimization in Power based on improved CSO algorithms, the algorithm is a kind of based on the swarm intelligence searching algorithm for improving crossover algorithm (ICSO) in length and breadth, mainly including lateral cross, crossed longitudinally and three operators of differential variation.Lateral cross is neither repeatedly matched by all particles in population are carried out with two, between pairing particle and its outer rim is carried out search and updated；Crossed longitudinally is that all dimensions are matched, and then carries out arithmetic crossover；Differential variation finally carries out optimum selecting by entering row variation disturbance to each particle, intersecting, and three operators not only accelerate convergence rate, maintain the multiformity of population by selection operation Population Regeneration.The beneficial effects of the present invention is：The improvement of the present invention between crossover algorithm (ICSO) fast convergence rate, population Inner are individual in length and breadth communication for information completely, global convergence ability by force, particle multiformity it is good, for reactive power optimization of power system, this high-dimensional, multiple constraint, nonlinear complicated practical problem have the good suitability.
Description
Technical field
The present invention relates to a kind of Method for Reactive Power Optimization in Power, is calculated based on improvement CSO (ICSO) more particularly, to a kind of
The Method for Reactive Power Optimization in Power of method.
Background technology
Reactive power optimization of power system has extremely important meaning to the efficient, stable of power system, economical operation.As
The typical optimization problem of of power system, idle work optimization are referred on the basis of various constraintss are met, are sent out by control
The extreme voltage output of motor, reactive compensation capacity, on-load transformer tap changer carry out the reactive power flow of reasonable disposition electrical network
Distribution so that network outages are minimum, improves system voltage quality, and then improves safety and the economy of system.This is
One non-linear, the optimization problem that various dimensions, multiple constraint, continuous and discontinuous variable combine.
In recent years, with the development of intelligent algorithm, more and more it is used based on the optimized algorithm of biocenose intelligence
Come above reactive power optimization of power system problem, such as：Particle cluster algorithm (PSO), evolution algorithm (EA), genetic algorithm (GA), group search
Rope algorithm (GSO), artificial bee colony algorithm (ACO), ant group algorithm (ACO), difference algorithm (DE) etc..
Though these algorithms respectively have feature, and face has made some progress to varying degrees, for power system without
Work(optimizes this multiple constraint, the complicated optimum problem of non-linear, high latitude still has many drawbacks.Such as：Genetic algorithm (GA)
Long operational time；Particle cluster algorithm (PSO) is easily trapped into local optimum；Evolution algorithm (EA), artificial bee colony algorithm (ACO), group
The particle multiformity such as searching algorithm (GSO) are poor, and control parameter is more etc..In addition, many algorithm global convergence abilities, for idle
The global convergence ability for optimizing this complex model is not strong enough, is easily trapped into local optimum.
Therefore, how to enable a kind of algorithm efficiently, quickly and accurately to converge to optimal solution and be always a technology difficulty
Topic.
The content of the invention
The technical problem to be solved, is just to provide and a kind of is ensureing particle multiformity and algorithm the convergence speed
Under the premise of, while and the Method for Reactive Power Optimization in Power based on improvement CSO algorithms with powerful global convergence ability.
Above-mentioned technical problem is solved, the technical solution used in the present invention is：
A kind of Method for Reactive Power Optimization in Power based on improvement CSO algorithms, it is characterised in that comprise the following steps：
S1, with the minimum object function of system active power loss, it is considered to equality constraint and inequality constraints, sets up power system
Idle work optimization model；
S2, with reference to reactive power optimization of power system model initialization population；
S3, enters selection operation after execution is crossed longitudinally；
S4, enters selection operation after performing lateral cross；
S5, enters selection operation after performing differential variation；
S6, it is no that judgement reaches end condition：If reaching maximum iteration time, end loop, output result；Otherwise, turn
Step 3.
In described step S1, the minimum target function type of system active power loss is：
Wherein, Ploss is system active power loss；Ui, UjIt is node i respectively, the voltage magnitude of j；Gij, θijRespectively node
Conductance and phase difference of voltage between i, j；N refers to Population Size；
Equality constraint formula (2) is：
Wherein, PgiAnd PdiThe respectively active output and burden with power of node i；QgiAnd QdiThe idle output of node i respectively
And load or burden without work；BijFor node i, the susceptance between j；
Inequality constraints is：
Wherein, Ugi.minAnd Ugi.maxThe respectively voltage upper and lower limit of electromotor i；Qci.minAnd Qci.maxRespectively electric capacity is mended
Repay the switching group number upper and lower limit of device i；KTi.minAnd KTi.maxThe respectively tap position upper and lower limit of ULTC i；
ULi.minAnd ULi.maxThe respectively voltage upper and lower limit of load bus i；QGi.minAnd QGi.maxRespectively electromotor i is idle to exert oneself
Upper and lower limit；
Formula (1), (2) and (3) constitutes reactive power optimization of power system model.
Described step S2 initialization population is specially：
Setting Population Size N, control variable number D, D is dimension, including：Generator terminal voltage, UgReactive compensation capacity
Qc, ULTC no-load voltage ratio KT；Maximum iteration time Maxgen, crossed longitudinally rate pvc, lateral cross rate phc, difference intersection
Rate CR, random initializtion population X in the search space of problem is tieed up in D, wherein, it is X that i-th is individuali=[Xi1,Xi2,…XiD]。
In step S3, execution is crossed longitudinally to specifically include following sub-step：
S3-1 obtains parent population, namely the solution obtained after differential variation, and first on behalf of initial population；
S3-2 is normalized to the every one-dimensional of parent individuality, and formula is as follows：
Wherein, i ∈ (1, N), j ∈ (1, D)；PjminAnd PjmaxRespectively jth ties up the upper and lower limit of control variable；K is current
Algebraically；
S3-3 carries out two to all dimensions in population and neither repeatedly matches, and has N/2 pair, and two adjacent numbers are the sequence of pairing
Number；
S3-4 is taken out per a pair in order successively, if d1, d2Bidimensional is selected；
S3-5 is in crossed longitudinally rate pvcUnder, the d of particle X (i)1And d2Dimension execution is crossed longitudinally, and formula is as follows：
MSvc(i,d1)=rX (i, d1)+(1-r)·X(i,d2)i∈N(1,M),d1,d2∈N(1,D)；
Wherein, r is the uniform random number on [0,1]；MSvc(i,d1) for the d of particle X (i)1Dimension filial generation；
S3-6 repeats sub-step S3-4 and step S3-5d/2 time；
S3-7 is to MDvcRenormalization is carried out, golden mean of the Confucian school solution is obtained, formula is as follows：
MSvc(i, j)=MSvc(i,j)·(Pjmax-Pjmin)+Pjmin；
S3-8 performs selection operation, the solution DS that is dominant after acquisition is crossed longitudinallyvc。
In step S4, perform lateral cross and specifically include following sub-step：
S4-1 obtains parent population, namely the crossed longitudinally solution DS that is dominantvc；
S4-2 to population in all individualities carry out two and neither repeatedly match, have N/2 pair, two adjacent numbers are pairing
Sequence number；
S4-3 is taken out per a pair in order successively, if particle X (i) and X (j) are removed；
S4-4 is in lateral cross probability PhcUnder horizontal line performed to the d dimensions of particle X (i) and X (j) intersect, formula is as follows：
MShc(i, d)=r1·X(i,d)+(1-r1)·X(j,d)+c1·(X(i,d)-X(j,d))
MShc(j, d)=r1·X(j,d)+(1-r2)·X(i,d)+c1·(X(j,d)-X(i,d))
Wherein, d ∈ (1, D)；r1,r2For the uniform random number on [0,1]；c1,c2For uniform between [- 1,1]
Distribution random numbers；X (i, d) and X (j, d) is respectively the d dimensions of particle i and j；MShc(i, d) and MShc(j, d) is respectively horizontal
Golden mean of the Confucian school solution after intersection；
S4-5 repeat steps S4-3 and step S4-4N/2 time；
S4-6 performs selection operation, obtains the solution DS that is dominant after lateral crosshc。
In step S5, perform differential variation and specifically include following sub-step：
S5-1 obtains the solution DS that is dominant of parent population, i.e. lateral crosshc；
S5-2 is individual to each in kth generationI ∈ (1, N), carry out mutation operation as the following formula：
Wherein, r1∈(1,N),r2∈(1,N),r3∈ (1, N), and i ≠ r1≠r2≠r3；MF ∈ [0.05,2], are scaling
The factor；For individualityVariant.
S5-3 is to each individuality in populationCrossover operation is carried out, formula is as follows：
Wherein, CRFor crossover probability, 0.2 is set to；Rnb (i) is a random integers between [1, N]；After intersection
The experiment vector for obtaining；
S5-4 performs selection operation.Calculate trial vectorAdaptive value, if compared withIt is more excellent, thenIt is no
Then,
Described selection operation is as follows：
If the adaptive value of golden mean of the Confucian school solution MS (i) is better than its parent X (i), DS (i) ← MS (i)；Otherwise, DS (i) ← X (i).
Compared with prior art, the present invention has following significant effect：
(1) ICSO algorithms proposed by the present invention are simple and clear, and control parameter is few, it is easy to operate, when reducing algorithm operation
Between；
(2) the crossed longitudinally and lateral cross operator used by ICSO algorithms proposed by the present invention causes population changing every time
All particles or all dimensions is all made to be matched two-by-two in generation, compromise updates, competition takes excellent, takes full advantage of each in colony
Individual information.And crossed longitudinally and lateral cross is alternateed and is carried out, and the optimal way of this interlock type causes particle
Individual information rapidly can be spread in population, reached good alternating current function, drastically increase the complete of algorithm
Office's convergence capabilities；
(3) the differential variation operator used by ICSO algorithms proposed by the present invention causes particle in quick renewal process
Population diversity can be kept well, improve the ability that algorithm jumps out local optimum.
Description of the drawings
Fig. 1 is the power system embodiment IEEE-57 node system figure using the present invention；
Fig. 2 is the flow chart based on the Method for Reactive Power Optimization in Power for improving CSO algorithms of the present invention；
Fig. 3 is ICSO algorithms and original CSO algorithms and PSO convergence of algorithm curve charts.
Specific embodiment
It is illustrated in figure 1 the power system embodiment IEEE-57 node system figure using the present invention.
The system includes 7 electromotors, 3 reactive-load compensation points and 15 ULTCs.Voltage adjusting range is
[0.95,1.1], reactive-load compensation are located at 18,25,53 nodes, and point 10 grades of regulations, the upper limit are respectively 0.2,0.1,0.1, have load to adjust
Pressure transformer has 17 grades of regulation stalls, and no-load voltage ratio range of accommodation is [0.9,1.1].Reference power is 100MW, initial active of system
Network loss is 0.3030pu.
As carried out by Fig. 2, the Method for Reactive Power Optimization in Power based on improvement CSO algorithms of the present invention is in IEEE57 nodes system
Flow chart in system example, comprises the following steps：
Step 1, with the minimum object function of system active power loss, it is considered to equality constraint and inequality constraints, sets up electric power
System Reactive Power Optimized model；
Step 2, with reference to reactive power optimization of power system model initialization population；
Step 3, enters selection operation after execution is crossed longitudinally；
Step 4, enters selection operation after performing lateral cross；
Step 5, enters selection operation after performing differential variation；
Step 6, end condition：If reaching maximum iteration time, end loop, output result；Otherwise, 3 are gone to step；
In the step 1, reactive power optimization of power system model includes object function and constraints, and object function is to be
System active power loss, constraints include equality constraint and inequality constraints.
Concrete form with the minimum object function of system active power loss is：
Wherein, Ploss is system active power loss；Ui, UjIt is node i respectively, the voltage magnitude of j；Gij, θijRespectively node
Conductance and phase difference of voltage between i, j.
Equality constraint is：
Wherein, PgiAnd PdiThe respectively active output and burden with power of node i；QgiAnd QdiThe idle output of node i respectively
And load or burden without work；BijFor node i, the susceptance between j.
Inequality constraints is：
Wherein, Ugi.minAnd Ugi.maxThe respectively voltage upper and lower limit of electromotor i；Qci.minAnd Qci.maxRespectively electric capacity is mended
Repay the switching group number upper and lower limit of device i；KTi.minAnd KTi.maxThe respectively tap position upper and lower limit of ULTC i；
ULi.minAnd ULi.maxThe respectively voltage upper and lower limit of load bus i；QGi.minAnd QGi.maxRespectively electromotor i is idle to exert oneself
Upper and lower limit；
In the step 2, initialization is specially：
Setting Population Size N=40, control variable number D=25 (i.e. dimension, including：Electromotor UgSet end voltage, idle benefit
Repay capacity Qc, ULTC no-load voltage ratio KT), maximum iteration time Maxgen=200, crossed longitudinally rate pvc=1, laterally hand over
Fork rate phc=0.6, difference crossing-over rate CR=0.2, random initializtion population X in the receipts rope space of problem is tieed up in D, wherein, i-th
Individual is Xi=[Xi1,Xi2,...XiD]。
In the step 3, crossed longitudinally concretely comprising the following steps is performed：
(1) obtain parent population (solution for obtaining after differential variation, first on behalf of initial population)；
(2) the every one-dimensional of parent individuality is normalized, formula is as follows：
Wherein, i ∈ (1, N), j ∈ (1, D)；PjminAnd PjmaxRespectively jth ties up the upper and lower limit of control variable；K is current
Algebraically.
(3) two neither repeatedly pairings (total N/2 to) are carried out to all dimensions in population, two adjacent numbers are the sequence of pairing
Number；
(4) take out per a pair successively in order, if d1, d2Bidimensional is selected；
(5) in crossed longitudinally rate pvcUnder, the d of particle X (i)1And d2Dimension execution is crossed longitudinally, and formula is as follows：
MSvc(i,d1)=rX (i, d1)+(1-r)·X(i,d2)i∈N(1,M),d1,d2∈N(1,D)
Wherein, r is the uniform random number on [0,1]；MSvc(i,d1) for the d of particle X (i)1Dimension filial generation；
(6) repeat step (4) and step (5) d/2 time；
(7) to MDvcRenormalization is carried out, golden mean of the Confucian school solution is obtained, formula is as follows：
MSvc(i, j)=MSvc(i,j)·(Pjmax-Pjmin)+Pjmin
(8) selection is performed, the solution DS that is dominant after acquisition is crossed longitudinallyvc。
In the step 4, concretely comprising the following steps for lateral cross is performed：
(1) obtain parent population (the i.e. crossed longitudinally solution DS that is dominantvc)；
(2) to population in all individualities carry out two neither repeatedly pairings (total N/2 to), two adjacent numbers are pairing
Sequence number；
(3) take out per a pair successively in order, if particle X (i) and X (j) are removed；
(4) in lateral cross probability PhcUnder horizontal line performed to the d dimensions of particle X (i) and X (j) intersect, formula is as follows：
MShc(i, d)=r1·X(i,d)+(1-r1)·X(j,d)+c1·(X(i,d)-X(j,d))
MShc(j, d)=r1·X(j,d)+(1-r2)·X(i,d)+c1·(X(j,d)-X(i,d))
Wherein, d ∈ (1, D)；r1,r2For the uniform random number on [0,1]；c1,c2For uniform between [- 1,1]
Distribution random numbers；X (i, d) and X (j, d) is respectively the d dimensions of particle i and j；MShc(i, d) and MShc(j, d) is respectively horizontal
Golden mean of the Confucian school solution after intersection；
(5) repeat step (3) and step (4) N/2 time；
(6) selection operation is performed, obtains the solution DS that is dominant after lateral crosshc。
In the step 5, comprising the following steps that for differential variation is performed：
(1) obtain parent population (the i.e. solution DS that is dominant of lateral crosshc)；
(2) it is individual to each in kth generationI ∈ (1, N), carry out mutation operation as the following formula：
Wherein, r1∈(1,N),r2∈(1,N),r3∈ (1, N), and i ≠ r1≠r2≠r3；MF ∈ [0.05,2], are scaling
The factor；For individualityVariant.
(3) to each individuality in populationCrossover operation is carried out, formula is as follows：
Wherein, CRFor crossover probability, 0.2 is set to；Rnb (i) is a random integers between [1, N]；After intersection
The experiment vector for obtaining；
(4) perform selection operation.Calculate trial vectorAdaptive value, if compared withIt is more excellent, thenIt is no
Then,
Selection operation described in the step 3 and step 4, operates as follows：
If golden mean of the Confucian school solution MS (i), adaptive value be better than its parent X (i), then DS (i) ← MS (i)；Otherwise, DS (i) ← X (i).
As shown in Figure 3 ICSO algorithms and original CSO algorithms and PSO convergence of algorithm curve charts, it can be seen that ICSO algorithms compared with
Other two kinds of algorithms all have found more excellent solution.Although preconvergence speed is not so good as PSO algorithms, the later stage is calculated than original CSO
Method and PSO algorithms will be fast, and embody more preferable global convergence performance, it is to avoid Premature Convergence.
Most there is value to contrast for IEEE57 node systems idle work optimization as shown in table 1, can be seen that by contrast, it is various meeting
On the premise of constraints, after ICSO optimizations, resulting system losses are minimum, and in the reasonable scope, optimum results make us full
Meaning.
Table 1
Claims (5)
1. it is a kind of based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth, it is characterised in that to comprise the following steps：
S1, with the minimum object function of system active power loss, it is considered to equality constraint and inequality constraints, sets up power system reactive power
Optimized model；
S2, with reference to reactive power optimization of power system model initialization population；
S3, enters selection operation after execution is crossed longitudinally；
S4, enters selection operation after performing lateral cross；
S5, enters selection operation after performing differential variation；
S6, it is no that judgement reaches end condition：If reaching maximum iteration time, end loop, output result；Otherwise, go to step
S3；
In described step S1, the minimum target function type of system active power loss is：
Wherein, Ploss is system active power loss；Ui, UjIt is the voltage magnitude of i-th and j-th node respectively；Gij, θijRespectively
For conductance and phase difference of voltage between i-th and j-th node；N refers to Population Size；
Equality constraint formula is：
Wherein, PgiAnd PdiThe active output of respectively i-th node and burden with power；QgiAnd QdiThe nothing of respectively i-th node
Work(is exported and load or burden without work；BijFor the susceptance between i-th node and j-th node；
Inequality constraints is：
Wherein, Ugi.minAnd Ugi.maxThe voltage upper and lower limit of respectively i-th electromotor；Qci.minAnd Qci.maxRespectively i-th electric
Hold the switching group number upper and lower limit of compensator；KTi.minAnd KTi.maxOn the tap position of respectively i-th ULTC,
Lower limit；ULi.minAnd ULi.maxThe voltage upper and lower limit of respectively i-th load bus；QGi.minAnd QGi.maxRespectively i-th generating
The upper and lower limit that machine is idle to exert oneself；
Formula (1), (2) and (3) constitutes reactive power optimization of power system model.
2. according to claim 1 based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth, its feature exists
In：Described step S2 initialization population is specially：
Setting Population Size N, control variable number D, D is dimension, including：Generator terminal voltage Ug, reactive compensation capacity Qc, have
Voltage adjustment of on-load transformer voltage ratio KT；Maximum iteration time Maxgen, crossed longitudinally rate pvc, lateral cross rate phc, difference crossing-over rate CR,
Random initializtion population X in the search space of problem is tieed up in D, wherein, it is X that i-th is individuali=[Xi1,Xi2,...XiD], i ∈
(1,N)。
3. according to claim 2 based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth, its feature exists
In：In step S3, execution is crossed longitudinally to specifically include following sub-step：
S3-1 obtains parent population, namely the solution obtained after differential variation, and first on behalf of initial population；
S3-2 is normalized to the every one-dimensional of parent individuality, and formula is as follows：
Wherein, i ∈ (1, N), j ∈ (1, D)；PjminAnd PjmaxRespectively jth ties up the upper and lower limit of control variable；K is current algebraically；
S3-3 carries out two to all dimensions in population and neither repeatedly matches, and has D/2 pair, and two adjacent numbers are the sequence number of pairing；
S3-4 is taken out per a pair in order successively, if d1, d2Bidimensional is selected；
S3-5 is in crossed longitudinally rate pvcUnder, kth for population in i-th it is individualD1And d2Dimension performs crossed longitudinally, public affairs
Formula is as follows：
MSvc(i,d1)=rXk(i,d1)+(1-r)·Xk(i,d2)i∈(1,N),d1,d2∈(1,D)
Wherein, r is the uniform random number on [0,1]；MSvc(i,d1) for kth it is individual for i-th in populationD1Dimension
Filial generation；
S3-6 repeats sub-step S3-4 and step S3-5D/2 time；
S3-7 is to MSvcRenormalization is carried out, golden mean of the Confucian school solution is obtained, formula is as follows：
MSvc(i, j)=MSvc(i,j)·(Pjmax-Pjmin)+Pjmin；
S3-8 performs selection operation, the solution DS that is dominant after acquisition is crossed longitudinallyvc。
4. according to claim 3 based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth, its feature exists
In：In step S4, perform lateral cross and specifically include following sub-step：
S4-1 obtains parent population, namely the crossed longitudinally solution DS that is dominantvc；
S4-2 to population in all individualities carry out two and neither repeatedly match, have N/2 pair, two adjacent numbers are the sequence of pairing
Number；
S4-3 is taken out in order successively per a pair, if kth for population in i-th it is individualIt is individual with j-thIt is removed；
S4-4 is in lateral cross probability PhcUnder to individualityWithD dimensions perform lateral cross, formula is as follows：
MShc(i, d)=r1·Xk(i,d)+(1-r1)·(j,d)+c1(Xk(i,d)-Xk(j,d))
MShc(j, d)=r2·Xk(j,d)+(1-r2)·Xk(i,d)+c2(Xk(j,d)-Xk(i,d))
Wherein, d ∈ (1, D)；r1,r2For the uniform random number on [0,1]；c1,c2For between [- 1,1] be uniformly distributed with
Machine number；Xk(i, d) and Xk(j, d) is respectively d dimension of the kth for i-th and j-th individuality in population；MShc(i, d) and MShc
(j, d) is respectively i-th and j-th golden mean of the Confucian school solution MS (i) after lateral cross, the d dimensions of MS (j)；
S4-5 repeat steps S4-3 and step S4-4N/2 time；
S4-6 performs selection operation, obtains the solution DS that is dominant after lateral crosshc。
5. according to claim 4 based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth, its feature exists
In：In step S5, perform differential variation and specifically include following sub-step：
S5-1 obtains the solution DS that is dominant of parent population, i.e. lateral crosshc；
S5-2 is individual for each in population to kthI ∈ (1, N), carry out mutation operation as the following formula：
Wherein, r1∈(1,N),r2∈(1,N),r3∈ (1, N), and i ≠ r1≠r2≠r3；MF ∈ [0.05,2], are zoom factor；For individualityVariant；
S5-3 is to kth for each individuality in populationCrossover operation is carried out, formula is as follows：
Wherein, CRFor difference crossover probability, 0.2 is set to；Rnb (i) is a random integers between [1, N]；After intersection
The trial vector for obtainingJth dimension；It is individual for i-th variation in kth+1 generation populationJth dimension；For kth
It is individual for i-th in populationJth dimension；
S5-4 performs selection operation：Calculate trial vectorAdaptive value, if compared withIt is more excellent, then assignmentGiveIt is no
Then, assignmentGive
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410469609.3A CN104348173B (en) | 2014-09-15 | 2014-09-15 | It is a kind of based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410469609.3A CN104348173B (en) | 2014-09-15 | 2014-09-15 | It is a kind of based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104348173A CN104348173A (en) | 2015-02-11 |
| CN104348173B true CN104348173B (en) | 2017-03-29 |
Family
ID=52503228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410469609.3A Active CN104348173B (en) | 2014-09-15 | 2014-09-15 | It is a kind of based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104348173B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105608507A (en) * | 2015-12-22 | 2016-05-25 | 广东电网有限责任公司江门供电局 | Transmission line inspection path optimization method based on differential crisscross optimization algorithm |
| CN106410816B (en) * | 2016-09-12 | 2019-01-22 | 广东工业大学 | Based on the multi-objective reactive optimization method and system for improving crossover algorithm in length and breadth |
| CN107145982B (en) * | 2017-05-05 | 2020-02-21 | 广东工业大学 | Unit combination optimization method and device based on longitudinal and transverse intersection algorithm |
| CN107276121A (en) * | 2017-06-23 | 2017-10-20 | 广东工业大学 | A kind of family grid-connected collaboration economic load dispatching optimization method of meter and uncertain factor |
| CN107591843B (en) * | 2017-08-31 | 2020-09-01 | 南京理工大学 | Double-fed wind field reactive power output optimization method in system recovery process |
| CN108915927B (en) * | 2018-07-03 | 2020-08-11 | 广东工业大学 | Method, device and equipment for determining parameters of wave power generation device |
| CN110224410B (en) * | 2019-03-27 | 2020-12-04 | 广东电网有限责任公司 | Power grid dynamic reactive power optimization configuration method based on transient voltage safety constraint |
| CN111062170A (en) * | 2019-12-03 | 2020-04-24 | 广东电网有限责任公司 | Transformer top layer oil temperature prediction method |
| CN112329328A (en) * | 2020-07-07 | 2021-02-05 | 广东电网有限责任公司 | Transformer loss calculation method based on CSO optimization neural network |
| CN112100907A (en) * | 2020-08-24 | 2020-12-18 | 广东工业大学 | Power grid optimal power flow problem solution method based on distributed criss-cross algorithm |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102222919B (en) * | 2011-05-19 | 2013-04-10 | 西南交通大学 | Power system reactive power optimization method based on improved differential evolution algorithm |
| CN102611119B (en) * | 2012-03-14 | 2014-07-09 | 华北电力大学 | Multi-target reactive power optimization method for electric system |
| CN102820662B (en) * | 2012-08-17 | 2014-08-06 | 华北电力大学 | Distributed power source contained power system multi-target reactive-power optimization method |
| CN103199544A (en) * | 2013-03-26 | 2013-07-10 | 上海理工大学 | Reactive power optimization method of electrical power system |
-
2014
- 2014-09-15 CN CN201410469609.3A patent/CN104348173B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN104348173A (en) | 2015-02-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104348173B (en) | It is a kind of based on the Method for Reactive Power Optimization in Power for improving crossover algorithm in length and breadth | |
| CN102820662B (en) | Distributed power source contained power system multi-target reactive-power optimization method | |
| CN105140907B (en) | Direct-current micro-grid multiple agent adaptively sagging uniformity control method for coordinating and device | |
| CN105790282B (en) | The idle work optimization analysis system and method for a kind of power network containing UPFC | |
| CN105956706A (en) | Distribution network reactive power optimization method based on improved artificial bee colony algorithm | |
| CN103746388B (en) | Electric distribution network reactive-voltage three grades of control method for coordinating | |
| CN104734153A (en) | Method of reconstructing power distribution network containing distributed power supply | |
| CN104734548B (en) | A kind of control method of photovoltaic combining inverter and photovoltaic combining inverter | |
| CN108306298A (en) | A kind of design method of flexibility multimode switch access power distribution network | |
| CN108847680A (en) | A kind of alternating current-direct current mixing power distribution network hierarchical control method based on flexible looped network device | |
| CN106712552B (en) | A kind of aviation more electric engin VIENNA rectifier control method | |
| CN103199544A (en) | Reactive power optimization method of electrical power system | |
| CN113644693B (en) | Distributed operation control method for renewable energy/hydrogen-containing alternating current-direct current hybrid system | |
| CN111049173A (en) | Self-organizing droop control method for multi-terminal direct-current distribution network | |
| Karimianfard et al. | An initial-point strategy for optimizing distribution system reconfiguration | |
| CN106159955B (en) | Electric system distributed optimal power flow method based on continuous punishment Duality Decomposition | |
| CN105633955B (en) | A kind of multifunctional solid current limiter Optimal Configuration Method | |
| CN107546773A (en) | A kind of more micro-capacitance sensor dynamic networking methods in region based on graph theory | |
| CN109687472A (en) | A kind of reactive voltage control method for coordinating based on flexible multimode switch | |
| CN109617049A (en) | A kind of UPFC configuration method of wind-powered electricity generation pooling zone | |
| CN110365050A (en) | The grid-connected Multipurpose Optimal Method of DWF based on difference cellular genetic algorithm | |
| Feng et al. | Stability constrained reinforcement learning for decentralized real-time voltage control | |
| CN106777521A (en) | Generating set based on double-chain quantum genetic algorithm relates to network parameters optimization method | |
| CN107346885B (en) | A kind of DC/DC bi-directional inverter stable DC busbar voltage method for optimally controlling | |
| CN107256440A (en) | A kind of carbon energy combined-flow method for solving based on imperial competition algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address | ||
| CP03 | Change of name, title or address |
Address after: 522000 west section of Donger Road, Rongcheng District, Jieyang City, Guangdong Province (power dispatching building) Patentee after: JIEYANG POWER SUPPLY BUREAU, GUANGDONG POWER GRID Co.,Ltd. Patentee after: GUANGDONG University OF TECHNOLOGY Address before: 522000 west section of Donger Road, Rongcheng District, Jieyang City, Guangdong Province (power dispatching building) Patentee before: JIEYANG POWER SUPPLY BUREAU, GUANGDONG STATE GRID Co.,Ltd. Patentee before: GUANGDONG University OF TECHNOLOGY |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210119 Address after: 522000 west section of Donger Road, Rongcheng District, Jieyang City, Guangdong Province (power dispatching building) Patentee after: JIEYANG POWER SUPPLY BUREAU, GUANGDONG POWER GRID Co.,Ltd. Address before: 522000 west section of Donger Road, Rongcheng District, Jieyang City, Guangdong Province (power dispatching building) Patentee before: JIEYANG POWER SUPPLY BUREAU, GUANGDONG POWER GRID Co.,Ltd. Patentee before: GUANGDONG University OF TECHNOLOGY |