CN109886836A - A kind of dynamic partition Prices Calculation based on partition clustering analysis - Google Patents
A kind of dynamic partition Prices Calculation based on partition clustering analysis Download PDFInfo
- Publication number
- CN109886836A CN109886836A CN201910155848.4A CN201910155848A CN109886836A CN 109886836 A CN109886836 A CN 109886836A CN 201910155848 A CN201910155848 A CN 201910155848A CN 109886836 A CN109886836 A CN 109886836A
- Authority
- CN
- China
- Prior art keywords
- power
- partition
- data
- clustering
- cluster
- 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
Abstract
The invention discloses a kind of dynamic partition Prices Calculations based on partition clustering analysis, it constructs direct current optimal power flow optimization aim and determines Operation of Electric Systems constraint condition, the water power power output after being optimized is solved, then solution node electricity price, conventional power unit and new energy unit output, flow data, the active antithesis factor and node injecting power；Generate effectively cluster attribute；Determine optimum clustering number；Dissimilarity matrix, partition clustering algorithm are generated by standardization, euclidean distance method to realize cluster；Each Price zone is divided according to cluster result, finally determines each region unified price；Power tariff calculated result is fed back again, the participant in the market in each subregion is settled accounts according to Power tariff.The present invention can determine the electricity price of each subregion, finally be settled accounts by the way of unified price to each node inside the same area, improve practical operability, meets most of user and wishes to obtain the demand of opposite price stabilization.
Description
Technical field
The invention belongs to technical field of electric power, and in particular to a kind of based on the dynamic partition electricity price of partition clustering analysis
Calculation method.
Background technique
With the development of electricity market, clearing method and type of transaction out in market gradually increase, to adapt to different regions
Electrical network feature.Deploying node is that most common spot market goes out clear mode, can consider the side of each generating set simultaneously
The physical characteristic of border cost and electric power networks, including Congestion and network loss obtain a certain moment meter and power supply to solve
The least cost of the newly-increased specific load power supply of certain node of resource, electric power networks feature.Using node limit in electricity market
Electricity price valuates to electric flux, the actual power cost of electric system can be reflected in terms of time and space two, to embody electricity
The degree of scarcity of power resource (including power generation, transmission line of electricity and loss).It is clear that, this price mechanism being capable of conduct
" invisible hand " provides accurate economic signals for each participant in market, guides the economic behaviour rationalization of main market players, from
And electric power resource service efficiency is improved, it optimizes allocation of resources.
However it is pointed out that and is lacked in the practical application interior joint Marginal Pricing mechanism of electric system there is also some
Point.Deploying node mechanism can embody the marginal cost that electric energy is supplied under short period scale, but consider electric energy supply and demand
Have the characteristics that be different from other customary commercials, including electric energy supply and demand while property, rapidity, real-time and randomness, use
Deploying node mechanism may generate the Spot Price result of sequence big ups and downs at any time；On the other hand, node limit
Price Mechanisms cover influence of the electric system constraint to electricity price, this also causes its generated electricity price result to have complexity
Spatial distribution characteristic, the Marginal Pricing of each node is for power load distributing and electric system constraint all quite sensitives.It can be seen that
In electric system practical application, deploying node will frequently be changed with space at any time, and shortage is predictable, certain
The performance of price signal effect is hindered in degree, participant in the market will have in face of the node side of thousands of acute variations
Border electricity price, the information exchange between each node is at high cost, and practical operability is poor；In addition, deploying node is purely by phase
The mathematical model answered is obtained by optimization algorithm, and with the sustainable development of electric system, electric power networks topological structure is more multiple
Miscellaneous, the amount of calculation of node electricity price will continue to increase, and the obtained electricity price result transparency is poor, to derivation algorithm
Degree of dependence it is high, it is difficult to received jointly by each side participant in market, this also becomes deploying node mechanism practical application
In one of important restriction factor.
In general, Congestion is generally only frequently apparent occurs in certain geography in practical power systems
In region, and the probability that other region Congestions occur is smaller, and situation is also smaller.By taking the Northwest's power grid as an example,
It assembles ground as new forms of energy resource, and the load level in the region as locating for new energy power supply is relatively low, and renewable energy disappears
Receive face long-distance and large-capacity power transmission, the problems such as power grid construction relatively lags behind, partial electric grid structural weak, in the practical fortune of system
Surrounding area locating for extensive new energy base in row, such as Hexi Corridor, Haixi and other regions of Qinghai Province area, it would be possible to
There is serious Congestion problem, influence new energy power output and send outside, and some areas submitting occurs other regions are then less
The case where scarce capacity.
For above-mentioned phenomenon, if the electric power networks of whole market can actually be divided to according to certain zoning ordinance
Dry isolated area, so that all nodes inside the same area have close or identical node electricity price (inside the region not
Congestion occurs), and then combine division result and node electricity price as a result, determining the electricity of each subregion by zone pricing model
Valence is finally settled accounts each node inside the same area by the way of unified price, so that improving can actually operate
Property.It is this to be based on deploying node mechanism, electricity market steering is simplified to complicated electricity price pricing method with regional relation
It is exactly locational marginal price mechanism, is applied to North Europe electricity market earliest and always, at present in Britain, U.S. Buddhist sieve
In Dazhou City be also applied successfully.
Other than the above-mentioned complete actual market application case for implementing zone pricing mechanism, other majorities use node
The actual market of electricity price also more or less applies zone pricing theory.For example, as node electricity price typical case case it
One U.S. market PJM, there is numerous small-scale price partition, and specifically, the market PJM includes the transaction of four classes altogether
Node type, physical node are only one type, other three classes are all according to the different zonules for requiring to divide, i.e., a large amount of objects
Manage the set of node.In other words, in the market PJM, the Trading Model of power grid is not fully overlapped with physical model.With collection
For node (hub), setting basic standard as defined in PJM market rules has: ensuring operation robustness, the price of electric system
Stability and electricity price it is predictable.The physical node quantity that maximum hub contains is more than 100, this similar area
The market setting that domain divides obviously embodies the price signal effect of Power tariff.Although in addition, New York electricity market is being fixed a price
On used deploying node completely, but subregion clearing are but used to the clearing of load, by entire control area point
At 15 areas, the settlement price of each period load is the average nodal electricity price of the load region period.It draws in this region
Point mode is also applied to New York zonal reserve model.The clearing form embodies most of user and wishes that acquisition is relatively stable
The requirement of price equally illustrates that the meaning of Power tariff is to be capable of providing the price signal of more simple and clear.
By the applicable cases of the Power tariff theory in above-mentioned actual node electricity price market it can be seen that although making completely
Come from scheduling process with considering that the deploying node of the Optimal Power Flow Problems physical model calculating of rack constraint out is only
, real actual node energy value, just there are accurate physics and economic meanings, and Power tariff is then needed to node
Marginal Pricing carries out more or less simplification, it is possible to it will lead to a small amount of price distortion phenomenon, however it is rationally average and appropriate
Benefit possessed by approximation can offset the secondary of price distortion in practical applications and make such as the transparency of electricity price and predictable
With.So being rarely employed in actual market and the identical deploying node mechanism of physical node meaning, majority
Market has all carried out simplification to a certain degree or has used mixed mechanism.
By the basic principle of locational marginal price mechanism it is found that the node electricity price for being divided in different zones is different, this
Mean the division in region by the value of each node electricity price of the system that directly affects.Therefore, locational marginal price mechanism is actually being answered
The critical issue solved is needed to be how to formulate rationally effective, strong applicability region division criterion with middle.It considers
The complexity of power transmission network topological structure and operation states of electric power system, the given of partition boundaries is all a ratio all the time
More difficult and controversial problem.If region division result can not accurately reflect the backlog of running
Situation, it is possible to it is less even without Congestion to will appear Congestion between each subregion, and occurs inside each subregion tight
The case where weight Congestion, lead to the price signal serious distortion of locational marginal price, generates the market orientation of mistake.In electricity
In power Market Design, a kind of deploying node reduced form or suitable locational marginal price are designed, makes it in practical fortune
It is effective and easy to operate particularly significant in row.
Summary of the invention
It is based on drawing in view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by the present invention is that providing one kind
The dynamic partition Prices Calculation of point clustering has both efficiency and practical, is particularly suitable for complex space point
The large-scale power system of cloth feature.
The invention adopts the following technical scheme:
A kind of dynamic partition Prices Calculation based on partition clustering analysis, acquisition electric system basic technical data,
Operation of Electric Systems constraint condition data, quote data；It constructs direct current optimal power flow optimization aim and determines Operation of Electric Systems
Constraint condition solves the water power after being optimized using direct current optimal power flow optimization object function according to the data of acquisition and contributes,
Then to minimize the power purchase expense of thermoelectricity, water power, new energy as objective function solution node electricity price, conventional power unit and new energy
Source unit output, flow data, the active antithesis factor and node injecting power；According to node electricity price calculated result and trend because
Son generates effectively cluster attribute；Density index, range index, given threshold are calculated, determines optimum clustering number；Pass through standardization
Processing, euclidean distance method generate dissimilarity matrix, partition clustering algorithm to realize cluster；Each electricity price is divided according to cluster result
Region finally determines each region unified price；Power tariff calculated result is fed back again, according to Power tariff to each point
Participant in the market in area settles accounts.
Specifically, it is as follows to establish optimization aim using the power purchase expense for minimizing thermoelectricity and water power as objective function:
Wherein, T represents the quotation period, and C is conventional power unit, and H is Hydropower Unit, and G_C is conventional power unit power output, and G_H is water
Electric unit output.
Specifically, Operation of Electric Systems constraint condition includes node power Constraints of Equilibrium, the constraint of line transmission power limit
It is constrained with unit output, solves the Unit Combination model of building, the water power power output GH0 after being optimized.
Further, Operation of Electric Systems constraint condition calculates as follows:
S.T.eT(PG_C+PG_H-PD)=0
Wherein, eTFor unit row vector, PG_CFor conventional power unit power output, PG_HFor Hydropower Unit power output, PDFor load, T generation
The statistical tables and reports valence period,For the tidal current limit of route.
Specifically, specific as follows as objective function using the power purchase expense for minimizing thermoelectricity, water power and new energy:
Min∑G_C*Bid_C+∑G_H*Bid_H+∑G_R*Bid_R
Wherein, G_C is conventional power unit power output, and Bid_C is conventional power unit quotation, and G_H is Hydropower Unit power output, and Bid_H is
Hydropower Unit quotation, G_R are new energy unit output, and Bid_R is new energy Bidding.
Specifically, effectively cluster attribute includes the following:
Based on existing system rack data, the power flow transfer matrix of system is calculated, as a kind of cluster attribute；It reads
The node electricity price data for entering each node, as second of cluster attribute；
The power flow transfer matrix of system calculates as follows:
Wherein, T represents quotation period, PGFor each node power generation injecting power, PDFor each node load power.
Specifically, the data local density index ρ of each nodeiIt calculates as follows:
Wherein, f is the function that value is 0 or 1, as d (xi,xj) < dcWhen, functional expression f (d (xi,xj)-dc) value be 1,
On the contrary then value is 0；Lumped parameter ΘiCharacterize data object xiIn close region immediate k number according to object set；
d(xi,xj) indicate data object xiAnd xjBetween metric range；
The packing density range index δ of each nodeiIt calculates as follows:
。
Specifically, clustering attribute based on gained, it is as follows to generate data matrix, and is standardized unified dimension
Euclidean distance calculates as follows:
Wherein, xitT-th of cluster attribute of characterize data object；
It is as follows to calculate generation dissimilarity matrix:
Based on data matrix and dissimilarity matrix, under obtained optimum clustering number, using the realization pair of partition clustering algorithm
The clustering of system node.
Further, the step of partition clustering algorithm is as follows:
A) k number is chosen from raw data set N according to as initial clustering cluster center；
B) each data object is divided at a distance from each clustering cluster center by each data object in metric data collection N
The clustering cluster center closest with it generates clustering cluster；
C) based on the assessment formula of clustering cluster compactness, objective function Z corresponding to current cluster result is obtained；
Wherein, n is each clustering cluster internal data in data set, ciFor cluster centre, Z is characterized as n and corresponding cluster centre
ciThe sum of metric range, α is distance metric parameter, NiThe data set covered for characterization clustering cluster i；
D) center of each clustering cluster, return step are successively updated based on weighted average criterion according to current cluster result
b；
E) the above iterative step is repeated, until objective function Z or clustering cluster center ciUntil no longer changing.
Specifically, electric system basic technical data includes number of nodes NodeN, route number LineN, 24 period systems
Load SystemLoad；Transmission line data include resistance R, reactance X, susceptance B；Transformer data include reactance X, no-load voltage ratio K；Electricity
It is each generating set power output upper lower limit value, including conventional power unit GenBound, Hydropower Unit that Force system, which runs constraint condition data,
HBound, new energy unit RBound；Branch Power Flow capacity limit FlowBound；Quote data is 24 periods of generator
Quotation information, including conventional power unit Bid_C, Hydropower Unit Bid_H, new energy unit Bid_R.
Compared with prior art, the present invention at least has the advantages that
A kind of dynamic partition Prices Calculation based on partition clustering analysis of the present invention, proposes dynamic partition electricity price meter
Calculation method efficient zoned can be realized to a certain extent to the simplification of deploying node, can offset valence in practical applications
The side effect of lattice distortion.On the other hand, the dynamic partition Prices Calculation based on partition clustering of proposition, can pass through conjunction
Office and appropriate approximation are patted, the transparent and predictable of electricity price is enhanced.Can overcome existing partition method it is cumbersome, it is not transparent enough,
Not readily understood, not very practical disadvantage is realized quick, efficient zoned.The electricity of each subregion is determined by zone pricing model
Valence is finally settled accounts each node inside the same area by the way of unified price, and can be improved can actually operate
Property, more meet most of user and wishes to obtain the requirement of opposite price stabilization.
Further, it establishes optimization aim and enables to finally obtained the result is that meeting expectation target.
Further, setting Operation of Electric Systems constraint condition can guarantee that calculated result meets the operation of electric system
Demand.
Further, it using the power purchase expense for minimizing thermoelectricity, water power and new energy as objective function, enables to
Under final calculation result, total power purchase expense is minimum.
Further, setting flow transferring relativity factor is effectively to cluster attribute, can consider each node to a certain extent
Distribution of electricity prices feature has more apparent economic meanings.And power flow transfer matrix T only with the topological structure of power grid and
The electrical relating to parameters of route, therefore in the case that topological structure is constant, influence suffered by region division result will be compared with
It is few, it can guarantee the relative stability of subregion to a certain extent.
Further, each region can effectively be identified by data local density index and packing density range index being arranged
Density maximum point, so that it is determined that optimum clustering number, that is, the number of density maximum point.
Further, dissimilarity matrix is generated, each element can characterize the distinctiveness ratio measured value between data sample two-by-two,
The distinctiveness ratio of two data samples is smaller, then the value is closer to 0；Distinctiveness ratio is bigger, then the value is bigger.
Further, partition clustering algorithm has strong applicability, the fireballing benefit of cluster.
In conclusion the present invention can overcome the spy that existing method is cumbersome, not transparent enough, not readily understood, not very practical
Point, realization is quick, efficient zoned, and the zone pricing calculation method proposed through the invention can determine the electricity price of each subregion, most
Each node inside the same area is settled accounts by the way of unified price eventually, can be improved practical operability, more
Meet most of user to wish to obtain the demand of opposite price stabilization.
Below by drawings and examples, technical scheme of the present invention will be described in further detail.
Detailed description of the invention
Fig. 1 is flow chart of the present invention.
Specific embodiment
Referring to Fig. 1, a kind of dynamic partition Prices Calculation based on partition clustering analysis of the present invention, including it is following
Step:
S1, electric system basic technical data, Operation of Electric Systems constraint condition data, quotation are obtained from relevant departments
Data；
Electric system basic technical data:
Number of nodes NodeN, route number LineN, 24 period system loading SystemLoad, transmission line data: resistance R,
Reactance X, susceptance B；Transformer data: reactance X, no-load voltage ratio K.
Operation of Electric Systems constraint condition data:
Each generating set power output upper lower limit value: conventional power unit GenBound, Hydropower Unit HBound, new energy unit
RBound；Branch Power Flow capacity limit FlowBound.
Quote data:
The quotation information of 24 periods of generator: conventional power unit Bid_C, Hydropower Unit Bid_H, new energy unit Bid_
R。
S2, building optimal load flow model
S201, building direct current optimal power flow optimization aim
Power purchase expense to minimize thermoelectricity and water power establishes optimization aim as objective function, and optimization aim is by formula (1)
It obtains:
Wherein, T represents the quotation period, generally takes 24.C is conventional power unit, and H is Hydropower Unit, and G_C goes out for conventional power unit
Power, G_H are Hydropower Unit power output.
S202, building Operation of Electric Systems constraint condition
Node power Constraints of Equilibrium；
The constraint of line transmission power limit；
Unit output constraint, comprising: fired power generating unit is contributed, and bound constrains, Hydropower Unit power output bound constrains；
Wherein, eTFor unit row vector, PG_CFor conventional power unit power output, PG_HFor Hydropower Unit power output, PDFor load, T generation
The statistical tables and reports valence period,For the tidal current limit of route.
The Unit Combination model for solving building, the water power power output GH0 after being optimized.
S3, by the electric system basic technical data obtained by step S1, Operation of Electric Systems constraint condition data, report
Valence mumber solves the water power power output after being optimized according to being input in the optimal load flow model of step S2 building；Then to minimize
Thermoelectricity, water power, new energy power purchase expense be that objective function solves corresponding optimal load flow model；
Shown in objective function such as formula (3):
Min∑G_C*Bid_C+∑G_H*Bid_H+∑G_R*Bid_R (3)
Wherein, G_C is conventional power unit power output, and Bid_C is conventional power unit quotation, and G_H is Hydropower Unit power output, and Bid_H is
Hydropower Unit quotation, G_R are new energy unit output, and Bid_R is new energy Bidding.With minimize thermoelectricity, water power and
The power purchase expense of new energy is objective function, and the constraint of three will meet node power Constraints of Equilibrium, line transmission power simultaneously
Limit restraint, unit output constraint.The power output of Unit Combination before the power output of water power is equal to.According to conventional Lagrangian letter
Number:
Wherein,For Lagrange multiplier corresponding with constraint.Then most had ready conditions according to KKT:
In view of for unit of offering, quotation capacity will not get over bound operation, generallyThen the i-th node
Node electricity price be:
Last solving result includes: node electricity price, conventional power unit and new energy unit output, flow data, active antithesis
The factor, node injecting power.
S4, clustering obtain Power tariff；
S401, effectively cluster attribute is generated according to node electricity price calculated result and the trend factor；
Based on existing system rack data, the power flow transfer matrix of system is calculated, as a kind of cluster attribute:
Wherein, T represents quotation period, PGFor each node power generation injecting power, PDFor each node load power.
The node electricity price data for reading in each node, as second of cluster attribute；
S402, density index, range index, given threshold are calculated, determines optimum clustering number；
The data local density index of each node calculates as follows:
F is the function that value is 0 or 1, as d (xi,xj) < dcWhen, functional expression f (d (xi,xj)-dc) value be 1, it is on the contrary
Then value is 0；Lumped parameter ΘiCharacterize data object xiIn close region immediate k number according to object set；d
(xi,xj) indicate data object xiAnd xjBetween metric range.From formula as can be seen that a data object, if its data
Local density's value is larger, then means that the packing density in its locating region is higher, on the contrary then lower.
The packing density range index of each node calculates as follows:
Only when data object xi is regional area or the maximum data object of global density, packing density distance
Value just can be larger；Conversely, the value of its packing density distance will significantly reduce.In other words, each data pair are depended on
The packing density range index of elephant, we can effectively identify the density maximum point in each region, and these density maximum points,
As long as the cluster center of each clustering cluster can be considered as in Density Clustering method, which means that having navigated to these density maximums
Point, then the number of these density maximum points is exactly optimum clustering number.
S403, dissimilarity matrix, partition clustering algorithm are generated by standardization, euclidean distance method to realize cluster；
Attribute is clustered based on gained, is generated data matrix (10), and is standardized unified dimension
Euclidean distance calculates as follows:
Wherein, xitT-th of cluster attribute of characterize data object.It is pointed out that working as q in formula (11) is 1,2 and
When leveling off to infinite, then formula (11) will be reduced to absolute value distance, Euclidean distance and Chebyshev's distance.Wherein Euclidean away from
From be most widely used, have typical real space meaning.Since the Node distribution of electric system is that have true geography
Space meaning, therefore the present invention still uses the close degree between Euclidean distance calculate node.
It is as follows to calculate generation dissimilarity matrix:
Based on data matrix and dissimilarity matrix, under obtained optimum clustering number, using the realization pair of partition clustering algorithm
The clustering of system node；Steps are as follows:
A) k number is chosen from raw data set N according to as initial clustering cluster center；
B) each data object is divided at a distance from each clustering cluster center by each data object in metric data collection N
The clustering cluster center closest with it, to generate clustering cluster；
C) based on the assessment formula of clustering cluster compactness, objective function Z corresponding to current cluster result is obtained；
Wherein, n is each clustering cluster internal data in data set, ciFor cluster centre, usually choose in a clustering cluster
For the weighted average of all data as cluster centre, Z is characterized as n and corresponding cluster centre ciThe sum of metric range, α is
Distance metric parameter, general value are 2, NiThe data set covered for characterization clustering cluster i；
D) center of each clustering cluster, return step are successively updated based on weighted average criterion according to current cluster result
b；
E) above-mentioned didactic iterative step is repeated, until objective function Z or clustering cluster center ciUntil no longer changing.
S404, each Price zone is divided according to cluster result, final each region determines unified price.
Based on gained cluster result, the node collection for being located at same clustering cluster is divided to the same area, realizes that region is drawn
Point；Based on gained region division as a result, taking the weighted average of each regional nodes Marginal Pricing as the unified area in each region
Domain electricity price.
S5, above-mentioned Power tariff calculated result is fed back into relevant departments, and then according to Power tariff in each subregion
Participant in the market settles accounts.
Although traditional deploying node mechanism can embody the marginal cost that electric energy is supplied under short period scale,
But in view of electric energy supply and demand has the characteristics that be different from other customary commercials, including electric energy supply and demand while property, rapidity, reality
When property and randomness may generate the Spot Price result of sequence big ups and downs at any time using deploying node mechanism；
Dynamic partition Prices Calculation proposed by the present invention efficient zoned can be realized to a certain extent to deploying node
Simplify, the side effect of price distortion can be offset in practical applications；
Traditional deploying node mechanism covers influence of the electric system constraint to electricity price, this also causes it to be produced
Raw electricity price result has a complicated spatial distribution characteristic, the Marginal Pricing of each node for power load distributing and electric system about
Beam all quite sensitives.In electric system practical application, deploying node will frequently be changed with space at any time, be lacked
It is predictable, the performance of price signal effect is hindered to a certain extent, and participant in the market will have in face of thousands of plays
The deploying node of strong variation, the information exchange between each node is at high cost, and practical operability is poor.The present invention proposes
The dynamic partition Prices Calculation based on partition clustering, the saturating of electricity price can be enhanced by reasonable draw and appropriate approximate
Bright property and predictable.
Deploying node is purely obtained by corresponding mathematical model by optimization algorithm, with continuing for electric system
Development, the amount of calculation of electric power networks topological structure even more complex, node electricity price will continue to increase, and obtained electricity
The valence result transparency is poor, high to the degree of dependence of derivation algorithm, it is difficult to be received jointly by each side participant in market.This hair
The new price partition method based on Clustering Theory of bright proposition, can overcome existing partition method it is cumbersome, it is not transparent enough, be not easy
Understand, not very practical disadvantage, realizes quick, efficient zoned.The electricity price of each subregion is determined by zone pricing model, most
Each node inside the same area is settled accounts by the way of unified price eventually, can be improved practical operability, more
Meet most of user to wish to obtain the requirement of opposite price stabilization.
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.The present invention being described and shown in usually here in attached drawing is real
The component for applying example can be arranged and be designed by a variety of different configurations.Therefore, below to the present invention provided in the accompanying drawings
The detailed description of embodiment be not intended to limit the range of claimed invention, but be merely representative of choosing of the invention
Determine embodiment.Based on the embodiments of the present invention, those of ordinary skill in the art institute without creative efforts
The every other embodiment obtained, shall fall within the protection scope of the present invention.
It is the calculated result that price partition is carried out with the paddy lotus period under the present invention typical day big to certain power grid winter below:
Power tariff calculation method proposed by the present invention identifies that optimum partition number of the power grid at the moment is 6, will be electric
Web area is divided into six pieces, and the electricity price of subregion 1-6 is respectively 0.3694,0.3653,0.3814,0.3586,0.3734,0.3, single
Position is member/kWh.
With the power grid locating for compared with actual geographic information, which effectively has identified following information:
New energy rich region (subregion 6), coal electricity low price region (subregion 4), dominant eigenvalues injection zone (subregion 2),
Show effectiveness of the invention.
The above content is merely illustrative of the invention's technical idea, and this does not limit the scope of protection of the present invention, all to press
According to technical idea proposed by the present invention, any changes made on the basis of the technical scheme each falls within claims of the present invention
Protection scope within.
Claims (10)
1. a kind of dynamic partition Prices Calculation based on partition clustering analysis, which is characterized in that it is basic to obtain electric system
Technical data, Operation of Electric Systems constraint condition data, quote data；It constructs direct current optimal power flow optimization aim and determines electric power
System runs constraint condition, and the water after being optimized is solved using direct current optimal power flow optimization object function according to the data of acquisition
Electricity power output, then to minimize the power purchase expense of thermoelectricity, water power, new energy as objective function solution node electricity price, conventional power unit
With new energy unit output, flow data, the active antithesis factor and node injecting power；According to node electricity price calculated result and tide
It flows the factor and generates effectively cluster attribute；Density index, range index, given threshold are calculated, determines optimum clustering number；Pass through standard
Change processing, euclidean distance method generate dissimilarity matrix, partition clustering algorithm to realize cluster；Each electricity price is divided according to cluster result
Region finally determines each region unified price；Power tariff calculated result is fed back again, according to Power tariff to each subregion
In participant in the market settle accounts.
2. the dynamic partition Prices Calculation according to claim 1 based on partition clustering analysis, which is characterized in that with
Minimizing the power purchase expense of thermoelectricity and water power, to be that objective function establishes optimization aim as follows:
Wherein, T represents the quotation period, and C is conventional power unit, and H is Hydropower Unit, and G_C is conventional power unit power output, and G_H is hydroelectric machine
Group power output.
3. the dynamic partition Prices Calculation according to claim 1 or 2 based on partition clustering analysis, feature exist
In, Operation of Electric Systems constraint condition include node power Constraints of Equilibrium, the constraint of line transmission power limit and unit output about
Beam solves the Unit Combination model of building, the water power power output GH0 after being optimized.
4. the dynamic partition Prices Calculation according to claim 3 based on partition clustering analysis, which is characterized in that electricity
It is as follows that Force system runs constraint condition calculating:
S.T.eT(PG_C+PG_H-PD)=0
Wherein, eTFor unit row vector, PG_CFor conventional power unit power output, PG_HFor Hydropower Unit power output, PDFor load, T represents quotation
Period,For the tidal current limit of route.
5. the dynamic partition Prices Calculation according to claim 1 based on partition clustering analysis, which is characterized in that with
The power purchase expense for minimizing thermoelectricity, water power and new energy is that objective function is specific as follows:
Min∑G_C*Bid_C+∑G_H*Bid_H+∑G_R*Bid_R
Wherein, G_C is conventional power unit power output, and Bid_C is conventional power unit quotation, and G_H is Hydropower Unit power output, and Bid_H is water power
Bidding, G_R are new energy unit output, and Bid_R is new energy Bidding.
6. the dynamic partition Prices Calculation according to claim 1 based on partition clustering analysis, which is characterized in that have
Effect clusters attribute
Based on existing system rack data, the power flow transfer matrix of system is calculated, as a kind of cluster attribute；Read in each section
The node electricity price data of point, as second of cluster attribute；
The power flow transfer matrix of system calculates as follows:
Wherein, T represents quotation period, PGFor each node power generation injecting power, PDFor each node load power.
7. the dynamic partition Prices Calculation according to claim 1 based on partition clustering analysis, which is characterized in that each
The data local density index ρ of nodeiIt calculates as follows:
Wherein, f is the function that value is 0 or 1, as d (xi,xj) < dcWhen, functional expression f (d (xi,xj)-dc) value be 1, it is on the contrary
Then value is 0；Lumped parameter ΘiCharacterize data object xiIn close region immediate k number according to object set；d(xi,
xj) indicate data object xiAnd xjBetween metric range；
The packing density range index δ of each nodeiIt calculates as follows:
。
8. the dynamic partition Prices Calculation according to claim 1 based on partition clustering analysis, which is characterized in that base
Attribute is clustered in gained, it is as follows to generate data matrix, and is standardized unified dimension
Euclidean distance calculates as follows:
Wherein, xitT-th of cluster attribute of characterize data object；
It is as follows to calculate generation dissimilarity matrix:
Based on data matrix and dissimilarity matrix, under obtained optimum clustering number, realized using partition clustering algorithm to system
The clustering of node.
9. the dynamic partition Prices Calculation according to claim 8 based on partition clustering analysis, which is characterized in that draw
The step of dividing clustering algorithm is as follows:
A) k number is chosen from raw data set N according to as initial clustering cluster center；
B) through each data object in metric data collection N at a distance from each clustering cluster center, by each data object be divided to and its
Closest clustering cluster center generates clustering cluster；
C) based on the assessment formula of clustering cluster compactness, objective function Z corresponding to current cluster result is obtained；
Wherein, n is each clustering cluster internal data in data set, ciFor cluster centre, Z is characterized as n and corresponding cluster centre ciDegree
Sum of the distance is measured, α is distance metric parameter, NiThe data set covered for characterization clustering cluster i；
D) center of each clustering cluster, return step b are successively updated based on weighted average criterion according to current cluster result；
E) the above iterative step is repeated, until objective function Z or clustering cluster center ciUntil no longer changing.
10. the dynamic partition Prices Calculation according to claim 1 based on partition clustering analysis, which is characterized in that
Electric system basic technical data includes number of nodes NodeN, route number LineN, 24 period system loading SystemLoad；It passes
Defeated line number is according to including resistance R, reactance X, susceptance B；Transformer data include reactance X, no-load voltage ratio K；Operation of Electric Systems constraint condition
Data are each generating set power output upper lower limit value, including conventional power unit GenBound, Hydropower Unit HBound, new energy unit
RBound；Branch Power Flow capacity limit FlowBound；Quote data is the quotation information of 24 periods of generator, including routine
B of Unit id_C, Hydropower Unit Bid_H, new energy unit Bid_R.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910155848.4A CN109886836B (en) | 2019-03-01 | 2019-03-01 | Dynamic partition electricity price calculation method based on partition clustering analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910155848.4A CN109886836B (en) | 2019-03-01 | 2019-03-01 | Dynamic partition electricity price calculation method based on partition clustering analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109886836A true CN109886836A (en) | 2019-06-14 |
| CN109886836B CN109886836B (en) | 2021-01-19 |
Family
ID=66930240
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910155848.4A Active CN109886836B (en) | 2019-03-01 | 2019-03-01 | Dynamic partition electricity price calculation method based on partition clustering analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109886836B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717079A (en) * | 2019-09-23 | 2020-01-21 | 中国南方电网有限责任公司 | Electricity price partitioning method and device, computer equipment and storage medium |
| CN110910173A (en) * | 2019-11-25 | 2020-03-24 | 深圳市深电能售电有限公司 | Power price analysis method for power spot market node |
| CN110956493A (en) * | 2019-11-18 | 2020-04-03 | 远光软件股份有限公司 | Method and device for predicting node electricity price through virtual area node |
| CN111461409A (en) * | 2020-03-10 | 2020-07-28 | 国网山西省电力公司经济技术研究院 | Abnormal value processing method for medium and long-term load data |
| CN111951121A (en) * | 2020-07-20 | 2020-11-17 | 广东电力交易中心有限责任公司 | Electric power spot market quotation mode classification method, device and storage medium |
| CN112184009A (en) * | 2020-05-11 | 2021-01-05 | 国网宁夏电力有限公司 | Method and device for clearing paid reactive auxiliary service and storage medium |
| CN112200596A (en) * | 2020-09-01 | 2021-01-08 | 中国南方电网有限责任公司 | Method, system, device and medium for determining regional marginal electricity price of power system |
| CN114142521A (en) * | 2021-11-30 | 2022-03-04 | 国网陕西省电力公司 | Multi-objective optimization scheduling method and system for distributed new energy power distribution network |
| CN114172208A (en) * | 2021-11-29 | 2022-03-11 | 国网山东省电力公司郯城县供电公司 | New energy consumption optimization system based on multi-region interconnection |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007028720A (en) * | 2005-07-13 | 2007-02-01 | Hitachi Ltd | System for predicting power loss at adjustment of power supply, method of predicting power loss at adjustment of power supply, and program for predicting power loss at adjustment of power supply |
| CN105356450A (en) * | 2015-10-28 | 2016-02-24 | 国家电网公司西北分部 | Power grid subarea division method based on dynamic electricity prices |
| CN106469337A (en) * | 2016-09-30 | 2017-03-01 | 西安交通大学 | Consider the design of subregion power capacity market model and the pricing method of transregional interconnection constraint |
| CN106972532A (en) * | 2017-04-26 | 2017-07-21 | 华中科技大学 | A kind of wind-powered electricity generation Power tariff evaluation method compensated based on peak regulation assistant service |
| CN107017618A (en) * | 2017-03-15 | 2017-08-04 | 中国电力科学研究院 | A kind of active power distribution network division of the power supply area method and device |
| CN108648024A (en) * | 2018-06-08 | 2018-10-12 | 河海大学 | A kind of power distribution network distributed generation resource deploying node computational methods |
| CN109146553A (en) * | 2018-07-27 | 2019-01-04 | 东北电力大学 | Spot Price forecasting system and its method based on multi-density cluster and multicore SVM |
-
2019
- 2019-03-01 CN CN201910155848.4A patent/CN109886836B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007028720A (en) * | 2005-07-13 | 2007-02-01 | Hitachi Ltd | System for predicting power loss at adjustment of power supply, method of predicting power loss at adjustment of power supply, and program for predicting power loss at adjustment of power supply |
| CN105356450A (en) * | 2015-10-28 | 2016-02-24 | 国家电网公司西北分部 | Power grid subarea division method based on dynamic electricity prices |
| CN106469337A (en) * | 2016-09-30 | 2017-03-01 | 西安交通大学 | Consider the design of subregion power capacity market model and the pricing method of transregional interconnection constraint |
| CN107017618A (en) * | 2017-03-15 | 2017-08-04 | 中国电力科学研究院 | A kind of active power distribution network division of the power supply area method and device |
| CN106972532A (en) * | 2017-04-26 | 2017-07-21 | 华中科技大学 | A kind of wind-powered electricity generation Power tariff evaluation method compensated based on peak regulation assistant service |
| CN108648024A (en) * | 2018-06-08 | 2018-10-12 | 河海大学 | A kind of power distribution network distributed generation resource deploying node computational methods |
| CN109146553A (en) * | 2018-07-27 | 2019-01-04 | 东北电力大学 | Spot Price forecasting system and its method based on multi-density cluster and multicore SVM |
Non-Patent Citations (6)
| Title |
|---|
| HONGMING YANG ET AL: ""A RBFN Hierarchical Clustering Based Network Partitioning Method for Zonal"", 《2ND INTERNATIONAL CONFERENCE ON ELECTRICAL AND ELECTRONICS ENGINEERING (ICEEE) AND XI CONFERENCE ON ELECTRICAL ENGINEERING (CIE 2005)》 * |
| HONGMING YANG ET AL: ""Monte Carlo Simulation Based Price Zone Partitioning Considering Market Uncertainty"", 《9TH INTERNATIONAL CONFERENCE ON PROBABILISTIC METHODS APPLIED TO POWER SYSTEMS》 * |
| YANGYANG LIU ET AL: ""A Novel RLMP Model Considering Power System Operation Risk"", 《IEEE》 * |
| 付蓉 等: ""基于节点电价灵敏度模糊聚类的电价分区方法"", 《电力需求侧管理》 * |
| 宋嗣博 等: ""基于节点边际电价的电力市场分区策略研究"", 《电力建设》 * |
| 汤慧娣: ""电力市场分区电价理论与应用"", 《中国优秀硕士学位论文全文数据库 经济与管理科学辑》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717079A (en) * | 2019-09-23 | 2020-01-21 | 中国南方电网有限责任公司 | Electricity price partitioning method and device, computer equipment and storage medium |
| CN110717079B (en) * | 2019-09-23 | 2022-01-04 | 中国南方电网有限责任公司 | Electricity price partitioning method and device, computer equipment and storage medium |
| CN110956493A (en) * | 2019-11-18 | 2020-04-03 | 远光软件股份有限公司 | Method and device for predicting node electricity price through virtual area node |
| CN110910173A (en) * | 2019-11-25 | 2020-03-24 | 深圳市深电能售电有限公司 | Power price analysis method for power spot market node |
| CN111461409A (en) * | 2020-03-10 | 2020-07-28 | 国网山西省电力公司经济技术研究院 | Abnormal value processing method for medium and long-term load data |
| CN112184009A (en) * | 2020-05-11 | 2021-01-05 | 国网宁夏电力有限公司 | Method and device for clearing paid reactive auxiliary service and storage medium |
| CN111951121A (en) * | 2020-07-20 | 2020-11-17 | 广东电力交易中心有限责任公司 | Electric power spot market quotation mode classification method, device and storage medium |
| CN112200596A (en) * | 2020-09-01 | 2021-01-08 | 中国南方电网有限责任公司 | Method, system, device and medium for determining regional marginal electricity price of power system |
| CN114172208A (en) * | 2021-11-29 | 2022-03-11 | 国网山东省电力公司郯城县供电公司 | New energy consumption optimization system based on multi-region interconnection |
| CN114172208B (en) * | 2021-11-29 | 2022-08-19 | 国网山东省电力公司郯城县供电公司 | New energy consumption optimization system based on multi-region interconnection |
| CN114142521A (en) * | 2021-11-30 | 2022-03-04 | 国网陕西省电力公司 | Multi-objective optimization scheduling method and system for distributed new energy power distribution network |
| CN114142521B (en) * | 2021-11-30 | 2023-08-25 | 国网陕西省电力公司 | Multi-target optimal scheduling method and system for distributed new energy power distribution network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109886836B (en) | 2021-01-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109886836A (en) | A kind of dynamic partition Prices Calculation based on partition clustering analysis | |
| Lee et al. | Distributed energy trading in microgrids: A game-theoretic model and its equilibrium analysis | |
| Huang et al. | A clustering based grouping method of nearly zero energy buildings for performance improvements | |
| Yi et al. | Coordinated operation strategy for a virtual power plant with multiple DER aggregators | |
| Faqiry et al. | Distributed bilevel energy allocation mechanism with grid constraints and hidden user information | |
| He et al. | Water-filling exact solutions for load balancing of smart power grid systems | |
| Heymann et al. | Orchestrating incentive designs to reduce adverse system-level effects of large-scale EV/PV adoption–The case of Portugal | |
| Wang et al. | Autonomous energy community based on energy contract | |
| CN108717608A (en) | Million kilowatt beach photovoltaic plant accesses electric network synthetic decision-making technique and system | |
| Clastres et al. | An analytical approach to activating demand elasticity with a demand response mechanism | |
| ALSalloum et al. | Demand side management in smart grids: A stackelberg multi period multi provider game | |
| Hussain et al. | Flexibility: Literature review on concepts, modeling, and provision method in smart grid | |
| Seklos et al. | Designing a distribution level flexibility market using mechanism design and optimal power flow | |
| Tabors et al. | Distributed energy resources: New markets and new products | |
| Wang et al. | A real time peer-to-peer energy trading for prosumers utilizing time-varying building virtual energy storage | |
| Ahmadiahangar et al. | Demand-side flexibility in smart grid | |
| Alvina et al. | Smart community based solution for energy management: an experimental setup for encouraging residential and commercial consumers participation in demand response program | |
| Harirchi et al. | Optimal payment sharing mechanism for renewable energy aggregation | |
| Zheng et al. | Bargaining‐based cooperative game among multi‐aggregators with overlapping consumers in incentive‐based demand response | |
| Abdel-Raouf et al. | A survey of game theory applications in electrical power micro-grid systems | |
| Saini et al. | Multi-objective day-ahead localized reactive power market clearing model using HFMOEA | |
| Oleinikova et al. | Market design for electricity ensuring operational flexibility | |
| Knirsch et al. | Trust-less electricity consumption optimization in local energy communities | |
| Spasova et al. | Energy exchange strategy for local energy markets with heterogenous renewable sources | |
| Akter et al. | Impacts of random household participations on a transactive open energy market in residential microgrids |
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 |