CN104834963A - Distributed wind power integration optimization configuration device based on improved bee colony algorithm and method - Google Patents

Abstract

The invention relates to a distributed wind power integration optimization configuration device based on improved bee colony algorithm and a distributed wind power integration optimization configuration method. The device comprises a dispatching center data acquisition module, an OPC Server, a DSP, a MAX232 chip, an input module, a power supply module, and an upper computer. In order to solve the optimization configuration problem of the distributed wind power integration access point, the improved bee colony algorithm is adopted. The position and the capacity of the distributed wind power integration access point can be set as the unknown parameters, and the optimizing matrix can be formed. By adopting the main loop capable of guaranteeing the richness of the data and the reliability of the optimizing result and the nested loop aiming at the local searching of the maximum yield value, the searching progress and the convergence speed can be accelerated, and the searching accuracy can be guaranteed, and then the problems such as slow convergence speed and easy occurrence of local optimization during the solving of the optimizing problem by adopting the basic bee colony algorithm can be solved. The overall optimal solution can be acquired by the fast convergence, when the algorithm is applied to the optimization configuration of the distributed wind power integration access point.

Description

Device and method is distributed rationally based on the distributing wind-electricity integration improving ant colony algorithm

Technical field

The invention belongs to technical field of wind power generation, be specially a kind of distributing wind-electricity integration based on improving ant colony algorithm and distribute device and method rationally.

Background technology

Along with the develop rapidly of modern industry, the demand of the mankind to the energy obviously increases, and on the earth, available conventional energy resources is increasingly deficient.Wind-powered electricity generation has become the second largest clean energy resource of China after water power, and in recent years, the centralized large-scale development of China's wind-powered electricity generation is gradually saturated, and grid-connected difficult problem is always undecided simultaneously, and off-grid Frequent Accidents, is unfavorable for industry development.Compare the large wind-powered electricity generation in large base, distributing access Wind Power Project has various advantages, and the advantage of development distributing Wind Power Project is to have maximally utilised wind resource.In generating link, distributing generation mode requires lower to wind speed, floor area etc.; In transmission of electricity link, the transmission distance of distributing generation mode is the feature such as shorter usually.In the mode that European wind-power market much adopts distributing to access, this allows itself and traditional electric power and electrical network well merge into one.At present, domestic wind-power market is when facing the present situation of dissolving difficult, and also actively use for reference the experience in Europe, the exploitation of distributing Wind Power Project has had successful practice in some areas of China.

After decentralized power supply access power distribution network on the impact of the factors such as grid nodes voltage, network trend, network loss and decentralized power supply access capacity and position closely related.So before decentralized power supply access power distribution network, must plan it.Original simple idea is exactly the combination listing the position of all decentralized power supply, type and capacity, then attempts one by one finding allocation optimum scheme.But along with the increase of number of power sources, type and accessible position, allocation plan by there is the growth of fearful exponential form, so find a kind of reliably feasible scheme imperative in order to solve this problem.

In prior art document, China Patent No. is 201210009111.X, name is called in the patent of " the network optimization planing method that a kind of large-scale wind power is grid-connected " and discloses the grid-connected network optimization planing method of a kind of large-scale wind power, its weak point is that arithmetic speed is excessively slow, do not meet the feature of distributing wind-power electricity generation, do not carry out capacity and the siteselecting planning of wind power plant in conjunction with the feature of existing electrical network, the requirement that electric power is dissolved cannot be realized nearby.

Summary of the invention

For prior art Problems existing, the invention provides a kind of distributing wind-electricity integration based on improving ant colony algorithm and distribute device and method rationally.

The technical solution used in the present invention is:

Apparatus of the present invention comprise: dispatching center's data acquisition module, OPC Server, DSP, MAX232 chip, load module, power module, host computer.Dispatching center's data acquisition module is the acquisition of data; OPC Server is mainly used for the communication between DSP and dispatching center's data acquisition module, employing be OPC communication, OPC generally can provide OPC client, and the opc server being used for providing with equipment manufacturer carries out communication.Because the standard that OPC has Microsoft decided at the higher level but not officially announced, so OPC client carries out communication without the need to revising the opc server that just can provide with each family; Communication between MAX232 driving circuit primary responsibility DSP and host computer, employing be serial communication mode, utilization be that Transistor-Transistor Logic level is converted to RS232 level by MAX232 level shifting circuit; DSP module major function is to the process of power distribution network data analysis.That DSP module adopts is the digital signal processor TMS320VC5402 of TI company; Keyboard is mainly used in the input of some master datas; Power module mainly contains the power-switching circuit of 5V power supply, voltage conversion chip TPS767D318.Wherein dispatching center's data acquisition module, OPC Server, DSP, driving circuit MAX232 chip, host computer connect in turn, 5V power supply is connected with voltage conversion chip TPS767D318, MAX232 etc. respectively, and the kernel of DSP is connected with 1.8V with 3.3V respectively with I/O interface.

The inventive method comprises: gather distribution power flow data according to power distribution network framed structure isoboles and dispatching center, utilize forward-backward sweep method to calculate power attenuation on each node successively, then calculate the via net loss of whole power distribution network the whole network according to the power attenuation situation on each branch road; Again carry out the selection of Wind turbines access point and the calculating of access capacity according to the continuous Optimum search of the size of via net loss, utilize in the process selected and calculate and improve ant colony algorithm, finally obtain the allocation optimum scheme of Wind turbines access power distribution network.

The on-position of the present invention to distributing wind energy turbine set and the optimization problem of capacity, be converted into after existing power distribution network access distributing Wind turbines, whole network of network Dissipation change problem.Then set up objective optimization function, make position and the capacity of being determined access by the principle that via net loss is minimum, wherein set up objective function is:

$y={\mathrm{minP}}_{LOSS}=\underset{i=1}{\overset{n-1}{\Σ}}(\mathrm{\Δ}{\tilde{S}}_i^{\′}+\mathrm{\Δ}{\tilde{S}}_i^{\′\′})$

The present invention is when the optimal location accessed access distributing Wind turbines and optimum capacity carry out Optimum search, and the algorithm of employing is the ant colony algorithm improved.Because ant colony algorithm is simple, realize convenience, fast convergence rate, strong robustness, it has obvious superiority in complex combination optimization, and gets the nod in optimization calculating field, the world and be used widely in practical problems.Unknown parameter is established in the position of distributing Wind turbines access point and capacity by the present invention, forms a D × N ₀matrix A _k, wherein A _krepresent the optimizing matrix of gained after k iteration, the searching algorithm of improvement bee colony for search factor be namely D × N ₀dimension matrix A _kcolumn vector, finally can according to column vector A _kcorresponding fitness value decides optimum position and capacity.And column vector A _kcorresponding fitness function, the function of the income angle value in i. e. source is as follows:

$g\left(X_i\right)=\frac{1}{P_{\mathrm{Loss}}}$

The i.e. inverse of network total losses, will search maximum fitness value here.

Can be used as at the capacity carrying out network nodes all in the process of searching for is the column vector that D ties up, here by changing the parameter of this each component of vector, be about to wherein by the component of optimizing as unknown quantity, other component is as known quantity, carry out the calculating of maximum return angle value, then successively Optimum search is carried out to other each components, finally determine the vectorial X of whole optimum _o.

The matrix A that the present invention is utilizing improvement ant colony algorithm to carry out continuous Optimum search _keach column vector must meet following constraint condition:

A) constraint of distributing Wind turbines maximum node number is installed in power distribution network:

Count (X _iin each component be not 0 number)≤B

In formula, X _irepresent the on-position of distributing Wind turbines and a solution of capacity in search procedure, B is the maximum node number that distributing Wind turbines is installed.

B) the maximum installed capacity constraint of decentralized power supply:

$\underset{i=1}{\overset{n-1}{\mathrm{\Σ}}}{P}_{\mathrm{DGi}}\≤P_{\mathrm{DG}\max }$

In formula, P _dGiit is the capacity of the distributing Wind turbines of i-th node installation; P _dGmaxfor allowing the maximum installed capacity of distributing Wind turbines.

C) node distributing installed capacity of wind-driven power constraint to be selected

0≤P _DGi≤P _DGimax

In formula, P _dGimaxit is the max cap. i-th node to be selected allowing the distributing Wind turbines installed.

D) distributing Wind turbines penetration constraint

$\frac{\underset{i=1}{\overset{n-1}{\mathrm{\Σ}}}{P}_{\mathrm{DGi}}}{S_{\mathrm{Load}}^{\max }}\≤\mathrm{\ρ}$

In formula, ρ is penetration coefficient, for installing the total volume of Wind turbines in system, for whole power distribution network maximal workload.

E) node voltage constraint

U _imin≤U _i≤U _imax

In formula, U _imin, U _imaxbe respectively minimum permission magnitude of voltage and the maximum allowable voltage of i-th node.

Should make full use of the existing transformer station of electrical network and circuit according to distributing access Wind Power Project, not newly-built high pressure sends circuit and 110 kilovolts, 66 kv substations in principle, and the power transformating and supplying facility of other electric pressures not newly-built as far as possible.Therefore when carrying out distributing wind power integration capacity and choosing, carry out access grid-connected according to the demand of existing power distribution network, the demand capacity of power distribution network needs to calculate according to the structure of power distribution network.After the structure of power distribution network is determined, will carry out the line losses management of whole network according to the data of scheduling, this method is by meritorious, the idle iterative equation of power distribution network node, branch road loss equation, and constantly iteration is as follows altogether for network total losses system of equations:

$P_i^{\′}=\left\{\begin{array}{cc}{P}_i-P_{\mathrm{DGi}}& i=n\\ P_i-P_{\mathrm{DGi}}-{\mathrm{\ΔP}}_{i-1}^{\′}-P_{i-1}^{\′}-{\mathrm{\ΔP}}_{\mathrm{Bi}}& (i=\mathrm{1,2}\mathrm{Ln}-1)\end{array}\right.$ (the active power iterative equation of node)

$Q_i^{\′}=\left\{\begin{array}{cc}{Q}_i-Q_{\mathrm{DGi}}& i=n\\ Q_i-Q_{\mathrm{DGi}}-{\mathrm{\ΔQ}}_{i-1}^{\′}-Q_{i-1}^{\′}-{\mathrm{\ΔQ}}_{\mathrm{Bi}}& (i=\mathrm{1,2}\mathrm{Ln}-1)\end{array}\right.$ (the inactivity iterative equation of node)

${\mathrm{\ΔS}}_i^{\′}=\frac{P_i^{\′2}+Q_i^{\′2}}{U_i^2}(R_i+{\mathrm{jX}}_i)={\mathrm{\ΔP}}_i^{\′}+{\mathrm{j\ΔQ}}_i^{\′},(i=\mathrm{1,2,3,4}L)$ (loss of wall scroll defeated line branch road)

(on wall scroll branch road Transformer Winding variable loss)

$P_{LOSS}=\underset{i=1}{\overset{n-1}{\Σ}}(\mathrm{\Δ}{\tilde{S}}_i^{\′}+\mathrm{\Δ}{\tilde{S}}_i^{\′\′})$ (via net loss total on power distribution network)

Wherein the via net loss of transmission line of electricity on this branch road after i-th branch road access distributing Wind turbines, R _i, X _irepresent the impedance on transmission line of electricity and reactance respectively; P ' _i, Q ' _ithat after access distributing wind-powered electricity generation group, each node is gained merit and reactive power distribution situation, represent the variable power loss on the winding of transformer contained by node on i-th, Δ P _bi, Δ Q _bibe illustrated respectively in the power attenuation on contained transformer on the i-th node; with with with be illustrated respectively in the variable loss of the winding on the i-th node in contained transformer and constant iron loss, P _dGi, Q _dGirepresent the active power and the reactive power size that i-th node access distributing Wind turbines, Δ P ' _i-1, Δ Q ' _i-1represent that after i-th node accesses decentralized power supply, institute connects gaining merit and reactive loss on next branch road respectively, P _lOSSrepresent the total losses in power distribution network on transmission line of electricity and the total losses sum on transformer.

Based on an Optimal Configuration Method for the wind power-generating grid-connected access point of distributing of improvement ant colony algorithm, it is characterized in that the distributing ant colony algorithm that have employed improvement, the implementation step of ant colony algorithm is as follows:

Step 1: for n=0 initial time, the N number of feasibility solution of stochastic generation (X ₁, X ₂, K X _n), the distribution-search bee search of food source in i. e. source.The feasible solution X of random generation _i:

$X_i^j=X_{\min }^j+\mathrm{rand}\left(\mathrm{0,1}\right)(X_{\max }^j-X_{\min }^j)$

In formula, and j ∈ (1,2 ..., D), i representing matrix A _kthe i-th column vector.

Will give income angle value function, calculate the random income angle value producing each feasibility solution vector, and by forward for rank N ₀individual feasibility solution is designated as the initial food source A of gathering honey honeybee ₀.

Step 2: according to initial food source, bee colony has the process of self study, namely follows the selective beam search of honeybee

$\begin{array}{ccccc}\mathrm{IF}& g\left(X_i^j\right)>g\left(X_k^j\right)& \mathrm{and}& X_i^j>X_k^j& \mathrm{Then}\end{array}$

$X_i^j=X_k^j+\mathrm{rand}(-\mathrm{1,1})(X_i^j-X_k^j)$

Directed iteration is carried out successively, until obtain searching maximum return angle value according to step like this.

Step 3: in order to improve search efficiency and keep the diversity of food source, carrying out searching in the process of carrying out, bee colony carries out there is capacity of self-regulation in the process of searching for, namely

J ∈ in formula (l, 2 ..., D), Y _i ^jthe element in newly-generated feasibility node, and Y ∈ S, be random number, its scope is [-1,1], and its value is decided by value.

Individual search volume is S, and wherein S meets the constraint condition of optimizing, carry out searching probability only with the distribution X in current foodstuff source _in () is relevant, and with previous moment search for food source distribution and distribute the moment have nothing to do.

Step 4: form new food source distribution X according to Step 3 _in (), utilizes step 2 to carry out the search of the fitness of food source, constantly carries out the circulation of step 3 and step 2, finally obtain the food source that income degree is different.

Step 5: be in honeycomb and do not sorted from high to low according to income degree information by the honeybee of following of employing, according to the greedy characteristic of ant colony algorithm to food source, by the different food source vector Y searched _ibetween and with food source vector X _ifitness value compares, and the food source of following gathering honey is followed according to the size of income degree, chooses the reservation that in both, fitness value is high.

The choice principle of bee colony to optimum food source position is:

G (Y _i)>=g (X _i) then the feasibility solution of Optimal Distribution be Y _i;

G (Y _i) < g (X _i) then the feasibility solution of Optimal Distribution be X _i.

The greedy characteristic of ant colony algorithm ensure that bee colony can retain the highest food source of fitness value all the time as elite's individuality, can not be fallen back in the direction obtaining high yield angle value food source.

Step 6: follow honeybee to income degree information relatively and sequence will observe the variation tendency of maximum return angle value simultaneously, set a minimal circulation restriction frequency n ₀, the minimal circulation frequency n of optimum circulation ₁, largest loop limits number of times MaxCN.If income angle value search cycle index n is greater than minimum restriction cycle index n ₀, and follow time number of rings n at the minimum of optimum cycle index ₁in scope, maximum return value does not change, and namely meets the following conditions g (X _i) < g (X _j) (i ≠ j); Then can enter next step to improve search efficiency, otherwise continue to return circulation step 2 to step 6, in like manner when income angle value cycle index reaches the cycle index of maximum restriction, stop circulation equally, perform next step.

Step 7: according to the food source income angle value size received, carries out the selection of Bian honey food source, follows honeybee and now change gathering honey honeybee into, in the gathering honey number of times of regulation, carries out gathering honey operation, make every effort to find better gathering honey point to around selected food source.Be similar to the operation of investigation honeybee, but search volume reduces greatly, and be in the peripheral extent of delimiting food source.

Selection mode is by probability selection in all food sources, new probability formula:

$P\left(X_i\right)=\frac{g\left(X_i\right)}{\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}g\left(X_i\right)}$

P (X _i) represent at variable-value to be X _itime, lead honeybee to the select probability of food source, g (X _i) represent at variable-value to be X _itime, corresponding fitness value, all fitness value sums.The search volume of following honeybee is still S, obtain food source probability only with the current food source state A obtained _krelevant, and have nothing to do with food source state before and the moment that distributes.

Step 8: if setting Evolution of Population number of times completes, then stop all calculating, export optimum food source income angle value and the relevant parameter of last record.

After searching algorithm Optimum search, maximum adaptation angle value will be obtained, and the feasibility solution X corresponding to maximum adaptation angle value _o, its feasibility solution X _oin each point of vector just represent each node connect the capacity of distributing Wind turbines, this just differentiated which node need to access distributing Wind turbines and corresponding point connect the capacity that distributing Wind turbines accesses.

The present invention is directed to the optimization allocation of the wind power-generating grid-connected access point of distributing, adopt a kind of ant colony algorithm of improvement, the position of distributing Wind turbines access point and capacity are asserted unknown parameter, form optimizing matrix, adopt the major cycle of reliability and the nested circulation of the Local Search for maximum return angle value of the rich and optimizing result ensureing data, accelerate search progress, speed of convergence, ensure that the accuracy of search, solve speed of convergence when basic ant colony algorithm solves combinatorial optimization problem and slowly, be easily absorbed in the problem of local optimum.Algorithm application can converge to globally optimal solution rapidly in distributing rationally of the wind power-generating grid-connected access point of distributing.In addition, algorithm is simple, strong robustness, has obvious superiority.

This device, based on the ant colony algorithm improved, utilizes DSP to process the data that SCADA transmits, and whole device can carry out the integrated of data interaction and system easily, thus greatly can strengthen the informatization of wind-power electricity generation.

Accompanying drawing explanation

Fig. 1 is the structural drawing of distributing device rationally based on the wind power-generating grid-connected access point of distributing improving ant colony algorithm;

Fig. 2 is the power-switching circuit of TPS767D318;

Fig. 3 is PC serial port level chance-over circuit;

Fig. 4 is for improving ant colony algorithm search utility process flow diagram.

Embodiment:

As Figure 1-4, device is distributed rationally based on the distributing wind-electricity integration improving ant colony algorithm, comprise: dispatching center's data acquisition module, OPC Server, DSP, MAX232 chip, load module, power module and host computer, building of hardware module is as follows: the data acquisition module of dispatching center is connected with DSP TMS320VC5402 data acquisition interface D15-D8 through OPC Server; DSP TMS320VC5402 comprising clock circuit, reset circuit is connected on corresponding interface respectively, the power supply of DSP is provided by the power-switching circuit of 5V integrated regulator, voltage conversion chip TPS767D318, wherein the work core voltage of TMS320VC5402 is 1.8V, I/O interface voltage is 3.3V; Serial line interface Transistor-Transistor Logic level being converted to 232 level through MAX232 chip when carrying out data transmission is uploaded to host computer.At DSP, all through communication module, host computer is uploaded to all data cases in data handling procedure.Dispatching center's data acquisition module is the acquisition of data source; OPC Server is responsible for the communication between DSP and dispatching center's data acquisition module, what it adopted is OPC communication, OPC has OPC client, the opc server being used for providing with equipment manufacturer carries out communication, because the standard that OPC has Microsoft decided at the higher level but not officially announced, so OPC client carries out communication without the need to revising the opc server that just can provide with each family.MAX232 driving circuit refers to communication between DSP and host computer, employing be serial communication mode, this module mainly contains MAX232 level shifting circuit, Transistor-Transistor Logic level is converted to RS232 level; DSP module major function is to the process of power distribution network data analysis, and that DSP module adopts is the digital signal processor TMS320VC5402 of TI company; Input through keyboard module is mainly used in the input of some master datas; 1.8V and the 3.3V power supply that power module mainly contains 5V power supply, voltage conversion chip TPS767D318 produces.Wherein dispatching center's data acquisition module, OPC Server, DSP, MAX232 chip, host computer connect in turn.5V power supply is connected with voltage conversion chip TPS767D318, MAX232 etc. respectively, and the kernel of DSP is connected with 1.8V with 3.3V respectively with I/O interface.

The collocation method of distributing device rationally of the described wind power-generating grid-connected access point of distributing based on improvement ant colony algorithm, comprises the following steps:

The first step: the data that SCADA system is monitored are sent to the data acquisition port of DSP by the external communication interface of dispatching center by OPC Server interface, waits for DSP process;

Second step: DSP carries out data screening according to processing demands, and main screening mainly contains the operational factor of the voltage of each node in power distribution network, load condition, branch impedance and branch road transformer, the power flow direction on network;

3rd step: DSP can according to the power flow direction figure of equivalence, utilizes forward-backward sweep method to calculate power attenuation on each node, finally calculates total via net loss situation;

4th step: the variable in power attenuation is set, the capacity of namely added on node Wind turbines.Namely the via net loss of the whole network, node power change, and again utilize the value of forward-backward sweep method rated output loss;

5th step: utilize DSP to carry out the Optimum search of the power attenuation minimum value in the 3rd step based on ant colony algorithm;

6th step: be mainly divided into major cycle and nested circulation in the search procedure of carrying out the 4th step, major cycle is exactly the reliability in order to ensure the rich of data and optimizing result, constantly global search is carried out to maximum return angle value, namely in search procedure, constantly change initial ranging vector; Nested circulation for be the Local Search of maximum return angle value, namely search procedure carries out beam search according to the actual change of income degree, thus accelerates search progress and ensure that the accuracy of search;

7th step: repeat the circulation in the 5th step, until the error percentage δ obtaining the difference of two optimum two maximum adaptation angle value is less than limiting error percentage ε, wherein

$\mathrm{\&delta;}=\left|\frac{g\left(X_{o1}\right)-g\left(X_{o2}\right)}{g\left(X_{o2}\right)}\right|*100\%$

G (X _o1), g (X _o2) be two values maximum in current maximum return angle value respectively, if δ < is ε, stops circulation, enter next step; Otherwise return and continue circulation, until reach largest loop limited number of times MaxCN, stop circulation.Wherein ε is limiting error percentage, and general value is less than 0.01%.

8th step: to the maximum return degree obtained in step 5 at every turn all by communication module is sent to host computer, finally can real-time monitored change to search;

9th step: the distribution overview exporting last maximum return angle value, and finally export the feasibility solution X corresponding to overall maximum return angle value _o.Its feasibility solution X _oin each point of vector just represent each node connect the capacity of distributing Wind turbines, if the component value in solution vector is 0, illustrates that this point is not allowable and add distributing Wind turbines; If the component in solution vector is not 0, just access the wind-powered electricity generation capacity of corresponding numerical value at this point.

The Optimal Configuration Method of the described wind power-generating grid-connected access point of distributing based on improvement ant colony algorithm, utilizes improvement ant colony algorithm to carry out search variables in the process of Optimum search and will meet following constraint condition:

A) constraint of distributing Wind turbines maximum node number is installed in power distribution network:

Count (X _iin each component be not 0 number)≤B

In formula, X _irepresent the on-position of distributing Wind turbines and a solution of capacity in search procedure, even X _ithe access capacity of=0 i-th node is 0, otherwise is the access capacity of i-th node, and B is the maximum node number that distributing Wind turbines is installed.

B) the maximum installed capacity constraint of decentralized power supply:

$\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{DGi}}\&le;P_{\mathrm{DG}\max }$

In formula, P _dGiit is the capacity of the distributing Wind turbines of i-th node installation; P _dGmaxfor allowing the maximum installed capacity of distributing Wind turbines.

C) node distributing installed capacity of wind-driven power constraint to be selected:

0≤P _DGi≤P _DGimax

In formula, P _dGimaxit is the max cap. that i-th node to be selected allows the distributing Wind turbines installed.

D) distributing Wind turbines penetration constraint:

$\frac{\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{DGi}}}{S_{\mathrm{Load}}^{\max }}\&le;\mathrm{\&rho;}$

In formula, ρ is penetration coefficient, for installing the total volume of Wind turbines in system, for whole power distribution network maximal workload.

E) node voltage constraint:

U _imin≤U _i≤U _imax

In formula, U _imin, U _imaxbe respectively minimum permission magnitude of voltage and the maximum allowable voltage of i-th node.

The Optimal Configuration Method of the described wind power-generating grid-connected access point of distributing based on improvement ant colony algorithm, what adopt is that the ant colony algorithm of improvement is as optimized algorithm, optimized variable is the vector of food source, here via net loss food source related function represents, the size of via net loss is represented by the inverse of maximum return angle value, and implementation step is as follows:

Step one: parameter initialization;

Step 2: lead honeybee random search food source;

Step 3: follow honeybee and carry out choosing more rich N according to leading the income angle value in honeybee sharing of food source ₀individual food source, it is larger that its income angle value is more enriched in i. e. source;

Step 4: for the more rich N of food source ₀individual food source carries out beam search, finds the maximum food source of income angle value;

Step 5: honeybee carries out food source and explores in investigation, and compares with current maximum return angle value, chooses the larger food source of income angle value and return step 3 and carry out recycle, otherwise search bee is explored again;

Step 6: repeat step 3, step 5, until the error percentage δ obtaining the difference of two optimum two maximum adaptation angle value is less than limiting error percentage ε, wherein

$\mathrm{\&delta;}=\left|\frac{g\left(X_{o1}\right)-g\left(X_{o2}\right)}{g\left(X_{o2}\right)}\right|*100\%$

G (X _o1), g (X _o2) be two values maximum in current maximum return angle value respectively, if δ < is ε, stops circulation, enter next step; Otherwise return and continue circulation, until reach largest loop limited number of times MaxCN, stop circulation.Wherein ε is limiting error percentage, and general value is less than 0.01%;

Step 7: output obtains the food source of maximum return angle value.

The concrete execution step of described step 4 is as follows:

1.: choose the N that richness is larger ₀individual food source;

2.: follow honeybee according to the food source income angle value size leading honeybee to share and variation tendency, follow honeybee and directed selection search is carried out to a certain component variation of food source;

3.: beam search is carried out to a certain Parameters variation, and to income angle value change and compare statistics.Search principle is: $\begin{array}{ccccc}\mathrm{IF}& g\left(X_i^j\right)>g\left(X_k^j\right)& \mathrm{and}& X_i^j>X_k^j& \mathrm{Then}\end{array}$

$X_i^j=X_k^j+\mathrm{rand}(-\mathrm{1,1})(X_i^j-X_k^j)$

Record the income angle value of all food sources.When searching times reaches minimum searching times n ₀, and search optimal value is at cumulative frequency n ₁in secondary, maximum return angle value does not change, then the feasible solution exporting current maximum return angle value that stops search directly enters step 5, otherwise carries out cyclic search, until search local optimum financial value and enter step 5 according to search principle;

4.: change search component, return step 3 and carry out cyclic search;

: the angle value that bears interest in step 4 is compared, and export current maximum value angle value and obtain the feasibility solution of maximum return angle value 5..

Claims (5)

1. distribute device rationally based on the distributing wind-electricity integration improving ant colony algorithm, it is characterized in that comprising: dispatching center's data acquisition module, OPC Server, DSP, MAX232 chip, load module, power module and host computer, dispatching center's data acquisition module is the acquisition of data source; OPC Server is responsible for the communication between DSP and dispatching center's data acquisition module, and MAX232 chip drive circuit refers to communication between DSP and host computer, employing be serial communication mode; DSP module is to the process of power distribution network data analysis, and that DSP module adopts is the digital signal processor TMS320VC5402 of TI company; Keyboard is for inputting master data; 1.8V and the 3.3V power supply that power module mainly contains 5V power supply, voltage conversion chip TPS767D318 produces, wherein dispatching center's data acquisition module, OPC Server, DSP, MAX232 chip, host computer connect in turn, 5V power supply is connected with voltage conversion chip TPS767D318, MAX232 chip respectively, and the kernel of DSP is connected with 3.3V power supply with 1.8V power supply respectively with I/O interface.

2. the distributing wind-electricity integration based on improving ant colony algorithm according to claim 1 distributes the collocation method of device rationally, it is characterized in that comprising the following steps:

The first step: the data that SCADA system is monitored are sent to the data acquisition port of DSP by the external communication interface of dispatching center by OPC Server interface, waits for DSP process;

Second step: DSP carries out data screening according to processing demands, and main screening mainly contains the operational factor of the voltage of each node in power distribution network, load condition, branch impedance and branch road transformer, the power flow direction on network;

3rd step: DSP can according to the power flow direction figure of equivalence, utilizes forward-backward sweep method to calculate power attenuation on each node, finally calculates total via net loss situation;

4th step: the variable in power attenuation is set, the capacity of namely added on node Wind turbines.Namely the via net loss of the whole network, node power change, and again utilize the value of forward-backward sweep method rated output loss;

5th step: utilize DSP to carry out the Optimum search of the power attenuation minimum value in the 3rd step based on ant colony algorithm;

6th step: be mainly divided into major cycle and nested circulation in the search procedure of carrying out the 4th step, major cycle is exactly the reliability in order to ensure the rich of data and optimizing result, constantly global search is carried out to maximum return angle value, namely in search procedure, constantly change initial ranging vector; Nested circulation for be the Local Search of maximum return angle value, namely search procedure carries out beam search according to the actual change of income degree, thus accelerates search progress and ensure that the accuracy of search;

7th step: repeat the circulation in the 5th step, until the error percentage δ obtaining the difference of two optimum two maximum adaptation angle value is less than limiting error percentage ε, wherein

$\mathrm{\&delta;}=\left|\frac{g\left(X_{o1}\right)-g\left(X_{o2}\right)}{g\left(X_{o2}\right)}\right|*100\%$

G (X _o1), g (X _o2) be two values maximum in current maximum return angle value respectively, if δ < is ε, stops circulation, enter next step; Otherwise return and continue circulation, until reach largest loop limited number of times MaxCN, stop circulation.Wherein ε is limiting error percentage, and general value is less than 0.01%.

8th step: to the maximum return degree obtained in step 5 at every turn all by communication module is sent to host computer, finally can real-time monitored change to search;

9th step: the distribution overview exporting last maximum return angle value, and finally export the feasibility solution X corresponding to overall maximum return angle value _o.Its feasibility solution X _oin each point of vector just represent each node connect the capacity of distributing Wind turbines, if the component value in solution vector is 0, illustrates that this point is not allowable and add distributing Wind turbines; If the component in solution vector is not 0, just access the wind-powered electricity generation capacity of corresponding numerical value at this point.

3. the distributing wind-electricity integration Optimal Configuration Method based on improving ant colony algorithm according to claim 2, is characterized in that utilizing improvement ant colony algorithm to carry out search variables in the process of Optimum search will meet following constraint condition:

A) constraint of distributing Wind turbines maximum node number is installed in power distribution network:

Count (X _iin each component be not 0 number)≤B

In formula, X _irepresent the on-position of distributing Wind turbines and a solution of capacity in search procedure, even X _ithe access capacity of=0 i-th node is 0, otherwise is the access capacity of i-th node, and B is the maximum node number that distributing Wind turbines is installed.

B) the maximum installed capacity constraint of decentralized power supply:

$\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{DGi}}\&le;P_{\mathrm{DG}\max }$

In formula, P _dGiit is the capacity of the distributing Wind turbines of i-th node installation; P _dGmaxfor allowing the maximum installed capacity of distributing Wind turbines.

C) node distributing installed capacity of wind-driven power constraint to be selected:

0≤P _DGi≤P _DGimax

In formula, P _dGimaxit is the max cap. that i-th node to be selected allows the distributing Wind turbines installed.

D) distributing Wind turbines penetration constraint:

$\frac{\underset{i=1}{\overset{n-1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{DGi}}}{S_{\mathrm{Load}}^{\max }}\&le;\mathrm{\&rho;}$

In formula, ρ is penetration coefficient, for installing the total volume of Wind turbines in system, for whole power distribution network maximal workload.

E) node voltage constraint:

U _imin≤U _i≤U _imax

In formula, U _imin, U _imaxbe respectively minimum permission magnitude of voltage and the maximum allowable voltage of i-th node.

4. the distributing wind-electricity integration Optimal Configuration Method based on improving ant colony algorithm according to claim 2, what it is characterized in that employing is that the ant colony algorithm of improvement is as optimized algorithm, optimized variable is the vector of food source, here via net loss food source related function represents, the size of via net loss is represented by the inverse of maximum return angle value, and implementation step is as follows:

Step one: parameter initialization;

Step 2: lead honeybee random search food source;

Step 3: follow honeybee and carry out choosing more rich N according to leading the income angle value in honeybee sharing of food source ₀individual food source, it is larger that its income angle value is more enriched in i. e. source;

Step 4: for the more rich N of food source ₀individual food source carries out beam search, finds the maximum food source of income angle value;

Step 5: honeybee carries out food source and explores in investigation, and compares with current maximum return angle value, chooses the larger food source of income angle value and return step 3 and carry out recycle, otherwise search bee is explored again;

Step 6: repeat step 3, step 5, until the error percentage δ obtaining the difference of two optimum two maximum adaptation angle value is less than limiting error percentage ε, wherein

$\mathrm{\&delta;}=\left|\frac{g\left(X_{o1}\right)-g\left(X_{o2}\right)}{g\left(X_{o2}\right)}\right|*100\%$

G (X _o1), g (X _o2) be two values maximum in current maximum return angle value respectively, if δ < is ε, stops circulation, enter next step; Otherwise return and continue circulation, until reach largest loop limited number of times MaxCN, stop circulation.Wherein ε is limiting error percentage, and general value is less than 0.01%;

Step 7: output obtains the food source of maximum return angle value.

5. the distributing wind-electricity integration Optimal Configuration Method based on improving ant colony algorithm according to claim 4, is characterized in that the concrete execution step of described step 4 is as follows:

1.: choose the N that richness is larger ₀individual food source;

2.: follow honeybee according to the food source income angle value size leading honeybee to share and variation tendency, follow honeybee and directed selection search is carried out to a certain component variation of food source;

3.: beam search is carried out to a certain Parameters variation, and to income angle value change and compare statistics.Search principle is: $\begin{array}{ccccc}\mathrm{IF}& g\left(X_i^j\right)>g\left(X_k^j\right)& \mathrm{and}& X_i^j>X_k^j& \mathrm{Then}\end{array}$

$X_i^j=X_k^j+\mathrm{rand}(-\mathrm{1,1})(X_i^j-X_k^j)$

Record the income angle value of all food sources.When searching times reaches minimum searching times n ₀, and search optimal value is at cumulative frequency n ₁in secondary, maximum return angle value does not change, then the feasible solution exporting current maximum return angle value that stops search directly enters step 5, otherwise carries out cyclic search, until search local optimum financial value and enter step 5 according to search principle;

4.: change search component, return step 3 and carry out cyclic search;

: the angle value that bears interest in step 4 is compared, and export current maximum value angle value and obtain the feasibility solution of maximum return angle value 5..