CN103699950A - Electric vehicle charging station planning method considering traffic network flow - Google Patents

Abstract

The invention relates to an electric vehicle charging station planning method considering traffic network flow. The method comprises the following steps: referencing an intercepting location model in a traffic field, taking largest captured traffic flow, minimum network loss of a power distribution system and minimum node voltage deviation as an objective, establishing a multi-objective decision model, then determining a reasonable weight coefficient of each objective function after being normalized by using a super efficiency data envelopment analysis evaluation method, and transforming a multi-objective optimization problem to a single objective optimization problem, later putting forward an improved binary particle swarm optimization algorithm to solve a single objective optimization model. The method considers the urban traffic network traffic, electric car driving mileage and location problem of an electric vehicle charging station which affects a distribution network. The example results of the method show that the planned charging station location can provide convenient and fast charging services for more electric cars, and further improve the safety and reliability of the system operation.

Description

A kind of electric automobile charging station planing method of considering transportation network flow

Technical field

The invention belongs to electric automobile charging station planning technology field, relate to the candidate site method of planning charging station by information such as transportation network and electric automobile during traveling mileages.

Background technology

Up to the present, the research aspect charging station planning is both at home and abroad also more preliminary, does not still form charging station plan model and method complete and system.Existing pertinent literature is most to be regarded common electricity consumption facility as and plans filling (changing) power station, and ignored it, provides the characteristic of public service.In fact, electric automobile charging station has two aspect attributes.First, charging station is a kind of urban transportation public service facility, and its function class is similar to refuelling station, and being all provides service for vehicle; Secondly, charging station is a kind of electricity consumption facility, and its planning construction will be considered on all many-sided impacts of distribution system, as electric power networks reconstruction, network loss, node voltage quality etc.Therefore, the present invention proposes in the situation that meets related constraint condition, the charging station of take can be more car owner's service of facilitating, and reduces the charging station planing method that is target to the negative effect of distribution system as far as possible.

Summary of the invention

Technical matters to be solved by this invention is distance travelled and the location problem on the electric automobile charging station of power distribution network impact of considering urban traffic network flow, electric automobile.

The present invention adopts following scheme to realize: a kind of electric automobile charging station planing method of considering transportation network flow, is characterized in that comprising following performing step:

Step S01: establishing target function;

Step S011: the calculating of the magnitude of traffic flow of capturing: its objective function f ₁and gravity-space interactive model expression formula is respectively suc as formula shown in (1) and (2):

$\max f_1=\left[\underset{q\∈Q}{\mathrm{\Σ}}{f}_q{y}_q\right]---\left(1\right)$

f _d＝W _OW _D/(d _p×1.5) (2)

In formula: q is that electric automobile arrives the shortest path between the D of destination from departure place O; Q is the complete or collected works of shortest path, and supposing has n node in traffic network system, considers symmetry, and does not consider loop, total n (n-1)/2 shortest path; f _qfor the magnitude of traffic flow on the q of path; W _oweight for path q starting point O; W _dweight for terminal D; d _qlength for path q; y _qrepresent that whether the magnitude of traffic flow on the q of path is captured, and if so, is 1, otherwise is 0;

Step S012: distribution system network loss: distribution system loss minimization target f ₂represent, its mathematical description is as follows:

$\min {\mathrm{<figure-callout\; id="2"\; label="f"\; filenames="130907145822.png"\; state="\{\{state\}\}">f</figure-callout>}}_2=\left[P_{\mathrm{Loss}}(U_{\mathrm{ik}}{P}_{\mathrm{Si}}^k,U_{\mathrm{ik}}{Q}_{\mathrm{Si}}^k)\right]---\left(3\right)$

In formula:

with

be respectively meritorious capacity and the reactive capability of the k type of charging station of building at i node: U _ikrepresent whether in node i, build k type of electric automobile charging station, 1 and 0 represents respectively build and do not build, P _lossfor distribution system Power loss corresponding under this construction scheme, injecting power and total load is poor, and constraint condition that can through type (13) is tried to achieve;

Step S013: node voltage skew: node voltage skew is designated as f ₃, it is for computing system node voltage and balance node voltage deviation, and computing formula is as follows:

$\mathrm{<figure-callout\; id="3"\; label="min\; f"\; filenames="130907145822.png"\; state="\{\{state\}\}">min</figure-callout>}{f}_{}{}_3=\left[\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{\mathrm{\&gamma;}}_i\left(\frac{|V_i-V_0|}{V_0}\right)\right]---\left(4\right)$

In formula: N is interstitial content in system; V ₀for balance node voltage; V _ivoltage for node i; γ _ifor weight factor, according to the importance of node, determine, define γ here _i=P _i/ P _load, think the importance of node and the positive correlation of this node load size, P _ifor the original loads of node i, P _loadfor distribution system total load;

Step S02: build constraint condition and comprise:

Node combination constraint $y_q=\left\{\begin{array}{cc}1& \underset{h\&Element;H_q}{\mathrm{\&Sigma;}}{a}_{\mathrm{qh}}{v}_h\&GreaterEqual;1\\ 0& \underset{h\&Element;H_q}{\mathrm{\&Sigma;}}{a}_{\mathrm{qh}}{v}_h=0\end{array}\right.---\left(7\right)$

In formula: a _qhrepresent that whether the node combination h on q is effective, value 1 and 0 represents respectively effectively and is invalid; v _hrepresent whether h combination is opened, value 1 and 0 represents respectively open and not open; H _qrepresent the upper all effective node combinations of circuit q;

Charging station on-load ability constraint: the charging station rated capacity in planning region should be not less than the maximum charge workload demand of electric automobile,

$\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}{P}_{\mathrm{Si}}^k-W_{\max }\&GreaterEqual;0---\left(8\right)$

In formula: W _maxthe maximum charge workload demand having calculated for having considered car owner's trip characteristics and electric automobile quantity, battery capacity factor; M is the kind of charging station;

Charging station is built number restriction: $\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}=M---\left(9\right)$

In formula: the charging station number of M for allowing to build;

In reality, a node can only be built at most a kind of type of charge station, that is: $\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}\&le;1---\left(10\right)$

The constraint of circuit transmitted power:

$P_l\&le;P_l^{\max }---\left(11\right)$

In formula: P _lthrough-put power for distribution line l; represent the through-put power upper limit that circuit l allows;

Adopt penalty function to process inequality constrain:

$K_L\left(P_l\right)=\left\{\begin{array}{cc}{K}_l{(P_l-P_{l\max })}^2,& P_l>P_{l\max }\\ 0,& P_l\&le;P_{l\max }\end{array}\right.---\left(12\right)$

In formula: K _lfor given abundant large constant;

Power-balance constraint:

$P_{\mathrm{SUB}}=P_{\mathrm{Loss}}+\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{Di}}+\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}{P}_{\mathrm{Si}}^k---\left(13\right)$

In formula: P _sUBfor balance node injecting power; P _dioriginal loads for node i;

System load flow equality constraint:

$\left\{\begin{array}{c}-P_{\mathrm{Di}}-U_{\mathrm{ik}}{P}_{\mathrm{Si}}^k=V_i\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{V}_j(G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}})\\ -Q_{\mathrm{Di}}-U_{\mathrm{ik}}{Q}_{\mathrm{Si}}^k=V_i\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{V}_j(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}})\end{array}\right.---\left(14\right)$

In formula: G _ijand B _ijbe respectively real part and the imaginary part of bus admittance matrix; θ _ijphase angle difference for branch road ij; P _diand Q _difor the original meritorious and load or burden without work of node i; V _ivoltage for node i;

Step S03: solving model;

Step S031: adopt Floyd algorithm to calculate all starting point O in transportation network and, to the shortest path q of terminal D, identify the node that path q passes through; Adopt gravity-space interactive model to calculate the magnitude of traffic flow on q;

Step S032: each of usining respectively in formula (1), (3) and (4) three objective functions, as single target function, is not considered two other objective function, asks for the optimum solution in these three kinds of situations;

Step S033: the one group of weight vectors that adopts pseudorandom number generator to produce, multi-objective optimization question is converted into single goal problem, adopt BPSO Algorithm for Solving;

Step S033: the optimized variable value of trying to achieve is brought in formula (1), (3), (4) objective function, obtain three desired values; Adopt super DEA to evaluate, minimize desired value as the input of decision package two, maximization desired value is as the output of decision package; According to final appraisal results, from one group of weight vectors, select a weight vectors the most effective as the final weight coefficient of each target;

Step S034: optimum addressing and the capacity of determining electric automobile charging station.

Can sum up following four aspects that mainly comprise of the present invention:

A. charging station candidate site determines.Generally speaking, important transport node (as railway station, bus station, large supermarket and large-scale residential block etc.) is also the important load node of electric system.Because charging station need to meet the charging needs in Vehicle Driving Cycle, be also larger load bus simultaneously, therefore the position candidate of charging station is set in to the coincidence node of transportation network and distribution system.

The calculating of the magnitude of traffic flow that b. charging pile is captured.Suppose when electric automobile is when the charging station, no matter battery dump energy level now, car owner can accept charging service until be full of.Consider automobile GPS navigation and car owner's driving experience, suppose that the shortest path of electric automobile car owner between can selecting from starting point to destination is as daily vehicle line here

C. objective function determines.The magnitude of traffic flow that the charging station of usining is captured, distribution system network loss and node voltage skew are set up Model for Multi-Objective Optimization as objective function, then adopt super efficiency DEA to determine each objective function weight, and multiple objective programming is converted to objective programming.

D. the present invention is model constrained is Constrained Nonlinear planning problem, adopts improved scale-of-two population intelligent optimization algorithm.

Numerical results of the present invention shows, the charging station addressing of planning provide convenient charging service to improve the safety and reliability of system operation can to more electric automobiles simultaneously.

Accompanying drawing explanation

Fig. 1 distribution system and transportation network node coincidence schematic diagram.

Fig. 2 is 33 node distribution systems and 25 junction traffic networks.

Embodiment

The present invention solves Optimized model that electric automobile charging station planning adopts and comprises determining of the determining of objective function, constraint condition and solve three steps.

Step 1: establishing target function

1. the calculating of the magnitude of traffic flow of capturing

Suppose when electric automobile is when the charging station, no matter battery dump energy level now, car owner can accept charging service until be full of.Consider automobile GPS navigation and car owner's driving experience, suppose that the shortest path of electric automobile car owner between can selecting from starting point to destination is as daily vehicle line here, shortest path and can obtaining by Floyd algorithm by node; The magnitude of traffic flow that on every paths, electric automobile during traveling produces can be tried to achieve by gravity-space interactive model.Objective function and gravity-space interactive model expression formula f ₁respectively suc as formula shown in (1) and (2):

$\max f_1=\left[\underset{q\&Element;Q}{\mathrm{\&Sigma;}}{f}_q{y}_q\right]---\left(1\right)$

f _q＝W _OW _D/(d _p×1.5) (2)

In formula: q is that electric automobile arrives the shortest path between the D of destination from departure place O; Q is the complete or collected works of shortest path, and supposing has n node in traffic network system, considers symmetry, and does not consider loop, total n (n-1)/2 shortest path; f _qfor the magnitude of traffic flow on the q of path; W _oweight for path q starting point O; W _dweight for terminal D; d _qlength for path q; y _qrepresent that whether the magnitude of traffic flow on the q of path is captured, and if so, is 1, otherwise is 0.

2. distribution system network loss

After electric automobile charging station access distribution system, the trend of system distributes and will change, and will cause that distribution system via net loss changes thereupon.Via net loss is not only relevant with original loads, also relevant with installation location and the amount of capacity of electric automobile charging station.Distribution system loss minimization target f ₂represent, its mathematical description is as follows:

$\mathrm{<figure-callout\; id="2"\; label="min\; f"\; filenames="130907145822.png"\; state="\{\{state\}\}">min</figure-callout>}{f}_{}{}_2=\left[P_{\mathrm{Loss}}(U_{\mathrm{ik}}{P}_{\mathrm{Si}}^k,U_{\mathrm{ik}}{Q}_{\mathrm{Si}}^k)\right]---\left(3\right)$

In formula:

with be respectively meritorious capacity and the reactive capability of the k type of charging station of building at i node; U _ikrepresent whether in node i, build k type of electric automobile charging station, 1 and 0 represents respectively build and do not build, P _lossfor distribution system Power loss corresponding under this construction scheme, injecting power and total load is poor, and constraint condition that can through type (13) is tried to achieve.

3. node voltage skew

The access of charging station will cause the node voltage profile in distribution system to change.Variation index f is proposed here ₃, for computing system node voltage and balance node voltage deviation.Computing formula is as follows:

$\mathrm{<figure-callout\; id="3"\; label="min\; f"\; filenames="130907145822.png"\; state="\{\{state\}\}">min</figure-callout>}{f}_{}{}_3=\left[\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{\mathrm{\&gamma;}}_i\left(\frac{|V_i-V_0|}{V_0}\right)\right]---\left(4\right)$

In formula: N is interstitial content in system; V ₀for balance node voltage; V _ivoltage for node i; γ _ifor weight factor, according to the importance of node, determine, define γ here _i=P _i/ P _load, think the importance of node and the positive correlation of this node load size, P _ifor the original loads of node i, P _loadfor distribution system total load.

4. multiple objective programming is converted to objective programming

Employing method is the mode to above-mentioned 3 objective function weighted sums, adopt super efficiency DEA to determine weight coefficient, super efficiency DEA is a kind of decision package relative efficiency that compares multiple-input and multiple-output by mathematical programming, and then decision package is carried out to the method for efficiency rating (evaluating decision package quality).Weight coefficient vector [α with objective function _d, β _d] be decision package, wherein α _dfor the weight coefficient vector of input quantity, β _dfor the weight coefficient vector of output quantity, under this weight vectors, optimize minimizing desired value and maximizing desired value respectively as the input and output of decision package of obtaining.For charging station Multiobjective Programming herein, its weight coefficient vector can be expressed as:

$[{\mathrm{\&alpha;}}_d,{\mathrm{\&beta;}}_d]=\left[\begin{array}{ccc}{\mathrm{\&alpha;}}_{11}& {\mathrm{\&alpha;}}_{12}& {\mathrm{\&beta;}}_{11}\\ {\mathrm{\&alpha;}}_{21}& {\mathrm{\&alpha;}}_{22}& {\mathrm{\&beta;}}_{21}\\ ...& ...& ...\\ {\mathrm{\&alpha;}}_{d1}& {\mathrm{\&alpha;}}_{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="130907145822.png"\; state="\{\{state\}\}">d</figure-callout>}2}& {\mathrm{\&beta;}}_{d1}\end{array}\right]---\left(5\right)$

The magnitude of traffic flow of capturing, distribution system via net loss and node voltage skew are made to normalized,

as the input quantity of decision package,

output quantity as decision package.Based on super efficiency DEA Model, evaluate this d decision package, adopt linear programming for solution.According to their relative efficiency E _ksize sequence, the corresponding decision package of maximal value is more effective with respect to other decision package, also under equivalent drops into, this decision package can have more output.

Choose E _kthe corresponding decision package of maximal value as normalization after the weight coefficient of each objective function.The single-goal function expression formula that adopts linear weighted function conversion to obtain is:

$\min F=[{\mathrm{\&alpha;}}_{k1}\frac{{\mathrm{<figure-callout\; id="2"\; label="f"\; filenames="130907145822.png"\; state="\{\{state\}\}">f</figure-callout>}}_2}{\min \left(f_2\right)}+{\mathrm{\&alpha;}}_{k2}\frac{{\mathrm{<figure-callout\; id="3"\; label="f"\; filenames="130907145822.png"\; state="\{\{state\}\}">f</figure-callout>}}_3}{\min \left(f_3\right)}+{\mathrm{\&beta;}}_{k1}\frac{\min (-f_1)}{-f_1}]---\left(6\right)$

Step 2: build constraint condition

1. node combination constraint

Suppose electric automobile in the process of moving, can not depart from its shortest path q, and be full of electricity after electric automobile maximum range be L.If starting point has charging station, residual running mileage is set to L; If do not had, can suppose that residual running mileage is 0.5L, reason is as follows: if residual running mileage is 0.5L, and can before exhausting, electric weight arrive from the nearest charging station charging of starting point, while returning, still can be full of electricity at this charging station, and turn back to after starting point, residual running mileage is not less than 0.5L.

Here provide following two definition: 1) if node combination h (all nodes that now hypothesis is included in combination h are the built charging station of establishing all) can provide charging service to the electric automobile on shortest path q; can be from arrival destination, departure place; and have sufficient electric weight to turn back to departure place, claim this to be combined as effective node combination; 2), when effectively node combines included all nodes and all builds charging station, claim that this effective node combination is open.Clearly, only have and on circuit q, at least have an effective node combination h, and this combination is open, the magnitude of traffic flow on circuit q is just captured.

$y_q=\left\{\begin{array}{cc}1& \underset{h\&Element;H_q}{\mathrm{\&Sigma;}}{a}_{\mathrm{qh}}{v}_h\&GreaterEqual;1\\ 0& \underset{h\&Element;H_q}{\mathrm{\&Sigma;}}{a}_{\mathrm{qh}}{v}_h=0\end{array}\right.---\left(7\right)$

In formula: a _qhrepresent that whether the node combination h on q is effective, value 1 and 0 represents respectively effectively and is invalid; v _hrepresent whether h combination is opened, value 1 and 0 represents respectively open and not open; H _qrepresent the upper all effective node combinations of circuit q.

2. charging station on-load ability constraint

In order to meet the electric automobile user demand of charging, the charging station rated capacity in planning region should be not less than the maximum charge workload demand of electric automobile.

$\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}{P}_{\mathrm{Si}}^k-W_{\max }\&GreaterEqual;0---\left(8\right)$

In formula: W _maxfor having considered car owner's trip characteristics, and the maximum charge workload demand that calculates of the factor such as electric automobile quantity, battery capacity; M is the kind of charging station.

3. charging station is built number restriction

Consider that urban land resource is generally more nervous, for fear of the caused waste of resources idle, should limit charging station number in planning region:

$\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}=M---\left(9\right)$

In formula: the charging station number of M for allowing to build.

4. in reality, a node can only be built at most a kind of type of charge station.

$\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}\&le;1---\left(10\right)$

5. circuit transmitted power constraint

$P_l\&le;P_l^{\max }---\left(11\right)$

In formula: P _lthrough-put power for distribution line l; represent the through-put power upper limit that circuit l allows.Adopt penalty function to process inequality constrain:

$K_L\left(P_l\right)=\left\{\begin{array}{cc}{K}_l{(P_l-P_{l\max })}^2,& P_l>P_{l\max }\\ 0,& P_l\&le;P_{l\max }\end{array}\right.---\left(12\right)$

In formula: K _lfor given abundant large constant.

6. power-balance constraint:

$P_{\mathrm{SUB}}=P_{\mathrm{Loss}}+\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{Di}}+\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}{P}_{\mathrm{Si}}^k---\left(13\right)$

In formula: P _sUBfor balance node injecting power; P _dioriginal loads for node i.

7. system load flow equality constraint

$\left\{\begin{array}{c}-P_{\mathrm{Di}}-U_{\mathrm{ik}}{P}_{\mathrm{Si}}^k=V_i\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{V}_j(G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}})\\ -Q_{\mathrm{Di}}-U_{\mathrm{ik}}{Q}_{\mathrm{Si}}^k=V_i\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{V}_j(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}})\end{array}\right.---\left(14\right)$

In formula: G _ijand B _ijbe respectively real part and the imaginary part of bus admittance matrix; θ _ijphase angle difference for branch road ij.P _diand Q _difor the original meritorious and load or burden without work of node i; V _ivoltage for node i;

Step 3: solving model

1. adopt Floyd algorithm to calculate all starting point O in transportation network and, to the shortest path q of terminal D, identify the node that path q passes through; Adopt gravity-space interactive model to calculate the magnitude of traffic flow on q.

2. each of usining respectively in aforesaid three objective functions, as single target function, is not considered two other objective function, asks for the optimum solution in these three kinds of situations.

4. the one group of weight vectors that adopts pseudorandom number generator to produce, is converted into single goal problem multi-objective optimization question, adopts BPSO Algorithm for Solving.

5. the optimized variable value of trying to achieve is brought in three objective functions, obtain three desired values.Adopt super DEA to evaluate, minimize desired value as the input of decision package two, maximization desired value is as the output of decision package; According to final appraisal results, from one group of weight vectors, select a weight vectors the most effective as the final weight coefficient of each target.

6). determine optimum addressing and the capacity of electric automobile charging station.

Example:

Application the present invention illustrates feasibility and the validity of institute's progressions model and method to 33 node distribution systems and 25 junction traffic networks.Here adopting following hypothesis: transportation network node 1-25 overlaps with distribution system node 1-25, is also that in distribution system, the real road between 1-25 node is communicated with situation as shown in the transportation network in Fig. 2.

Need explanation, the node not overlapping for distribution system and transportation network, transportation network node and distribution system node are not coupled on geographic position, and this of being equivalent to model of the present invention simplifies situation, and the algorithm proposing stands good.

Total 0.1MW, 0.2MW, 0.3MW and tetra-kinds of charging station candidate capacity of 0.4MW; Require the charging station number of building to be given as 4; Given electric automobile maximum range L is 120km; According to the busy extent of each transport node, node weights is set, as shown in table 1.

Table 1 transport node weight coefficient

Node	Weight	Node	Weight	Node	Weight
						1	0.54	10	0.54	19	0.80
2	0.80	11	0.05	20	0.27
						3	0.27	12	0.54	21	0.27
4	0.27	13	0.05	22	0.54
						5	0.27	14	0.54	23	0.05
6	0.07	15	0.27	24	1.34

[0111]

7	0.05	16	0.27	25	0.05
						8	0.54	17	0.27
9	0.27	18	1.07

In the situation that not considering two other objective function, optimize separately successively three targets, the magnitude of traffic flow maximal value that can to obtain loss minimization value and be 0.1907MW, node voltage skew minimum value be 0.0448 (perunit value), capture be 1.3947(account for total magnitude of traffic flow 45.8%).

The weight variation range of supposing each objective function is 0.1-0.8, and step-length is 0.1, amounts to like this 36 groups of weight vectors, adopts linear weighted function summation approach that decision-making problem of multi-objective is converted to single goal problem; Afterwards, utilize improvement BPSO to solve the single-object problem after conversion, and an optimized variable value of trying to achieve is brought in each objective function, obtain respectively the value of each objective function under different weight vectors.Finally, adopt super DEA to evaluate the validity of 36 groups of weight vectors, to determine the most effective one group of weight vectors.Result of calculation and the evaluation result of several groups of typical case's decision packages are listed in table 2, and the complete result of 36 groups of decision packages is shown in the Table A 1 in appendix.

The typical decision package of table 2 calculates and DEA evaluation result

As can be seen from Table 2, decision package 9 and 22 super efficiency DEA evaluation of estimate are all greater than 1, be Effective Decision-making Units, but the relative efficiency of No. 9 decision packages are higher, therefore adopt the corresponding weight coefficient 0.2,0.1 and 0.7 of decision package No. 9, respectively as objective function f ₁, f ₂and f ₃weight coefficient the most reasonable.Solving the optimal programming result that the single-object problem that forms after three objective function linear weighted functions obtains is: the node 18,24,8 and 14 in distribution system respectively construction capacity is 0.1,0.1,0.2 and the charging station of 0.4MW.From distribution system (No. 26 for balance node), can find out, charging station capacious build in distribution system feeder line compared with front end, can effectively reduce like this via net loss and variation; From Fig. 2, transportation network can be found out, charging station many places are in position, transport hub, and the magnitude of traffic flow of passing through is larger, and the charging station of planning provides convenient charging service can to more electric automobiles.

The foregoing is only preferred embodiment of the present invention, all equalizations of doing according to the present patent application the scope of the claims change and modify, and all should belong to covering scope of the present invention.

Claims (2)

1. an electric automobile charging station planing method of considering transportation network flow, is characterized in that comprising following performing step:

Step S01: establishing target function;

Step S011: the calculating of the magnitude of traffic flow of capturing: its objective function f ₁and gravity-space interactive model expression formula is respectively suc as formula shown in (1) and (2):

${\max f}_1=\left[\underset{q\&Element;Q}{\mathrm{\&Sigma;}}{f}_q{y}_q\right]---\left(1\right)$

f _q＝W _OW _D/(d _q×1.5) (2)

In formula: q is that electric automobile arrives the shortest path between the D of destination from departure place O; Q is the complete or collected works of shortest path, and supposing has n node in traffic network system, considers symmetry, and does not consider loop, total n (n-1)/2 shortest path; f _qfor the magnitude of traffic flow on the q of path; W _oweight for path q starting point O; W _dweight for terminal D; d _qlength for path q; y _qrepresent that whether the magnitude of traffic flow on the q of path is captured, and if so, is 1, otherwise is 0;

Step S012: distribution system network loss: distribution system loss minimization target f ₂represent, its mathematical description is as follows:

${\min f}_2=\left[P_{\mathrm{Loss}}(U_{\mathrm{ik}}{P}_{\mathrm{Si}}^k,U_{\mathrm{ik}}{Q}_{\mathrm{Si}}^k)\right]---\left(3\right)$

In formula:

with

be respectively meritorious capacity and the reactive capability of the k type of charging station of building at i node: U _ikrepresent whether in node i, build k type of electric automobile charging station, 1 and 0 represents respectively build and do not build, P _lossfor distribution system Power loss corresponding under this construction scheme, injecting power and total load is poor, and constraint condition that can through type (13) is tried to achieve;

Step S013: node voltage skew: node voltage side-play amount f ₃represent, it is for computing system node voltage and balance node voltage deviation, and computing formula is as follows:

${\min f}_3=\left[\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{\mathrm{\&gamma;}}_i\left(\frac{|V_i-V_0|}{V_0}\right)\right]---\left(4\right)$

In formula: N is interstitial content in system; V ₀for balance node voltage; V _ivoltage for node i; γ _ifor weight factor, according to the importance of node, determine, define γ here _i=P _i/ P _load, think the importance of node and the positive correlation of this node load size, P _ifor the original loads of node i, P _loadfor distribution system total load;

Step S02: build constraint condition and comprise:

Node combination constraint $y_q=\left\{\begin{array}{cc}1& \underset{{h\&Element;H}_q}{\mathrm{\&Sigma;}}{a}_{\mathrm{qh}}{v}_h\&GreaterEqual;1\\ 0& \underset{{h\&Element;H}_q}{\mathrm{\&Sigma;}}{a}_{\mathrm{qh}}{v}_h=0\end{array}\right.---\left(7\right)$

In formula: a _qhrepresent that whether the node combination h on q is effective, value 1 and 0 represents respectively effectively and is invalid; v _hrepresent whether h combination is opened, value 1 and 0 represents respectively open and not open; H _qrepresent the upper all effective node combinations of circuit q;

Charging station on-load ability constraint: the charging station rated capacity in planning region should be not less than the maximum charge workload demand of electric automobile,

$\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}{P}_{\mathrm{Si}}^k-W_{\max }\&GreaterEqual;0---\left(8\right)$

In formula: W _maxthe maximum charge workload demand having calculated for having considered car owner's trip characteristics and electric automobile quantity, battery capacity factor; M is the kind of charging station;

Charging station is built number restriction: $\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}=M---\left(9\right)$

In formula: the charging station number of M for allowing to build; In reality, a node can only be built at most a kind of type of charge station, that is: $\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}\&le;1---\left(10\right)$

The constraint of circuit transmitted power:

$P_l\&le;{P_l}^{\max }---\left(11\right)$

In formula: P _lthrough-put power for distribution line 1; P _l ^maxrepresent the through-put power upper limit that circuit 1 allows;

Adopt penalty function K _l(P _l) processing inequality constrain:

$K_L\left(P_l\right)=\left\{\begin{array}{cc}{K}_l{(P_l-P_{l\max })}^2,& P_l>P_{l\max }\\ 0,& P_l\&le;P_{l\max }\end{array}---\left(12\right)\right.$

In formula: K _lfor given abundant large constant;

Power-balance constraint:

$P_{\mathrm{SUB}}=P_{\mathrm{Loss}}+\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{Di}}+\underset{i=1}{\overset{N-1}{\mathrm{\&Sigma;}}}{U}_{\mathrm{ik}}{P}_{\mathrm{Si}}^k---\left(13\right)$

In formula: P _sUBfor balance node injecting power; P _dioriginal loads for node i;

System load flow equality constraint:

$\left\{\begin{array}{c}-P_{\mathrm{Di}}-U_{\mathrm{ik}}{P}_{\mathrm{Si}}^k=V_i\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{V}_j(G_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}}+B_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}})\\ -Q_{\mathrm{Di}}-U_{\mathrm{ik}}{Q}_{\mathrm{Si}}^k=V_i\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{V}_j(G_{\mathrm{ij}}\sin {\mathrm{\&theta;}}_{\mathrm{ij}}-B_{\mathrm{ij}}\cos {\mathrm{\&theta;}}_{\mathrm{ij}})\end{array}\right.---\left(14\right)$

In formula: G _ijand B _ijbe respectively real part and the imaginary part of bus admittance matrix; θ _ijphase angle difference for branch road ij; P _diand Q _difor the original meritorious and load or burden without work of node i; V _ivoltage for node i;

Step S03: solving model;

Step S031: adopt Floyd algorithm to calculate all starting point O in transportation network and, to the shortest path q of terminal D, identify the node that path q passes through; Adopt gravity-space interactive model to calculate the magnitude of traffic flow on q;

Step S032: each of usining respectively in formula (1), (3) and (4) three objective functions, as single target function, is not considered two other objective function, asks for the optimum solution in these three kinds of situations;

Step S033: the one group of weight vectors that adopts pseudorandom number generator to produce, multi-objective optimization question is converted into single goal problem, adopt BPSO Algorithm for Solving;

Step S033: the optimized variable value of trying to achieve is brought in formula (1), (3), (4) objective function, obtain three desired values; Adopt super DEA to evaluate, minimize desired value as the input of decision package two, maximization desired value is as the output of decision package; According to final appraisal results, from one group of weight vectors, select a weight vectors the most effective as the final weight coefficient of each target;

Step S034: optimum addressing and the capacity of determining electric automobile charging station.

2. the electric automobile charging station planing method of consideration transportation network flow according to claim 1, is characterized in that: in described step S033 with the weight coefficient vector [α of objective function _d, β _d] be decision package, wherein α _dfor the weight coefficient vector of input quantity, β _dfor the weight coefficient vector of output quantity, under this weight vectors, optimize minimizing desired value and maximizing desired value respectively as the input and output of decision package of obtaining; Its weight coefficient vector can be expressed as:

$[{\mathrm{\&alpha;}}_d,{\mathrm{\&beta;}}_d]=\left[\begin{array}{ccc}{\mathrm{\&alpha;}}_{11}& {\mathrm{\&alpha;}}_{12}& {\mathrm{\&beta;}}_{11}\\ {\mathrm{\&alpha;}}_{21}& {\mathrm{\&alpha;}}_{22}& {\mathrm{\&beta;}}_{21}\\ \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;\\ {\mathrm{\&alpha;}}_{d1}& {\mathrm{\&alpha;}}_{d2}& {\mathrm{\&beta;}}_{d1}\end{array}\right]---\left(5\right)$

The magnitude of traffic flow of capturing, distribution system via net loss and node voltage skew are made to normalized,

as the input quantity of decision package,

output quantity as decision package; Based on super efficiency DEA Model, evaluate this d decision package, adopt linear programming for solution;

Choose relative efficiency E _kthe corresponding decision package of maximal value as normalization after the weight coefficient of each objective function; Adopt linear weighted function to transform the weighting single goal F obtaining, its function expression is:

$\min F=[{\mathrm{\&alpha;}}_{k1}\frac{f_2}{\min \left(f_2\right)}+{\mathrm{\&alpha;}}_{k2}\frac{f_3}{\min \left(f_3\right)}+{\mathrm{\&beta;}}_{k1}\frac{\min (-f_1)}{-f_1}]---\left(6\right).$

Images