CN103136585B - Weighting Voronoi diagram substation planning method based on chaotic and genetic strategy - Google Patents

Abstract

The invention relates to the field of electric systems and discloses a weighting Voronoi diagram substation planning method based on a chaotic and genetic strategy. The method aims at solving the problems that a prior algorithm is low in rate of convergence, poor in capacity of local optimization and sensitive in initial value, premature convergence exists, the unreasonable phenomenon caused by division of power-supply districts according to the principle of proximity exists, the load rate of a planned substation can not be controlled, and the like, and optimizing site selection of the substation and division of the power-supply districts by means of certain algorithms. The method comprises the steps of setting parameters; chaotic initialization and generating initial population including N individuals; carrying out the site selection of the substation and load distribution on the N individuals; judging whether end criterion is satisfied; calculating the fitness variance sigma 2 of the population; chaotic search; and executing and saving an optimized genetic algorithm and then returning to the fourth step. The weighting Voronoi diagram substation planning method based on the chaotic and genetic strategy is mainly applicable to the electric systems.

Description

Based on the weighted Voronoi diagrams figure substation planning method of Chaos-Genetic strategy

Technical field

The present invention relates to field of power, particularly relate to the weighted Voronoi diagrams figure substation planning method based on Chaos-Genetic strategy.

Background technology

In urban distribution network development plan, the load level that the position of High Voltage Distribution Substations, capacity and service area planning are related to the space layout of following area to be planned electrical network, grid structure and can provide, when China's land resource is more and more becoming tight, it is the major issue being related to area to be planned Sustainable Development in Future that scientific and reasonable transformer station's Planning of spatial arrangement and corresponding soil, site are reserved, and tool is of great significance.

Substation planning problem is a multivariate, multiobject large-scale nonlinear optimization problem, and existing existing method is divided into 3 classes such as Mathematics Optimization Method, Heuristic Method and modern intelligent optimization method substantially.In recent years, Intelligent Optimization Technique obtained application gradually in this problem, as TS algorithm, simulated annealing, genetic algorithm, particle cluster algorithm etc.The deficiencies such as these methods not only exist that speed of convergence is slow to some extent, local optimal searching ability, computing time are long, but also there is service area and divide only to adopt and distribute nearby and the problem such as transformer station's load factor that is unreasonable and planning of causing is uncontrollable.

In recent years, computational geometry Voronoi figure is introduced into electric system and solves substation planning problem, its in solution load density uneven situation in service area classifying rationally effect fine, but single Voronoi drawing method has certain susceptibility to initial value, the weight of weighted Voronoi diagrams is chosen and is also had much room for improvement, and these class methods only can obtain local extremum solution.And based on the substation site selection of genetic algorithm, make full use of the robustness of genetic algorithm, by a series of genetic manipulations such as selection, intersection, variations, evolve colony gradually, make colony close to or comprise the state of optimum solution, but there is precocious phenomenon.

Summary of the invention

The present invention is intended to overcome the deficiencies in the prior art, overcome that previous algorithm the convergence speed is slow, local optimal searching ability, initial value responsive and precocious phenomenon and service area divide to adopt and distribute and the problem such as transformer station's load factor that is unreasonable and planning of causing is uncontrollable nearby, by certain algorithm, solve the optimization of substation site selection and division of the power supply area, for achieving the above object, the technical scheme that the present invention takes is, based on the weighted Voronoi diagrams figure substation planning method of Chaos-Genetic strategy, comprise the steps:

Step 1: setup parameter, comprises setting:

Maxeranum: maximum iteration time; Popsize: population scale; N: number of individuals in initial population; PCross crossover probability; PMutation mutation probability; PInversion inversion probability;

Step 2: chaos intialization, produces the initial population comprising individuality, and using the coding of each individuality as substation location;

Step 3: realize weighted Voronoi diagrams figure substation site selection based on Automatic adjusument weight and load distribution thereof to individual, and calculate the adaptive value f of each individuality, the good Popsize of a selectivity solution is as initial solution from N number of initial population;

Step 4: judge whether to meet termination criterion, if meet, algorithm stops, otherwise performs step 5;

Step 5: the adaptive value variances sigma calculating population ²if, σ ²< ε, ε are adaptive value variance threshold values given in advance, occur precocious, carry out Chaos Search; Otherwise, jump to step 7;

Step 6: Chaos Search;

Step 7: perform and preserve optimum genetic algorithm, return step 4 after terminating.

Chaos intialization is further specially:

2.1 produce n-dimensional vector z at random ₀, its each component values is between-1 ~ 1

Z ₀=(z ₀₁, z ₀₂..., z _0n), n is transformer station's number, z ₀₁, z ₀₂..., z _0nfor the z of stochastic generation ₀each component of vector;

2.2 get chaos sequence: according to cube mapping function obtain z ₁, z ₂, z ₃..., z _n, obtain chaos sequence thus

z＝(z ₁,z ₂,z ₃,…,z _N)

Z ₁, z ₂, z ₃..., z _nfor each component of the z phasor that the vector of initial random generation is obtained by cube mapping function;

2.3 according to

$x_{\mathrm{ij}}=\frac{a_j+b_j}{2}+\frac{(b_j-a_j)}{2}{z}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,N;j=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,n$

By chaos sequence z _ia jth component carrier to the span (a of optimized variable _j, b _j) in, thereby is achieved N number of initial position in n transformer station x direction, a _j, b _jbe respectively lower limit and the higher limit in x direction, transformer station's actual power region, z _ijfor each component z of z vector ₁, z ₂, z ₃..., z _ncomponent; Obtain N number of initial position in n transformer station y direction in the same way;

2.4 initial population being obtained individuality respectively by N number of initial position of x, y according to individual UVR exposure principle.

The determination of initial weight and the Automatic adjusument of weight are specially:

The determination of 3.1 initial weights: according to

Determine initial weight, wherein: l (i) is i-th substation's on-load; E (i) is the rated capacity of i-th transformer station; γ is that transformer station is meeting the most high capacity rate under main transformer " N-1 " principle; for transformer station's power factor;

The Automatic adjusument of 3.2 each transformer station weights.

Chaos Search is specially:

6.1 produce a n-dimensional vector cr at random ₀, by cr _i+1=4cr _i ³-3cr _iproduce cr=(cr ₁, cr ₂, cr ₃..., cr _i...), i=1,2 ..., Popsize, according to Δ x _ij=m+ (n-m) × cr _ijby cr _ieach component carrier in chaotic disturbance scope [m, n], obtain disturbance quantity Δ x=(Δ x ₁, Δ x ₂..., Δ x _i...), cr ₀for the vector produced at random, cr ₁, cr ₂, cr ₃..., cr _i... for each component of cr vector;

Chaotic disturbance amount component is loaded on each individuality by 6.2, forms the position, chaotic disturbance x direction of each individuality

x′＝(x′ ₁,x′ ₂,…,x′ _i,…)，i＝1,2,…,Popsize

Wherein

$\left\{\begin{array}{c}{x}_i^{\&prime;}=(x_{i1}^{\&prime;},x_{i2}^{\&prime;},\&CenterDot;\&CenterDot;\&CenterDot;,x_{\mathrm{in}}^{\&prime;})\\ x_{\mathrm{ij}}^{\&prime;}=x_{\mathrm{ij}}+{\mathrm{\&Delta;x}}_{\mathrm{ij}}\end{array}\right.$

X ' for chaos sequence add disturbance quantity after obtain, x ' ₁, x ' ₂..., x ' _i... for its each component, x ' _i1, x ' _i2..., x ' _infor x ' _icomponent; The position, chaotic disturbance y direction of each individuality is formed according to 6.1,6.2;

Substation site selection based on the weighted Voronoi diagrams figure transformer station addressing algorithm of Automatic adjusument weight and load distribution thereof are carried out in the 6.3 pairs of chaotic disturbance positions, calculate the adaptive value f ' of each individuality;

Adaptive value before and after 6.4 more each individual chaotic disturbances, if the feasible solution x ' that Chaos Search arrives _ibe better than the feasible solution x of corresponding individuality _i, then x ' is used _isubstitute x _i, perform step 7.

Perform the optimum genetic algorithm of preservation to be specially:

7.1 individual UVR exposure

Individual UVR exposure adopts floating-point encoding mode, transformer station's number that individual code length equals 2 times, and individual each gene is substation location in planning region, i.e. the x direction coordinate of locus or y direction coordinate, and site is floating number;

7.2 fitness function

Transformer station and network investment and annual running cost is selected to be used as fitness function to evaluate each individual corresponding programme;

7.3.1 operation is selected

Select operation adopt optimum maintaining strategy and determine that formula system of selection of sampling combines, determine that formula is sampled system of selection:

(1). calculate the expectation existence number N of each individuality in colony of future generation in colony _i, i=1,2 ..., M;

(2). use N _iintegral part determine each corresponding individual existence number in colony of future generation;

(3). according to N _ifraction part descending sort is carried out to individuality, order get before individuality joins in colony of future generation;

7.4 crossover and mutation operations

The employing intersection that counts carries out interlace operation, replaces protogene value complete mutation operation by a random number in genes of individuals value span;

7.5 inversion operations

Put upside down the sequence in the gene order in individual UVR exposure string between random two locus formulated, thus the better individuality that formation one is new;

7.6 transformer station's load areas divide

If d transformer station collection S={s in planning region ₁, s ₂..., s _d, then arbitrary load x belongs to transformer station s _ibe defined as:

R(s _i,x)＝{d(x,s _i)≤d(x,s _j),j＝1,2,…,d,j≠i}

D (x, s in formula _i), d (x, s _j) represent arbitrary load x and s in plane respectively _iand s _jbetween Euclidean distance;

7.7 preserve optimum genetic algorithm.

Technical characterstic of the present invention and effect:

1. calculate weight according to substation capacity and power load distributing thereof, on the basis of improving weight calculation, add load factor adaptive control flow process, forming weight can the substation planning method of weighted Voronoi diagrams figure of self-adaptative adjustment.

2. weighted Voronoi diagrams figure method and genetic algorithm are combined, achieve the weighted Voronoi diagrams figure substation site selection preserving optimum genetic algorithm.

3. introduce chaos thought, solve the premature convergence problem of genetic algorithm, make program results more be tending towards excellent, realize the weighted Voronoi diagrams figure transformer station site selecting method based on Chaos Genetic Algorithm.

By the weighted Voronoi diagrams figure substation planning method based on Chaos-Genetic strategy, be applied in different examples, result shows that no matter the method is from investment cost, transformer station's load distribution region or transformer station's actual loading rate aspect are all better than simple genetic algorithm and simple weighted Voronoi diagrams figure transformer station site selecting method, and the two weighted Voronoi diagrams drawing method based on genetic algorithm combined, this fully shows the validity of the method to load density uneven situation power supply area classifying rationally, and no matter in the determination of substation location, or it is all reliable than single method in the division of transformer station's service area, rationally.

Accompanying drawing explanation

Fig. 1 weight can the weighted Voronoi diagrams figure substation planning method of self-adaptative adjustment;

Expense: 11041.43 ten thousand yuan;

1 ~ No. 9 transformer station's load factor: 0.7800.6500.6130.6070.5740.5610.5420.5310.822.

Fig. 2 is based on the weighted Voronoi diagrams figure transformer station site selecting method of Chaos Genetic Algorithm;

Expense: 11017.79 ten thousand yuan;

1 ~ No. 9 transformer station's load factor: 0.6960.6460.6210.6200.5590.5820.5800.5640.767.

Fig. 3 weight can the weighted Voronoi diagrams figure substation planning method of self-adaptative adjustment;

Expense: 4736.98 ten thousand yuan

1 ~ No. 7 transformer station's load factor: 0.7740.6900.6060.6020.5300.5100.451.

Fig. 4 is based on the weighted Voronoi diagrams figure transformer station site selecting method of Chaos Genetic Algorithm;

Expense: 4736.02 ten thousand yuan;

1 ~ No. 7 transformer station's load factor: 0.6710.6330.6330.6230.600.5990.536.

Fig. 5 preserves optimum genetic algorithm process flow diagram.

Embodiment

The present invention:

1. calculate weight according to substation capacity and power load distributing thereof, on the basis of improving weight calculation, add load factor adaptive control flow process, forming weight can the substation planning method of weighted Voronoi diagrams figure of self-adaptative adjustment.

2. weighted Voronoi diagrams figure method and genetic algorithm are combined, achieve the weighted Voronoi diagrams figure substation site selection preserving optimum genetic algorithm.

3. introduce chaos thought, solve genetic algorithm ground premature convergence problem, make program results more be tending towards excellent, realize the weighted Voronoi diagrams figure transformer station site selecting method based on Chaos Genetic Algorithm.

The present invention is further illustrated below in conjunction with the drawings and specific embodiments.

1. the most high capacity rate that meets under " N-1 " principle of transformer station is as shown in table 1.

Most high capacity rate under table 1 transformer station " N-1 " principle

Transformer station is containing main transformer number	Load factor
		Two main transformers	0.65
Three main transformers	0.87

When the actual loading rate of transformer station is higher than most high capacity rate under " N-1 " principle, program results does not meet the demands.

2. example 1 is the transformer station of 83 × 50MVA, 13 × 40MVA.Fig. 1 and Fig. 2 represents that utilization can the substation planning method of weighted Voronoi diagrams figure of self-adaptative adjustment and the program results of the weighted Voronoi diagrams figure transformer station site selecting method based on Chaos Genetic Algorithm based on weight respectively.

Step 1: setup parameter

Maxeranum: maximum iteration time; Popsize: population scale number; N: number of individuals in initial population; PCross crossover probability; PMutation mutation probability; PInversion inversion probability.

Step 2: chaos intialization, produces the initial population comprising individuality.

2.1 produce n-dimensional vector z at random ₀, its each component values is between-1 ~ 1

Z ₀=(z ₀₁, z ₀₂..., z _0n), n is transformer station's number, z ₀₁, z ₀₂..., z _0nfor the z of stochastic generation ₀each component of vector;

2.2 get chaos sequence: according to cube mapping function obtain z ₁, z ₂, z ₃..., z _n, obtain chaos sequence thus

z＝(z ₁,z ₂,z ₃,…,z _N)

Z ₁, z ₂, z ₃..., z _nfor each component of the z phasor that the vector of initial random generation is obtained by cube mapping function;

2.3 according to

$x_{\mathrm{ij}}=\frac{a_j+b_j}{2}+\frac{(b_j-a_j)}{2}{z}_{\mathrm{ij}},i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,N;j=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,n$

By chaos sequence z _ia jth component carrier to the span (a of optimized variable _j, b _j) in, thereby is achieved N number of initial position in n transformer station x direction, a _j, b _jbe respectively lower limit and the higher limit in x direction, transformer station's actual power region, z _ijfor each component z of z vector ₁, z ₂, z ₃..., z _ncomponent; Obtain N number of initial position in n transformer station y direction in the same way;

2.4 initial population being obtained individuality respectively by N number of initial position of x, y according to individual UVR exposure principle.

Step 3: realize weighted Voronoi diagrams figure substation site selection based on Automatic adjusument weight and load distribution thereof to individual, and calculate the adaptive value f of each individuality, the good Popsize of a selectivity solution is as initial solution from N number of initial population.The key of the method is the determination of initial weight and the Automatic adjusument of weight.

The determination of 3.1 initial weights: according to

Determine initial weight, wherein: l (i) is i-th substation's on-load; E (i) is the rated capacity of i-th transformer station; γ is that transformer station is meeting the most high capacity rate under main transformer " N-1 " principle; for transformer station's power factor;

The Automatic adjusument of 3.2 each transformer station weights.In order to avoid because initial weight cannot meet that transformer station actual loading rate requires, the load factor that causes is too high or too low, and principle carries out the Automatic adjusument of weight as shown in Table 2.

Table 2 weight Automatic adjusument principle

In table: a and b is respectively higher limit and the lower limit of transformer station's actual loading rate; Δ is weight adjusting amount, self-adaptative adjustment; it is the weight after i-th transformer station, the t time iteration.When the change of actual loading rate does not meet table 2 condition, weight remains unchanged.Weight increases, and the power supply area of corresponding transformer station diminishes, otherwise corresponding power supply area becomes large.

Step 4: judge whether to meet termination criterion, if meet, algorithm stops, otherwise performs step 5;

Step 5: the adaptive value variances sigma calculating population ²if, σ ²< ε (ε is adaptive value variance threshold values given in advance), occurs precocious, carries out Chaos Search; Otherwise, jump to step 7.

Step 6: Chaos Search.

6.1 produce a n-dimensional vector cr at random ₀, by cr _i+1=4cr _i ³-3cr _iproduce cr=(cr ₁, cr ₂, cr ₃..., cr _i...), i=1,2 ..., Popsize, according to Δ x _ij=m+ (n-m) × cr _ijby cr _ieach component carrier in chaotic disturbance scope [m, n], obtain disturbance quantity Δ x=(Δ x ₁, Δ x ₂..., Δ x _i...), cr ₀for the vector produced at random, cr ₁, cr ₂, cr ₃..., cr _i... for each component of cr vector;

Chaotic disturbance amount component is loaded on each individuality by 6.2, forms the position, chaotic disturbance x direction of each individuality

x′＝(x′ ₁,x′ ₂,…,x′ _i,…)(i＝1,2,…,Popsize)，

Wherein

$\left\{\begin{array}{c}{x}_i^{\&prime;}=(x_{i1}^{\&prime;},x_{i2}^{\&prime;},\&CenterDot;\&CenterDot;\&CenterDot;,x_{\mathrm{in}}^{\&prime;})\\ x_{\mathrm{ij}}^{\&prime;}=x_{\mathrm{ij}}+{\mathrm{\&Delta;x}}_{\mathrm{ij}}\end{array}\right.$

X ' for chaos sequence add disturbance quantity after obtain, x ' ₁, x ' ₂..., x ' _i... for its each component, x ' _i1, x ' _i2..., x ' _infor x ' _icomponent; The position, chaotic disturbance y direction of each individuality is formed according to 6.1,6.2;

Substation site selection based on the weighted Voronoi diagrams figure transformer station addressing algorithm of Automatic adjusument weight and load distribution thereof are carried out in the 6.3 pairs of chaotic disturbance positions, calculate the adaptive value f ' of each individuality;

Adaptive value before and after 6.4 more each individual chaotic disturbances, if the feasible solution x ' that Chaos Search arrives _ibe better than the feasible solution x of corresponding individuality _i, then x ' is used _isubstitute x _i, perform step 7.

Step 7: perform and preserve optimum genetic algorithm, return step 4 after terminating.

7.1 individual UVR exposure

Individual UVR exposure adopts floating-point encoding mode, transformer station's number that individual code length equals 2 times, and individual each gene is substation location in planning region, i.e. the x direction coordinate of locus or y direction coordinate, and site is floating number;

7.2 fitness function

Transformer station and network investment and annual running cost is selected to be used as fitness function to evaluate the programme corresponding to each individuality;

7.3.1 operation is selected

Select operation adopt optimum maintaining strategy and determine that formula system of selection of sampling combines, determine that formula is sampled system of selection:

(1). calculate the expectation existence number N of each individuality in colony of future generation in colony _i, i=1,2 ..., M;

(2). use N _iintegral part determine each corresponding individual existence number in colony of future generation;

(3). according to N _ifraction part descending sort is carried out to individuality, order get before individuality joins in colony of future generation;

7.4 crossover and mutation operations

The employing intersection that counts carries out interlace operation, replaces protogene value complete mutation operation by a random number in genes of individuals value span;

7.5 inversion operations

Put upside down the sequence in the gene order in individual UVR exposure string between random two locus formulated, thus the better individuality that formation one is new;

7.6 transformer station's load areas divide

If d transformer station collection S={s in planning region ₁, s ₂..., s _d, then arbitrary load x belongs to transformer station s _ibe defined as:

R(s _i,x)＝{d(x,s _i)≤d(x,s _j),j＝1,2,…,d,j≠i}

D (x, s in formula _i), d (x, s _j) represent arbitrary load x and s in plane respectively _iand s _jbetween Euclidean distance;

7.7 preserve optimum genetic algorithm.

Claims (4)

1., based on a weighted Voronoi diagrams figure substation planning method for Chaos-Genetic strategy, it is characterized in that, comprise the steps:

Step 1: setup parameter, comprises setting:

Maxeranum: maximum iteration time; Popsize: population scale; N: number of individuals in initial population; PCross crossover probability; PMutation mutation probability; PInversion inversion probability;

Step 2: chaos intialization, produces the initial population comprising individuality, and using the coding of each individuality as substation location;

Step 3: realize weighted Voronoi diagrams figure substation site selection based on Automatic adjusument weight and load distribution thereof to individual, and calculate the adaptive value f of each individuality, the good Popsize of a selectivity solution is as initial solution from N number of initial population;

Step 4: judge whether to meet termination criterion, if meet, algorithm stops, otherwise performs step 5;

Step 5: the adaptive value variances sigma calculating population ²if, σ ²< ε, ε are adaptive value variance threshold values given in advance, occur precocious, carry out Chaos Search; Otherwise, jump to step 7;

Step 6: Chaos Search;

Step 7: perform and preserve optimum genetic algorithm, return step 4 after terminating;

The determination of initial weight and the Automatic adjusument of weight are specially:

1) determination of initial weight: according to

Determine initial weight, wherein: l (i) is i-th substation's on-load; E (i) is the rated capacity of i-th transformer station; γ is that transformer station is meeting the most high capacity rate under main transformer " N-1 " principle; for transformer station's power factor;

2) Automatic adjusument of each transformer station weight.

2. a kind of weighted Voronoi diagrams figure substation planning method based on Chaos-Genetic strategy as claimed in claim 1, it is characterized in that, chaos intialization is further specially:

2.1 produce n-dimensional vector z at random ₀, its each component values is between-1 ~ 1

Z ₀=(z ₀₁, z ₀₂..., z _0n), n is transformer station's number, z ₀₁, z ₀₂..., z _0nfor the z of stochastic generation ₀each component of vector;

2.2 get chaos sequence: according to cube mapping function obtain z ₁, z ₂, z ₃..., z _n, obtain chaos sequence thus

z＝(z ₁,z ₂,z ₃,…,z _N)

Z ₁, z ₂, z ₃..., z _nfor each component of the z phasor that the vector of initial random generation is obtained by cube mapping function;

2.3 according to

$x_{\mathrm{ij}}=\frac{a_j+b_j}{2}+\frac{(b_j-a_j)}{2}{z}_{\mathrm{ij}},i=\mathrm{1,2},...,N;j=\mathrm{1,2},...,n$

By chaos sequence z _ia jth component carrier to the span (a of optimized variable _j, b _j) in, thereby is achieved N number of initial position in n transformer station x direction, a _j, b _jbe respectively lower limit and the higher limit in x direction, transformer station's actual power region, z _ijfor each component z of z vector ₁, z ₂, z ₃..., z _ncomponent; Obtain N number of initial position in n transformer station y direction in the same way;

2.4 initial population being obtained individuality respectively by N number of initial position of x, y according to individual UVR exposure principle.

3. a kind of weighted Voronoi diagrams figure substation planning method based on Chaos-Genetic strategy as claimed in claim 1, it is characterized in that, Chaos Search is specially:

6.1 produce a n-dimensional vector cr at random ₀, by cr _i+1=4cr _i ³-3cr _iproduce cr=(cr ₁, cr ₂, cr ₃..., cr _i...), i=1,2 ..., Popsize, according to Δ x _ij=m+ (n-m) × cr _ijby cr _ieach component carrier in chaotic disturbance scope [m, n], obtain disturbance quantity Δ x=(Δ x ₁, Δ x ₂..., Δ x _i...), cr ₀for the vector produced at random, cr ₁, cr ₂, cr ₃..., cr _i... for each component of cr vector;

Chaotic disturbance amount component is loaded on each individuality by 6.2, forms the position, chaotic disturbance x direction of each individuality

x′＝(x′ ₁,x′ ₂,…,x′ _i,…)，i＝1,2,…,Popsize

Wherein

$\left\{\begin{array}{c}{x}_i^{\&prime;}=(x_{i1}^{\&prime;},x_{i2}^{\&prime;},...,x_{\mathrm{in}}^{\&prime;})\\ x_{\mathrm{ij}}^{\&prime;}=x_{\mathrm{ij}}+\mathrm{\&Delta;}{x}_{\mathrm{ij}}\end{array}\right.$

X ' for chaos sequence add disturbance quantity after obtain, x ' ₁, x ' ₂..., x ' _i... for its each component, x ' _i1, x ' _i2..., x ' _infor x ' _icomponent; The position, chaotic disturbance y direction of each individuality is formed according to 6.1,6.2;

Substation site selection based on the weighted Voronoi diagrams figure transformer station addressing algorithm of Automatic adjusument weight and load distribution thereof are carried out in the 6.3 pairs of chaotic disturbance positions, calculate the adaptive value f ' of each individuality;

Adaptive value before and after 6.4 more each individual chaotic disturbances, if the feasible solution x ' that Chaos Search arrives _ibe better than the feasible solution x of corresponding individuality _i, then x ' is used _isubstitute x _i, perform step 7.

4. a kind of weighted Voronoi diagrams figure substation planning method based on Chaos-Genetic strategy as claimed in claim 1, is characterized in that, performs to preserve optimum genetic algorithm and be specially:

7.1 individual UVR exposure

Individual UVR exposure adopts floating-point encoding mode, transformer station's number that individual code length equals 2 times, and individual each gene is substation location in planning region, i.e. the x direction coordinate of locus or y direction coordinate, and site is floating number;

7.2 fitness function

Transformer station and network investment and annual running cost is selected to be used as fitness function to evaluate each individual corresponding programme;

7.3.1 operation is selected

Select operation adopt optimum maintaining strategy and determine that formula system of selection of sampling combines, determine that formula is sampled system of selection:

(1). calculate expectation existence number N i, the i=1 of each individuality in colony of future generation in colony, 2 ..., M;

(2). use N _iintegral part determine each corresponding individual existence number in colony of future generation;

(3). according to N _ifraction part descending sort is carried out to individuality, order get before individuality joins in colony of future generation;

7.4 crossover and mutation operations

The employing intersection that counts carries out interlace operation, replaces protogene value complete mutation operation by a random number in genes of individuals value span;

7.5 inversion operations

Put upside down the sequence in the gene order in individual UVR exposure string between random two locus formulated, thus the better individuality that formation one is new;

7.6 transformer station's load areas divide

If d transformer station collection S={s in planning region ₁, s ₂..., s _d, then arbitrary load x belongs to transformer station s _ibe defined as:

R(s _i,x)＝{d(x,s _i)≤d(x,s _j),j＝1,2,…,d,j≠i}

D (x, s in formula _i), d (x, s _j) represent arbitrary load x and s in plane respectively _iand s _jbetween Euclidean distance;

7.7 preserve optimum genetic algorithm.