CN109830990A - A kind of energy storage Optimal Configuration Method based on Congestion access containing scene - Google Patents
A kind of energy storage Optimal Configuration Method based on Congestion access containing scene Download PDFInfo
- Publication number
- CN109830990A CN109830990A CN201910017454.2A CN201910017454A CN109830990A CN 109830990 A CN109830990 A CN 109830990A CN 201910017454 A CN201910017454 A CN 201910017454A CN 109830990 A CN109830990 A CN 109830990A
- Authority
- CN
- China
- Prior art keywords
- energy storage
- power
- energy
- formula
- cost
- 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
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A kind of energy storage Optimal Configuration Method based on Congestion access containing scene, includes the following steps: to obtain each unit data；Establish stored energy capacitance two stages Optimized model；Model is constructed by target of system performance driving economy in the case of no energy storage, most easy generating system obstruction transmission line of electricity is selected, determines installation node set to be selected；Energy storage position and capacity are encoded, first generation population is randomly formed；Using population as known quantity, each moment power output of energy storage in first stage function and system operation cost G are found out using gradient method and finds out system integrated operation cost F using energy storage maximum output as the rated power of energy storage；Calculate population at individual fitness；Whether function restrains or whether reaches maximum number of iterations, if not going in next step, otherwise exports optimal result；New population is formed, the 5th step is gone to.This invention has comprehensively considered operating cost and energy storage cost of investment, helps to improve validity, the economy of energy storage and capacity configuration.
Description
Technical field
The present invention relates to the technical fields of energy storage Optimal Configuration Method, more particularly to based on Congestion access containing scene
Energy storage Optimal Configuration Method technical field.
Background technique
The intermittence and randomness of wind-power electricity generation and photovoltaic power generation bring great challenge to the safe operation of electric system,
The anti-tune peak character of wind-powered electricity generation expands the peak-valley difference of load indirectly, increases a possibility that transmission line of electricity blocks, photovoltaic hair
The uncertain power output of electricity will cause the voltage fluctuation of electric system.Traditional dispatching method is according to load prediction curve and new energy
Source generated energy prediction curve is adjusted by changing the generated energy of thermoelectricity and water power, but thermoelectricity adjustable range is fairly limited,
It is be easy to cause abandonment, is unfavorable for the maximization that clean energy resource uses, and adjustment cost is high；Although water power comparison is flexible, by
Regional impact is larger.Backlog is prevented while to maximally utilise clean energy resource, and can reduce scheduling cost,
Therefore energy storage technology access power grid response is fast, energy is high is of great significance.
Summary of the invention
Goal of the invention: in order to overcome the deficiencies in the prior art, the present invention provides one kind based on Congestion containing wind
The energy storage Optimal Configuration Method of soft exchange can be accomplished to reduce abandonment and abandon light quantity, increase the consumption rate to new energy, improve system
The economy of system operation.
To achieve the above object, the present invention adopts the following technical scheme:
A kind of energy storage Optimal Configuration Method based on Congestion access containing scene, includes the following steps:
S1, the four seasons typical case power output for obtaining new energy unit and each node typical load curve, obtain each unit data；
S2, stored energy capacitance two stages Optimized model is established；
S3, model is constructed by target of system performance driving economy in the case of no energy storage, it is defeated selects most easy generating system obstruction
Electric line determines installation node set to be selected；
S4, it is based on genetic algorithm, energy storage position and capacity is encoded, first generation population is randomly formed；
S5, using population as known quantity, find out each moment power output of energy storage in first stage function and system using gradient method
Operating cost G finds out system integrated operation cost F using energy storage maximum output as the rated power of energy storage；
S6, population at individual fitness is calculated；
Whether S7, function restrain or whether reach maximum number of iterations, if NO go to step S8；If yes it exports most
It is excellent as a result, configuration terminate；
S8, using optimal individual conserving method selection operator, and replicated, intersected, mutation operation, forming new kind
Group, goes to step S5.
The present invention is minimum with system operation cost by considering Congestion, system moving model when foundation is without energy storage
The route for being easy to happen Congestion is calculated in target, determines energy storage node to be selected.Then establish with system operation cost and
The two-stage model of the minimum target of the sum of energy storage investment.By the way that energy storage node screening strategy is added in a model, preferably protect
The convergence calculated has been demonstrate,proved, the solution efficiency of model is improved.
The utility model has the advantages that the present invention, which can be realized, can effectively cope with Congestion in the case where wind/soft exchange power grid,
Preferably consumption scene power output, reduces the operating cost of system entirety.First stage model of the invention is acquired using gradient method
The optimal power output of each unit and energy storage, so that it is determined that the maximum output of energy storage, second stage function model is obtained using genetic algorithm
To the node and combined capacity of energy storage, the optimal installation node of energy storage is finally obtained by genetic manipulation iterative solution and capacity is big
It is small.
Detailed description of the invention
Fig. 1 is overview flow chart of the invention.
Fig. 2 is the flow chart of determination energy storage node to be installed of the invention.
Specific embodiment
As shown in Figure 1 to Figure 2, technical solution of the present invention does further detailed description.
The overall procedure of the method for the present invention is described with reference to Fig. 1, the specific steps are as follows:
As shown in Figure 1, a kind of energy storage Optimal Configuration Method based on Congestion access containing scene, steps are as follows:
S1. the four seasons typical case for obtaining new energy unit contributes and each node typical load curve, obtains each unit data, right
Each variable is initialized；
S2. stored energy capacitance two stages Optimized model is established, specifically: using system operation cost as target, construct the first rank
Section Optimized model constructs second stage optimization to include that system runs the system integrated operation cost with energy storage investment as target
Model establishes stored energy capacitance two stages Optimized model；Angularly established from power constraint, Climing constant, Congestion constraint etc.
Formula or inequality constraints；
S3. model is constructed by target of system performance driving economy in the case of no energy storage, it is defeated selects most easy generating system obstruction
Electric line determines installation node set to be selected；
S4. energy storage position and capacity are encoded, is randomly formed first generation population；
S5. using population as known quantity, each moment power output of energy storage in first stage function and system are found out using gradient method
Operating cost G finds out system integrated operation cost F using energy storage maximum output as the rated power of energy storage；；
S6. population at individual fitness is calculated；
S7. whether function restrains or whether reaches maximum number of iterations, if NO goes to step S8；If yes it exports most
It is excellent as a result, configuration terminate；
S8. optimal individual conserving method selection operator is used, and replicated, intersected, mutation genetic operation, is formed new
Population goes to S5.
Stored energy capacitance two stages Optimized model is established in step S2 of the present invention, specific as follows:
S21, using system operation cost as target, construct first stage Optimized model:
G=min (F1+F2+F3+F4) (1)
S22, using include system operation and energy storage investment system integrated operation cost as target, building second stage optimization
Model:
F=min (F1+F2+F3+F4+F5) (2)
Wherein:
Wherein:
F1For the fuel cost of conventional power unit, adk、bdk、cdkRespectively indicate the normal operation that unit k runs day at d-th
Cost coefficient, Pk(t) indicate that unit k runs the prediction power generating value a few days ago of the t moment of day at d-th, N, which is represented, removes photovoltaic and wind
Conventional power unit sum outside electricity, D are operation day sum in 1 year；F2Indicate receipts caused by the peak load shifting of energy storage in 1 year
Benefit, herein generally less than 0, Pdis.d(t) and Pch.dIt (t) is the electric discharge for the t moment that energy storage runs day at d-th and the function that charges respectively
Rate；ξdis.d(t) it indicates that energy storage runs the discharge condition of the t moment of day at d-th, takes 0 or 1, energy storage is not at electric discharge shape when taking 0
State, discharge condition can be in by going out when taking 1；ξch.d(t) indicate energy storage d-th run day t moment charged state, take 0 or
1, energy storage is not at charged state when taking 0, and going out when taking 1 can be in charged state, md(t) the t moment electricity of d-th of operation day is represented
Valence；F3For the penalty price of new energy unit abandonment light, ρdro.n(t) abandoning of day is run at d-th for t moment new energy unit n
Wind penalty coefficient, PWP.dn(t) and PW.dn(t) be illustrated respectively in d-th operation day t moment new energy unit maximum output and reality
Border power output, NWFor honourable new energy unit sum；F4Indicate cutting load punishment cost, λdtIt indicates to run day t moment at d-th
Cutting load penalty coefficient；PLC.di(t) the cutting load amount of t moment at node i is indicated, I is node total number；F5For the average annual throwing of energy storage
Provide cost, CEAnd CPRespectively indicate energy storage unit capacity cost coefficient and energy storage unit power cost coefficient, ENAnd PNIt respectively refers to store up
The maximum output capacity and peak power output of energy device, r is allowance for depreciation, YrFor battery service life, YaFor Project design year
Limit, λ are the annual maintenance cost coefficient of stored energy mechanism, CrepFor the displacement cost of energy-storage battery in the project time limit, size and storage
The actual life Y of batteryrRelated, calculation formula is as follows:
Wherein,For the specified number of 100% charge and discharge of battery, d is depth of discharge,For in 1 year
Depth of discharge is the charge and discharge cycles number in the case where d,For in 1 year the case where depth of discharge is d it is inferior
Effect is the charge and discharge cycles number of 100% depth of discharge, kpFor the index coefficient of different type energy-storage battery cycle life.
S23, equation or inequality constraints are angularly established specifically such as from power constraint, Climing constant, Congestion constraint
Under:
0≤PW.dn(t)≤PWP.dn(t) (13)
0≤|Pdk(t)-Pdk(t-1)|≤RDk (14)
0≤Pch.d(t)≤ξch.d(t)PN (16)
0≤Pdis.d(t)≤ξdis.d(t)PN (17)
ξdis.d(t)+ξch.d(t)≤1 (18)
|Pdl(t)|≤PlMax, l take 1,2,3 ... L. (19)
Ed(t)=Ed(t-1)·(1-σ)+Pch.d(t)ξch.d(t)ηch-Pdis.d(t)ξdis.d(t)/ηdis (20)
SOCminEN≤Ed(t)≤SOCmaxEN (21)
Wherein formula (11) is power-balance constraint, PLO.di(t) load that the i-node of the t moment of day is run at d-th is indicated
Value, ∑ Ploss.dIt (t) is power loss sum；Formula (12), (13) are generator output restriction；Formula (14) is climbed for generator
Slope constraint, RDkIndicate the permitted maximum creep speed of conventional power generation unit k；Formula (15) is system reserve rotation condition, Ru(t)
For load spinning reserve；Formula (16), (17), (18) are the constraint of energy storage charge-discharge electric power；Formula (19) is transmission line Congestion function
Rate constraint, Pdl(t) it indicates to run transmission power of the day transmission line of electricity l in t moment, P at d-thl maxIndicate transmission line of electricity l most
Big transmission power, L are transmission line of electricity sum；Formula (20), (21) are energy storage Constraint, and σ is battery self-discharge rate；Formula (22)
For the constraint of energy storage energy balance.
It calculates line power and line power is calculated using DC power flow algorithm:
Pdl(t)=AlPin.d(t) (23)
Wherein: AlFor the sensitivity coefficient vector matrix of the l articles branch, 1 × (n-1)；Pin.d(t) d-th of operation day is indicated
The node injecting power vector matrix in addition to balance nodes of t moment, (n-1) × 1.
Foundation described in step S3 of the invention constructs model by target of system economy in the case of no energy storage, specifically such as
Under:
The typical load curve of S31, the new energy unit power curve in selection 1 year and each node of each season；
S32, model is constructed by target of system performance driving economy in the case of no energy storage:
C=min (F1+F3+F4) (24)
In formula: optimized operation cost of the C for system in the case of no energy storage, F1、F3、F4Phase is defined with formula (3), (5), (6)
Together.
Equation or inequality constraints are angularly established from power constraint, Climing constant, spinning reserve constraint, Congestion:
0≤PW.dn(t)≤PWP.dn(t) (27)
0≤|Pdk(t)-Pdk(t-1)|≤RDk (28)
|Pdl(t)|≤Pl max (30)
Wherein: formula (25) is power-balance constraint, and formula (26), the power output that (27) are conventional power unit and new energy unit are about
Beam, formula (28) are the Climing constant of conventional power unit, and formula (29) is the constraint of system spinning reserve, and formula (30) is that multi-line power transmission blocks about
Beam.
S33, the transmission power that each branch under Optimum Economic operating condition is calculated with gradient method；
S34, Congestion risk indicator is defined are as follows:
HlThe degree of risk that route l blocks is represented, the smaller expression of value is more easy to happen backlog.By institute's route selection
Two end nodes on road install node as energy storage to be selected, and find out each transmission line of electricity Congestion value-at-risk；
If S35, choosing the smallest main line of the whole network Congestion risk indicator, the feasible node at selected branch both ends is made
Node is installed for energy storage to be selected.
The optimal adam algorithm contributed and used in gradient descent method is solved in step S33 of the present invention, specific as follows:
For the minimum optimization problem of objective function J (θ), it is only necessary to which parameter θ advances along the opposite direction of gradient
One step-length, that is, learning rate, so that it may the decline of function to achieve the objective.Parameter more new formula is as follows:
WhereinIt is the gradient of parameter, η is for learning rate, it is proposed that be set as 0.001, super ginseng value suggestion: β1=
0.9,β2=0.999, ε=1e-8.
It is based on genetic algorithm described in step S4 of the present invention, energy storage position and capacity are encoded, specific as follows:
Using the integer coding of elongated degree, the information information of energy storage is to (Ni,Ei) indicate, wherein NiIndicate energy storage installation
Node location number, EiIndicate stored energy capacitance number, the position of energy storage and volume solutions individual are by several one-to-one correspondence
NiAnd EiComposition, is shown below:
N={ N1,N2,...,Nn}
C={ C1,C2,...,Cn}
Code length selects number and stored energy capacitance related with the node of energy storage.
Population at individual fitness is calculated described in step S6 of the present invention, specific as follows:
Establish fitness function are as follows:
Wherein f (x) is objective function.
The condition of convergence described in step S7 of the present invention, specific as follows:
The variety rate of fitness of optimum individual is in convergence range, i.e., full in two generation populations of continuous front and back in genetic manipulation
Foot:
In formula: CnewFor optimum individual fitness in newly generated group；ColdFor optimum individual fitness in former generation group,
ε is a lesser positive.
Described in step S8 of the present invention use optimal individual conserving method selection operator, and replicated, intersected, make a variation something lost
Operation is passed, new population is formed, specific as follows:
The selection of S81, genetic operator: optimum maintaining strategy employed herein, i.e., fitness is highest several in current group
Individual directly replaces the equivalent individual that fitness is minimum in current group, ensures that optimal in current group in this way
The fitness of body is not less than the fitness of former generation group, and selected operator is copied to new population；
S82, genetic operator crossover operation: it is swapped by the way of uniformity crossover herein, such as there are A=
(Ni,Ei), B=(Nj,Ej) two parent individualities, it is random to generate mask word ω identical with genes of individuals seat length1,ω2,
ω3,...,ω2nIf ωi=1, then the genic value on i-th of locus of offspring individual A ' of A inherits the corresponding genic value of A, B
I-th of locus of offspring individual B ' on genic value inherit the corresponding genic value of B；If ωi=0, then on i-th of locus of A '
Genic value inherit the corresponding genic value of B, the genic value on i-th of locus of B ' inherits the corresponding genic value of A.
S83, genetic operator mutation operation: for an energy storage position and volume solutions individual, if being produced at random between 10
Raw number is no more than mutation probability, then to any two node N in the individuali、NjOn stored energy capacitance Ei、EjIt swaps, obtains
To the individual of lower generation.
Illustrate the present invention in order to clearer, expansion explanation will be carried out to related content below.
(1) model preprocessing method
Convergence is increased using DC power flow.DC power flow constraint is increased to primal problem in the initial stage, DC power flow can
Approximation is seen as a kind of simplification of AC power flow, and DC power flow calculating is more relatively easy, in the constraint for not increasing primal problem
Under the premise of condition, network security trend constraint problem is considered compared to direct solution, solves be more readily available satisfaction in this way
The solution of safe trend constraint；
(2) foundation of the energy storage two stages Optimized model based on system performance driving economy
Step 1. constructs first stage Optimized model using system operation cost as target:
G=min (F1+F2+F3+F4)
For step 2. to include that system runs the system integrated operation cost with energy storage investment as target, building second stage is excellent
Change model:
F=min (F1+F2+F3+F4+F5)
Wherein:
Wherein:
F1For the fuel cost of conventional power unit, adk、bdk、cdkRespectively indicate the normal operation that unit k runs day at d-th
Cost coefficient, Pk(t) indicate that unit k runs the prediction power generating value a few days ago of the t moment of day at d-th, N, which is represented, removes photovoltaic and wind
Conventional power unit sum outside electricity, D are operation day sum in 1 year；F2Indicate receipts caused by the peak load shifting of energy storage in 1 year
Benefit, herein generally less than 0, Pdis.d(t) and Pch.dIt (t) is the electric discharge for the t moment that energy storage runs day at d-th and the function that charges respectively
Rate；ξdis.d(t) it indicates that energy storage runs the discharge condition of the t moment of day at d-th, takes 0 or 1, energy storage is not at electric discharge shape when taking 0
State, discharge condition can be in by going out when taking 1；ξch.d(t) indicate energy storage d-th run day t moment charged state, take 0 or
1, energy storage is not at charged state when taking 0, and going out when taking 1 can be in charged state, md(t) the t moment electricity of d-th of operation day is represented
Valence；F3For the penalty price of new energy unit abandonment light, ρdro.n(t) abandoning of day is run at d-th for t moment new energy unit n
Wind penalty coefficient, PWP.dn(t) and PW.dn(t) be illustrated respectively in d-th operation day t moment new energy unit maximum output and reality
Border power output, NWFor honourable new energy unit sum；F4Indicate cutting load punishment cost, λdtIt indicates to run day t moment at d-th
Cutting load penalty coefficient；PLC.di(t) the cutting load amount of t moment at node i is indicated, I is node total number；F5For the average annual throwing of energy storage
Provide cost, CEAnd CPRespectively indicate energy storage unit capacity cost coefficient and energy storage unit power cost coefficient, ENAnd PNIt respectively refers to store up
The maximum output capacity and peak power output of energy device, r is allowance for depreciation, YrFor battery service life, YaFor Project design year
Limit, λ are the annual maintenance cost coefficient of stored energy mechanism, CrepFor the displacement cost of energy-storage battery in the project time limit, size and storage
The actual life Y of batteryrRelated, calculation formula is as follows:
Wherein,For the specified number of 100% charge and discharge of battery, d is depth of discharge,For in 1 year
Depth of discharge is the charge and discharge cycles number in the case where d,For in 1 year the case where depth of discharge is d it is inferior
Effect is the charge and discharge cycles number of 100% depth of discharge, kpFor the index coefficient of different type energy-storage battery cycle life.
Step 3. angularly establishes equation or inequality constraints is specific from power constraint, Climing constant, Congestion constraint
It is as follows:
0≤PW.dn(t)≤PWP.dn(t)
0≤|Pdk(t)-Pdk(t-1)|≤RDk
0≤Pch.d(t)≤ξch.d(t)PN
0≤Pdis.d(t)≤ξdis.d(t)PN
ξdis.d(t)+ξch.d(t)≤1
Pdl(t)≤Pl max, l takes 1,2,3 ... L.
Ed(t)=Ed(t-1)·(1-σ)+Pch.d(t)ξch.d(t)ηch-Pdis.d(t)ξdis.d(t)/ηdis
SOCminEN≤Ed(t)≤SOCmaxEN
Wherein, PLO.di(t) load value that the i-node of the t moment of day is run at d-th, ∑ P are indicatedloss.dIt (t) is power
Loss sum, RDkIndicate the permitted maximum creep speed of conventional power generation unit k, RuIt (t) is load spinning reserve, Pdl(t) table
Show and runs transmission power of the day transmission line of electricity l in t moment, P at d-thl maxIndicate that the maximum delivery power of transmission line of electricity l, L are
Transmission line of electricity sum.
It calculates line power and line power is calculated using DC power flow algorithm:
Pdl(t)=AlPin.d(t)
Wherein: AlFor the sensitivity coefficient vector matrix of the l articles branch, 1 × (n-1)；Pin.d(t) d-th of operation day is indicated
The node injecting power vector matrix in addition to balance nodes of t moment, (n-1) × 1.
(3) selection method of energy storage node to be installed
As shown in Fig. 2, since system network nodes number is more, being calculated most using the method for exhaustion for complicated electric system
Excellent allocation plan calculation amount is larger, and time-consuming, constructs model by target of system performance driving economy in the case of no energy storage, selects most
Easy generating system blocks transmission line of electricity, determines installation node set to be selected, specific as follows:
Step 1. is using the four seasons typical case power output and four seasons typical load of the scene in above-mentioned energy storage Optimized model as no storage
The four seasons typical case power output and four seasons typical load of the scene of energy system.
Step 2. establishes the optimal operation model without energy-storage system:
C=min (F1+F3+F4)
0≤PW.dn(t)≤PWP.dn(t)
0≤|Pdk(t)-Pdk(t-1)|≤RDk
|Pdl(t)|≤Pl max
Step 3. calculates the transmission power of each branch under Optimum Economic operating condition；
Step 4. defines Congestion risk indicator are as follows:
HlThe degree of risk that route l blocks is represented, the smaller expression of value is more easy to happen backlog.By institute's route selection
Two end nodes on road install node as energy storage to be selected, and find out each transmission line of electricity Congestion value-at-risk；
If step 5. chooses the smallest main line of the whole network Congestion risk indicator, by the feasible node at selected branch both ends
Node is installed as energy storage to be selected.
(4) method for solving of the Optimized model of energy storage
The present invention solves first stage model using gradient adam method, and second stage model is solved using genetic algorithm.
The minimum optimization problem of adam algorithm in gradient descent method for objective function J (θ), it is only necessary to by parameter θ
Along opposite one step-length of direction advance of gradient, that is, learning rate, so that it may the decline of function to achieve the objective.Parameter is more
New formula is as follows:
WhereinIt is the gradient of parameter, η is for learning rate, it is proposed that be set as 0.001, super ginseng value suggestion: β1=
0.9,β2=0.999, ε=1e-8.
Genetic algorithm is the integer coding using elongated degree, and the information information of energy storage is to (Ni,Ei) indicate, wherein NiTable
Show the node location number of energy storage installation, EiIndicate stored energy capacitance number, the position of energy storage and volume solutions individual are by several
A one-to-one NiAnd EiComposition, is shown below:
N={ N1,N2,...,Nn}
C={ C1,C2,...,Cn}
Code length selects number and stored energy capacitance related with the node of energy storage.
Detailed process is as follows for fitness function, genetic manipulation and the condition of convergence of genetic algorithm:
1) fitness function is established are as follows:
Wherein f (x) is objective function.
2) selection of genetic operator: optimum maintaining strategy employed herein, i.e., fitness is highest several in current group
Individual directly replaces the equivalent individual that fitness is minimum in current group, ensures that the optimum individual in current group in this way
Fitness be not less than former generation group fitness；
3) genetic operator crossover operation: being swapped by the way of uniformity crossover herein, such as there are A=(Ni,
Ei), B=(Nj,Ej) two parent individualities, it is random to generate mask word ω identical with genes of individuals seat length1,ω2,ω3,...,
ω2nIf ωi=1, then the genic value on i-th of locus of offspring individual A ' of A inherits the corresponding genic value of A, the filial generation of B
Genic value on i-th of locus of body B ' inherits the corresponding genic value of B；If ωi=0, then the genic value on i-th of locus of A '
The corresponding genic value of B is inherited, the genic value on i-th of locus of B ' inherits the corresponding genic value of A.
4) genetic operator mutation operation: for an energy storage position and volume solutions individual, if being randomly generated between 10
Number be no more than mutation probability, then to any two node N in the individuali、NjOn stored energy capacitance Ei、EjIt swaps, obtains
One individual of lower generation.
5) variety rate of fitness of optimum individual is in convergence range, i.e., full in two generation populations of continuous front and back in genetic manipulation
Foot:
In formula: CnewFor optimum individual fitness in newly generated group；ColdFor optimum individual fitness in former generation group,
ε is a lesser positive.
The above is only a preferred embodiment of the present invention, it should be pointed out that: for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (8)
1. a kind of energy storage Optimal Configuration Method based on Congestion access containing scene, characterized by the following steps:
S1, the four seasons typical case power output for obtaining new energy unit and each node typical load curve, obtain each unit data；
S2, stored energy capacitance two stages Optimized model is established；
S3, model is constructed by target of system performance driving economy in the case of no energy storage, selects most easy generating system obstruction power transmission line
Road determines installation node set to be selected；
S4, it is based on genetic algorithm, energy storage position and capacity is encoded, first generation population is randomly formed；
S5, using population as known quantity, find out each moment power output of energy storage in first stage function using gradient method and system run
Cost G finds out system integrated operation cost F using energy storage maximum output as the rated power of energy storage；
S6, population at individual fitness is calculated；
Whether S7, function restrain or whether reach maximum number of iterations, if NO go to step S8；If yes optimal knot is exported
Fruit, configuration terminate；
S8, using optimal individual conserving method selection operator, and replicated, intersected, mutation operation, forming new population, being turned
To step S5.
2. the energy storage Optimal Configuration Method according to claim 1 based on Congestion access containing scene, it is characterised in that:
In the step S2, stored energy capacitance two stages Optimized model process is specific as follows:
S21, using system operation cost as target, construct first stage Optimized model:
G=min (F1+F2+F3+F4) (1)
S22, using include system operation and energy storage investment system integrated operation cost as target, building second stage optimize mould
Type:
F=min (F1+F2+F3+F4+F5) (2)
Wherein:
Wherein:
F1For the fuel cost of conventional power unit, adk、bdk、cdkRespectively indicate the cost that unit k runs the normal operation of day at d-th
Coefficient, Pk(t) indicate that unit k runs the prediction power generating value a few days ago of the t moment of day at d-th, N is represented in addition to photovoltaic and wind-powered electricity generation
Conventional power unit sum, D are operation day sum in 1 year；F2Indicate income caused by the peak load shifting of energy storage in 1 year, herein
Generally less than 0, Pdis.dIt (t) is discharge power of the energy storage in the t moment of d-th of operation day, Pch.dIt (t) is energy storage respectively in d
The charge power of the t moment of a operation day；ξdis.d(t) indicate energy storage d-th run day t moment discharge condition, take 0 or
1, energy storage is not at discharge condition when taking 0, and energy storage is in discharge condition when taking 1；ξch.d(t) indicate energy storage d-th of operation day
T moment charged state, take 0 or 1, energy storage is not at charged state when taking 0, and energy storage is in charged state, m when taking 1d(t)
Represent the t moment electricity price of d-th of operation day；F3For the penalty price of new energy unit abandonment light, ρdro.nIt (t) is the new energy of t moment
Source unit n runs the abandonment penalty coefficient of day, P at d-thWP.dn(t) and PW.dn(t) be illustrated respectively in d-th operation day t moment
The maximum output of new energy unit and practical power output, NWFor honourable new energy unit sum；F4Indicate cutting load punishment cost, λdt
Indicate the cutting load penalty coefficient that day t moment is run at d-th；PLC.di(t) the cutting load amount of t moment at node i is indicated, I is
Node total number；F5For the average annual cost of investment of energy storage, CEAnd CPRespectively indicate energy storage unit capacity cost coefficient and energy storage specific work
Rate cost coefficient, ENAnd PNThe maximum output capacity and peak power output of energy storage device are respectively referred to, r is allowance for depreciation, YrFor battery
Service life, YaFor the Project design time limit, λ is the annual maintenance cost coefficient of stored energy mechanism, CrepFor energy storage electricity in the project time limit
The actual life Y of the displacement cost in pond, size and batteryrRelated, calculation formula is as follows:
Wherein,For the specified number of 100% charge and discharge of battery, d is depth of discharge,For in 1 year in charge and discharge
Depth is the charge and discharge cycles number in the case where d,To be equivalent in the case where depth of discharge is d in 1 year
The charge and discharge cycles number of 100% depth of discharge, kpFor the index coefficient of different type energy-storage battery cycle life；
S23, equation is angularly established or inequality constraints is specific as follows from power constraint, Climing constant, Congestion constraint:
0≤PW.dn(t)≤PWP.dn(t) (13)
0≤|Pdk(t)-Pdk(t-1)|≤RDk (14)
0≤Pch.d(t)≤ξch.d(t)PN (16)
0≤Pdis.d(t)≤ξdis.d(t)PN (17)
ξdis.d(t)+ξch.d(t)≤1 (18)
Pdl(t)≤Pl max, l takes 1,2,3 ... L. (19)
Ed(t)=Ed(t-1)·(1-σ)+Pch.d(t)ξch.d(t)ηch-Pdis.d(t)ξdis.d(t)/ηdis (20)
SOCminEN≤Ed(t)≤SOCmaxEN (21)
Wherein formula (11) is power-balance constraint, PLO.di(t) load value that the i-node of the t moment of day is run at d-th is indicated,
∑Ploss.dIt (t) is power loss sum；Formula (12), (13) are generator output restriction；Formula (14) is that generator is climbed about
Beam, RDkIndicate the permitted maximum creep speed of conventional power generation unit k；Formula (15) is system reserve rotation condition, Ru(t) it is negative
Lotus spinning reserve；Formula (16), (17), (18) are the constraint of energy storage charge-discharge electric power；Formula (19) be transmission line Congestion power about
Beam, Pdl(t) it indicates to run transmission power of the day transmission line of electricity l in t moment, P at d-thl maxIndicate that the maximum of transmission line of electricity l is defeated
Power is sent, L is transmission line of electricity sum；Formula (20), (21) are energy storage Constraint, and σ is battery self-discharge rate；Formula (22) is
The constraint of energy storage energy balance；
It calculates line power and line power is calculated using DC power flow algorithm:
Pdl(t)=AlPin.d(t) (23)
Wherein: AlFor the sensitivity coefficient vector matrix of the l articles branch, 1 × (n-1)；Pin.d(t) when indicating d-th of operation day t
The node injecting power vector matrix in addition to balance nodes carved, (n-1) × 1.
3. the energy storage Optimal Configuration Method according to claim 1 based on Congestion access containing scene, it is characterised in that:
In the step S3, model is constructed by target of system performance driving economy in the case of no energy storage, selects most easy generating system obstruction
Transmission line of electricity determines installation node set to be selected, and detailed process is as follows:
The typical load curve of S31, the new energy unit power curve in selection 1 year and each node of each season；
S32, model is constructed by target of system performance driving economy in the case of no energy storage:
C=min (F1+F3+F4) (24)
In formula: optimized operation cost of the C for system in the case of no energy storage, F1For the fuel cost of conventional power unit, F3For new energy source machine
The penalty price of group abandonment light, F4For cutting load punishment cost；
Constraint condition is specific as follows:
0≤PW.dn(t)≤PWP.dn(t) (27)
0≤|Pdk(t)-Pdk(t-1)|≤RDk (28)
Pdl(t)≤Pl max (30)
Wherein: formula (25) is power-balance constraint, formula (26), the units limits that (27) are conventional power unit and new energy unit, formula
It (28) is the Climing constant of conventional power unit, formula (29) is the constraint of system spinning reserve, and formula (30) is multi-line power transmission obstruction constraint；
S33, the transmission power that each branch under Optimum Economic operating condition is calculated with gradient method；
S34, Congestion risk indicator is defined are as follows:
HlThe degree of risk that route l blocks is represented, the smaller expression of value is more easy to happen backlog；By selected route
Two end nodes install node as energy storage to be selected, and find out each transmission line of electricity Congestion value-at-risk；
If S35, choose the smallest main line of the whole network Congestion risk indicator, using the feasible node at selected branch both ends as to
Select energy storage that node is installed.
4. the energy storage Optimal Configuration Method according to claim 3 based on Congestion access containing scene, it is characterised in that:
The step S33 calculates the transmission power of each branch under Optimum Economic operating condition with gradient method are as follows:
For the minimum optimization problem of objective function J (θ), parameter θ is advanced a step-length along the opposite direction of gradient,
It is exactly learning rate, the decline of function to achieve the objective；Parameter more new formula is as follows:
WhereinIt is the gradient of parameter, η is to be set as 0.001, super ginseng value suggestion: β for learning rate1=0.9, β2=
0.999, ε=1e-8.
5. the energy storage Optimal Configuration Method according to claim 1 based on Congestion access containing scene, it is characterised in that:
It is based on genetic algorithm in the step S4, energy storage position and capacity are encoded, specific as follows:
Using the integer coding of elongated degree, the information information of energy storage is to (Ni,Ei) indicate, wherein NiIndicate the section of energy storage installation
Point Position Number, EiIndicate stored energy capacitance number, the position of energy storage and volume solutions individual are by several one-to-one NiWith
EiComposition, is shown below:
N={ N1,N2,...,Nn}
C={ C1,C2,...,Cn}
Code length selects number and stored energy capacitance related with the node of energy storage.
6. the energy storage Optimal Configuration Method according to claim 1 based on Congestion access containing scene, it is characterised in that:
Population at individual fitness is calculated described in the step S6, specific as follows:
Establish fitness function are as follows:
Wherein f (x) is objective function.
7. the energy storage Optimal Configuration Method according to claim 1 based on Congestion access containing scene, it is characterised in that:
The condition of convergence described in the step S7, specific as follows:
In two generation populations of continuous front and back the variety rate of fitness of optimum individual meets in convergence range in genetic manipulation:
In formula: CnewFor optimum individual fitness in newly generated group；ColdFor optimum individual fitness in former generation group, ε is
One lesser positive.
8. the energy storage Optimal Configuration Method according to claim 1 based on Congestion access containing scene, it is characterised in that:
Optimal individual conserving method selection operator is used described in the step S8, and is replicated, intersected, mutation genetic operation, shape
The population of Cheng Xin, specific as follows:
The selection of S81, genetic operator: optimum maintaining strategy employed herein, i.e., in current group fitness it is highest it is several each and every one
Body directly replaces the equivalent individual that fitness is minimum in current group, ensures that the optimum individual in current group in this way
Fitness is not less than the fitness of former generation group, and selected operator is copied to new population；
S82, genetic operator crossover operation: being swapped by the way of uniformity crossover herein, such as there are A=(Ni,
Ei), B=(Nj,Ej) two parent individualities, it is random to generate mask word ω identical with genes of individuals seat length1,ω2,ω3,...,
ω2nIf ωi=1, then the genic value on i-th of locus of offspring individual A ' of A inherits the corresponding genic value of A, the filial generation of B
Genic value on i-th of locus of body B ' inherits the corresponding genic value of B；If ωi=0, then the genic value on i-th of locus of A '
The corresponding genic value of B is inherited, the genic value on i-th of locus of B ' inherits the corresponding genic value of A；
S83, genetic operator mutation operation: for an energy storage position and volume solutions individual, if being randomly generated between 10
Number is no more than mutation probability, then to any two node N in the individuali、NjOn stored energy capacitance Ei、EjIt swaps, obtains one
A individual of lower generation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910017454.2A CN109830990B (en) | 2019-01-08 | 2019-01-08 | Energy storage optimization configuration method based on wind and light access contained in output resistor plug |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910017454.2A CN109830990B (en) | 2019-01-08 | 2019-01-08 | Energy storage optimization configuration method based on wind and light access contained in output resistor plug |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109830990A true CN109830990A (en) | 2019-05-31 |
| CN109830990B CN109830990B (en) | 2022-09-20 |
Family
ID=66860141
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910017454.2A Active CN109830990B (en) | 2019-01-08 | 2019-01-08 | Energy storage optimization configuration method based on wind and light access contained in output resistor plug |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109830990B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110071505A (en) * | 2019-06-04 | 2019-07-30 | 清华大学 | The power transmission network enlarging of the access containing large-scale wind power configures joint planing method with energy storage |
| CN111628492A (en) * | 2020-04-13 | 2020-09-04 | 四川大学 | Power grid blocking management and control method with cooperation of high-voltage distribution network partition reconstruction and energy storage |
| CN112117772A (en) * | 2020-09-21 | 2020-12-22 | 南京工程学院 | Energy storage fast optimization configuration method for considering output resistor plug under new energy access |
| CN112232983A (en) * | 2020-10-15 | 2021-01-15 | 国网上海市电力公司 | Active power distribution network energy storage optimal configuration method, electronic equipment and storage medium |
| CN112564183A (en) * | 2020-12-02 | 2021-03-26 | 中国电力工程顾问集团华北电力设计院有限公司 | Capacity optimization configuration method for wind, light and fire storage in power grid planning |
| CN113346526A (en) * | 2021-05-24 | 2021-09-03 | 国网综合能源服务集团有限公司 | Multi-node energy storage system configuration method based on discrete-continuous hybrid method |
| CN113656988A (en) * | 2021-10-20 | 2021-11-16 | 中国能源建设集团湖南省电力设计院有限公司 | Optimization method for improving wind power consumption energy storage power |
| CN113725916A (en) * | 2021-08-31 | 2021-11-30 | 南京邮电大学 | DPFC optimal configuration method for promoting high-permeability new energy consumption |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013174145A1 (en) * | 2012-05-23 | 2013-11-28 | 国家电网公司 | Large-scale wind power grid-integration reactive voltage optimization method based on improved artificial fish swarm hybrid optimization algorithm |
| CN104795828A (en) * | 2015-04-24 | 2015-07-22 | 清华大学 | Wind storage capacity configuration method based on genetic algorithm |
| CN106096757A (en) * | 2016-05-31 | 2016-11-09 | 天津天大求实电力新技术股份有限公司 | Based on the microgrid energy storage addressing constant volume optimization method improving quantum genetic algorithm |
| CN107681656A (en) * | 2017-09-27 | 2018-02-09 | 华中科技大学 | A kind of congestion cost bi-level programming method for considering real time execution risk |
| CN108288861A (en) * | 2018-02-01 | 2018-07-17 | 福州大学 | The method of wind farm group wind storage system addressing constant volume combined optimization |
-
2019
- 2019-01-08 CN CN201910017454.2A patent/CN109830990B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013174145A1 (en) * | 2012-05-23 | 2013-11-28 | 国家电网公司 | Large-scale wind power grid-integration reactive voltage optimization method based on improved artificial fish swarm hybrid optimization algorithm |
| CN104795828A (en) * | 2015-04-24 | 2015-07-22 | 清华大学 | Wind storage capacity configuration method based on genetic algorithm |
| CN106096757A (en) * | 2016-05-31 | 2016-11-09 | 天津天大求实电力新技术股份有限公司 | Based on the microgrid energy storage addressing constant volume optimization method improving quantum genetic algorithm |
| CN107681656A (en) * | 2017-09-27 | 2018-02-09 | 华中科技大学 | A kind of congestion cost bi-level programming method for considering real time execution risk |
| CN108288861A (en) * | 2018-02-01 | 2018-07-17 | 福州大学 | The method of wind farm group wind storage system addressing constant volume combined optimization |
Non-Patent Citations (1)
| Title |
|---|
| 吴俊玲等: "减少弃风损失的储能容量和布局优化研究", 《电力建设》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110071505A (en) * | 2019-06-04 | 2019-07-30 | 清华大学 | The power transmission network enlarging of the access containing large-scale wind power configures joint planing method with energy storage |
| CN111628492A (en) * | 2020-04-13 | 2020-09-04 | 四川大学 | Power grid blocking management and control method with cooperation of high-voltage distribution network partition reconstruction and energy storage |
| CN111628492B (en) * | 2020-04-13 | 2021-11-09 | 四川大学 | Power grid blocking management and control method with cooperation of high-voltage distribution network partition reconstruction and energy storage |
| CN112117772A (en) * | 2020-09-21 | 2020-12-22 | 南京工程学院 | Energy storage fast optimization configuration method for considering output resistor plug under new energy access |
| CN112232983A (en) * | 2020-10-15 | 2021-01-15 | 国网上海市电力公司 | Active power distribution network energy storage optimal configuration method, electronic equipment and storage medium |
| CN112564183A (en) * | 2020-12-02 | 2021-03-26 | 中国电力工程顾问集团华北电力设计院有限公司 | Capacity optimization configuration method for wind, light and fire storage in power grid planning |
| CN112564183B (en) * | 2020-12-02 | 2022-11-15 | 中国电力工程顾问集团华北电力设计院有限公司 | Capacity optimization configuration method for wind, light and fire storage in power grid planning |
| CN113346526A (en) * | 2021-05-24 | 2021-09-03 | 国网综合能源服务集团有限公司 | Multi-node energy storage system configuration method based on discrete-continuous hybrid method |
| CN113346526B (en) * | 2021-05-24 | 2022-05-20 | 国网综合能源服务集团有限公司 | Multi-node energy storage system configuration method based on discrete-continuous hybrid method |
| CN113725916A (en) * | 2021-08-31 | 2021-11-30 | 南京邮电大学 | DPFC optimal configuration method for promoting high-permeability new energy consumption |
| CN113725916B (en) * | 2021-08-31 | 2024-01-12 | 南京邮电大学 | DPFC (differential pressure filter) optimal configuration method for promoting new energy consumption with high permeability |
| CN113656988A (en) * | 2021-10-20 | 2021-11-16 | 中国能源建设集团湖南省电力设计院有限公司 | Optimization method for improving wind power consumption energy storage power |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109830990B (en) | 2022-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109830990A (en) | A kind of energy storage Optimal Configuration Method based on Congestion access containing scene | |
| CN106951985B (en) | Multi-objective optimal scheduling method for cascade reservoir based on improved artificial bee colony algorithm | |
| WO2023274425A1 (en) | Multi-energy capacity optimization configuration method for wind-solar-water-fire storage system | |
| CN104701871B (en) | One kind is containing the honourable complementary microgrid hybrid energy-storing capacity optimum proportioning method of water multi-source | |
| CN105977991B (en) | A kind of self microgrid Optimal Configuration Method for considering price type demand response | |
| CN108599206B (en) | Power distribution network hybrid energy storage configuration method under high-proportion uncertain power supply scene | |
| CN110071505A (en) | The power transmission network enlarging of the access containing large-scale wind power configures joint planing method with energy storage | |
| CN106887841A (en) | A kind of genetic particle group optimizing method on multiple populations of the capacity configuration of micro-capacitance sensor containing electric automobile | |
| CN103310065A (en) | Intelligent distribution network reconstruction method concerning distributed power generation and energy storage unit | |
| CN107437824A (en) | A kind of computational methods of the Area distribution formula power supply allowed capacity based on genetic algorithm | |
| CN108448620A (en) | High permeability distributed generation resource assemblage classification method based on integrated performance index | |
| CN108206543A (en) | A kind of energy source router and its running optimizatin method based on energy cascade utilization | |
| CN108242807B (en) | A kind of reconstruction method of power distribution network containing photovoltaic considering multidimensional security constraint | |
| CN106503466B (en) | Equipment capacity configuration method and device of electric boiler and solar combined heating system | |
| Sediqi et al. | Optimal economical sizing of grid-connected hybrid renewable energy system | |
| CN105932690A (en) | Distribution network operation optimization method integrating reactive power optimization and network reconstruction | |
| CN109742796A (en) | A kind of Multiple Time Scales coordinated scheduling method of scene extreme misery storage association system | |
| CN117689178B (en) | Method and device for dispatching and optimizing long-period operation of combined type ground source heat pump system | |
| CN110943465B (en) | Energy storage system site selection and volume fixing optimization method | |
| CN104616071B (en) | A kind of wind-light storage complementary power generation system Optimal Configuration Method | |
| CN109347139A (en) | A kind of Distributed Generation in Distribution System maximum penetration level Optimal Configuration Method | |
| CN105574681A (en) | Multi-time-scale community energy local area network energy scheduling method | |
| CN106712060A (en) | Multi-agent-based hundred-megawatt level battery energy storage system control method and system | |
| CN117134409A (en) | Micro-grid system considering electro-hydro-thermal complementation and multi-objective optimal configuration method thereof | |
| CN117610828A (en) | Step-size dense-by-dense multi-target cascade hydropower station group optimal scheduling method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |