CN105552892A - Distribution network reconfiguration method - Google Patents

Abstract

The invention discloses a distribution network reconfiguration method, which comprises the following steps: building a distribution network load optimization model; and optimizing the distribution network load optimization model on the basis of a genetic algorithm, and obtaining a distribution network reconfiguration scheme. The reconfiguration is carried out on a distribution network on the basis of a genetic method; and load reconfiguration is carried out by changing the state of a switch, so that line loss of the distribution network is reduced; the benefits are improved; and the power supply reliability and safety are improved.

Description

A distribution network reconfiguration method

technical field

The invention relates to the field of distribution networks, in particular to a distribution network reconfiguration method.

Background technique

The line loss of the distribution network, especially the loss of the 10kV low-voltage side line, is one of the links with more losses in the distribution network, and it is one of the key points of the line loss management work. In order to ensure the reliability and safety of the power supply network, the wiring and operation methods of the distribution network on the medium and low voltage side have become more and more complex, and gradually tend to be diversified. Multiple lines are connected to each other, and the power supply mode of the ring network is implemented. At present, except for a few simple heuristic algorithms based on physical models, most of the power distribution network reconfiguration uses intelligent algorithms to solve this problem. The power distribution network has the mandatory requirement of radial open-loop operation. The existing methods generally have disadvantages in the coding design that cannot guarantee the feasibility of the network. In the calculation process, a large number of infeasible networks with rings or isolated islands will inevitably be generated. In the calculation process, it is necessary to frequently check the network and repeatedly operate and repair the invalid network in order to obtain a valid network. At the same time, the operators constructed by the existing encoding methods are not closely related to the changes of the actual distribution network structure. The above-mentioned shortcomings lead to the problems of many iterations and long time consumption in the existing intelligent algorithms, which are difficult to be applied to solve the reconstruction of large-scale power distribution networks.

Contents of the invention

The invention provides a distribution network reconfiguration method, which reconfigures the distribution network based on a genetic method, balances the load rate of a heavy-duty transformer, reduces network loss, and improves power supply reliability.

In order to achieve the above object, the present invention provides a distribution network reconfiguration method, which is characterized in that the method includes:

Establish distribution network load optimization model;

The distribution network load optimization model is optimized based on the genetic algorithm, and the distribution network reconfiguration scheme is obtained.

The distribution network load optimization model established above includes:

Establish distribution network load topology;

Create an objective function;

Identify constraints.

The above methods for establishing the objective function include:

Take the minimum network loss as the objective function;

The calculation formula PLoss of distribution network loss is shown in formula (1);

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; k</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&Center\; Dot;</span>}{\mathrm{<span\; class="notranslate">\; I</span>}}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

In formula (1): nb is the number of branches of the distribution network; ri is the resistance of the i-th branch; is the current flowing through the i-th branch; ki is the state of the switch (node) i, ki=0 means breaking, ki=1 means closing;

The distribution network loss PLoss is obtained through power flow calculation.

The above constraints include: branch capacity constraints and node voltage constraints;

The above-mentioned branch capacity constraint is that the power flowing through the branch cannot be greater than the maximum power of the branch, as shown in formula (2):

P _ij ≤ P _ijmax

Q _ij ≤ Q _ijmax (2)

In formula (2), P _ij and P _ijmax are respectively the active power of branch ij and the maximum active power of branch ij; Q _ij and Q _ijmax are respectively the reactive power of branch ij and the maximum reactive power of branch ij value;

The above node voltage constraint is that the voltage of the node cannot exceed the upper and lower limits of the voltage allowed by the node, as shown in formula (3):

U _imin ≤ U _i ≤ U _imax (3)

In formula (3), U _i , U _imin , and U _imax are the voltage at node i, the minimum voltage at node i, and the maximum voltage at node i, respectively.

The method for optimizing the distribution network load optimization model based on the genetic algorithm includes:

The initial network coding of the distribution network generates chromosomes;

Decode the chromosome to obtain the fitness of the chromosome;

According to the fitness, select the parent individual for breeding offspring;

The parent individual performs crossover operation and mutation operation successively;

Repeat iterations to output the chromosome with the highest fitness and the corresponding reconstruction scheme.

The initial network coding generation chromosome of the above distribution network contains:

The distribution network adopts the breadth-first spanning tree algorithm to generate an initial network;

The initial network encodes a chromosome;

The chromosome generates N chromosomes through the mutation operator, and N chromosomes constitute the initial population of the genetic algorithm, and N is an even number;

Among them, each chromosome corresponds to a radial distribution network that can realize operation.

The above-mentioned decoded chromosome to obtain the fitness of the chromosome includes:

All chromosomes are decoded according to the reverse process of encoding;

All chromosomes obtain the network loss of the initial network through power flow calculation; the reciprocal of the network loss obtained by each chromosome is taken as the fitness of the chromosome;

Among them, if the node voltage and line power of the initial network exceed the limit, the fitness of the chromosome is taken as zero.

The above-mentioned paternal individuals selected for breeding offspring according to fitness include:

Chromosomes whose fitness is higher than the preset threshold are directly copied to offspring in the way of elite retention, and the number of elite retention is M, and M is an even number;

Chromosomes whose fitness is lower than the preset threshold select (N-M)/2 paternal individuals for reproduction in a semi-random and semi-optimal method.

The above-mentioned parent individuals successively perform crossover operation and mutation operation including:

The even-numbered parent individuals perform the crossover operation pair by pair according to the crossover operator;

After the crossover operation, the individuals obtained by the crossover are mutated according to the mutation operator, and the reproduction is completed to obtain the offspring chromosome group;

Among them, the crossover rate is set to 0.5; the mutation rate is set to 1.

The above repeated iterations output the chromosome with the highest fitness and the corresponding reconstruction scheme includes:

Repeatedly decode the chromosome to obtain the fitness of the chromosome, select the paternal individual for breeding offspring according to the fitness, and the paternal individual successively performs crossover and mutation operations to iterate;

Determine whether the chromosome with the highest fitness keeps its offspring unchanged, and if so, stop iterating and output the chromosome with the highest fitness and its corresponding reconstruction plan; otherwise, continue to iterate.

Compared with the prior art, the distribution network reconfiguration method of the present invention has the advantage that the distribution network is reconfigured based on the genetic method, and the load reconfiguration is performed by changing the state of the switch to reduce the line loss of the distribution network , Improve efficiency, improve power supply reliability and security.

Description of drawings

Fig. 1 is the method flowchart of distribution network reconfiguration method of the present invention;

Figure 2 is a topological diagram of IEEE standard 3-feeder 16-node power distribution system;

Fig. 3 is a flow chart of optimizing the distribution network load optimization model based on genetic algorithm.

detailed description

Specific embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

As shown in Figure 1, the present invention discloses a distribution network reconfiguration method based on a genetic method, which specifically includes the following steps:

S1. Establish a distribution network load optimization model.

S1.1. Establish the distribution network load topology.

Firstly, data collection of distribution network: determine the object of analysis, the number of nodes, the number of lines, the number of switches, and the parameters of lines and nodes, including line resistance, node active load and reactive load.

Then analyze and draw the topological structure diagram of the system, determine the position of the switch, each node and the state of the switch, as well as the line and node parameters.

As shown in Figure 2, it is a topological diagram of an IEEE standard 3-feeder 16-node power distribution system. Among them, serial numbers (1)-(16) represent loads, and serial numbers 1-16 represent switches. The switch state before reconstruction is: switches 4, 11 and 13 are off, and the rest of the switches are closed.

S1.2, establishing an objective function, which includes:

S1.2.1. Take the minimum network loss as the objective function.

S1.2.2. The calculation formula PLoss of distribution network loss is shown in formula (1);

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; k</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&CenterDot;</span>}{\mathrm{<span\; class="notranslate">\; I</span>}}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

In formula (1): nb is the number of branches of the distribution network; ri is the resistance of the i-th branch; is the current flowing through the i-th branch; ki is the state of the switch (node) i, ki=0 means breaking, ki=1 means closing;

S1.2.3. Obtain distribution network loss PLoss through power flow calculation.

S1.3. Determine the constraints. Constraints include: branch capacity constraints and node voltage constraints.

The branch capacity constraint is that the power flowing through the branch cannot be greater than the maximum power of the branch, as shown in formula (2):

P _ij ≤ P _ijmax

Q _ij ≤ Q _ijmax (2)

In formula (2), P _ij and P _ijmax are respectively the active power of branch ij and the maximum active power of branch ij; Q _ij and Q _ijmax are respectively the reactive power of branch ij and the maximum reactive power of branch ij value;

The node voltage constraint is that the voltage of the node cannot exceed the upper and lower limits of the voltage allowed by the node, as shown in formula (3):

U _imin ≤ U _i ≤ U _imax (3)

In formula (3), U _i , U _imin , and U _imax are the voltage at node i, the minimum voltage at node i, and the maximum voltage at node i, respectively.

S2. As shown in FIG. 3 , the distribution network load optimization model is optimized based on the genetic algorithm, and a distribution network reconfiguration scheme is obtained.

S2.1. The initial network coding of the distribution network generates chromosomes.

S2.1.1. Close all the switches of the distribution network to be reconfigured, and use the breadth-first spanning tree algorithm to generate an initial network for the distribution network.

S2.1.2. The initial network is encoded to generate a chromosome.

S2.1.3. The chromosome is used as a blueprint to generate N chromosomes through mutation operators, and N chromosomes constitute the initial population of the genetic algorithm, and N is an even number; each chromosome corresponds to a radial distribution network that can realize operation.

S2.2. Decoding the chromosome to obtain the fitness of the chromosome.

S2.2.1. All chromosomes are decoded according to the reverse process of encoding;

S2.2.2. All chromosomes obtain the network loss of the initial network through power flow calculation; the reciprocal of the network loss obtained by each chromosome is taken as the fitness of the chromosome. Here, the node voltage and line power are checked at the same time. If the node voltage and line power of the initial network exceed the limit, the fitness of the chromosome is taken as zero.

S2.3. Select paternal individuals for breeding offspring according to fitness.

S2.3.1. First, chromosomes whose fitness is higher than the preset threshold are directly copied to offspring in the form of elite retention without cross-mutation, which is the widely used elite retention strategy. The number of elite retention is M, and M is even.

S2.3.2. For chromosomes whose fitness is lower than the preset threshold, a semi-random and semi-optimal method is used to select (N-M)/2 paternal individuals for breeding offspring.

S2.4. The father individual performs crossover operation and mutation operation successively.

S2.4.1. Perform the crossover operation on the (N-M)/2 father individuals selected in S2.3.2 pair by pair with a certain probability according to the crossover operator.

S2.4.2. After the crossover operation, mutate the individuals obtained by the crossover according to a certain probability according to the mutation operator, and complete the reproduction to obtain the offspring chromosome group. Among them, the probability of crossover and mutation satisfies the principle of mutation as the main and crossover as supplementary: the crossover rate is set to 0.5; the mutation rate is set to 1.

S2.5. Iteratively output the chromosome with the highest fitness and the corresponding reconstruction scheme.

Repeatedly decode the chromosome to obtain the fitness of the chromosome, select the paternal individual for breeding offspring according to the fitness, and the paternal individual performs crossover operation and mutation operation successively, and iterates. The convergence condition is to keep the chromosome with the highest fitness in the population unchanged for several generations.

In the iterative process, it is judged whether the chromosome with the highest fitness keeps the offspring unchanged for several generations. If so, the convergence condition is satisfied, and the iterative output of the chromosome with the highest fitness and its corresponding reconstruction scheme and network loss is stopped. The result It is optimal; if not, jump to S2.3 and continue to iterate.

S2.6. After the above calculation is completed, the optimal solution is selected according to Table 1 according to the calculation results, and the status of each switch is shown in Table 2.

Table 1. Network loss values under different switch states

switch number switch status switch number switch status switch number switch status 1 C 7 C 13 C 2 C 8 C 14 C 3 C 9 o 15 C 4 C 10 C 16 C 5 o 11 C 6 o 12 C

Table 2. The switch state of the optimal network after reconstruction

In Table 2, C means that the switch is closed, and O means that the switch is turned off.

Although the content of the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description should not be considered as limiting the present invention. Various modifications and alterations to the present invention will become apparent to those skilled in the art upon reading the foregoing disclosure. Therefore, the protection scope of the present invention should be defined by the appended claims.

Claims (10)

1. a reconstruction method of power distribution network, is characterized in that, the method comprises:

Set up distribution network load Optimized model;

Based on genetic algorithm, distribution network load Optimized model is optimized, obtains power distribution network reconfiguration scheme.

2. reconstruction method of power distribution network as claimed in claim 1, it is characterized in that, described distribution network load Optimized model of setting up comprises:

Set up distribution network load topological structure;

Set up target function;

Determine constraints.

3. reconstruction method of power distribution network as claimed in claim 2, it is characterized in that, the described method setting up target function comprises:

Take loss minimization as target function;

The computing formula P of distribution network loss _lossshown in (1);

$P_{Loss}=\underset{i=1}{\overset{n_b}{\&Sigma;}}{k}_i{r}_i|{\stackrel{\&CenterDot;}{I}}_i{|}^2---\left(1\right)$

In formula (1): n _bfor the circuitry number of power distribution network; r _ibe the resistance of the i-th branch road; for flowing through the electric current of i-th branch road; k _ifor the state of switch (node) i, ki=0 represents disjunction, and ki=1 represents closed;

Distribution network loss P is obtained by Load flow calculation _loss.

4. reconstruction method of power distribution network as claimed in claim 2, it is characterized in that, described constraints comprises: tributary capacity constraint and node voltage constraint;

Described tributary capacity is constrained to the maximum power that the power flowing through branch road can not be greater than this branch road, shown in (2):

P _ij≤P _ijmax

Q _ij≤Q _ijmax(2)

In formula (2), P _ij, P _ijmaxbe respectively the active power of branch road i-j and the meritorious maximum of branch road i-j; Q _ij, Q _ijmaxbe respectively the reactive power of branch road i-j and the idle maximum of branch road i-j;

The voltage that described node voltage is constrained to node can not exceed the voltage bound that this node allows to pass through, shown in (3):

U _imin≤U _i≤U _imax(3)

In formula (3), U _i, U _imin, U _imaxbe respectively the voltage max of the voltage of node i, the voltage minimum of node i and node i.

5. reconstruction method of power distribution network as claimed in claim 1, is characterized in that, describedly comprises the method that distribution network load Optimized model is optimized based on genetic algorithm:

The initial network coding of power distribution network generates chromosome;

Decoding chromosome obtains chromosomal fitness;

Select the male parent for breeding filial generation individual according to fitness;

Male parent is individual successively carries out interlace operation and mutation operation;

Iteration exports the highest chromosome of fitness and corresponding reconfiguration scheme.

6. reconstruction method of power distribution network as claimed in claim 5, is characterized in that, the initial network coding of described power distribution network generates chromosome and comprises:

Distribution network adopts breadth First spanning tree algorithm to produce an initial network;

This initial network carries out coding generation chromosome;

This chromosome generates N number of chromosome by mutation operator, and N number of chromosome forms the initial population of genetic algorithm, and N is even number;

Wherein, each chromosome correspondence one can realize the radial distribution networks of operation.

7. reconstruction method of power distribution network as claimed in claim 5, it is characterized in that, described decoding chromosome obtains chromosomal fitness and comprises:

All chromosome is decoded according to the inverse process of coding;

All chromosome obtains the network loss of initial network through Load flow calculation; Each chromosome obtain network loss inverse be taken as this chromosomal fitness;

Wherein, if the node voltage of initial network and line power out-of-limit, then chromosomal fitness is taken as zero.

8. reconstruction method of power distribution network as claimed in claim 6, is characterized in that, describedly selects to comprise for the male parent individuality of breeding filial generation according to fitness:

The mode that fitness retains with elite higher than the chromosome of predetermined threshold value directly copies to filial generation, and the number that elite retains is M, M is even number;

Fitness is individual for the male parent of breeding filial generation with half random half optimum method choice (N-M)/2 lower than the chromosome of predetermined threshold value.

9. the reconstruction method of power distribution network as described in claim 5 or 8, is characterized in that, the individual priority of described male parent carries out interlace operation and mutation operation comprises:

Even number male parent individuality carries out interlace operation by by crossover operator;

After interlace operation, the individuality that intersection obtains is made a variation by mutation operator, completes breeding and obtain child chromosome group;

Wherein crossing-over rate is set to 0.5; Aberration rate is set to 1.

10. reconstruction method of power distribution network as claimed in claim 5, is characterized in that, described iteration exports the highest chromosome of fitness and corresponding reconfiguration scheme comprises:

Repeat decoding chromosome obtaining chromosomal fitness, selecting the male parent for breeding filial generation individual according to fitness, and the individual priority of male parent carries out interlace operation and mutation operation, carries out iteration;

Judge whether the chromosome that fitness is the highest continues to keep filial generation constant, if then stop iteration exporting the reconfiguration scheme of the highest chromosome of fitness and correspondence thereof; Then proceed iteration if not.