CN106849162A - Consider the grid-connected active distribution network ADAPTIVE ROBUST optimization method of a large amount of regenerative resources - Google Patents
Consider the grid-connected active distribution network ADAPTIVE ROBUST optimization method of a large amount of regenerative resources Download PDFInfo
- Publication number
- CN106849162A CN106849162A CN201710072474.0A CN201710072474A CN106849162A CN 106849162 A CN106849162 A CN 106849162A CN 201710072474 A CN201710072474 A CN 201710072474A CN 106849162 A CN106849162 A CN 106849162A
- Authority
- CN
- China
- Prior art keywords
- active
- distribution network
- power source
- adaptive
- distributed power
- 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
- 238000009826 distribution Methods 0.000 title claims abstract description 69
- 238000005457 optimization Methods 0.000 title claims abstract description 67
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000001172 regenerating effect Effects 0.000 title claims abstract description 23
- 238000005520 cutting process Methods 0.000 claims abstract description 12
- 238000013507 mapping Methods 0.000 claims description 11
- 230000035939 shock Effects 0.000 claims description 6
- 238000013459 approach Methods 0.000 claims description 3
- 230000017105 transposition Effects 0.000 claims description 3
- 238000013486 operation strategy Methods 0.000 abstract description 2
- 239000000243 solution Substances 0.000 description 16
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000002347 injection Methods 0.000 description 2
- 239000007924 injection Substances 0.000 description 2
- 238000012886 linear function Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000001568 sexual effect Effects 0.000 description 1
Classifications
-
- H02J3/382—
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a kind of active distribution network ADAPTIVE ROBUST optimization method for considering that a large amount of regenerative resources are grid-connected, including：Robust feasible zone is estimated；Build nonlinear adaptive function；Build ADAPTIVE ROBUST active reactive Optimized model；Kernel function space reflection；Cutting plane algorithm is solved and active distribution network Optimal Operation Strategies；The method is uncertain for the height of regenerative resource large-scale grid connection, introduces the thought of robust optimization, has ensured the safe operation of power system under optimizing decision；Adaptive technique is introduced, realizes optimizing decision with uncertain variables adaptive change；Establish active distribution network ADAPTIVE ROBUST active reactive Coordination and Optimization Model, it is ensured that the economy and high efficiency of distribution network operation under optimizing decision；Efficient, the rapid solving of former problem are realized using KERNEL FUNCTION METHOD and cutting plane solution strategies；Ensure that reliability in active distribution network running, economy and security.
Description
Technical field
It is especially a kind of to consider that a large amount of regenerative resources are grid-connected the present invention relates to active distribution network optimization progress control method
Active distribution network ADAPTIVE ROBUST optimization method.
Background technology
The large-scale grid connection of regenerative resource, the operation to power distribution network brings wide influence and huge challenge.Point
Cloth power supply is Typical Representative therein, and regenerative resource is mostly intermittent energy source, with randomness and probabilistic spy
Point.At present in active distribution network running optimizatin, mainly there are three kinds of processing methods to the uncertainty of regenerative resource：(1) with
Machine is planned；(2) fuzzy programming；(3) robust optimization.The uncertain probability distribution that stochastic programming is taken often has with actual conditions
Certain deviation, economy and the security of distribution network operation cannot get Reliable guarantee；Fuzzy programming often will be according to decision-making
The personal experience of person, decision-making has larger subjectivity, it is difficult to ensure safety, the economical operation of power distribution network；Robust optimization will not
Determine that variable is expressed as the form in interval, it is ensured that security constraint can meet under condition of uncertainty, simultaneously because considering institute
Situation about occurring is possible to, decision-making often has certain conservative, and the economy of distribution network operation is not also high.
Therefore, the existing active distribution network for considering that a large amount of regenerative resources are grid-connected is deposited at aspects such as economy, securities
Deficiency, it is impossible to the problems such as realizing the active control and real-time optimization of power distribution network.
The content of the invention
Goal of the invention：To overcome the deficiencies in the prior art, the present invention provides one kind to be ensured actively to match somebody with somebody under condition of uncertainty
The active distribution network ADAPTIVE ROBUST optimization side for considering that a large amount of regenerative resources are grid-connected of economic, the safe efficient operation of power network
Method.
Technical scheme：A kind of active distribution network ADAPTIVE ROBUST optimization method for considering that a large amount of regenerative resources are grid-connected, bag
Include following steps：
(1) robust feasible zone is estimated
For traditional tide model, carry out optimum linearity and approach, build linear approximation tide model；Based on the linear approximation
Tide model, asks for the approximate mapping relations between node injecting power vector and node voltage vector；Based on the mapping relations,
Ask for power distribution network it is active-exert oneself space, i.e. robust feasible zone of the decision variable of idle coordination optimization estimate；
(2) nonlinear adaptive function is built
When active distribution network reactive capability is not enough, distributed power source can not always run on MPPT patterns, its active power output
Control need to be coordinated and optimized with idle exerting oneself；The auto-adaptive function and distributed power source of distributed power source active power output are built respectively
The idle auto-adaptive function exerted oneself, uncertain variables are distributed power source maximum active power output；
(3) build ADAPTIVE ROBUST it is active-idle work optimization model
Based on active distribution network distributed power source active power output and the idle auto-adaptive function exerted oneself, active distribution network is built
ADAPTIVE ROBUST is active-idle Coordination and Optimization Model；
(4) kernel function space reflection
For power distribution network ADAPTIVE ROBUST it is active-idle Coordination and Optimization Model the characteristics of, introduce KERNEL FUNCTION METHOD by low-dimensional not
Determine variable mappings to higher-dimension kernel function space, the nonlinear optimal problem that will be difficult to resolve under lower dimensional space is converted under higher dimensional space
The linear optimization problem that can be solved；
(5) cutting plane algorithm is solved
For the linear adaption robust optimization problem under higher dimensional space, using cutting plane solution strategies, by primal problem
Two subproblems are resolved into, the solution of primal problem is obtained by two alternating iterations of subproblem；
(6) active distribution network Optimal Operation Strategies.
Further, in the step (1), based on traditional tide model, approached using optimum linearity, structure it is linear near
Like tide model, it is shown below：
S=T (V) (1)
Wherein, S is node injecting power vector, and V is node voltage vector；
Equivalent transformation is carried out to above-mentioned model, the proximal line between node voltage vector and node injecting power vector is asked for
Sexual intercourse, is shown below：
V=T-1(S) (2)
So as to the constraint to node injecting power will be converted into the constraint of voltage, further realize that robust feasible zone is estimated
Meter.
Further, in the step (2), the grid-connected active distribution network of extensive regenerative resource, distributed power source
Maximum active power output has uncertainty, is expressed as affine form, is shown below：
Wherein, pmaxIt is distributed power source maximum active power output,It is the predicted value of distributed power source maximum active power output,It is maximum perturbation amount, ε exerts oneself Discontinuous Factors for distributed power source, and Ω is uncertain collection；
Traditional robust Optimal Decision-making often has problem high of spending of guarding, therefore uses ADAPTIVE ROBUST optimisation technique,
Realize that optimizing decision changes with the change of disturbance quantity, so as to greatly reduce the conservative of decision-making.
When active distribution network reactive capability is not enough, distributed power source can not run on MPPT patterns, and need to simultaneously regulate and control it has
Work(is exerted oneself and is exerted oneself with idle, that is, carry out active-idle coordination optimization of active distribution network；Therefore build respectively its active power output and
The idle auto-adaptive function exerted oneself.
The auto-adaptive function of distributed power source active power output, is shown below：
P=p0+p(ε) (4)
P (ε)=pα1ε+pα2εTε+… (5)
Wherein, p is distributed power source active power output, p0It is the optimal active power output of distributed power source in the case of undisturbed, p
(ε) is auto-adaptive function of the distributed power source active power output with shock wave, pα1For the single order of distributed power source active power output is disturbed
Momentum, pα2It is the second order disturbance quantity of distributed power source active power output, εTIt is the transposition of ε；Other formula (5) is polynomial function,
The specific number of its item number is related to the expression of Optimized model；
The auto-adaptive function that distributed power source OPTIMAL REACTIVE POWER is exerted oneself, is shown below：
Q=q0+q(ε) (6)
Q (ε)=qα1ε+qα2εTε+… (7)
Wherein, Q exerts oneself for distributed power source OPTIMAL REACTIVE POWER, q0For the OPTIMAL REACTIVE POWER of distributed power source in the case of undisturbed goes out
Power, q (ε) is the idle auto-adaptive function exerted oneself with shock wave of distributed power source, qα1One exerted oneself for distributed power source is idle
Rank disturbance quantity, qα2The second order disturbance quantity exerted oneself for distributed power source is idle；In addition, formula (7) is polynomial function, its item number tool
Body number is related to the expression of Optimized model.
Further, in the step (3), traditional robust is optimized for a min-max problem, and ADAPTIVE ROBUST optimizes
Thought be decision making under certainty value in replacing former robust to optimize with adaptive decision-making function；Active distribution network ADAPTIVE ROBUST
Active-idle Coordination and Optimization Model is：
Wherein,It is the object function under undisturbed state,To there is state of disturbance
Under object function, g (P, Q)=0 represent equality constraint, h (P, Q)≤0 represent inequality constraints.
Further, active distribution network ADAPTIVE ROBUST it is active-idle Coordination and Optimization Model addition of nonlinear equation about
Beam, introduces non-convex link, and model is difficult to effectively solution.For the problem, KERNEL FUNCTION METHOD is introduced, low-dimensional uncertain variables are reflected
Higher-dimension kernel function space is incident upon, new uncertain set is built in higher-dimension kernel function space, by uncertain variables under lower dimensional space
Non-linear type auto-adaptive function be converted into the linear function of new uncertain variables under kernel function space, so as to by lower dimensional space
Under nonlinear adaptive robust optimization problem be converted into higher dimensional space lower linear ADAPTIVE ROBUST optimization problem.
Linear adaption robust optimization problem under higher-dimension kernel function space is a min-max problem, and it has decision-making
Variable P, Q and the class variables of Discontinuous Factors ε two, the problem are in the nature the saddle point for seeking two class variables.Using the solution of cutting plane
Strategy, by former min-max problems, resolves into one on decision variable P, the subproblem of Q, and one on Discontinuous Factors ε
Subproblem, obtain the solutions of primal problem by the alternating iterations of two subproblems.
Beneficial effect：Compared with prior art, a kind of active distribution for considering that a large amount of regenerative resources are grid-connected of the invention
Net ADAPTIVE ROBUST optimization method, with advantages below：(1) taken into full account that the grid-connected height for bringing of a large amount of regenerative resources is not true
It is qualitative, by robust feasible zone estimation technique, the constraint to decision variable will be converted into the security constraint of voltage, it is ensured that excellent
Change under decision-making, the safe operation of power distribution network；(2) adaptive technique is introduced, the guarantor of traditional robust Optimal Decision-making is greatly reduced
Keeping property, for the economy of power distribution network, Effec-tive Function provide powerful guarantee；(3) Optimal Decision-making is realized real with disturbance quantity change
Shi Bianhua, realizes the real-time optimization of active distribution network；(4) direct solution is difficult under processing lower dimensional space using KERNEL FUNCTION METHOD
Nonlinear adaptive robust optimization problem, the linear adaption robust optimization that becoming can solve under higher-dimension kernel function space is asked
Topic；For the linear adaption robust optimization problem under higher-dimension kernel function space, using cutting plane algorithm, primal problem is resolved into
Two subproblems, last solution is obtained by two alternating iterations of subproblem, realizes efficient, the rapid solving of former problem,
For the solution of other such problem provides a kind of thinking；The present invention is applied to the active for considering that a large amount of regenerative resources are accessed
Power distribution network running optimizatin.
Brief description of the drawings
Fig. 1 is the FB(flow block) of the inventive method；
Fig. 2 is that robust feasible zone estimates isoboles；
Fig. 3 is kernel function mapping isoboles.
Specific embodiment
Further detailed description is done to the present invention below in conjunction with the accompanying drawings.
As shown in figure 1, a kind of active distribution network ADAPTIVE ROBUST optimization method for considering that a large amount of regenerative resources are grid-connected, bag
Include step in detail below：
(1) robust feasible zone is estimated
For traditional tide model, carry out optimum linearity and approach, build linear approximation tide model；Based on the linear trend
Model, asks for the approximate mapping relations between node injecting power vector and node voltage vector；Based on the mapping relations, ask for
Power distribution network is active-and exert oneself space, i.e. robust feasible zone of the decision variable of idle coordination optimization estimate；
(2) when active distribution network reactive capability is not enough, distributed power source can not always run on MPPT patterns, its it is active go out
Power need to coordinate and optimize control with idle exerting oneself；The auto-adaptive function and distributed electrical of distributed power source active power output are built respectively
The auto-adaptive function that source is idle to exert oneself, uncertain variables are distributed power source maximum active power output；
(3) based on active distribution network distributed power source active power output and the idle auto-adaptive function exerted oneself, build and actively match somebody with somebody
Power network ADAPTIVE ROBUST is active-idle Coordination and Optimization Model；
(4) for power distribution network ADAPTIVE ROBUST it is active-idle Coordination and Optimization Model the characteristics of, introduce KERNEL FUNCTION METHOD by low-dimensional
Uncertain variables map to higher-dimension kernel function space, and the nonlinear optimal problem that will be difficult to resolve under lower dimensional space is converted into higher dimensional space
The linear optimization problem that can be solved down；
(5) for the linear adaption robust optimization problem under higher dimensional space, using cutting plane solution strategies, asked original
Topic resolves into two subproblems, and the solution of primal problem is obtained by two alternating iterations of subproblem；
Built on linear approximation tide model in step (1), traditional tide model is shown below：
S=IV (9)
In formula：S is node injecting power vector, and I is node Injection Current vector, and V is node voltage vector.
Approached using linear optimal, ask for approximately linear power flow equation, be shown below：
S=T (V) (10)
Linear approximate relationship between power distribution network interior joint voltage vector and node injecting power vector, it is as follows：
V=T-1(S) (11)
As shown in Fig. 2 can obtain right between node power Injection Space and node voltage space by linear transformation T
Should be related to, according to the linear approximate relationship between node voltage vector and node injecting power vector, by the section of power distribution network
Point voltage constraint is converted into the constraint to node injecting power.
The node voltage constraint of power distribution network is as follows：
h(V)≤0 (12)
The constraint of node injecting power is as follows：
h(T-1(S))≤0 (13)
Based on node injecting power constraint, you can obtain the active power output of distributed power source and idle space of exerting oneself,
As active distribution network robust feasible zone is estimated.
On in step (2), when active distribution network reactive capability is sufficient, distributed power source can run on maximum power point with
Track (Maximum Power Point Tracking, MPPT) pattern, it has actual active power output for its maximum active power output, only
Need to its idle be exerted oneself.When active distribution network reactive capability is not enough, distributed power source can not run on MPPT patterns,
Its active power output and idle active-idle coordination optimization exerted oneself, that is, carry out active distribution network need to simultaneously be regulated and controled.
Traditional robust optimum theory considers all of possibility, including the extremely low extreme case of probability of occurrence, therefore its
Provide decision-making and often spend problem high with conservative.Adaptive technique is introduced, by the change direct reaction of disturbance quantity to decision-making
Amount, realizes decision value adaptive change with the change of disturbance quantity, greatly reduces the conservative of decision-making, is the warp of power distribution network
Ji, Effec-tive Function provide strong guarantee.
The grid-connected active distribution network of extensive regenerative resource, the maximum active power output of distributed power source is with uncertain
Property, affine algebra theory is introduced, affine form is denoted as, it is shown below：
In formula, pmaxIt is distributed power source maximum active power output,It is the prediction of distributed power source maximum active power output
Value,It is maximum perturbation amount, ε exerts oneself Discontinuous Factors for distributed power source, and Ω is uncertain collection.
When active distribution network reactive capability is not enough, distributed power source can not run on MPPT patterns, and its active reactive is exerted oneself
It is both needed to be run according to adaptive mode.In the case, simple linear adaption rule can not well be fitted active distribution
The optimized operation point of net, need to use more accurate auto-adaptive function so that optimization instruction is more nearly under every kind of possible operating mode
Global optimum's operating point.
The auto-adaptive function of distributed power source active power output is a polynomial function, and it is expressed as follows shown in formula：
P=p0+p(ε) (15)
P (ε)=pα1ε+pα2εTε+… (16)
Wherein, p is distributed power source active power output, p0It is the optimal active power output of distributed power source in the case of undisturbed, p
(ε) is auto-adaptive function of the distributed power source active power output with shock wave, pα1For the single order of distributed power source active power output is disturbed
Momentum, pα2It is the second order disturbance quantity of distributed power source active power output, εTIt is the transposition of ε；
The auto-adaptive function that distributed power source OPTIMAL REACTIVE POWER is exerted oneself is similarly a nonlinear polynomial function, its expression
Formula is as follows：
Q=q0+q(ε) (17)
Q (ε)=qα1ε+qα2εTε+… (18)
Wherein, Q exerts oneself for distributed power source OPTIMAL REACTIVE POWER, q0For the OPTIMAL REACTIVE POWER of distributed power source in the case of undisturbed goes out
Power, q (ε) is the idle auto-adaptive function exerted oneself with shock wave of distributed power source, qα1One exerted oneself for distributed power source is idle
Rank disturbance quantity, qα2The second order disturbance quantity exerted oneself for distributed power source is idle.
In on step (3), compared with traditional robust optimizes, the thought of ADAPTIVE ROBUST optimization is being asked for disturbance variable
Change and the decision-making of adaptive change, i.e., what it was solved is a decision rule, rather than a decision value.Based on nonlinear
Distributed power source active power output and idle auto-adaptive function of exerting oneself, build active distribution network ADAPTIVE ROBUST it is active-idle coordination
Optimized model, expression formula is as follows：
Wherein,It is the object function under undisturbed state,To there is state of disturbance
Under object function, g (P, Q)=0 represent equality constraint, h (P, Q)≤0 represent inequality constraints.
On step (4), active distribution network ADAPTIVE ROBUST is active-and idle Coordination and Optimization Model addition of nonlinear equation
Constraint, introduces non-convex link, and model is difficult to effectively solution.As shown in figure 3, KERNEL FUNCTION METHOD is introduced, by low-dimensional uncertain variables
Higher-dimension kernel function space is mapped to by F, new uncertain set is built in higher-dimension kernel function space, will be not true under lower dimensional space
The non-linear type auto-adaptive function for determining variable is converted into the linear function of new uncertain variables under kernel function space, so that will be low
Nonlinear adaptive robust optimization problem under dimension space is converted into higher dimensional space lower linear ADAPTIVE ROBUST optimization problem.
The introducing of kernel function avoids " dimension disaster ", substantially reduces amount of calculation；And the dimension of the input space is to core letter
Matrix number without influence, therefore, Kernel-Based Methods can effectively process higher-dimension input；The form of non-linear transform function need not be known
And parameter；The change of the form and parameter of kernel function can implicitly change the mapping from the input space to feature space, and then right
The property of feature space produces influence, finally changes the performance of various Kernel-Based Methods；Kernel-Based Methods can be with different calculations
Method is combined, and forms various different methods based on kernel function technology, and this two-part design can be carried out individually, and can
Think the different kernel function of different application selections and algorithm
The specific selection of Kernel Function of the present invention is related to the expression of model, can be Polynomial kernel function, branch
Vector machine (SVM) kernel function and Radial basis kernel function etc. are held, is illustrated by taking Polynomial kernel function as an example, Polynomial kernel function
Expression formula is as follows:
F(x,xi)=[(xxi)+1]d (20)
Wherein：X and xiIt is the vector in former lower dimensional space, d is polynomial order, d=1,2 ..., N, F (x, xi) it is core letter
Number, is also called d rank graders.
On step (5), linear adaption robust optimization problem under higher-dimension kernel function space is asked for a min-max
Topic, there is decision variable P, Q and the class variables of Discontinuous Factors ε two in it, the problem is in the nature the saddle point for seeking two class variables.Using
The solution strategies of cutting plane, by former min-max problems, resolve into one on decision variable P, the subproblem of Q, and a pass
In the subproblem of Discontinuous Factors ε, the solution of primal problem is obtained by two alternating iterations of subproblem.Specific decomposable process is as follows
It is shown：
In the t times iteration, ask on decision variable P, during the subproblem of Q, Discontinuous Factors ε takes t-1 times as known quantity
The value of iteration, its object function is as follows：
Wherein, εt-1For the Discontinuous Factors that last iteration is asked for；
When asking for the subproblem on Discontinuous Factors ε, decision variable P, Q take the t times value of iteration as known quantity, its
Object function is as follows：
Wherein, pt(ε) and qt(ε) is respectively the t times active power output and idle power generating value of iteration；
By the decomposition, then former problem becomes two subproblems as implied above, and is comprised only in two subproblems
The unknown quantity of one class, is a certainty optimization problem for legibility, can be obtained not by as above two alternating iterations of subproblem
The solution of robust optimization under the conditions of certainty.
Distributed power source can simultaneously to power distribution network provide active power output and it is idle exert oneself, therefore consider its operational mode, need
To its active power output and idle exerting oneself control of optimize, as active-idle coordination optimization；For the big rule of renewable power supply
The grid-connected height of mould is uncertain, introduces the thought of robust optimization, and going out for decision variable is determined using robust feasible zone estimation technique
Power space, to ensure the safe operation of power system；For the problem that traditional robust optimization conservative is too high, self adaptation skill is introduced
Art, realizes optimizing decision with uncertain variables adaptive change so that the conservative of optimizing decision is substantially reduced, it is ensured that distribution
The economy of network operation；The nonlinear adaptive letter that the optimal active power output of distributed power source and OPTIMAL REACTIVE POWER are exerted oneself is built respectively
Number, and build power distribution network ADAPTIVE ROBUST it is active-idle Coordination and Optimization Model；Using KERNEL FUNCTION METHOD, will be difficult to resolve under lower dimensional space
Nonlinear adaptive robust optimization problem be converted into the linear adaption robust optimization problem that can be solved under higher dimensional space, and propose
Cutting plane solution strategies, under kernel function space, two subproblems are resolved into by primal problem, by two alternatings of subproblem
Iteration obtains last solution.
Claims (6)
1. a kind of active distribution network ADAPTIVE ROBUST optimization method for considering that a large amount of regenerative resources are grid-connected, it is characterised in that bag
Include following steps：
(1) robust feasible zone is estimated
For traditional tide model, carry out optimum linearity and approach, build linear approximation tide model；Based on the linear approximation trend
Model, asks for the approximate mapping relations between node injecting power vector and node voltage vector；Based on the mapping relations, ask for
Power distribution network is active-and the decision variable of idle coordination optimization exerts oneself space；
(2) nonlinear adaptive function is built
When active distribution network reactive capability is not enough, distributed power source can not always run on MPPT patterns, its active power output and nothing
Work(is exerted oneself need to coordinate and optimize control；The auto-adaptive function and distributed power source for building distributed power source active power output respectively are idle
The auto-adaptive function exerted oneself, uncertain variables are distributed power source maximum active power output；
(3) build ADAPTIVE ROBUST it is active-idle work optimization model
Based on active distribution network distributed power source active power output and the idle auto-adaptive function exerted oneself, active distribution network is built adaptive
Answer robust it is active-idle Coordination and Optimization Model；
(4) kernel function space reflection
For power distribution network ADAPTIVE ROBUST it is active-idle Coordination and Optimization Model the characteristics of, introduce KERNEL FUNCTION METHOD low-dimensional is not known
To higher-dimension kernel function space, the nonlinear optimal problem that will be difficult to resolve under lower dimensional space can be solved variable mappings under being converted into higher dimensional space
Linear optimization problem；
(5) cutting plane algorithm is solved
For the linear adaption robust optimization problem under higher dimensional space, using cutting plane solution strategies, primal problem is decomposed
Into two subproblems, the solution of primal problem is obtained by two alternating iterations of subproblem；
(6) active distribution network optimization operation.
2. a kind of active distribution network ADAPTIVE ROBUST optimization for considering that a large amount of regenerative resources are grid-connected according to claim 1
Method, it is characterised in that in the step (1), linear approximation tide model such as following formula：
S=T (V)
Wherein, S is node injecting power vector, and V is node voltage vector；
Mapping relations such as following formula between node injecting power vector and voltage vector：
V=T-1(S)。
3. a kind of active distribution network ADAPTIVE ROBUST optimization for considering that a large amount of regenerative resources are grid-connected according to claim 1
Method, it is characterised in that in the step (2), the maximum active power output of distributed power source is：
Wherein, pmaxIt is distributed power source maximum active power output,It is the predicted value of distributed power source maximum active power output,
It is maximum perturbation amount, ε exerts oneself Discontinuous Factors for distributed power source, and Ω is uncertain collection；
The auto-adaptive function of distributed power source active power output, is shown below：
P=p0+p(ε)
P (ε)=pα1ε+pα2εTε+…
Wherein, p is distributed power source active power output, p0It is the optimal active power output of distributed power source in the case of undisturbed, p (ε) is
Distributed power source active power output with shock wave auto-adaptive function, pα1It is the first-order perturbation amount of distributed power source active power output,
pα2It is the second order disturbance quantity of distributed power source active power output, εTIt is the transposition of ε；
The auto-adaptive function that distributed power source OPTIMAL REACTIVE POWER is exerted oneself, is shown below：
Q=q0+q(ε)
Q (ε)=qα1ε+qα2εTε+…
Wherein, Q exerts oneself for distributed power source OPTIMAL REACTIVE POWER, q0For the OPTIMAL REACTIVE POWER of distributed power source in the case of undisturbed is exerted oneself, q
(ε) is the idle auto-adaptive function exerted oneself with shock wave of distributed power source, qα1The single order exerted oneself for distributed power source is idle is disturbed
Momentum, qα2The second order disturbance quantity exerted oneself for distributed power source is idle.
4. a kind of active distribution network ADAPTIVE ROBUST optimization for considering that a large amount of regenerative resources are grid-connected according to claim 1
Method, it is characterised in that in the step (3), active distribution network ADAPTIVE ROBUST is active-and idle Coordination and Optimization Model is：
G (P, Q)=0
h(P,Q)≤0
P=p0+ p (ε) p (ε)=pα1ε+pα2εTε+…
Q=q0+ q (ε) q (ε)=qα1ε+qα2εTε+…
ε∈Ω
Wherein,It is the object function under undisturbed state,For under having a state of disturbance
Object function, g (P, Q)=0 represents equality constraint, and h (P, Q)≤0 represents inequality constraints.
5. a kind of active distribution network ADAPTIVE ROBUST optimization for considering that a large amount of regenerative resources are grid-connected according to claim 1
Low-dimensional uncertain variables are mapped to higher-dimension kernel function sky by method, it is characterised in that in the step (4) by KERNEL FUNCTION METHOD
Between, so as to the nonlinear adaptive robust optimization problem that will be difficult to resolve under lower dimensional space be converted into can be solved under higher dimensional space it is linear from
Adapt to robust optimization problem.
6. a kind of active distribution network ADAPTIVE ROBUST optimization for considering that a large amount of regenerative resources are grid-connected according to claim 1
Method, it is characterised in that in the step (5), in kernel function space, using cutting plane solution strategies, by linear adaption
Robust optimization problem resolves into a subproblem for asking for decision variable P, Q and a subproblem for asking for Discontinuous Factors ε, passes through
Two alternating iterations of subproblem obtain last solution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710072474.0A CN106849162B (en) | 2017-02-10 | 2017-02-10 | The grid-connected active distribution network ADAPTIVE ROBUST optimization method of a large amount of renewable energy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710072474.0A CN106849162B (en) | 2017-02-10 | 2017-02-10 | The grid-connected active distribution network ADAPTIVE ROBUST optimization method of a large amount of renewable energy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106849162A true CN106849162A (en) | 2017-06-13 |
| CN106849162B CN106849162B (en) | 2019-06-21 |
Family
ID=59122325
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710072474.0A Active CN106849162B (en) | 2017-02-10 | 2017-02-10 | The grid-connected active distribution network ADAPTIVE ROBUST optimization method of a large amount of renewable energy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106849162B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107508283A (en) * | 2017-08-14 | 2017-12-22 | 东南大学 | A kind of distributed power source operation domain method for solving based on affine arithmetic |
| CN112733316A (en) * | 2019-10-29 | 2021-04-30 | 新奥数能科技有限公司 | Energy system modeling optimization method and device |
| CN112928751A (en) * | 2021-02-02 | 2021-06-08 | 山东大学 | New energy power grid connection equipment configuration method based on robust optimization |
| CN113256446A (en) * | 2021-05-31 | 2021-08-13 | 华南理工大学 | Power distribution network robust optimization scheduling method considering data loss and achieving information physical fusion |
| CN113394771A (en) * | 2021-06-07 | 2021-09-14 | 广西大学 | Method for solving feasible domain of nonlinear power system tie line |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103036245A (en) * | 2012-11-30 | 2013-04-10 | 中国南方电网有限责任公司 | Novel transmission loss reducing method by alternating current or direct current (AC / DC) synergy used in interconnected network and novel transmission loss reducing system by AC / DC synergy used in interconnected network |
| CN105119320A (en) * | 2015-09-15 | 2015-12-02 | 东北大学 | Distributed wind power plant fan optimized arrangement system and method |
| CN106300426A (en) * | 2016-09-19 | 2017-01-04 | 上海电力学院 | The self adaptation passivity PI control method of grid-connected inverting system based on MMC |
-
2017
- 2017-02-10 CN CN201710072474.0A patent/CN106849162B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103036245A (en) * | 2012-11-30 | 2013-04-10 | 中国南方电网有限责任公司 | Novel transmission loss reducing method by alternating current or direct current (AC / DC) synergy used in interconnected network and novel transmission loss reducing system by AC / DC synergy used in interconnected network |
| CN105119320A (en) * | 2015-09-15 | 2015-12-02 | 东北大学 | Distributed wind power plant fan optimized arrangement system and method |
| CN106300426A (en) * | 2016-09-19 | 2017-01-04 | 上海电力学院 | The self adaptation passivity PI control method of grid-connected inverting system based on MMC |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107508283A (en) * | 2017-08-14 | 2017-12-22 | 东南大学 | A kind of distributed power source operation domain method for solving based on affine arithmetic |
| CN112733316A (en) * | 2019-10-29 | 2021-04-30 | 新奥数能科技有限公司 | Energy system modeling optimization method and device |
| CN112733316B (en) * | 2019-10-29 | 2024-04-19 | 新奥数能科技有限公司 | Optimization method and device for modeling of energy system |
| CN112928751A (en) * | 2021-02-02 | 2021-06-08 | 山东大学 | New energy power grid connection equipment configuration method based on robust optimization |
| CN112928751B (en) * | 2021-02-02 | 2023-08-22 | 山东大学 | New energy power grid-connected equipment configuration method based on robust optimization |
| CN113256446A (en) * | 2021-05-31 | 2021-08-13 | 华南理工大学 | Power distribution network robust optimization scheduling method considering data loss and achieving information physical fusion |
| CN113394771A (en) * | 2021-06-07 | 2021-09-14 | 广西大学 | Method for solving feasible domain of nonlinear power system tie line |
| CN113394771B (en) * | 2021-06-07 | 2023-01-10 | 广西大学 | Method for solving feasible domain of nonlinear power system tie line |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106849162B (en) | 2019-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106849162A (en) | Consider the grid-connected active distribution network ADAPTIVE ROBUST optimization method of a large amount of regenerative resources | |
| Patwal et al. | A novel TVAC-PSO based mutation strategies algorithm for generation scheduling of pumped storage hydrothermal system incorporating solar units | |
| Tabanjat et al. | Energy management hypothesis for hybrid power system of H2/WT/PV/GMT via AI techniques | |
| Mohammadi et al. | A review on application of artificial intelligence techniques in microgrids | |
| Tavoosi et al. | A machine learning approach for active/reactive power control of grid-connected doubly-fed induction generators | |
| Chen et al. | Data-augmented sequential deep learning for wind power forecasting | |
| Ju et al. | Ultra-short-term photovoltaic power prediction based on self-attention mechanism and multi-task learning | |
| Dong et al. | Machine-learning-based real-time economic dispatch in islanding microgrids in a cloud-edge computing environment | |
| Yang et al. | Hybrid policy-based reinforcement learning of adaptive energy management for the Energy transmission-constrained island group | |
| CN106527568A (en) | Maximum power tracker control method based on generalized dynamic fuzzy neural network | |
| Panda et al. | Hybrid neuro fuzzy approach for automatic generation control of two–area interconnected power system | |
| Li et al. | Day-ahead optimal dispatch strategy for active distribution network based on improved deep reinforcement learning | |
| CN107623337A (en) | A kind of energy management method for micro-grid | |
| Yin et al. | Expandable deep width learning for voltage control of three-state energy model based smart grids containing flexible energy sources | |
| Zheng et al. | Deep reinforcement learning based active disturbance rejection load frequency control of multi-area interconnected power systems with renewable energy | |
| CN113780622B (en) | Multi-agent reinforcement learning-based distributed scheduling method for multi-microgrid power distribution system | |
| Masero et al. | A fast implementation of coalitional model predictive controllers based on machine learning: Application to solar power plants | |
| Aburiyana et al. | Direct net load forecasting using adaptive neuro fuzzy inference system | |
| Wang et al. | Transfer-Reinforcement-Learning-Based rescheduling of differential power grids considering security constraints | |
| Liu et al. | Deep reinforcement learning based approach for dynamic optimal power flow in active distribution network | |
| Morato et al. | Advanced control for energy management of grid-connected hybrid power systems in the sugar cane industry | |
| CN106849119A (en) | Active distribution network ADAPTIVE ROBUST idle work optimization method | |
| Shi et al. | Day‐ahead optimal dispatching of hybrid power system based on deep reinforcement learning | |
| López-Cardona et al. | Proximal policy optimization with graph neural networks for optimal power flow | |
| Akhmetova et al. | Reliability assurance of the thermal energy sources using the neural network modelling |
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 |