CN105896528A - Power distribution network reconstruction method based on isolation ecological niche genetic algorithm - Google Patents
Power distribution network reconstruction method based on isolation ecological niche genetic algorithm Download PDFInfo
- Publication number
- CN105896528A CN105896528A CN201610255192.XA CN201610255192A CN105896528A CN 105896528 A CN105896528 A CN 105896528A CN 201610255192 A CN201610255192 A CN 201610255192A CN 105896528 A CN105896528 A CN 105896528A
- Authority
- CN
- China
- Prior art keywords
- sub
- group
- distribution network
- power distribution
- scale
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/04—Circuit arrangements for ac mains or ac distribution networks for connecting networks of the same frequency but supplied from different sources
- H02J3/06—Controlling transfer of power between connected networks; Controlling sharing of load between connected networks
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Power Engineering (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)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to a power distribution network reconstruction method based on an isolation ecological niche genetic algorithm. The power distribution network reconstruction method is characterized in: researching to use a target normalization method to obtain the efficient solution; taking minimum network loss as the main target; taking load balancing as the constrained condition to process and establish a compromise model; and bringing forward a chromosome integer coding mode to ensure one-to-one correspondence between the chromosome and the feasible solution. The power distribution network reconstruction method based on an isolation ecological niche genetic algorithm can effectively solve the premature convergence problem of the genetic algorithm and can quickly converge to the optimal solution by introducing the isolation ecological niche genetic algorithm, and at the same time, aiming at that binary coding cannot effectively reflect the structural features of the power distribution network reconstruction problem, the power distribution network reconstruction method provides the chromosome integer coding mode, thus greatly reducing the chromosome length and improving the operation speed; and by means of introduction of the isolation ecological niche technique, the power distribution network reconstruction method effectively solves the premature convergence problem of a traditional genetic algorithm on solving the power distribution network reconstruction problem.
Description
Technical field
The present invention relates to power distribution network optimisation technique field, particularly a kind of reconstruction method of power distribution network.
Background technology
Power distribution network is a power system link from generating to electricity consumption, and it is distributed in load center region, for each use
Family directly provides power supply.In the comprehensive line loss of power system electrical network at different levels, the line loss of low and medium voltage distribution network account for the biggest ratio
Example.Therefore, under conditions of meeting medium voltage distribution network operational reliability, by network reconfiguration reduce line loss be one the most worth
The problem of research
Power distribution network reconfiguration is a multi-target non-linear hybrid optimization problem, and existing algorithm is mostly with simple target function
For model.Due to the nonlinear characteristic of Distribution Networks Reconfiguration, Optimized Iterative is required to carry out a distribution Load flow calculation each time,
Continuous print distribution Load flow calculation is necessarily required to calculate in a large number the time.In order to improve calculating speed, it is ensured that draw optimum or suboptimum
Distributing network structure, people have attempted different methods to the problem solving multiple target Distribution system.Branch exchange method and taboo
Search method is too dependent on the initial configuration of distribution network structure and distribution network, it is impossible to ensure global optimum.Optimal flow pattern
Method[4]Being required for calculating a trend to conjunction by opening each time, amount of calculation is bigger.Genetic algorithm[5]It is to carry out excellent from group's point
Changing search rather than the change of single on off state, the speed of such global convergence is accelerated, but Premature Convergence and later stage search are late
Blunt problem is difficult to overcome.Generally on off state (0/1) the direct chromosome of branch road represents, but the switch of distribution is not to appoint
Meaning independent assortment, it is affected by distribution radial pattern structure and each load has the actual motion constraints such as supply of electric power,
Therefore, use binary coding will necessarily produce a large amount of infeasible solution, thus reduce the efficiency of algorithm.
Within 1975, Holland be first proposed niche evolutionary algorithm by the inspiration of ecosystem evolution[13]Thought.?
In niche evolutionary algorithm, a population is similar to an ecosystem, and one group of individuality with certain similitude is similar to thing
Kind.One species is regarded as a sub-population, and therefore, species, sub-population, microhabitat are man-to-man relations.Microhabitat is evolved
The elementary object of algorithm is exactly to form and maintain stable various beggar population, enters concurrently in the zones of different of search volume
Change search, thus overcome the uniform convergence trend of genetic shift, it is achieved multimodal, the optimization of multi-objective problem.Quarantine items is lost
Propagation algorithm is exactly the geographic isolation technology according to nature, and the initial population of genetic algorithm is divided into several sub-group, sub-group
Between independent evolve, the evolution speed of each sub-group and scale depend on the average adaptation level of each sub-group.Due to every
Sub-group after from is independent of one another, and the evolutionary process of each sub-group can be controlled by clear-cut flexibly, and this is microhabitat skill
The unexistent feature of art.So, algorithm can not only ensure the diversity solved in colony effectively, and has the strongest guiding
Evolvability.
Summary of the invention
It is an object of the invention to provide a kind of reconstruction method of power distribution network based on genetic algorithm based on isolation niche, it is by net
Damaging minimum as major heading, balancing the load, as constraints, by introducing genetic algorithm based on isolation niche, efficiently solves
Genetic algorithm premature problem also rapidly converges to optimal solution.
It is an object of the invention to be realized by such technical scheme, specifically comprise the following steps that
1) according to power distribution network actual conditions, it is thus achieved that the parameter in genetic algorithm based on isolation niche;
2) Mathematical Modeling of power distribution network reconfiguration is built;
3) power distribution network is carried out chromosome integer coding;
4) the producing and the isolation of initial subgroup of initial individuals;
5) determination of individual fitness;
6) determination of sub-group scale;
7) protection releases and judges;
8) inferior strain is not lived and is judged;
9) mutual exclusion of the same race judges;
10) update evolution and convergence judges.
Further, step 2) described in Mathematical Modeling be to take into account balancing the load and loss minimization turns to the mathematics of object function
Model:
In formula, f is the active loss of system, can try to achieve for trend method by front pushing back;kiRepresent the state of switch i,
Being 0-1 discrete magnitude, 0 represents disconnection, and 1 represents Guan Bi;riRepresent the resistance of branch road i;Pi、QiRepresent the active power flowing through branch road i
And reactive power;UiRepresenting the voltage of branch road i endpoint node, constraints is:
2-1) tributary capacity constraint Si≤Si,max (4)
2-2) node voltage constraint Ui,min≤Ui≤Ui,max (5)
2-3) account load balancing constraints
2-4) network topology constraint unpowered isolated island;
In formula, Ui、Ui,min、Ui,maxIt is respectively voltage and the bound thereof of node i.
Further, step 3) described in power distribution network carried out in chromosome integer coding below all control variables need to meet
Three times principles:
3-1) the branch switch on any loop must not close, and is not involved in chromosome coding;
3-2) switch being connected with power supply point also should close, and is also not involved in chromosome coding;
If 3-3) common switch be opened twice and more than, then this solution is infeasible solution, removes this chromosome.
Further, step 4) described in the producing and the isolation of initial subgroup of initial individuals method particularly includes: produce at random
Life randomly generates N number of initial individuals;N number of initial individuals is all given K sub-group, and the number of individuals that each sub-group contains is
N/K。
Further, step 5) described in the determination of individual fitness method particularly includes:
Calculate all individual fitnesses in colony, and preserve the individuality that adaptive value is the highest.It is provided with t for kth sub-group
In jth individual, by calling trend forward-backward sweep method calculation procedure, the target function value i.e. line loss trying to achieve its correspondence is Fkj
(t).Use penalty function method to retrain the branch power of formula (4), the node voltage landing of formula (5) retrains and the load of formula (6) is put down
Weighing apparatus constraint processes, then this individual fitness is:
In formula: kU、kS、kBAnd φUi、φsi、φBiBe respectively voltage, power and account load balancing constraints penalty factor with penalize
Function;
For overcoming voltage, power and balancing the load rate dimension disunity and numerical value difference is bigger etc. that calculating is caused by problem
Impact, with formula (8), (9) and (10) calculate φUi、φsi, and φBi:
Further, step 6) described in the concrete grammar of determination of sub-group scale as follows:
The scale of sub-group is relevant with the average adaptive value of sub-group, and the average adaptive value of sub-group is the biggest, and it is at next
The individuality having in Dai is the most;Otherwise, the individuality having is the fewest.But number must is fulfilled for maximum allowable scale and Minimal Protective
The restriction of scale;If t+1 is for scale n of kth sub-groupk(t+1) M is metmin≤nk(t+1)≤Mmax, wherein Mmin, Mmax
It is respectively Minimal Protective scale and maximum allowable scale, setting according to the actual conditions of network and the scale of initial subgroup of its value
Depending on size;
Being defined below of sub-group scale:
6-1) give average fitness assignment M of each sub-groupminIndividuality, average according to sub-group of remaining individuality
Adaptive value utilizes roulette method choice, until total Population reaches N;
6-2) the average adaptive value of sub-group can be taken as such as following formula:
In formula: fk aT () is the t average adaptive value for kth sub-group;fki(t) for t in kth sub-group i-th
Individual individual fitness;nkT () is the t scale for kth sub-group;
6-3) scale n in sub-group k t+1 generationk(t+1) it is:
Further, step 7) described in protection release judge concrete grammar as follows:
Make t for scale n of kth sub-groupkT () meets nk(t)≤Mmin, then promoter colony defence program, force
Make nk(t)=Mmin.If evolve to t+1 for time this sub-group scale nk(t+1) meet: Mmin≤nk(t+1)≤Mmax, then solve
Divisor group body is protected.
Further, step 8) described in inferior strain live judge concrete grammar as follows:
Making kth sub-group at continually evolving i for interior, its population size is satisfied by: Mmin≤nk(t+i)≤Mmax, and
Average adaptive value fk a(t+i) the average adaptive value of all sub-than other population is little, then start inferior strain and do not live program, regenerate one
Sub-group etc. scale.Newly generated sub-group is retrained by above-mentioned chromosome coding mode, it is ensured that the individuality in sub-group with
The one_to_one corresponding of efficient solution.
Further, step 9) described in mutual exclusion of the same race judge concrete grammar as follows:
Pick out two sub-groups at random, judge its similarity degree according to certain rule, to meeting two of condition of similarity
Sub-group, removes one of them, produces the new explanation of same size;With a certain constant as base two words of definition equal length
The quantity according with the corresponding position of string different is broad sense Hamming distances between the two.When Swarm Evolution to t for time, randomly draw two sons
Colony p, q, if its population size meets: np(t)=nqT each individuality in (), and sub-group p can in sub-group q
Find the corresponding broad sense Hamming distances constant less than a certain setting, then start exclusive of the same race, reject and adapt to
It is worth less sub-group, and regenerates the sub-group of the scale such as.
Further, step 10) described in update evolve and convergence judge specifically comprise the following steps that
10-1) succession of the old by the new judges: judge whether Xie Qunzhong exists oneself through the sub-group stagnated of evolving, and carries out it newly
Old replacement, produces the new explanation of same size, but to retain the sub-group comprising optimum individual;
10-2) recalculate adaptive value: newborn sub-group is calculated adaptive value, and applies young weak safeguard measure;
10-3) sub-group is evolved: owing to the scale of sub-group is associated with its mean apparent level in colony, therefore sub
The scale of colony is continually changing.The breeding selecting sub-group is individual, utilizes intersection and mutation operator to produce the next generation and solves
Group;
10-4) convergence test, if meeting convergence conditions, then terminates evolutionary process, completes power distribution network reconfiguration;Otherwise
Return step 5).
Owing to have employed technique scheme, present invention have the advantage that:
The present invention have studied target method for normalizing and seeks efficient solution, using loss minimization as major heading, balancing the load conduct
Restriction condition treat sets up compromise model.Propose the mode of chromosome integer coding to guarantee between chromosome and feasible solution one
One is corresponding.By introducing genetic algorithm based on isolation niche, efficiently solve genetic algorithm premature problem Fast Convergent
To optimal solution.Meanwhile, the present invention is directed to binary coding and can not effectively reflect the architectural feature of Distribution Networks Reconfiguration problem, carry
Go out chromosome integer coding mode, highly shortened chromosome length, improve arithmetic speed.By introducing isolation your pupil
Border technology, efficiently solves traditional genetic algorithm at the Premature Convergence solved in Distribution Networks Reconfiguration problem.
Other advantages, target and the feature of the present invention will be illustrated to a certain extent in the following description, and
And to a certain extent, will be apparent to those skilled in the art based on to investigating hereafter, or can
To be instructed from the practice of the present invention.The target of the present invention and other advantages can be wanted by description below and right
Ask book to realize and obtain.
Accompanying drawing explanation
The accompanying drawing of the present invention is described as follows.
Fig. 1 is IEEE typical case's three feeder systems;
Fig. 2 is IEEE33 Node power distribution system;
Fig. 3 is the comparison of the node voltage before and after Distribution Networks Reconfiguration;
Fig. 4 is the schematic flow sheet of the present invention.
Detailed description of the invention
The invention will be further described with embodiment below in conjunction with the accompanying drawings.
A kind of reconstruction method of power distribution network based on genetic algorithm based on isolation niche, achieves power distribution network weight by following steps
Structure:
One, the Mathematical Modeling of Distribution Networks Reconfiguration
The optimization object function of Distribution Networks Reconfiguration has a variety of, and conventional target has: active loss is minimum, load balancing
Change and improve power supply quality, the stability improving system and reliability.
The Mathematical Modeling with balanced load as object function proposed by M.A.Kashem is:
In formula, BSYSBalancing the load pointer for system;NbFor system branch number summation;SiFor flowing through the complex power of branch road;
SimaxNominal transmission capacity for branch road i.
Branch road balancing the load pointerMathematical description be:
Say from mathematical meaning, balancing the load[12]Refer to the balancing the load pointer B of arbitrary branch road iLiIt is equal or approximate to
In system loading balanced pointers BSYS.I.e.Wherein ε is for be manually set according to distribution network concrete condition
Arbitrarily small number.
Take into account balancing the load and loss minimization turn to the Mathematical Modeling of object function and is:
In formula, f is the active loss of system, can try to achieve for trend method by front pushing back;kiRepresent the state of switch i,
Being 0-1 discrete magnitude, 0 represents disconnection, and 1 represents Guan Bi;riRepresent the resistance of branch road i;Pi、QiRepresent the active power flowing through branch road i
And reactive power;UiRepresent the voltage of branch road i endpoint node.
Constraints:
1) tributary capacity constraint Si≤Si,max(4)
2) node voltage constraint Ui,min≤Ui≤Ui,max(5)
3) account load balancing constraints
4) network topology constraint unpowered isolated island
In formula, Ui、Ui,min、Ui,maxIt is respectively voltage and the bound thereof of node i.
Two, power distribution network reconfiguration genetic algorithm based on isolation niche
1, the integer coding of chromosome
Coding is for particular problem, selects suitable encoding scheme, completes solution space to genetic algorithm solution space
Conversion.In order to preferably reflect the architectural feature of power distribution network network reconstruction, it is simple to the exploitation of genetic operator, use herein
Integer coding mode.
As a example by IEEE typical case's three feeder systems, as Fig. 1 shows, coding method is described.
(1) in order to improve efficiency of algorithm, meet power supply network power constraint under conditions of, all control variables must are fulfilled for
Three rules.Rule one: the branch switch on any loop must not close, and is not involved in chromosome coding;Rule two: with electricity
The switch that source point is connected also should close, and is also not involved in chromosome coding;Rule three: if common switch is opened twice and with
On, then this solution is infeasible solution, removes this chromosome.
(2) looped network that is made up of interconnection 5,11,16 of definition be respectively defined as 1,2,3 ring net.As a example by looped network 1,1.
2. switch is not involved in coding, and is the most again numbered the switch in looped network, respectivelyThe most 5. number open
Pass is respectively defined as the most 4. number switch, by that analogy.
(3) as a example by Fig. 1,3 genes of chromosome respectively 1,2,3 ring net is opened the sequence number of switch, its change
Scope is [1,4].So all of on off state with the rule coverage of integer coding, meets the unique of all chromosome and solution
, there is not infeasible solution in mapping relations.
2, the generation of initial individuals and the isolation of initial subgroup
Randomly generate N number of initial individuals;N number of initial individuals is all given K sub-group, the individuality that each sub-group contains
Number is N/K.The size of population size directly influences convergence or the computational efficiency of genetic algorithm, the too small easy convergence of scale
To locally optimal solution, scale is crossed conference and is caused calculating speed to reduce.Therefore quantity N of initial individuals and initial sub-population quantity N/
K should be different according to actual varying in size of distribution network scale, typically selected between 10~200.
3, the determination of individual fitness
Calculate all individual fitnesses in colony, and preserve the individuality that adaptive value is the highest.It is provided with t for kth sub-group
In jth individual, by calling trend forward-backward sweep method calculation procedure, the target function value i.e. line loss trying to achieve its correspondence is Fkj
(t).Use penalty function method to retrain the branch power of formula (4), the node voltage landing of formula (5) retrains and the load of formula (6) is put down
Weighing apparatus constraint processes, then this individual fitness is:
In formula: kU、kS、kBAnd φUi、φsi、φBiBe respectively voltage, power and account load balancing constraints penalty factor with penalize
Function.
For overcoming voltage, power and balancing the load rate dimension disunity and numerical value difference is bigger etc. that calculating is caused by problem
Impact, with formula (8), (9) and (10) calculate φUi、φsi, and φBi:
4, the determination of sub-group scale
The scale of sub-group is relevant with the average adaptive value of sub-group, and the average adaptive value of sub-group is the biggest, and it is at next
The individuality having in Dai is the most;Otherwise, the individuality having is the fewest.But number must is fulfilled for maximum allowable scale and Minimal Protective
The restriction of scale.
If t+1 is for scale n of kth sub-groupk(t+1) M is metmin≤nk(t+1)≤Mmax, wherein Mmin, MmaxRespectively
For Minimal Protective scale and maximum allowable scale, its value set the actual conditions according to network and the scale of initial subgroup
Depending on.It is too small that Minimal Protective scale sets, the easy Premature Convergence of evolution of sub-population;It is excessive that maximum allowable scale sets,
The evolution of sub-population is difficult to convergence, and it is more to expend resource, and cost is higher.Therefore, sub-group Minimal Protective scale MminAnd
Big permission scale MmaxValue should be close to initial sub-population scale[15]。
Being defined below of sub-group scale:
(1) average fitness assignment M of each sub-group is givenminIndividuality, remaining individuality is fitted according to the average of sub-group
Should be worth and utilize roulette method choice, until total Population reaches N.
(2) the average adaptive value of sub-group can be taken as such as following formula:
In formula: fk aT () is the t average adaptive value for kth sub-group;fki(t) for t in kth sub-group i-th
Individual individual fitness;nkT () is the t scale for kth sub-group.
(3) scale n in sub-group k t+1 generationk(t+1) it is:
5, protection releases and judges
In order to keep population diversity, it is desirable to have the sub-group that average adaptive value is relatively low is protected on consciousness ground, so it does not mistake
Early it is eliminated, and keeps certain evolvability, the colony of satisfied protection condition subsequent is removed protection;Therefore, in program
Introduce protection during design to judge and protect to release to judge.
Make t for scale n of kth sub-groupkT () meets nk(t)≤Mmin, then promoter colony defence program, force
Make nk(t)=Mmin.If evolve to t+1 for time this sub-group scale nk(t+1) meet: Mmin≤nk(t+1)≤Mmax, then solve
Divisor group body is protected.
6, inferior strain is not lived and is judged
To solving unprotected in group and that constant generations performance is worst colony, is rejected and the scale such as generation new the most sub
Colony.
Making kth sub-group at continually evolving i for interior, its population size is satisfied by: Mmin≤nk(t+i)≤Mmax, and
Average adaptive value fk a(t+i) the average adaptive value of all sub-than other population is little, then start inferior strain and do not live program, regenerate one
Sub-group etc. scale.Newly generated sub-group is retrained by above-mentioned chromosome coding mode, it is ensured that the individuality in sub-group with
The one_to_one corresponding of efficient solution.
7, mutual exclusion of the same race judges
Pick out two sub-groups at random, judge its similarity degree according to certain rule, to meeting two of condition of similarity
Sub-group, removes one of them, produces the new explanation of same size.
The different quantity in a certain constant as base two character string correspondence positions of definition equal length is between the two wide
Justice Hamming distances[14].When Swarm Evolution to t for time, randomly draw two sub-groups p, q, if its population size meet: np(t)
=nqEach individuality in (t), and sub-group p can find in sub-group q a corresponding broad sense hamming away from
From the constant less than a certain setting, then start exclusive of the same race, reject the sub-group that adaptive value is less, and regenerate one
Sub-group etc. scale.
8, evolution and convergence deterministic process are updated
(1) succession of the old by the new judges: judge whether Xie Qunzhong exists oneself through the sub-group stagnated of evolving, and it is carried out the old and new
Substitute, produce the new explanation of same size, but the sub-group comprising optimum individual to be retained.
(2) adaptive value is recalculated: newborn sub-group is calculated adaptive value, and applies young weak safeguard measure.
(3) sub-group is evolved: owing to the scale of sub-group is associated with its mean apparent level in colony, therefore subgroup
The scale of body is continually changing.The breeding selecting sub-group is individual, utilizes and intersects and mutation operator generation Xie Qun of future generation.
(4) convergence test, if meeting convergence conditions, then terminates evolutionary process;Otherwise return 3.
Embodiment:
IEEE33 Node power distribution system as in figure 2 it is shown, this system has 37 branch roads, 33 nodes, 5 interconnection switches:
TS7-20, TS8-14, TS11-21, TS17-32, TS24-28Rated voltage is 12.66kV.Total meritorious, the load or burden without work of system is respectively as follows:
3715kW and 2300kvar.
In genetic algorithm based on isolation niche, parameter chooses, according to distribution network actual conditions employing heuristic[8]Obtain;
Owing to from the point of view of the meaning and optimum results of operating parameter, they are the most independent each other, it is possible to first suppose other parameters
Immobilize, the optimal choice value of research single parameter, the most comprehensively.In example: chromosome coding a length of 5, initial population
Being 60, sub-group number is 5, and the maximum allowable scale of sub-group is 18, and Minimal Protective scale is 6, and crossing-over rate is 0.618, variation
Rate is 0.05.Using the program of Matlab7.0 establishment herein in research, program is run 50 times continuously, and 92% evolved to for 18 generations,
8% evolved to for 19 generations obtains optimum results in table 1.
Before and after table 1 reconstruct, active loss is compared
Fig. 3 is the comparison of the node voltage before and after Distribution Networks Reconfiguration.System node minimum voltage before reconstruct, per unit value is
0.9182, system node minimum voltage after reconstruct, per unit value is 0.9384.Other each node voltage amplitude has had to a certain degree
Raising, thus improve power supply quality.
Finally illustrating, above example is only in order to illustrate technical scheme and unrestricted, although with reference to relatively
The present invention has been described in detail by good embodiment, it will be understood by those within the art that, can be to the skill of the present invention
Art scheme is modified or equivalent, and without deviating from objective and the scope of the technical program, it all should be contained in the present invention
Right in the middle of.
Claims (10)
1. a reconstruction method of power distribution network based on genetic algorithm based on isolation niche, it is characterised in that specifically comprise the following steps that
1) according to power distribution network actual conditions, it is thus achieved that the parameter in genetic algorithm based on isolation niche;
2) Mathematical Modeling of power distribution network reconfiguration is built;
3) power distribution network is carried out chromosome integer coding;
4) the producing and the isolation of initial subgroup of initial individuals;
5) determination of individual fitness;
6) determination of sub-group scale;
7) protection releases and judges;
8) inferior strain is not lived and is judged;
9) mutual exclusion of the same race judges;
10) update evolution and convergence judges.
2. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 1, it is characterised in that step
2) Mathematical Modeling described in is to take into account balancing the load and loss minimization turns to the Mathematical Modeling of object function:
In formula, f is the active loss of system, can try to achieve for trend method by front pushing back;kiRepresent the state of switch i, be 0-1
Discrete magnitude, 0 represents disconnection, and 1 represents Guan Bi;riRepresent the resistance of branch road i;Pi、QiRepresent active power and the nothing flowing through branch road i
Merit power;UiRepresenting the voltage of branch road i endpoint node, constraints is:
2-1) tributary capacity constraint Si≤Si,max (4)
2-2) node voltage constraint Ui,min≤Ui≤Ui,max (5)
2-3) account load balancing constraints
2-4) network topology constraint unpowered isolated island;
In formula, Ui、Ui,min、Ui,maxIt is respectively voltage and the bound thereof of node i.
3. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 2, it is characterised in that step
3) described in, power distribution network is carried out all control variables in chromosome integer coding and need to meet following three times principles:
3-1) the branch switch on any loop must not close, and is not involved in chromosome coding;
3-2) switch being connected with power supply point also should close, and is also not involved in chromosome coding;
If 3-3) common switch be opened twice and more than, then this solution is infeasible solution, removes this chromosome.
4. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 3, it is characterised in that step
4) generation of initial individuals described in and the isolation of initial subgroup method particularly includes: randomly generate N number of initial
Body;N number of initial individuals is all given K sub-group, and the number of individuals that each sub-group contains is N/K.
5. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 4, it is characterised in that step
5) determination of individual fitness described in method particularly includes:
Calculate all individual fitnesses in colony, and preserve the individuality that adaptive value is the highest.It is provided with t in kth sub-group
Jth is individual, and by calling trend forward-backward sweep method calculation procedure, the target function value i.e. line loss trying to achieve its correspondence is Fkj(t)。
The balancing the load of the branch power of formula (4) is retrained by employing penalty function method, the node voltage of formula (5) lands constraint and formula (6) is about
Shu Jinhang process, then this individual fitness is:
In formula: kU、kS、kBAnd φUi、φsi、φBiBe respectively voltage, power and account load balancing constraints penalty factor with penalize letter
Number;
For overcoming voltage, power and balancing the load rate dimension disunity and the shadow that calculating causes by problem such as numerical value difference is bigger thereof
Ring, calculate φ by formula (8), (9) and (10)Ui、φsi, and φBi:
6. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 5, it is characterised in that step
6) concrete grammar of the determination of sub-group scale described in is as follows:
The scale of sub-group is relevant with the average adaptive value of sub-group, and the average adaptive value of sub-group is the biggest, and it is in the next generation
The individuality having is the most;Otherwise, the individuality having is the fewest.But number must is fulfilled for maximum allowable scale and Minimal Protective scale
Restriction;If t+1 is for scale n of kth sub-groupk(t+1) M is metmin≤nk(t+1)≤Mmax, wherein Mmin, MmaxRespectively
For Minimal Protective scale and maximum allowable scale, its value set the actual conditions according to network and the scale of initial subgroup
Depending on;
Being defined below of sub-group scale:
6-1) give average fitness assignment M of each sub-groupminIndividuality, remaining individuality is according to the average adaptation of sub-group
Value utilizes roulette method choice, until total Population reaches N;
6-2) the average adaptive value of sub-group can be taken as such as following formula:
In formula: fk aT () is the t average adaptive value for kth sub-group;fkiT () is individual for i-th in kth sub-group for t
Adaptive value;nkT () is the t scale for kth sub-group;
6-3) scale n in sub-group k t+1 generationk(t+1) it is:
7. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 6, it is characterised in that step
7) it is as follows that protection described in releases the concrete grammar judged:
Make t for scale n of kth sub-groupkT () meets nk(t)≤Mmin, then promoter colony defence program, force to make nk
(t)=Mmin.If evolve to t+1 for time this sub-group scale nk(t+1) meet: Mmin≤nk(t+1)≤Mmax, then son is released
Colony protects.
8. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 7, it is characterised in that step
8) concrete grammar judged alive of inferior strain described in is as follows:
Making kth sub-group at continually evolving i for interior, its population size is satisfied by: Mmin≤nk(t+i)≤Mmax, and the suitableeest
F should be worthk a(t+i) the average adaptive value of all sub-than other population is little, then start inferior strain and do not live program, regenerates the scale such as
Sub-group.Newly generated sub-group is retrained by above-mentioned chromosome coding mode, it is ensured that the individuality in sub-group and efficient solution
One_to_one corresponding.
9. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 8, it is characterised in that step
9) concrete grammar that mutual exclusion of the same race described in judges is as follows:
Pick out two sub-groups at random, judge its similarity degree, to two subgroups meeting condition of similarity according to certain rule
Body, removes one of them, produces the new explanation of same size;With a certain constant as base two character strings of definition equal length
The different quantity in corresponding position is broad sense Hamming distances between the two.When Swarm Evolution to t for time, randomly draw two sub-groups
P, q, if its population size meets: np(t)=nqT each individuality in (), and sub-group p can find in sub-group q
One corresponding broad sense Hamming distances less than the constant of a certain setting, then starts exclusive of the same race, rejects adaptive value relatively
Little sub-group, and regenerate the sub-group of the scale such as.
10. reconstruction method of power distribution network based on genetic algorithm based on isolation niche as claimed in claim 9, it is characterised in that step
Rapid 10) update described in and evolve and restrain specifically comprising the following steps that of judgement
10-1) succession of the old by the new judges: judge whether Xie Qunzhong exists oneself through the sub-group stagnated of evolving, and it is carried out the old and new more
Replace, produce the new explanation of same size, but the sub-group comprising optimum individual to be retained;
10-2) recalculate adaptive value: newborn sub-group is calculated adaptive value, and applies young weak safeguard measure;
10-3) sub-group is evolved: owing to the scale of sub-group is associated with its mean apparent level in colony, therefore sub-group
Scale be continually changing.The breeding selecting sub-group is individual, utilizes and intersects and mutation operator generation Xie Qun of future generation;
10-4) convergence test, if meeting convergence conditions, then terminates evolutionary process, completes power distribution network reconfiguration;Otherwise return
Step 5).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610255192.XA CN105896528A (en) | 2016-04-21 | 2016-04-21 | Power distribution network reconstruction method based on isolation ecological niche genetic algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610255192.XA CN105896528A (en) | 2016-04-21 | 2016-04-21 | Power distribution network reconstruction method based on isolation ecological niche genetic algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105896528A true CN105896528A (en) | 2016-08-24 |
Family
ID=56704408
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610255192.XA Pending CN105896528A (en) | 2016-04-21 | 2016-04-21 | Power distribution network reconstruction method based on isolation ecological niche genetic algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105896528A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106602557A (en) * | 2017-02-24 | 2017-04-26 | 三峡大学 | Multi-period optimization reconstruction method of active power distribution network comprising electric automobiles |
CN107064794A (en) * | 2016-12-16 | 2017-08-18 | 南阳师范学院 | A kind of fire-proof motor fault detection method based on genetic algorithm |
CN107590572A (en) * | 2017-10-24 | 2018-01-16 | 广东电网有限责任公司电力调度控制中心 | A kind of complicated Directional protection in loops MBPS acquiring methods based on IBBO |
CN109038575A (en) * | 2018-09-05 | 2018-12-18 | 东北大学 | Based on the reconstructing method containing distributed power distribution network for improving the raw algorithm that goes out of species |
CN109255142A (en) * | 2018-05-16 | 2019-01-22 | 浙江大学 | Tensegrity torus Topology Optimization Method based on niche genetic algorithm |
CN109672185A (en) * | 2019-01-14 | 2019-04-23 | 中国电力科学研究院有限公司 | A kind of distribution network voltage control method and system |
CN116187723A (en) * | 2023-04-26 | 2023-05-30 | 佰聆数据股份有限公司 | Resource scheduling method and device applied to distribution line loss reduction scene |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103903062A (en) * | 2014-03-12 | 2014-07-02 | 国家电网公司 | Method for reconstructing power distribution network based on double-ant-colony optimization |
JP2015061429A (en) * | 2013-09-19 | 2015-03-30 | 株式会社エヌエフ回路設計ブロック | Power storage system and control method therefor |
-
2016
- 2016-04-21 CN CN201610255192.XA patent/CN105896528A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015061429A (en) * | 2013-09-19 | 2015-03-30 | 株式会社エヌエフ回路設計ブロック | Power storage system and control method therefor |
CN103903062A (en) * | 2014-03-12 | 2014-07-02 | 国家电网公司 | Method for reconstructing power distribution network based on double-ant-colony optimization |
Non-Patent Citations (1)
Title |
---|
宫林等: "隔离小生境遗传算法在配电网络重构中的应用", 《电力系统及其自动化学报》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107064794A (en) * | 2016-12-16 | 2017-08-18 | 南阳师范学院 | A kind of fire-proof motor fault detection method based on genetic algorithm |
CN106602557A (en) * | 2017-02-24 | 2017-04-26 | 三峡大学 | Multi-period optimization reconstruction method of active power distribution network comprising electric automobiles |
CN106602557B (en) * | 2017-02-24 | 2019-09-24 | 三峡大学 | A kind of multi-period optimal reconfiguration method of active distribution network containing electric car |
CN107590572A (en) * | 2017-10-24 | 2018-01-16 | 广东电网有限责任公司电力调度控制中心 | A kind of complicated Directional protection in loops MBPS acquiring methods based on IBBO |
CN107590572B (en) * | 2017-10-24 | 2021-05-04 | 广东电网有限责任公司电力调度控制中心 | IBBO-based complex looped network direction protection MBPS (network Block switch) solving method |
CN109255142A (en) * | 2018-05-16 | 2019-01-22 | 浙江大学 | Tensegrity torus Topology Optimization Method based on niche genetic algorithm |
CN109255142B (en) * | 2018-05-16 | 2022-10-14 | 浙江大学 | Ring-shaped tension overall structure topology optimization method based on niche genetic algorithm |
CN109038575A (en) * | 2018-09-05 | 2018-12-18 | 东北大学 | Based on the reconstructing method containing distributed power distribution network for improving the raw algorithm that goes out of species |
CN109672185A (en) * | 2019-01-14 | 2019-04-23 | 中国电力科学研究院有限公司 | A kind of distribution network voltage control method and system |
CN116187723A (en) * | 2023-04-26 | 2023-05-30 | 佰聆数据股份有限公司 | Resource scheduling method and device applied to distribution line loss reduction scene |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105896528A (en) | Power distribution network reconstruction method based on isolation ecological niche genetic algorithm | |
Mahdavi et al. | A comprehensive review of metaheuristic methods for the reconfiguration of electric power distribution systems and comparison with a novel approach based on efficient genetic algorithm | |
Nguyen et al. | Multi-objective electric distribution network reconfiguration solution using runner-root algorithm | |
CN108734202A (en) | A kind of Fault Diagnosis for HV Circuit Breakers method based on improved BP | |
CN104362623B (en) | Multi-target network reestablishing method for active power distribution network | |
Chandrasekaran et al. | Network and reliability constrained unit commitment problem using binary real coded firefly algorithm | |
Yang et al. | A hybrid evolutionary algorithm by combination of PSO and GA for unconstrained and constrained optimization problems | |
Abdelaziz | Genetic algorithm-based power transmission expansion planning | |
Singh et al. | Performance of a hybrid EA-DE-memetic algorithm on CEC 2011 real world optimization problems | |
CN105098762A (en) | Island dividing method for power distribution network having distributed power sources | |
Rani et al. | Self adaptive harmony search algorithm for optimal capacitor placement on radial distribution systems | |
CN106920015A (en) | Suitable for the most short loop method for dynamically partitioning of power distribution network reconfiguration representation | |
CN109217284A (en) | A kind of reconstruction method of power distribution network based on immune binary particle swarm algorithm | |
CN105184415B (en) | A kind of Distribution Networks Reconfiguration design method | |
CN106777449A (en) | Distribution Network Reconfiguration based on binary particle swarm algorithm | |
CN105323166A (en) | Cloud computing-oriented routing method based on network efficiency priority | |
Ying et al. | An adaptive stochastic ranking mechanism in MOEA/D for constrained multi-objective optimization | |
CN112994099B (en) | High-proportion distributed photovoltaic grid-connected digestion capacity analysis method | |
CN107706907A (en) | A kind of Distribution Network Reconfiguration and device | |
Chandrasekaran et al. | Multi-objective unit commitment problem using Cuckoo search Lagrangian method | |
Dahalan et al. | Optimum DNR and DG sizing for power loss reduction using improved meta-heuristic methods | |
Li et al. | Distribution network reconfiguration based on niche binary particle swarm optimization algorithm | |
Torres-Jimenez et al. | Reconfiguration of power distribution systems using genetic algorithms and spanning trees | |
Rathore et al. | Impact of distributed generation in transmission network expansion planning problem | |
Choi et al. | An application of genetic algorithms to the network reconfiguration in distribution for loss minimization and load balancing problem. II |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160824 |
|
RJ01 | Rejection of invention patent application after publication |