CN109034563A - A kind of increment power distribution network source net lotus collaborative planning method of multi-agent Game - Google Patents
A kind of increment power distribution network source net lotus collaborative planning method of multi-agent Game Download PDFInfo
- Publication number
- CN109034563A CN109034563A CN201810745212.0A CN201810745212A CN109034563A CN 109034563 A CN109034563 A CN 109034563A CN 201810745212 A CN201810745212 A CN 201810745212A CN 109034563 A CN109034563 A CN 109034563A
- Authority
- CN
- China
- Prior art keywords
- power
- game
- load
- investment
- node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The present invention relates to a kind of increment power distribution network source net lotus collaborative planning methods of multi-agent Game, step: establish the Decision Model of multiple investment subjects respectively；According to the Static Game behavior between the transitive relation of three investment subjects analysis DG investment operation quotient and distribution company, game reaches equilibrium state, establishes static game model based on this；The uncertainty contributed using robust optimization processing DG, and virtual game person " the Nature " is introduced, according to its dynamic game behavior between distribution company, game reaches equilibrium state, establishes Dynamic Game Model based on this；Consider DG investment operation quotient, distribution company, power consumer and " the Nature " multi-agent Game, forms the static-dynamic state joint game pattern towards increment distribution network planning, establish static-dynamic state joint game model based on this；Using iterative search method Solving Nash Equilibrium point, the Decision Model based on theory of games is solved, final programme is obtained.
Description
Technical field
The present invention relates to a kind of Power System Planning research fields, especially with regard to a kind of consideration uncertainty and multiagent
The increment power distribution network source net lotus collaborative planning method of game.
Background technique
With moving forward steadily for increment distribution business pilot reform work, increment distribution business starts to open social capital
It puts.On the one hand, the distributed generation resource investor, participate in Demand Side Response power consumer initially as independent subject participate in power distribution network
Investment and operation so that the diversification of investment subjects becomes one of most significant feature of China's increment power distribution network；On the other hand, divide
Cloth power supply accesses on a large scale is filled with more uncertain factors to increment power distribution network.In this context, research considers more
Interest subject and probabilistic increment distribution network planning method have important theoretical and practical significance.
In fact, in increment power distribution network, different subjects in planning often using number one as starting point, it is less to examine
Consider the overall interests in whole market, i.e., is all individual rationality.These investment subject independent of each other meetings in decision process
By constantly analyzing the planning strategy of other main bodys, to adjust the strategy of itself, to realize the maximization of number one.It can
See, the above-mentioned plan model based on individual rationality has the following problems: 1) due to having ignored between autonomous investment main body
Game Relationship, this planing method can not embody the operating mechanism in practical increment power distribution network market, thus programmed decision-making is accurate
Property and validity is not high；2) above-mentioned planing method only carries out decision from the angle of global optimum, cannot be considered in terms of each in market
The Interest demands of a investment subject are possible to the interests of damage Individual Investors while pursuing overall income highest, thus
The market vitality is reduced, the development of increment power distribution network is restricted.
Therefore, on the basis of game mechanism between studying independent subject, the increment distribution network planning based on individual rationality is constructed
Model is drawn, is the effective thinking for more meeting market mechanism to seek the programmed decision-making Nash Equilibrium point of multiagent participation.
Summary of the invention
In view of the above-mentioned problems, the object of the present invention is to provide a kind of collaborations of the increment power distribution network source net lotus of multi-agent Game to advise
The method of drawing realizes the depth integration of theory of games and robust optimization by introducing virtual game person's " the Nature ".
To achieve the above object, the present invention takes following technical scheme: a kind of increment power distribution network source net of multi-agent Game
Lotus collaborative planning method, which comprises the following steps: 1) Decision Model of multiple investment subjects is respectively established,
Investment subject includes DG investment operation quotient, distribution company and the power consumer for participating in DSR；2) according to the transmitting of three investment subjects
Static Game behavior between relationship analysis DG investment operation quotient and distribution company, in gambling process as DG investment operation quotient and
When distribution company either side change strategy can not all obtain more incomes, game reaches equilibrium state, builds based on this
Vertical static game model；3) using the uncertainty of robust optimization processing DG power output, and virtual game person " the Nature ", root are introduced
According to its dynamic game behavior between distribution company, when " the Nature " and distribution company either side change in gambling process
Strategy can not all obtain when more preferably paying, and game reaches equilibrium state, establishes Dynamic Game Model based on this；4) consider
DG investment operation quotient, distribution company, power consumer and " the Nature " multi-agent Game are formed towards increment distribution network planning
Static-dynamic state joint game pattern: Static Game is constituted between distribution company and DG investment operation quotient, while between " the Nature "
Dynamic game is constituted, establishes static-dynamic state joint game model based on this；5) iterative search method Solving Nash Equilibrium is used
Point solves the Decision Model based on static-dynamic state joint game model, obtains final programme.
Further, in the step 1), the Decision Model of DG investment operation quotient is responsible for advising distributed generation resource
It draws, target is self benefits maximization, and decision variable is position and the capacity of distributed generation resource；
Objective function are as follows:
In formula,For DG sale of electricity income,For DG cost of investment,For DG O&M cost,It is government for can
The power generation of the renewable sources of energy is subsidized；
Constraint condition includes DG node access number limitation to be selected, the constraint of DG permeability and DG units limits:
1. DG node access number limitation to be selected:
Ni.min≤Ni≤Ni.max
In formula, Ni.minFor the lower limit value for accessing DG number in node i to be selected, Ni.maxTo access DG number in node i to be selected
Upper limit value；NiThe number of units of DG is accessed for node i to be selected；
2. DG permeability constrains:
In formula, xiFor 0-1 variable, xi=0 i-th of node to be selected of expression does not access DG, xi=1 indicates i-th of node to be selected
Access DG；For the rated power of separate unit DG；δ is the permeability after DG is grid-connected, PloadFor node total load；
3. DG units limits:
In formula,For DG t moment total active power output；For DG power output lower limit value,For the upper limit of DG power output
Value.
Further, in the step 1), the Decision Model of distribution company plans power grid, and target is itself receipts
Benefit maximizes, and decision variable is route newly-built project；
Objective function are as follows:
In formula,For distribution company sale of electricity income,For new route cost of investment,For Web-based exercise,For
Failure cost,Based on net purchases strategies,For to DG investment operation quotient's purchases strategies；
Constraint condition includes new route investment and recovery, Branch Power Flow constraint and security constraint:
1. new route investment and recovery
In formula, ΩkTo increase load bus set newly；yj,kTo be k node j-th strip new route；
2. Branch Power Flow constrains
In formula, Pi.tFor the active power of t moment node i, Qi.tFor the reactive power of t moment node i；Ui.tFor t moment section
The voltage magnitude of point i, Uj.tFor the voltage magnitude of t moment node j；GijFor the conductance of branch ij, BijFor the susceptance of branch ij；θij
For the phase angle difference between node i and node j voltage；
3. security constraint
In formula, Ui.minFor the lower limit value of node i voltage magnitude, Ui.maxFor the upper limit value of node i voltage magnitude；Pij.tFor branch
The transimission power of road ij, Pij.maxFor the upper limit value of the transimission power of branch ij.
Further, in the step 1), the objective function of the Decision Model of power consumer are as follows:
In formula,For the interrupt power of t moment interruptible load；The power being transferred out of for t moment load；When for t
Carve the power that load is transferred into；For the electric cost expenditure of reduction after participation Demand Side Response；For interruptible load compensation
With；
Constraint condition includes transfer load power constraint and interruptible load power constraint:
1. transfer load power constraint
In formula, λminThe lower limit value of power coefficient, λ are transferred out of for t moment loadmaxPower train is transferred out of for t moment load
Several upper limit values；μminThe lower limit value of power coefficient, μ are transferred into for t moment loadmaxPower coefficient is transferred into for t moment load
Upper limit value；
2. interruptible load power constraint
In formula,For load can interrupt power lower limit value,For load can interrupt power upper limit value.
Further, in the step 2), Static Game action process are as follows: in a game bout, power consumer is being connect
The power of transfer load and interruptible load is determined when contracture after electricity price information and interruptible load excitation information, and with equivalent negative
The form of lotus feeds back to distribution company；Distribution company is according to last round of DG investment operation quotient to the decision and electric power of DG addressing constant volume
The information on load of user feedback makes distribution total benefit maximum by adjusting the newly-built project of route, while DG investment operation quotient
According to last round of distribution company to the decision of new route, the investment operation of DG is made by adjusting the addressing of DG and constant volume scheme
Benefit is maximum；After updating the addressing constant volume of network topology and DG, into next game bout.
Further, in the step 3), dynamic game process are as follows: " the Nature " will be directed to current electric grid topological structure first
Distributed generation resource power output is constantly adjusted in indeterminacy section, the power loss level and via net loss, minimum for increasing power distribution network are matched
Net total benefit；Thereafter, it based on distribution company most badly will plan scene caused by " the Nature ", is opened up by optimizing network
It flutters, maximizes distribution total benefit, network topology is then updated, into next game bout.
Further, in the step 4), game process are as follows: distribution company receives the equivalent load of power consumer feedback, root
The on-position of the DG gone out according to decision in last round of game and capacity formulate route after the uncertainty for considering DG power output
Newly-built project；Network topology structure after the robust optimization that DG investment operation quotient goes out according to decision in last round of game simultaneously carries out
The addressing constant volume of DG；After updating the addressing constant volume scheme of network topology and DG, into next game bout.
Further, the static-dynamic state joint game model are as follows:
Strategy combination is optimum programming scheme under equilibrium state, and the program had both considered the maximization of benefits of different parties, together
When there is preferable robustness again.
Further, the transitive relation of three investment subjects are as follows: introduce " the Nature " and be used as corresponding virtual subject, DG
Investment operation quotient carries out the addressing constant volume of DG under current grid structure, and by the position of DG and capacity pass to distribution company and
" the Nature "；Power consumer formulates active response after receiving tou power price information and interruptible load excitation information from distribution company
Measure, that is, determine the power of transfer load and interruptible load, and distribution company is fed back in the form of equivalent load；" greatly certainly
Interference so " is made to the planning of distribution company in the case where knowing that DG layouts and combines current grid structure, " decision " goes out DG
It contributes and passes it to distribution company；Distribution company receives the transmitting information from other main bodys, and decision goes out new route, shape
The grid structure of Cheng Xin.
Further, in the step 5), 5.1) model solution method is the following steps are included: input initial data and parameter:
Parameter necessary to data needed for betting model is established in initialization and calculating participant's income；5.2) game participant strategy is generated
Space: the policy space of DG investment operation quotient is set f (x, N)={ f of DG node state waiting1,f2,…,fn, wherein fn
Indicate the accessible DG capacity of node n；The policy space of distribution company is the set y={ y of route yet to be built1,y2,…,yn,
Middle ynEach element is the selectable path of route；5.3) initialization: empty in the strategy of DG investment operation quotient and distribution company
Between in randomly select a class value f respectively0And y0, and according to the policy space for determining " the Nature " the case where value, as at the beginning of iteration
Value；5.4) each participant carries out independent optimization: in the n-th wheel gambling process, power consumer turns to target with number one maximum
Obtain optimal transfer power and can interrupt power, and result is fed back into distribution company；DG investment operation quotient matches according to the (n-1)th wheel
Grid structure under net company strategy obtains optimal DG access strategy by target of DG investment operation maximizing the benefits, and thus
Determine the policy space of " the Nature "；5.5) judge whether to reach balanced, if (fn,yn)=(fn-1,yn-1), then reach balanced, it is defeated
Programme (f out*,y*)；Conversely, then return step 5.4).
The invention adopts the above technical scheme, which has the following advantages: 1, the present invention passes through in accurate simulation market
The game behavior of main body, it is ensured that each main market players continues to optimize itself decision in gambling process, itself is received with realizing
The maximization of benefit, and promote the validity of the market vitality and programmed decision-making.2, the present invention is by introducing virtual game person " greatly certainly
So ", it can fully consider influence of the uncertain factor to programmed decision-making in the plan model based on theory of games, pass through
Active optimization planning decision carrys out lifting system income.
Detailed description of the invention
Fig. 1 is transitive relation figure between each investment subject；
Fig. 2 is Static Game flow chart；
Fig. 3 is dynamic game flow chart；
Fig. 4 is static-dynamic state joint game flow chart；
Fig. 5 is that the plan model based on theory of games solves flow chart；
Fig. 6 is electric system IEEE33 node distribution network systems figure；
Fig. 7 is that daily load curve compares figure.
Specific embodiment
The present invention is described in detail below with reference to the accompanying drawings and embodiments.
The present invention provides a kind of increment power distribution network source net lotus collaborative planning method of multi-agent Game comprising following step
It is rapid:
1) Decision Model of multiple investment subjects is established respectively；Investment subject includes DG (distributed generation resource) investment fortune
Seek the power consumer of quotient, distribution company and participation DSR (Demand Side Response).
It is different as the planning problem of single main body from tradition using distribution company, multiple investment subjects involved in the present invention and each
The mutual Interest demands of main body are not identical.For DG investment operation quotient, it is desirable to reduce DG investment construction and operation at
This, increases power selling income, to make maximum revenue；For distribution company, it would be desirable that the costs such as network loss, investment, power purchase are reduced,
Increase power selling income, so that number one be made to maximize；And for power consumer, then it is desirable to by adjusting electricity consumption row
To reduce electric cost expenditure.Different investment subjects target when participating in planning is biased to different, decision independent of one another, it is therefore desirable to point
Not Gou Jian above three investment subject plan model.
1.1) Decision Model of DG investment operation quotient:
DG investment operation quotient is mainly responsible in increment distribution network planning and plans distributed generation resource, and target is itself
Maximum revenue, decision variable are position and the capacity of distributed generation resource；In this example, it is assumed that distributed generation resource is photovoltaic
Power generation.
1.1.1) objective function:
The objective function of the Decision Model of DG investment operation quotient mainly includes DG sale of electricity incomeDG cost of investmentAnd DG O&M costIn addition to this, since increment power distribution network detailed rules for the implementation encourage the access of distributed generation resource, because
This is additionally contemplates that government subsidizes the power generation of renewable energyObjective function are as follows:
Wherein:
In formula, θesFor DG investment operation quotient's unit sale of electricity electricity price；For DG t moment total active power output；θgcFor can
Renewable sources of energy unit power generation subsidy expense；θsgFor unit capacity DG cost of investment；xiFor 0-1 variable, xi=0 indicate i-th to
Node is selected not access DG, xi=1 indicates that i-th of node to be selected accesses DG；For the rated power of separate unit DG；NiFor node to be selected
The number of units of i access DG；R is discount rate；LT is the life cycle of equipment；θomFor DG unit power generation O&M expense；TtFor period t's
Runing time；ΩtFor period sum；ΩiFor DG node road set yet to be built.
1.1.2) constraint condition: the constraint condition of the Decision Model of DG investment operation quotient includes DG node access to be selected
Number limitation, the constraint of DG permeability and DG units limits.
1. DG node access number limitation to be selected:
Ni.min≤Ni≤Ni.max (6)
In formula, Ni.minFor the lower limit value for accessing DG number in node i to be selected, Ni.maxTo access DG number in node i to be selected
Upper limit value.
2. DG permeability constrains:
In formula, δ is the permeability after DG is grid-connected, PloadFor node total load.
3. DG units limits:
In formula,For DG power output lower limit value,For the upper limit value of DG power output.
1.2) Decision Model of distribution company: distribution company mainly advises power grid in increment distribution network planning
It draws, target is self benefits maximization, and decision variable is route newly-built project.
1.2.1) objective function:
The objective function of the Decision Model of distribution company includes distribution company sale of electricity incomeNew route investment
CostWeb-based exerciseFailure costMajor network purchases strategiesAnd to DG investment operation quotient's purchases strategiesObjective function are as follows:
Wherein:
In formula, ψesFor the sale of electricity electricity price of distribution company；For the original loads of t moment；For t moment interruptible load
Interrupt power；The power being transferred out of for t moment load；The power being transferred into for t moment load；ψsgIndicate newly-built line
Road unit length expense；yjFor 0-1 variable, yj=0 expression j-th strip waits for that new route is not selected, yj=1 expression j-th strip waits for newly
It is selected to build route；ljFor the length of new route；For the active power loss of t moment；EENStFor t moment electricity not
Sufficient desired value；λbFor the failure rate of the b articles route；ψeb1For superior power grid purchase electricity price；ψeb2For to commercially available from DG investment operation
Electricity price；T is plant life cycle；ΩbFor route sum；ΩjFor line set yet to be built.
1.2.2) constraint condition: the constraint condition of the Decision Model of distribution company include new route investment and recovery,
Branch Power Flow constraint and security constraint.
1. new route investment and recovery
In formula, ΩkTo increase load bus set newly；yj,kTo be k node j-th strip new route；
2. Branch Power Flow constrains
In formula, Pi.tFor the active power of t moment node i, Qi.tFor the reactive power of t moment node i；Ui.tFor t moment section
The voltage magnitude of point i, Uj.tFor the voltage magnitude of t moment node j；GijFor the conductance of branch ij, BijFor the susceptance of branch ij；θij
For the phase angle difference between node i and node j voltage.
3. security constraint
In formula, Ui.minFor the lower limit value of node i voltage magnitude, Ui.maxFor the upper limit value of node i voltage magnitude；Pij.tFor branch
The transimission power of road ij, Pij.maxFor the upper limit value of the transimission power of branch ij.
1.3) Decision Model of power consumer: power consumer is rung in increment distribution network planning by participating in Demand-side
It answers, adjusts electricity consumption behavior to reduce electric cost expenditure.The present invention considers two kinds of Demand Side Response modes: the price based on tou power price
The stimulable type DSR of type DSR and interruptible load.The user of the price type DSR based on tou power price is participated in electricity price peak period
It is transferred out of load, is transferred into load in electricity price low ebb period；Participate in interruptible load stimulable type DSR user then by with electricity
Net company signs a contract, and interrupts in certain periods and reduction plans obtain corresponding compensation simultaneously.
1.3.1) objective function
The objective function of the Decision Model of power consumer mainly includes the electricity charge branch of reduction after participating in Demand Side Response
OutAnd interruptible load reimbursement for expensesObjective function are as follows:
Wherein:
In formula, ωebFor the purchase electricity price of user；ωgcFor interruptible load reimbursement for expenses.
1.3.2) constraint condition
Side response mode mainly includes transfer load power constraint to the constraint condition of power consumer plan model according to demand
And interruptible load power constraint.
1. transfer load power constraint
In formula, λminThe lower limit value of power coefficient, λ are transferred out of for t moment loadmaxPower train is transferred out of for t moment load
Several upper limit values；μminThe lower limit value of power coefficient, μ are transferred into for t moment loadmaxPower coefficient is transferred into for t moment load
Upper limit value.
2. interruptible load power constraint
In formula,For load can interrupt power lower limit value,For load can interrupt power upper limit value.
2) according to the Static Game between the transitive relation of three investment subjects analysis DG investment operation quotient and distribution company
Behavior can not all obtain more incomes when DG investment operation quotient and distribution company either side change strategy in gambling process
When, game reaches equilibrium state, establishes static game model based on this；
Since planning investment subject of the invention is DG investment operation quotient, distribution company and power consumer.Furthermore it is distributed
After formula plant-grid connection, power output uncertainty will influence whether the operational safety of power distribution network, and can make relevant cost increase from
And reduce the economy of programme.Therefore consider that the power output of distributed generation resource, which is considered as special decision variable, to be used to characterize it
Uncertainty, and introduce " the Nature " and be used as corresponding virtual subject.
As shown in Figure 1, the transitive relation of three investment subjects are as follows: introduce " the Nature " and be used as corresponding virtual subject, DG
Investment operation quotient carries out the addressing constant volume of DG under current grid structure, and by the position of DG and capacity pass to distribution company and
" the Nature "；Power consumer formulates active response after receiving tou power price information and interruptible load excitation information from distribution company
Measure, that is, determine the power of transfer load and interruptible load, and distribution company is fed back in the form of equivalent load；" greatly certainly
Interference so " is made to the planning of distribution company in the case where knowing that DG layouts and combines current grid structure, " decision " goes out DG
It contributes and passes it to distribution company；Distribution company receives the transmitting information from other main bodys, and decision goes out new route, shape
The grid structure of Cheng Xin.
By the transitive relation between each main body it is found that power consumer only obtains tou power price from distribution company and can interrupt sharp
Information is encouraged, is not directly affected by its decision, therefore when analyzing the game behavior of each main body, power consumer is not regarded
For the participant of game.
Static game model are as follows:
Due to needing to complete the planning and construction of increment power distribution network, DG investment operation quotient jointly under the premise of independent decision-making
Grasp whole policy informations of other side mutually in planning process with distribution company, and the two makes decisions simultaneously, is not present
Sequencing in the action, therefore static game of complete information pattern is formed between DG investment operation quotient and distribution company.Such as Fig. 2
Shown, Static Game behavior is as follows:
In a game bout, power consumer determines after receiving tou power price information and interruptible load excitation information
The power of transfer load and interruptible load, and distribution company is fed back in the form of equivalent load.Distribution company is according to upper one
DG investment operation quotient is taken turns to the information on load of decision and the power consumer feedback of DG addressing constant volume, by adjusting the newly-built side of route
Case makes distribution total benefit maximum, at the same DG investment operation quotient according to last round of distribution company to the decision of new route, pass through
The addressing and constant volume scheme for adjusting DG make the investment operation benefit of DG maximum.In the addressing constant volume for updating network topology and DG
Afterwards, into next game bout.
It can not all be obtained more in gambling process when DG investment operation quotient and distribution company either side change strategy
When income, game reaches equilibrium state, then static game model are as follows:
In formula, f*For the planning strategy of the DG investment operation quotient under equilibrium state, y*For the distribution company under equilibrium state
Planning strategy；Variables collection when argmax () is maximized for objective function；F is DG investment operation quotient's planning strategy；Y is
Distribution company programming strategy.
3) using the uncertainty of robust optimization processing DG power output, and introduce virtual game person " the Nature ", according to its with
Dynamic game behavior between distribution company, when " the Nature " and distribution company either side change strategy all in gambling process
It can not obtain when more preferably paying, game reaches equilibrium state, establishes Dynamic Game Model based on this；
For the distribution network planning problem containing distributed generation resource, fluctuated in a certain range for distributed generation resource power output
Bring interference, it is always desirable to design optimal strategy so that the cost allowance or operation risk that can suffer from reach minimum, maximum
Inhibit to degree adverse effect caused by uncertainty, the thought that this process optimizes with robust is mutually agreed with.Therefore, the present invention uses
The uncertainty of the method processing DG power output of robust optimization, passes through the fluctuation range of indeterminacy section description power output, the fluctuation model
It encloses and is determined by the installed capacity of DG:
In formula, αminThe lower limit value of DG capacity ratio, α are accounted for for DG power outputmaxThe upper limit value of DG capacity ratio is accounted for for DG power output.
In above-mentioned robust optimization problem, the uncertainty of DG power output will be such that the income of power distribution network reduces, and rack
Planning is then to turn to target with the Income Maximum of power distribution network, from game angle, two policymaker's " the Nature " and distribution company
Zero-sum game relationship is constituted, thus the game that robust optimization problem can be converted between " the Nature " and distribution company is asked
Topic.The game person as " the Nature ", best reply means are the plans first observed its worst interference, then construct reply,
The dynamic game process that the two action in game bout has sequencing is consequently formed, as shown in Figure 3:
In a game bout, " the Nature " will be continuous in indeterminacy section for current electric grid topological structure first
Distributed generation resource power output is adjusted, increases the power loss level and via net loss of power distribution network, minimizes distribution total benefit.Thereafter, distribution
Based on company most badly will plan scene caused by " the Nature ", by optimizing network topology, maximizes distribution and always imitate
Then benefit updates network topology, into next game bout.
It all can not more preferably be paid in gambling process when " the Nature " and distribution company either side change strategy
When, game reaches equilibrium state, Dynamic Game Model are as follows:
In formula, P*For the strategy of " the Nature " under equilibrium state；Change when argmin () is minimized for objective function
Duration set；P is the strategy of " the Nature ".
4) consider the multi-agent Games such as DG investment operation quotient, distribution company, power consumer and " the Nature ", formed towards
Static Game is constituted between the static-dynamic state joint game pattern of increment distribution network planning, i.e. distribution company and DG investment operation quotient,
Dynamic game is constituted between " the Nature " simultaneously, establishes static-dynamic state joint game model based on this；
As shown in figure 4, game process are as follows:
In a game bout, distribution company receives the equivalent load of power consumer feedback, according in last round of game
The on-position for the DG that decision goes out and capacity formulate the newly-built project of route after the uncertainty for considering DG power output；DG simultaneously
Network topology structure after the robust optimization that investment operation quotient goes out according to decision in last round of game carries out the addressing constant volume of DG.?
After updating network topology and the addressing constant volume scheme of DG, into next game bout.
Finally formed static-dynamic state joint game model are as follows:
Strategy combination is optimum programming scheme under equilibrium state, and the program had both considered the maximization of benefits of different parties, together
When there is preferable robustness again.
5) iterative search method Solving Nash Equilibrium point is used, (is based on static-dynamic state joint game mould to based on theory of games
Type) Decision Model solved, obtain final programme.
Planning problem under multi-agent Game environment is not a globality optimization problem, but each participant is based on respectively
From multiple independent optimization problems of target.Game equilibrium point is solved there are many method at present, such as iterative search method, is inversely returned
Nanofarad, maximum-minimum optimization, sequences and rejecting disadvantage Strategies Method etc..To above-mentioned game planning problem, the present invention is adopted
With iterative search method Solving Nash Equilibrium point.Solve process
As shown in figure 5, in conjunction with increment distribution network planning problem, model solution method the following steps are included:
5.1) input initial data and parameter: data needed for betting model is established in initialization and calculating participant's income must
The parameter needed；
Wherein, required data include DG relevant parameter (such as DG unit capacity investment cost, DG separate unit rated capacity,
DG unit sale of electricity electricity price, DG unit power generation O&M expense and DG power generation government subsidy etc.) and newly-increased load point and possibility access digit
It sets；
Parameter necessary to participant's income includes distribution relevant parameter, such as new route unit length expense, distribution public affairs
Take charge of sale of electricity electricity price and to major network purchase electricity price etc..
5.2) generate game participant policy space: the policy space of DG investment operation quotient is the collection of DG node state waiting
Close f (x, N)={ f1,f2,…,fn, wherein fnIndicate the accessible DG capacity of node n；The policy space of distribution company be to
Build the set y={ y of route1,y2,…,yn, wherein ynEach element is the selectable path of route.
5.3) initialization: a class value f is randomly selected respectively in the policy space of DG investment operation quotient and distribution company0
And y0, and according to the policy space for determining " the Nature " the case where value, as iterative initial value.
5.4) each participant carries out independent optimization: in the n-th wheel gambling process, power consumer is turned to number one maximum
Target obtain optimal transfer power and can interrupt power, and result is fed back into distribution company.DG investment operation quotient is according to (n-1)th
The grid structure under distribution company strategy is taken turns, obtains optimal DG access strategy by target of DG investment operation maximizing the benefits, and
Thereby determine that the policy space of " the Nature ".
5.5) judge whether to reach balanced, if (fn,yn)=(fn-1,yn-1), then reach balanced, exports programme (f*,
y*)；Conversely, then return step 5.4).
Embodiment:
1, parameter setting
The present embodiment selects example of the modified IEEE33 node distribution network systems as simulation analysis, structure such as Fig. 6
It is shown.The system includes 37 branches, total load 3715kW+2700kvar, reference voltage 12.66kV.
DG is thought of as photovoltaic power generation, and photovoltaic power generation on-position to be selected is { 7,20,24,32 }, other relevant parameters such as table 1
It is shown.
Table 1DG relevant parameter
| DG unit capacity investment cost (ten thousand yuan/kW) | 1 |
| DG separate unit rated capacity/kW | 50 |
| DG unit sale of electricity electricity price (member/kWh) | 0.4 |
| DG unit generates electricity O&M expense (member/kWh) | 0.2 |
| DG generates electricity government subsidy (member/kWh) | 0.2 |
Node 33~37 is to increase load bus newly, total capacity 460kW, and specific payload is as shown in table 2.Solid line indicates
Existing route, dotted line are the route to be selected of newly-increased load access.Other relevant parameters are as shown in table 3.
Table 2 increases load point newly and may on-position
| Load bus number | Payload (kW/kvar) | Possible on-position |
| 34 | 100/60 | 19,20,21,22 |
| 35 | 180/100 | 23,24,25,26 |
| 36 | 80/40 | 9,10,11 |
| 37 | 100/60 | 29,30,31,32 |
3 distribution relevant parameter of table
Price type DSR based on tou power price is by peak Pinggu Time segments division are as follows: peak period (10:00-12:00,20:00-
22:00)；Usually section (08:00-09:00,13:00-19:00,23:00-01:00)；The paddy period (02:00-07:00), and assume
The whole network user both participates in Demand Side Response.Stimulable type DSR based on interruptible load is by maximum 25 node of load in system
As interruptible load, the break period is annual summer (6,7, August), monthly interrupts 7 days, can be daily 10:00- the break period
22:00, user obtain interruptible load subsidy for 0.4 yuan/(kWh).
2, simulation result
(1) DG investment operation quotient and distribution company
By being analyzed above it is found that according to whether considering the uncertainty of DG power output between DG investment operation quotient and distribution company
It will form different game patterns.For the validity and correctness for verifying proposition method of the present invention, it is provided with following 3 kinds of planning field
Scape:
Scene one: for the increment distribution network planning for not using game theory；
Scene two: to use game theory but not considering the probabilistic increment distribution network planning of DG power output；
Scene three: to be established using game theory and the probabilistic increment distribution network planning of consideration DG power output, the present invention
Betting model.
Program results under three kinds of scenes are as shown in table 4.
4 program results of table
| DG investment operation quotient | Distribution company | |
| Scene one | 7(2),20(1),24(2),32(2) | 34-20,35-24,36-10,37-30 |
| Scene two | 7(1),20(1),24(2),32(4) | 34-20,35-26,36-11,37-30 |
| Scene three | 7(1),20(1),24(2),32(4) | 34-20,35-23,36-10,37-30 |
As shown in Table 4, program results of the DG investment operation quotient in scene one are to be respectively connected to 2 in node 7,24 and 32
Photovoltaic unit accesses 1 photovoltaic unit in node 20；Program results in scene two and scene three are in node 7 and 20 point
Not Jie Ru 1 photovoltaic unit, node 24 access 2 photovoltaic units, node 32 access 4 photovoltaic units.
Distribution company the program results in scene one be between node 34 and node 20, node 35 and node 24 it
Between, between node 36 and node 10, new route between node 37 and node 30；Program results in scene two are to exist respectively
Between node 34 and node 20, between node 35 and node 26, between node 36 and node 11, it is new between node 37 and node 30
Build route；It is respectively between node 34 and node 20, between node 35 and node 23, node in the program results in scene three
Between 36 and node 10, new route between node 37 and node 30.
By above-mentioned program results it is found that DG investment operation quotient is identical with the program results in scene three in scene two, with
Result in scene one is different, and program results of the distribution company in three kinds of scenes are different.The reason is that robust optimizes
Object is grid structure, is not related to the addressing constant volume of DG, therefore whether considers that the uncertain of DG power output only can be to distribution public affairs
The program results of department have an impact, the programme without will affect DG investment operation quotient.
(2) power consumer
Power consumer will be waited by two kinds of Demand Side Response means Load adjustment demands of transfer load and interruptible load
Feedback loading is imitated to distribution company.Per day load curve before and after implementation Demand Side Response is as shown in Figure 7.
As shown in Table 4, it is in electricity price in, 8. -9 points and 23. -24 points at 1 point and usually section and is not at interruptible load
Period, power consumer are not engaged in Demand Side Response, therefore it is completely the same to implement Demand Side Response front and back load curve；13
Although -19 point of point is in electricity price usually section but simultaneously in the interruptible load period, interruptible load node is by interrupting one
Divide load that load curve is offset downward；Being in the electricity price peak period in 10. -12 points and 20. -22 points and being in can interrupt
Load period, power consumer offset downward load curve by being transferred out of load and interruptible load；Electricity is in 2. -7 points
The valence paddy period but it is not at the interruptible load period, power consumer offsets up load curve by being transferred into load.
As the above analysis, power consumer can be adjusted load by two kinds of Demand Side Response modes, thus
It realizes " peak load shifting ", and influences the programmed decision-making of increment power distribution network in turn.
3, comparative analysis
(1) consider the analysis on Necessity of multi-agent Game
Illustrated by the comparison of scene one and the lower DG investment operation quotient of scene two and distribution company items cost and income
The necessity of the method for the present invention consideration multi-agent Game.Concrete outcome is as shown in table 5 and table 6.
Table 5DG investment operation quotient's items cost and income (unit: ten thousand yuan)
As shown in Table 5, every costs and benefits of the DG investment operation quotient in scene two are all increased compared to scene one
Add, wherein sale of electricity income is than more than one 10.65 ten thousand yuan of scene, and cost of investment is than more than one 5.09 ten thousand yuan of scene, O&M cost and can be again
Raw energy power generation subsidy is than more than one 5.32 ten thousand yuan of scene.The reason is that scene two fall into a trap and multi-agent Game after, DG's is grid-connected
Capacity increases, so that cost of investment increases, while DG power output increases so that sale of electricity income, O&M cost and renewable energy power generation
Subsidy all increases.
6 distribution company items cost of table and income (unit: ten thousand yuan)
As shown in Table 6, sale of electricity income of the distribution company in scene two is identical as scene one, the reason is that in two kinds of scenes
Under workload demand it is constant.
It cost of investment, Web-based exercise and increased to DG investment operation quotient purchases strategies compared to scene one, wherein throwing
Cost is provided than more than one 3.87 ten thousand yuan of scene, Web-based exercise is than more than one 3.43 ten thousand yuan of scene, to DG investment operation quotient purchases strategies ratio
More than one 9.31 ten thousand yuan of scene.The reason is that scene two fall into a trap and multi-agent Game after, the length of one side new route is more
Long, so that investment cost and via net loss increase, the grid connection capacity of another aspect DG increases, based on the principle of preferential consumption DG,
Distribution company increases to electricity commercially available from DG investment operation.
First failure cost and major network purchases strategies decrease compared to scene, wherein failure cost is fewer than scene one
0.03 ten thousand yuan, major network purchases strategies are than one few 2.05 ten thousand yuan of scene.The reason is that in scene two one side DG grid connection capacity
Increase so that when failure can power supply volume increase, expected loss of energy is reduced, and another aspect distribution company is to DG investment operation quotient
Power purchase increases, and reduces in the case where total purchase of electricity is certain to major network power purchase.
The net profit (unit: ten thousand yuan) of table 7DG investment operation quotient and distribution company
| CDG | CDN | CSUM | |
| Scene one | 38.86 | 295.31 | 334.17 |
| Scene two | 44.42 | 280.78 | 325.2 |
As shown in Table 7, in scene two the sum of net profit of DG investment operation quotient and distribution company than scene one few 8.97
Wan Yuan, but the net profit of DG investment operation quotient's individual is than more than one 5.56 ten thousand yuan of scene.The reason is that in scene one, planning
Optimization aim is to maximize the overall interests of DG investment operation quotient and distribution company, and under the programme of this scene, DG is thrown
Money the sum of operator and the interests of distribution company are the largest compared with other scenes.But in scene one, the maximization of overall efficiency
It is to sacrifice DG investment operation quotient interests as cost, on the one hand this planing method does not meet the actual motion machine of electricity market
System, because the DG investment operation quotient as autonomous investment main body can not make oneself benefit for the maximization and receiving of overall interests
The impaired programme of benefit；On the other hand, if this scheme is imposed on DG investment operation quotient, it will reduce increment power distribution network
The market vitality, this, which is undoubtedly, runs in the opposite direction with reform of current delta power distribution network.In scene two, programme is more
It is obtained after a continuous game of main body, the decision combinations of each investment subject form a kind of Nash Equilibrium point, i.e., any participant is not
It can be changed by independent strategy to obtain more preferably result.This method not only more meets market mechanism, but also makes overall plans
The interests of all participants in the market.
(2) probabilistic analysis on Necessity is considered in the plan model of multi-agent Game
Illustrated by the comparison of scene two and the lower DG investment operation quotient of scene three and distribution company items cost and income
Using the uncertain necessity in the increment distribution network planning for considering multi-agent Game of robust optimization processing DG power output.Tool
Body result is as shown in table 8 and table 9.
Table 8DG investment operation quotient's items cost and income (unit: ten thousand yuan)
As shown in Table 8, every cost and income of the DG investment operation quotient in scene three are identical as scene two.Its reason
It is to consider that DG power output uncertainty only will affect the decision of distribution company, and DG investment operation quotient is in scene two and scene three
Program results it is identical.
9 distribution company items cost of table and income (unit: ten thousand yuan)
As shown in Table 9, sale of electricity income of the distribution company in scene three is identical as scene two, the reason is that in two kinds of scenes
Under workload demand it is constant.
Cost of investment, Web-based exercise, failure cost and major network purchases strategies decrease compared to scene two, wherein investing
Cost than two few 3.06 ten thousand yuan of scene, Web-based exercise than two few 4.78 ten thousand yuan of scene, failure cost than two few 0.21 ten thousand yuan of scene,
Major network purchases strategies are than two few 400,000 yuan of scene.The reason is that passing through after the uncertainty that scene three is fallen into a trap and DG contributes
On the one hand robust optimization improves network topology structure, so that the length of new route is shorter, investment cost is reduced, on the other hand
Interference caused by uncertainty is inhibited, so that every operating cost reduces.
It second increased to DG investment operation quotient purchases strategies compared to scene, than more than 2 35.87 ten thousand yuan of scene.It is former
Because being, the whole power output that DG goes out DG under fluctuation worst-case scenario in scene three increases, to DG investment operation commercially available from electricity increase.
The net profit (unit: ten thousand yuan) of table 10DG investment operation quotient and distribution company
| CDG | CDN | CSUM | |
| Scene two | 44.42 | 280.78 | 325.2 |
| Scene three | 44.42 | 292.86 | 337.28 |
As shown in Table 10, in scene three net profit of distribution company than more than 2 12.08 ten thousand yuan of scene.The reason is that field
Jing Erzhong, due to not fully considering the uncertainty of DG power output during planning, in later period operation, DG power output
Random fluctuation will cause power grid via net loss and power loss level increase, thus make its operating cost increase, and invent propose
Method by introducing virtual game person's " the Nature ", sufficiently meter and DG power output in the plan model for considering multi-agent Game
Uncertainty reduce distribution company in the case where DG goes out fluctuation worst-case scenario by actively improving network topology structure
Every operating cost, to effectively increase planning net profit in the case where sale of electricity revenue unchangeable.
(3) comparison of scene method and robust optimization of the present invention
To prove to describe the probabilistic correctness of DG power output using robust Optimal methods, by the result of invention proposition method
It is compared with the program results based on scene method.Wherein the scene setting in scene method is to contribute to carry out to photovoltaic power generation timing
365 random scenes of whole year after sampling, otherwise processing are identical as the method for the present invention.Calculated result is as shown in table 11.
Solution time comparison under 11 Different Optimization method of table
| Method type | Solve time/s |
| Scene method | 8981 |
| Robust optimization | 1736 |
When as shown in Table 11, using the uncertainty of scene method processing DG power output, model solution is taken a long time, in contrast
It is greatly improved using solving speed after robust Optimal methods, there is significant advantage in terms of computational efficiency.The reason is that scene method
It needs to obtain the accurate probability distribution of uncertain parameter, and carries out scene sampling and reduction to it, and robust optimization is then basis
The fluctuation range of uncertain parameter optimizes the programme under the worst scene, and calculation amount is relatively small.
Above-mentioned simulation result shows: 1) method proposed by the present invention is contributed by investment subject in accurate simulation market and DG
Static-dynamic state joint game behavior between uncertainty, it is ensured that each main market players continues to optimize itself in gambling process
Decision promotes the validity of the market vitality and programmed decision-making to realize that self benefits maximize.2) method proposed by the present invention is logical
It crosses and introduces virtual game person " the Nature ", the uncertain of DG power output has been fully considered in the plan model for considering multi-agent Game
Property, by actively improving network topology structure, reduce every operation that distribution company goes out in DG under fluctuation worst-case scenario
Cost, to effectively increase planning net profit in the case where sale of electricity revenue unchangeable.3) compared with scene method, the present invention is adopted
It is uncertain with robust optimization processing, it can effectively improve the solution efficiency of model.
The various embodiments described above are merely to illustrate the present invention, and each step may be changed, in the technology of the present invention
On the basis of scheme, the improvement and equivalents that all principles according to the present invention carry out separate step should not be excluded in this hair
Except bright protection scope.
Claims (10)
1. a kind of increment power distribution network source net lotus collaborative planning method of multi-agent Game, which comprises the following steps:
1) establish the Decision Model of multiple investment subjects respectively, investment subject include DG investment operation quotient, distribution company and
Participate in the power consumer of DSR；
2) the Static Game behavior between DG investment operation quotient and distribution company is analyzed according to the transitive relation of three investment subjects,
In gambling process when DG investment operation quotient and distribution company either side, which change strategy, can not all obtain more incomes, win
It plays chess and reaches equilibrium state, establish static game model based on this；
3) using the uncertainty of robust optimization processing DG power output, and virtual game person " the Nature " is introduced, according to itself and distribution
Dynamic game behavior between company all can not in gambling process when " the Nature " and distribution company either side change strategy
When acquisition is more preferably paid, game reaches equilibrium state, establishes Dynamic Game Model based on this；
4) consider DG investment operation quotient, distribution company, power consumer and " the Nature " multi-agent Game, formation is matched towards increment
The static-dynamic state joint game pattern of Electric Power Network Planning: constituting Static Game between distribution company and DG investment operation quotient, while and
Dynamic game is constituted between " the Nature ", establishes static-dynamic state joint game model based on this；
5) use iterative search method Solving Nash Equilibrium point, to the Decision Model based on static-dynamic state joint game model into
Row solves, and obtains final programme.
2. method as described in claim 1, it is characterised in that: in the step 1), the Decision Model of DG investment operation quotient
It is responsible for planning distributed generation resource, target is self benefits maximization, and decision variable is position and the appearance of distributed generation resource
Amount；
Objective function are as follows:
In formula,For DG sale of electricity income,For DG cost of investment,For DG O&M cost,It is government for renewable
The power generation of the energy is subsidized；
Constraint condition includes DG node access number limitation to be selected, the constraint of DG permeability and DG units limits:
1. DG node access number limitation to be selected:
Ni.min≤Ni≤Ni.max
In formula, Ni.minFor the lower limit value for accessing DG number in node i to be selected, Ni.maxTo access the upper of DG number in node i to be selected
Limit value；NiThe number of units of DG is accessed for node i to be selected；
2. DG permeability constrains:
In formula, xiFor 0-1 variable, xi=0 i-th of node to be selected of expression does not access DG, xi=1 indicates i-th of node access to be selected
DG；For the rated power of separate unit DG；δ is the permeability after DG is grid-connected, PloadFor node total load；
3. DG units limits:
In formula, Pt DGFor DG t moment total active power output；For DG power output lower limit value,For the upper limit value of DG power output.
3. method as described in claim 1, it is characterised in that: in the step 1), the Decision Model of distribution company is to electricity
Net is planned that target is self benefits maximization, and decision variable is route newly-built project；
Objective function are as follows:
In formula,For distribution company sale of electricity income,For new route cost of investment,For Web-based exercise,For failure
Cost,Based on net purchases strategies,For to DG investment operation quotient's purchases strategies；
Constraint condition includes new route investment and recovery, Branch Power Flow constraint and security constraint:
1. new route investment and recovery
In formula, ΩkTo increase load bus set newly；yj,kTo be k node j-th strip new route；
2. Branch Power Flow constrains
In formula, Pi.tFor the active power of t moment node i, Qi.tFor the reactive power of t moment node i；Ui.tFor t moment node i
Voltage magnitude, Uj.tFor the voltage magnitude of t moment node j；GijFor the conductance of branch ij, BijFor the susceptance of branch ij；θijFor
Phase angle difference between node i and node j voltage；
3. security constraint
In formula, Ui.minFor the lower limit value of node i voltage magnitude, Ui.maxFor the upper limit value of node i voltage magnitude；Pij.tFor branch ij
Transimission power, Pij.maxFor the upper limit value of the transimission power of branch ij.
4. method as described in claim 1, it is characterised in that: in the step 1), the mesh of the Decision Model of power consumer
Scalar functions are as follows:
In formula, Pt itFor the interrupt power of t moment interruptible load；Pt outThe power being transferred out of for t moment load；Pt inFor t moment
The power that load is transferred into；For the electric cost expenditure of reduction after participation Demand Side Response；For interruptible load reimbursement for expenses；
Constraint condition includes transfer load power constraint and interruptible load power constraint:
1. transfer load power constraint
In formula, λminThe lower limit value of power coefficient, λ are transferred out of for t moment loadmaxPower coefficient is transferred out of for t moment load
Upper limit value；μminThe lower limit value of power coefficient, μ are transferred into for t moment loadmaxThe upper of power coefficient is transferred into for t moment load
Limit value；
2. interruptible load power constraint
In formula,For load can interrupt power lower limit value,For load can interrupt power upper limit value.
5. method as described in claim 1, it is characterised in that: in the step 2), Static Game action process are as follows: rich at one
It plays chess in bout, power consumer determines transfer load after receiving tou power price information and interruptible load excitation information and can interrupt
The power of load, and distribution company is fed back in the form of equivalent load；Distribution company is according to last round of DG investment operation quotient couple
The information on load of decision and the power consumer feedback of DG addressing constant volume, makes distribution total benefit by adjusting the newly-built project of route
Maximum, at the same DG investment operation quotient according to last round of distribution company to the decision of new route, by adjusting the addressing of DG and fixed
Appearance scheme makes the investment operation benefit of DG maximum；After updating the addressing constant volume of network topology and DG, into next game
Bout.
6. method as described in claim 1, it is characterised in that: in the step 3), dynamic game process are as follows: " the Nature " first
Distributed generation resource power output will be constantly adjusted in indeterminacy section for current electric grid topological structure, increases the power loss water of power distribution network
Gentle via net loss minimizes distribution total benefit；Thereafter, distribution company will be most badly to plan scene caused by " the Nature "
Based on, by optimizing network topology, distribution total benefit is maximized, network topology is then updated, into next game bout.
7. method as described in claim 1, it is characterised in that: in the step 4), game process are as follows: distribution company receives electric power
The equivalent load of user feedback is considering DG power output according to the on-position of the DG of decision in last round of game out and capacity
After uncertainty, the newly-built project of route is formulated；The robust that DG investment operation quotient goes out according to decision in last round of game simultaneously is excellent
Network topology structure after change carries out the addressing constant volume of DG；After updating the addressing constant volume scheme of network topology and DG, under
One game bout.
8. method as claimed in claim 7, it is characterised in that: the static-dynamic state joint game model are as follows:
Strategy combination is optimum programming scheme under equilibrium state, and the program had not only considered the maximization of benefits of different parties, but also
With preferable robustness.
9. such as claim 1 or 5 or 6 or 7 or 8 the methods, it is characterised in that: the transitive relation of three investment subjects
Are as follows: it introduces " the Nature " and is used as corresponding virtual subject, the addressing that DG investment operation quotient carries out DG under current grid structure is fixed
Hold, and the position of DG and capacity are passed into distribution company and " the Nature "；Power consumer receives tou power price from distribution company
Active response measure is formulated after information and interruptible load excitation information, that is, determines the power of transfer load and interruptible load,
And distribution company is fed back in the form of equivalent load；" the Nature " is knowing the case where DG layouts and combines current grid structure
Under interference is made to the planning of distribution company, " decision " goes out DG and contributes and pass it to distribution company；Distribution company, which receives, to be come
From the transmitting information of other main bodys, decision goes out new route, forms new grid structure.
10. such as any one of claim 1-9 the method, it is characterised in that: in the step 5), model solution method include with
Lower step:
5.1) initial data and parameter are inputted: necessary to data needed for betting model is established in initialization and calculating participant's income
Parameter；
5.2) generate game participant policy space: the policy space of DG investment operation quotient is the set f of DG node state waiting
(x, N)={ f1,f2,…,fn, wherein fnIndicate the accessible DG capacity of node n；The policy space of distribution company is line yet to be built
Set y={ the y on road1,y2,…,yn, wherein ynEach element is the selectable path of route；
5.3) initialization: a class value f is randomly selected respectively in the policy space of DG investment operation quotient and distribution company0And y0,
And according to the policy space for determining " the Nature " the case where value, as iterative initial value；
5.4) each participant carries out independent optimization: in the n-th wheel gambling process, power consumer turns to target with number one maximum
Obtain optimal transfer power and can interrupt power, and result is fed back into distribution company；DG investment operation quotient matches according to the (n-1)th wheel
Grid structure under net company strategy obtains optimal DG access strategy by target of DG investment operation maximizing the benefits, and thus
Determine the policy space of " the Nature "；
5.5) judge whether to reach balanced, if (fn,yn)=(fn-1,yn-1), then reach balanced, exports programme (f*,y*)；Instead
It, then return step 5.4).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810745212.0A CN109034563B (en) | 2018-07-09 | 2018-07-09 | Multi-subject game incremental power distribution network source-network-load collaborative planning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810745212.0A CN109034563B (en) | 2018-07-09 | 2018-07-09 | Multi-subject game incremental power distribution network source-network-load collaborative planning method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109034563A true CN109034563A (en) | 2018-12-18 |
| CN109034563B CN109034563B (en) | 2020-06-23 |
Family
ID=64641892
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810745212.0A Active CN109034563B (en) | 2018-07-09 | 2018-07-09 | Multi-subject game incremental power distribution network source-network-load collaborative planning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109034563B (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784554A (en) * | 2019-01-03 | 2019-05-21 | 山东科技大学 | A kind of electric system optimal scheduling method based on leader-followers games |
| CN109919452A (en) * | 2019-02-15 | 2019-06-21 | 三峡大学 | A kind of electric power based on multi-agent Game-Gas Comprehensive energy resource system joint planing method |
| CN109934487A (en) * | 2019-03-11 | 2019-06-25 | 国网福建省电力有限公司 | A kind of active distribution network coordinated planning method considering multiagent interest game |
| CN110135007A (en) * | 2019-04-19 | 2019-08-16 | 昆明理工大学 | A kind of electric energy supersedure effect analysis method based on multi-party cooperative game |
| CN110197439A (en) * | 2019-05-05 | 2019-09-03 | 三峡大学 | A kind of increment distribution network planning method of the polygon Incompletely information games of consideration source net lotus |
| CN110298493A (en) * | 2019-06-10 | 2019-10-01 | 华北电力大学 | A kind of power source planning method based on game and market trend automatic adjusument mechanism |
| CN110458363A (en) * | 2019-08-15 | 2019-11-15 | 西安建筑科技大学 | A kind of distributed generation resource Life cycle game management method |
| CN111695828A (en) * | 2020-06-17 | 2020-09-22 | 华润智慧能源有限公司 | Incremental power distribution network coordination planning method, device, equipment and medium |
| CN111969589A (en) * | 2020-07-21 | 2020-11-20 | 天津大学 | Active power distribution network voltage control method considering multi-subject game under market environment |
| CN112131695A (en) * | 2020-09-30 | 2020-12-25 | 国网河北省电力有限公司经济技术研究院 | Source network collaborative planning method and terminal equipment |
| CN112243281A (en) * | 2020-10-15 | 2021-01-19 | 广州大学 | Bayesian game routing design method based on rechargeable wireless sensor network |
| CN112381421A (en) * | 2020-11-17 | 2021-02-19 | 三峡大学 | Source network load multi-main-body game planning method considering power market full dimensionality |
| CN112560284A (en) * | 2020-12-24 | 2021-03-26 | 国网河北省电力有限公司经济技术研究院 | Power distribution network planning method for multi-subject game and terminal equipment |
| CN112651544A (en) * | 2020-11-24 | 2021-04-13 | 国网江苏省电力有限公司经济技术研究院 | Incremental power distribution multi-main-body coordination optimization method |
| CN112862352A (en) * | 2021-03-05 | 2021-05-28 | 三峡大学 | Joint planning method for wide-area comprehensive energy system |
| CN113222233A (en) * | 2021-04-30 | 2021-08-06 | 中国石油天然气股份有限公司 | Natural gas multi-main-body energy game analysis method and system |
| CN113743660A (en) * | 2021-08-30 | 2021-12-03 | 三峡大学 | Power distribution network planning method based on multilateral incomplete information evolution game |
| CN114123171A (en) * | 2021-11-12 | 2022-03-01 | 国网江苏省电力有限公司经济技术研究院 | Incremental power distribution network distributed optimization planning method and medium based on potential game |
| CN114188940A (en) * | 2021-11-30 | 2022-03-15 | 国网浙江杭州市萧山区供电有限公司 | Active power distribution network scheduling method considering benefits of multiple subjects |
| CN115395521A (en) * | 2022-10-25 | 2022-11-25 | 国网天津市电力公司营销服务中心 | Renewable energy, energy storage and charging pile collaborative planning method and system |
| CN116526447A (en) * | 2023-03-22 | 2023-08-01 | 国网山东省电力公司临沂供电公司 | Multi-main-body game-based light Chu Lubang configuration method |
| CN117039884A (en) * | 2023-08-22 | 2023-11-10 | 四川大学 | Flexible interconnection planning method of low-voltage distribution network suitable for improving power supply capacity of multiple scenes |
| CN117374996A (en) * | 2023-12-08 | 2024-01-09 | 国网天津市电力公司城南供电分公司 | Snowflake type power distribution network coordination optimization operation method, system, equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106684869A (en) * | 2017-03-17 | 2017-05-17 | 燕山大学 | Active distribution network failure recovery strategy considering inside and outside games |
| CN106803157A (en) * | 2017-02-17 | 2017-06-06 | 广东电网有限责任公司电力科学研究院 | A kind of quality of power supply ameliorative way of low-voltage network distributed energy storage system |
| CN107045655A (en) * | 2016-12-07 | 2017-08-15 | 三峡大学 | Wolf pack clan strategy process based on the random consistent game of multiple agent and virtual generating clan |
| CN107301470A (en) * | 2017-05-24 | 2017-10-27 | 天津大学 | A kind of power distribution network Expansion Planning stores up the dual blank-holder of addressing constant volume with light |
| CN107706921A (en) * | 2017-10-20 | 2018-02-16 | 山东建筑大学 | Micro-capacitance sensor voltage adjusting method and device based on Nash Game |
-
2018
- 2018-07-09 CN CN201810745212.0A patent/CN109034563B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107045655A (en) * | 2016-12-07 | 2017-08-15 | 三峡大学 | Wolf pack clan strategy process based on the random consistent game of multiple agent and virtual generating clan |
| CN106803157A (en) * | 2017-02-17 | 2017-06-06 | 广东电网有限责任公司电力科学研究院 | A kind of quality of power supply ameliorative way of low-voltage network distributed energy storage system |
| CN106684869A (en) * | 2017-03-17 | 2017-05-17 | 燕山大学 | Active distribution network failure recovery strategy considering inside and outside games |
| CN107301470A (en) * | 2017-05-24 | 2017-10-27 | 天津大学 | A kind of power distribution network Expansion Planning stores up the dual blank-holder of addressing constant volume with light |
| CN107706921A (en) * | 2017-10-20 | 2018-02-16 | 山东建筑大学 | Micro-capacitance sensor voltage adjusting method and device based on Nash Game |
Non-Patent Citations (4)
| Title |
|---|
| ZENG BO ETAL: "《A game-theoretic framework for active distribution under the electricity market》", 《TURKISH JOURNAL OF ELECTRICAL ENGINEERING&COMPUTER SCIENCES》 * |
| 张晨: "《售电放开政策下电力交易多方主体利益分析模型研究》", 《万方数据》 * |
| 曾嘉志等: "《用电侧市场放开下的电力市场多主体博弈》", 《电力系统自动化》 * |
| 李玟萱: "《主动配电网下多种博弈关系建模分析》", 《中国优秀硕士论文全文数据库信息科技辑》 * |
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784554A (en) * | 2019-01-03 | 2019-05-21 | 山东科技大学 | A kind of electric system optimal scheduling method based on leader-followers games |
| CN109784554B (en) * | 2019-01-03 | 2023-07-18 | 山西风行测控股份有限公司 | Power system optimal scheduling method based on master-slave game |
| CN109919452A (en) * | 2019-02-15 | 2019-06-21 | 三峡大学 | A kind of electric power based on multi-agent Game-Gas Comprehensive energy resource system joint planing method |
| CN109919452B (en) * | 2019-02-15 | 2023-11-28 | 三峡大学 | Multi-main-body game-based power-natural gas comprehensive energy system joint planning method |
| CN109934487A (en) * | 2019-03-11 | 2019-06-25 | 国网福建省电力有限公司 | A kind of active distribution network coordinated planning method considering multiagent interest game |
| CN110135007A (en) * | 2019-04-19 | 2019-08-16 | 昆明理工大学 | A kind of electric energy supersedure effect analysis method based on multi-party cooperative game |
| CN110197439A (en) * | 2019-05-05 | 2019-09-03 | 三峡大学 | A kind of increment distribution network planning method of the polygon Incompletely information games of consideration source net lotus |
| CN110197439B (en) * | 2019-05-05 | 2023-03-31 | 三峡大学 | Incremental distribution network planning method considering source network load multilateral incomplete information game |
| CN110298493A (en) * | 2019-06-10 | 2019-10-01 | 华北电力大学 | A kind of power source planning method based on game and market trend automatic adjusument mechanism |
| CN110298493B (en) * | 2019-06-10 | 2022-10-25 | 华北电力大学 | Power supply planning method based on game and market dynamic self-adaptive adjustment mechanism |
| CN110458363A (en) * | 2019-08-15 | 2019-11-15 | 西安建筑科技大学 | A kind of distributed generation resource Life cycle game management method |
| CN111695828A (en) * | 2020-06-17 | 2020-09-22 | 华润智慧能源有限公司 | Incremental power distribution network coordination planning method, device, equipment and medium |
| CN111695828B (en) * | 2020-06-17 | 2023-08-22 | 华润智慧能源有限公司 | Coordinated planning method, device, equipment and medium for incremental power distribution network |
| CN111969589A (en) * | 2020-07-21 | 2020-11-20 | 天津大学 | Active power distribution network voltage control method considering multi-subject game under market environment |
| CN111969589B (en) * | 2020-07-21 | 2023-06-06 | 天津大学 | Active power distribution network voltage control method considering multi-main game in market environment |
| CN112131695A (en) * | 2020-09-30 | 2020-12-25 | 国网河北省电力有限公司经济技术研究院 | Source network collaborative planning method and terminal equipment |
| CN112243281B (en) * | 2020-10-15 | 2022-06-21 | 广州大学 | Bayesian game routing design method based on rechargeable wireless sensor network |
| CN112243281A (en) * | 2020-10-15 | 2021-01-19 | 广州大学 | Bayesian game routing design method based on rechargeable wireless sensor network |
| CN112381421A (en) * | 2020-11-17 | 2021-02-19 | 三峡大学 | Source network load multi-main-body game planning method considering power market full dimensionality |
| CN112651544A (en) * | 2020-11-24 | 2021-04-13 | 国网江苏省电力有限公司经济技术研究院 | Incremental power distribution multi-main-body coordination optimization method |
| CN112651544B (en) * | 2020-11-24 | 2022-05-24 | 国网江苏省电力有限公司经济技术研究院 | Incremental power distribution multi-main-body coordination optimization method |
| CN112560284B (en) * | 2020-12-24 | 2022-04-08 | 国网河北省电力有限公司经济技术研究院 | Power distribution network planning method for multi-subject game and terminal equipment |
| CN112560284A (en) * | 2020-12-24 | 2021-03-26 | 国网河北省电力有限公司经济技术研究院 | Power distribution network planning method for multi-subject game and terminal equipment |
| CN112862352A (en) * | 2021-03-05 | 2021-05-28 | 三峡大学 | Joint planning method for wide-area comprehensive energy system |
| CN113222233A (en) * | 2021-04-30 | 2021-08-06 | 中国石油天然气股份有限公司 | Natural gas multi-main-body energy game analysis method and system |
| CN113743660A (en) * | 2021-08-30 | 2021-12-03 | 三峡大学 | Power distribution network planning method based on multilateral incomplete information evolution game |
| CN114123171A (en) * | 2021-11-12 | 2022-03-01 | 国网江苏省电力有限公司经济技术研究院 | Incremental power distribution network distributed optimization planning method and medium based on potential game |
| CN114188940A (en) * | 2021-11-30 | 2022-03-15 | 国网浙江杭州市萧山区供电有限公司 | Active power distribution network scheduling method considering benefits of multiple subjects |
| CN114188940B (en) * | 2021-11-30 | 2023-08-11 | 国网浙江杭州市萧山区供电有限公司 | Active power distribution network scheduling method considering benefits of multiple subjects |
| CN115395521A (en) * | 2022-10-25 | 2022-11-25 | 国网天津市电力公司营销服务中心 | Renewable energy, energy storage and charging pile collaborative planning method and system |
| CN116526447A (en) * | 2023-03-22 | 2023-08-01 | 国网山东省电力公司临沂供电公司 | Multi-main-body game-based light Chu Lubang configuration method |
| CN116526447B (en) * | 2023-03-22 | 2024-03-12 | 国网山东省电力公司临沂供电公司 | Multi-main-body game-based light Chu Lubang configuration method |
| CN117039884A (en) * | 2023-08-22 | 2023-11-10 | 四川大学 | Flexible interconnection planning method of low-voltage distribution network suitable for improving power supply capacity of multiple scenes |
| CN117374996A (en) * | 2023-12-08 | 2024-01-09 | 国网天津市电力公司城南供电分公司 | Snowflake type power distribution network coordination optimization operation method, system, equipment and storage medium |
| CN117374996B (en) * | 2023-12-08 | 2024-03-19 | 国网天津市电力公司城南供电分公司 | Snowflake type power distribution network coordination optimization operation method, system, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109034563B (en) | 2020-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034563A (en) | A kind of increment power distribution network source net lotus collaborative planning method of multi-agent Game | |
| Cheng et al. | Game-theoretic approaches applied to transactions in the open and ever-growing electricity markets from the perspective of power demand response: An overview | |
| Paridari et al. | Robust scheduling of smart appliances in active apartments with user behavior uncertainty | |
| CN109934487A (en) | A kind of active distribution network coordinated planning method considering multiagent interest game | |
| Wang et al. | Stochastic cooperative bidding strategy for multiple microgrids with peer-to-peer energy trading | |
| Kristiansen et al. | Merchant electricity transmission expansion: A European case study | |
| Yi et al. | Coordinated operation strategy for a virtual power plant with multiple DER aggregators | |
| Kronbak et al. | Sharing rules and stability in coalition games with externalities | |
| CN109787261A (en) | Grid side and user side energy storage system capacity configuration optimizing method | |
| CN104392394B (en) | A kind of detection method of micro-capacitance sensor energy storage nargin | |
| Jin et al. | Pricing and operation strategy for peer-to-peer energy trading using distribution system usage charge and game theoretic model | |
| CN102063682A (en) | Wealth management method | |
| Rettieva | Equilibria in dynamic multicriteria games | |
| CN107220889A (en) | The distributed resource method of commerce of microgrid community under a kind of many agent frameworks | |
| Bjørndal et al. | Energy storage operation and electricity market design: On the market power of monopolistic storage operators | |
| Cai et al. | Energy management in residential microgrid using model predictive control-based reinforcement learning and Shapley value | |
| Li et al. | A data-driven joint chance-constrained game for renewable energy aggregators in the local market | |
| CN103839177A (en) | Improved Shapley value method distribution method of micro-grid load gaming | |
| Leimbach et al. | Solution algorithms for regional interactions in large-scale integrated assessment models of climate change | |
| Cozzi | R&D cooperation and growth | |
| CN114943476A (en) | Incremental power distribution network-oriented collaborative planning method | |
| Karaca | On the theory and applications of mechanism design and coalitional games in electricity markets | |
| CN112886565A (en) | Power distribution network coordinated operation strategy formulation method considering multi-party benefit balance | |
| Wu et al. | Peer-to-peer energy trading optimization for community prosumers considering carbon cap-and-trade | |
| Wang et al. | Development of cooperative game model between urban landscape design and urban brand image recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant after: STATE GRID CORPORATION OF CHINA Applicant after: STATE GRID ECONOMIC AND TECHNOLOGICAL RESEARCH INSTITUTE CO. LTD.B412, STATE GRID OFFICE Address before: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant before: State Grid Corporation of China Applicant before: STATE GRID ECONOMIC AND TECHNOLOGICAL RESEARCH INSTITUTE CO. LTD.B412, STATE GRID OFFICE |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200319 Address after: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant after: STATE GRID CORPORATION OF CHINA Applicant after: STATE GRID ECONOMIC AND TECHNOLOGICAL RESEARCH INSTITUTE CO. LTD.B412, STATE GRID OFFICE Applicant after: STATE GRID HEBEI ELECTRIC POWER Co.,Ltd. Applicant after: ECONOMIC AND TECHNOLOGY RESEARCH INSTITUTE OF STATE GRID HEBEI ELECTRIC POWER Co.,Ltd. Applicant after: STATE GRID HEBEI ELECTRIC POWER Co.,Ltd. SHIJIAZHUANG POWER SUPPLY BRANCH Applicant after: CHINA THREE GORGES University Address before: 100031 Xicheng District West Chang'an Avenue, No. 86, Beijing Applicant before: STATE GRID CORPORATION OF CHINA Applicant before: STATE GRID ECONOMIC AND TECHNOLOGICAL RESEARCH INSTITUTE CO. LTD.B412, STATE GRID OFFICE |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |