CN112467803B - Distributed power supply and novel load typical scene generation method and system - Google Patents
Distributed power supply and novel load typical scene generation method and system Download PDFInfo
- Publication number
- CN112467803B CN112467803B CN202011298627.1A CN202011298627A CN112467803B CN 112467803 B CN112467803 B CN 112467803B CN 202011298627 A CN202011298627 A CN 202011298627A CN 112467803 B CN112467803 B CN 112467803B
- Authority
- CN
- China
- Prior art keywords
- load
- distributed power
- power supply
- output
- optimized
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000005457 optimization Methods 0.000 claims abstract description 76
- 238000005265 energy consumption Methods 0.000 claims abstract description 37
- 230000008859 change Effects 0.000 claims abstract description 14
- 239000002245 particle Substances 0.000 claims description 64
- 238000004422 calculation algorithm Methods 0.000 claims description 30
- 238000004146 energy storage Methods 0.000 claims description 17
- 230000011218 segmentation Effects 0.000 claims description 17
- 230000005611 electricity Effects 0.000 claims description 16
- 230000009467 reduction Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 230000007613 environmental effect Effects 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 34
- 238000004458 analytical method Methods 0.000 description 9
- 238000011161 development Methods 0.000 description 6
- 230000018109 developmental process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000007418 data mining Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000007599 discharging Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 230000002411 adverse Effects 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 2
- 238000005485 electric heating Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- CATMPQFFVNKDEY-AAEUAGOBSA-N gamma-Glu-Trp Chemical compound C1=CC=C2C(C[C@H](NC(=O)CC[C@H](N)C(O)=O)C(O)=O)=CNC2=C1 CATMPQFFVNKDEY-AAEUAGOBSA-N 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- XOFYZVNMUHMLCC-ZPOLXVRWSA-N prednisone Chemical compound O=C1C=C[C@]2(C)[C@H]3C(=O)C[C@](C)([C@@](CC4)(O)C(=O)CO)[C@@H]4[C@@H]3CCC2=C1 XOFYZVNMUHMLCC-ZPOLXVRWSA-N 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- 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
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
- H02J3/466—Scheduling the operation of the generators, e.g. connecting or disconnecting generators to meet a given demand
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- 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
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/12—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load
- H02J3/14—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load by switching loads on to, or off from, network, e.g. progressively balanced loading
- H02J3/144—Demand-response operation of the power transmission or distribution network
-
- 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
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/28—Arrangements for balancing of the load in a network by storage of energy
-
- 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]
-
- 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
- H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
- H02J2300/40—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation wherein a plurality of decentralised, dispersed or local energy generation technologies are operated simultaneously
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B70/00—Technologies for an efficient end-user side electric power management and consumption
- Y02B70/30—Systems integrating technologies related to power network operation and communication or information technologies for improving the carbon footprint of the management of residential or tertiary loads, i.e. smart grids as climate change mitigation technology in the buildings sector, including also the last stages of power distribution and the control, monitoring or operating management systems at local level
- Y02B70/3225—Demand response systems, e.g. load shedding, peak shaving
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S20/00—Management or operation of end-user stationary applications or the last stages of power distribution; Controlling, monitoring or operating thereof
- Y04S20/20—End-user application control systems
- Y04S20/222—Demand response systems, e.g. load shedding, peak shaving
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Power Engineering (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Educational Administration (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a distributed power supply and novel load typical scene generation method and system. The method comprises the following steps: acquiring the output capacity of a distributed power supply and the demand of a novel load; establishing a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target; optimizing the output capacity of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output capacity of the distributed power supply and the optimized demand of the novel load; clustering the optimized output capacity of the distributed power supply and the optimized novel load demand to obtain a clustered data set of the output capacity of the distributed power supply and the novel load; and (4) according to the output of the distributed power supply and a novel load clustering data set, carrying out scene classification by adopting a method of maximizing scene information entropy change quantity, and generating various typical scenes. By adopting the method and the system, the typical scene can be generated while the safety, the reliability and the environmental protection of the operation of the power grid are ensured.
Description
Technical Field
The invention relates to the technical field of power system optimization scheduling, in particular to a distributed power supply and novel load typical scene generation method and system.
Background
In order to alleviate the energy crisis and solve the environmental problem, the utilization rate of renewable energy in the power distribution network is increasing day by day, and the application range of novel loads represented by electric vehicles and electric heating in the power distribution network is expanding continuously. However, the randomness and volatility of new energy sources pose risks and challenges to the operation of power distribution grid dispatch. With the continuous development of the power distribution network technology, people not only can meet the requirement of realizing the safe and stable operation of the power distribution network, but also can pay more attention to the operation benefits brought by the power distribution network. In order to fully exert the efficiency of all assets and equipment in a power grid, meet the current situation that a large amount of renewable energy sources are accessed and the load demand is continuously improved, and improve the operation benefit of the power grid to the maximum extent, the renewable energy sources need to be reasonably integrated, regulated and controlled, and a power distribution network system needs to be reasonably scheduled.
In the current society, in order to realize promoting the degree of depth of electric wire netting and internet and merge, use the advanced technology means in traditional electric wire netting, couple traditional trade and emerging technique to further improve the operation efficiency of energy distribution, promote relevant technological level, and then promote the electric power industry development on the whole. In order to deal with new challenges brought by the rapid development of databases, data mining technologies are developed to search and mine valuable information from a large amount of complex data. The data mining technology is applied to the power system, so that the reliability of power supply, good power quality, high efficiency of power grid operation and high-quality power supply service can be ensured.
In recent years, with the continuous expansion and the rapid development of intellectualization of the power supply scale of the power distribution network, more and more data types are obtained from the power distribution network, and the trend of large data of the power distribution network is increasingly obvious. How to solve the problem of large-scale power distribution network scheduling and avoid unnecessary searching calculation so as to realize multi-scene planning is a problem to be solved urgently.
Disclosure of Invention
The invention aims to provide a distributed power supply and novel load typical scene generation method and system, which can generate typical scenes while ensuring the safety, reliability and environmental protection of the operation of a power grid.
In order to achieve the purpose, the invention provides the following scheme:
a representative scene generation method, comprising:
acquiring the output power of the distributed power supply and the demand of a novel load; the novel load comprises an electric automobile;
establishing a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target;
optimizing the output of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output of the distributed power supply and the optimized demand of the novel load;
clustering the optimized distributed power output amount and the optimized novel load demand amount to obtain a distributed power output amount and a novel load clustering data set;
and according to the distributed power supply output and the novel load clustering data set, carrying out scene classification by adopting a method of maximizing scene information entropy variation, and generating various typical scenes.
Optionally, the establishing of the power distribution network optimization model with the minimum power distribution network scheduling energy consumption as a target specifically includes:
determining an objective function by taking the minimum total electric quantity absorbed by a main network, the minimum electric quantity abandoned by a distributed power supply and the minimum load peak-valley difference punishment energy consumption as targets;
determining a constraint condition; the constraint conditions comprise power balance constraint, system operation constraint, distributed power supply output constraint, energy storage system constraint and load side constraint;
determining the objective function and the constraint condition as a power distribution network optimization model;
wherein,
the objective function is determined according to the following formula:
minF＝F Buy +F Cur +F LPV
P i Buy (t)＝P sum (t)-P i DG (t)
wherein F is an objective function; f Buy Total power absorbed from the main network; f Cur Discarding the electric quantity for the distributed power supply; f LPV Penalizing energy consumption for load peak-to-valley difference; a is Buy Is the unit amount of electricity absorbed from the main network; t is the total time period; n is the total number of nodes; p i Buy (t) is the active power input from the main network to the node i in the period of t; a is a Cur Discarding the electric quantity for a unit of the distributed power supply; p i DG* (t) is the theoretical value of the active power of the node i in the period of t; p i DG (t) is an active power actual value of the node i in the t time period after electricity is abandoned; p is sum (t) is the total load of the distribution network in the period t; p i EV (t) is the electric vehicle charging power of the node i in the time period t; p i spare (t) is the backup load power of node i in the time period t; p i involve (t) supplying power of a user side demand load to a node i in a period of t; a is a LPV Punishment of energy consumption for load peak-valley difference;
the maximum value of the total load of the power distribution network in the t period is obtained;
is the minimum value of the total load of the power distribution network in the period t.
Optionally, the optimizing the output of the distributed power source and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output of the distributed power source and the optimized demand of the novel load specifically includes:
according to the power distribution network optimization model, optimizing the output amount of the distributed power supply and the demand of the novel load by adopting an improved particle swarm algorithm to obtain the optimized output amount of the distributed power supply and the optimized demand of the novel load;
the improved particle swarm optimization is to calculate the inertia weight in the speed updating formula of the particle swarm optimization by adopting the following formula:
in the formula, w max 、w min The maximum value and the minimum value of the inertia weight w are respectively; d. d max Respectively the current iteration number and the maximum iteration number.
Optionally, the optimized distributed power output amount and the optimized novel load demand amount are clustered to obtain a distributed power output and novel load cluster data set, and the method specifically includes:
and clustering the optimized distributed power output amount and the optimized novel load demand amount by adopting a rapid search density peak value clustering method to obtain a distributed power output amount and a novel load clustering data set.
Optionally, the method for maximizing the scene information entropy change amount is adopted to perform scene classification according to the distributed power supply output and the novel load clustering data set, so as to generate a plurality of typical scenes, which specifically includes:
taking data information in the distributed power supply output and novel load clustering data set as leaf nodes, and calculating information entropies of the leaf nodes;
judging whether all leaf nodes are all inseparable; if yes, outputting all leaf nodes as typical scenes; if not, the divisible leaf nodes are used as root nodes, the information entropy reduction maximization is taken as a target, the root nodes are divided, and the step of 'judging whether all the leaf nodes are not divisible' is returned after the divided leaf nodes are obtained.
The present invention also provides a typical scene generation system, including:
the data acquisition module is used for acquiring the output quantity of the distributed power supply and the demand quantity of the novel load; the novel load comprises an electric automobile;
the model building module is used for building a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target;
the optimization module is used for optimizing the output of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output of the distributed power supply and the optimized demand of the novel load;
the clustering module is used for clustering the optimized distributed power output amount and the optimized novel load demand amount to obtain a distributed power output amount and a novel load clustering data set;
and the typical scene generation module is used for carrying out scene classification by adopting a method of maximizing scene information entropy change according to the distributed power supply output and the novel load clustering data set to generate various typical scenes.
Optionally, the model building module specifically includes:
the target function determining unit is used for determining a target function by taking the minimum total electric quantity absorbed by the main network, the minimum electric quantity abandoned by the distributed power supply and the load peak-valley difference punishment energy consumption as targets;
a constraint condition determining unit for determining a constraint condition; the constraint conditions comprise power balance constraint, system operation constraint, distributed power supply output constraint, energy storage system constraint and load side constraint;
the model establishing unit is used for determining the objective function and the constraint condition as a power distribution network optimization model;
wherein,
the objective function is determined according to the following formula:
minF＝F Buy +F Cur +F LPV
P i Buy (t)＝P sum (t)-P i DG (t)
wherein F is an objective function; f Buy Total power absorbed from the main network; f Cur Discarding the electric quantity for the distributed power supply; f LPV Penalizing energy consumption for load peak-to-valley difference; a is a Buy The unit amount of electricity absorbed from the main network; t is the total time period; n is the total number of nodes; p i Buy (t) is the active power input from the main network to the node i in the period of t; a is Cur Discarding the electric quantity for a unit of the distributed power supply; p i DG* (t) is a theoretical value of the active power of the node i in a period of t; p is i DG (t) is an actual value of active power after the node i abandons power in the t time period; p is sum (t) is the total load of the distribution network in the period t; p i EV (t) is the electric vehicle charging power of the node i in the time period t; p is i spare (t) is the backup load power of node i in the time period t; p is i involve (t) supplying power of a user side demand load to a node i in a period of t; a is LPV Punishment of energy consumption for load peak-valley difference;
the maximum value of the total load of the power distribution network in the t period is obtained;
is a period of tMinimum value of the total load of the internal distribution network.
Optionally, the optimization module specifically includes:
the optimization unit is used for optimizing the output capacity of the distributed power supply and the demand of the novel load by adopting an improved particle swarm algorithm according to the power distribution network optimization model to obtain the optimized output capacity of the distributed power supply and the optimized demand of the novel load;
the improved particle swarm optimization is to calculate the inertia weight in the speed updating formula of the particle swarm optimization by adopting the following formula:
in the formula, w max 、w min The maximum value and the minimum value of the inertia weight w are respectively; d. d max Respectively the current iteration number and the maximum iteration number.
Optionally, the clustering module specifically includes:
and the clustering unit is used for clustering the optimized distributed power output amount and the optimized novel load demand amount by adopting a fast search density peak value clustering method to obtain a distributed power output amount and a novel load clustering data set.
Optionally, the typical scene generating module specifically includes:
the information entropy calculation unit is used for taking the data information in the distributed power supply output and novel load clustering data set as leaf nodes and calculating the information entropy of the leaf nodes;
the judging unit is used for judging whether all leaf nodes are all inseparable or not; if yes, executing a typical scene generating unit; if not, executing the segmentation unit;
a typical scene generation unit for outputting all leaf nodes as typical scenes;
and the segmentation unit is used for taking the divisible leaf nodes as root nodes, taking the maximum information entropy reduction as a target, carrying out segmentation processing on the root nodes, obtaining the segmented leaf nodes, and then executing the judgment unit.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a distributed power supply and novel load typical scene generation method and system, which are used for acquiring the output power of a distributed power supply and the demand of a novel load; establishing a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target; optimizing the output capacity of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output capacity of the distributed power supply and the optimized demand of the novel load; clustering the optimized output capacity of the distributed power supply and the optimized novel load demand to obtain a clustered data set of the output capacity of the distributed power supply and the novel load; and according to the output of the distributed power supply and the novel load clustering data set, carrying out scene classification by adopting a method for maximizing scene information entropy change amount to generate various typical scenes. The invention can generate a typical scene while ensuring the safety, reliability and environmental protection of the operation of the power grid.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive labor.
FIG. 1 is a flowchart of a typical scenario generation method for a distributed power supply and a new load according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating exemplary scenario generation for a distributed power source and a new load based on optimized scheduling in an embodiment of the present invention;
FIG. 3 is a flow chart of an improved particle swarm optimization algorithm in an embodiment of the present invention;
FIG. 4 is a flowchart of a typical scene extraction method based on information entropy according to an embodiment of the present invention;
FIG. 5 is a diagram of an IEEE33 node network structure in an embodiment of the present invention;
fig. 6 is a diagram of a distributed power supply and a novel load clustering result in the embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a distributed power supply and novel load typical scene generation method and system, which can generate typical scenes while ensuring the safety, reliability and environmental protection of power grid operation.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Examples
Fig. 1 is a flowchart of a method for generating a typical scenario of a distributed power source and a novel load in an embodiment of the present invention, and as shown in fig. 1, the method for generating a typical scenario of a distributed power source and a novel load includes:
step 101: acquiring the output power of the distributed power supply and the demand of a novel load; the novel load comprises an electric automobile and electric heating equipment.
Step 102: and establishing a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target.
and determining an objective function by taking the minimum total electric quantity absorbed by the main network, the minimum electric quantity abandoned by the distributed power supply and the load peak-valley difference punishment energy consumption as targets.
Determining a constraint condition; the constraint conditions comprise power balance constraint, system operation constraint, distributed power supply output constraint, energy storage system constraint and load side constraint.
And determining the objective function and the constraint condition as a power distribution network optimization model.
Wherein,
the objective function is determined according to the following formula:
minF＝F Buy +F Cur +F LPV
P i Buy (t)＝P sum (t)-P i DG (t)
wherein F is an objective function; f Buy Total power absorbed from the main network; f Cur Discarding the electric quantity for the distributed power supply; f LPV Penalizing energy consumption for load peak-to-valley difference; a is Buy The unit amount of electricity absorbed from the main network; t is the total time period; n is the total number of nodes; p i Buy (t) is the active power input from the main network to the node i in the period of t; a is Cur The unit electricity abandonment quantity is the unit electricity abandonment quantity of the distributed power supply; p i DG* (t) is a theoretical value of the active power of the node i in a period of t; p i DG (t) is an actual value of active power after the node i abandons power in the t time period; p is sum (t) is the total load of the distribution network in the period of t; p i EV (t) is the electric vehicle charging power of the node i in the t time period; p i spare (t) is the backup load power of node i during the time period t; p is i involve (t) providing the user for node i during the t periodSide demand load power; a is LPV Punishment of energy consumption for load peak-valley difference;
the maximum value of the total load of the power distribution network in the t period is obtained;
is the minimum value of the total load of the distribution network in the period t.
Step 103: and optimizing the output capacity of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output capacity of the distributed power supply and the optimized demand of the novel load.
and optimizing the output force of the distributed power source and the demand of the novel load by adopting an improved particle swarm algorithm according to the power distribution network optimization model to obtain the optimized output force of the distributed power source and the optimized demand of the novel load.
The improved particle swarm optimization is to calculate the inertia weight in the speed updating formula of the particle swarm optimization by adopting the following formula:
in the formula, w max 、w min The maximum value and the minimum value of the inertia weight w are respectively; d. d max Respectively the current iteration number and the maximum iteration number.
Step 104: and clustering the optimized output capacity of the distributed power supply and the optimized novel load demand to obtain a clustered data set of the output capacity of the distributed power supply and the novel load.
and clustering the optimized output capacity of the distributed power supply and the optimized novel load demand by adopting a fast search density peak value clustering method to obtain a distributed power supply output capacity and a novel load clustering data set.
Step 105: and (4) according to the output of the distributed power supply and a novel load clustering data set, carrying out scene classification by adopting a method of maximizing scene information entropy change quantity, and generating various typical scenes.
and taking the data information in the distributed power output and novel load clustering data set as leaf nodes, and calculating the information entropy of the leaf nodes.
Judging whether all leaf nodes are all inseparable; if yes, outputting all leaf nodes as typical scenes; if not, the divisible leaf nodes are used as root nodes, the information entropy reduction maximization is taken as a target, the root nodes are divided, and the step of 'judging whether all the leaf nodes are inseparable' is returned after the divided leaf nodes are obtained.
With reference to fig. 2, the distributed power supply and the novel load typical scene generation method provided by the present invention are further explained.
Referring to fig. 2, a flowchart of a distributed power supply and novel load typical planning scenario generation technique based on optimized scheduling according to an embodiment of the present invention is shown, including the following steps:
step 1: data acquisition and preprocessing
The scene generation technology provided by the invention needs a large amount of data as a basis, and the quality of the data greatly influences the scene generation result. Therefore, the data to be processed must be reasonably collected and selected, unnecessary calculation errors are reduced, and the accuracy of the result is ensured.
The data of the invention is the annual real-time data of a certain power grid company. At present, the data of the power distribution network mainly come from an automatic system, an electric energy monitoring system, a load control system and the like.
Selecting an object to be processed, sampling the output capacity of the distributed power supply and the demand of a novel load within one year every other hour to obtain a data point, wherein each distributed power supply or load point has 8760 data points, and finally arranging two data sets to obtain:
in the formula, S dis 、S load Respectively collecting data points of a distributed power supply output value and a novel load demand value;
the output value and the required value at the moment j in the ith data point are respectively.
Due to the influence of factors such as operating environment and equipment state, some adverse data can appear in the power distribution network data, and the common causes of the adverse data mainly include:
(1) Due to the influence of actual operating environment and the like, partial equipment of the power distribution network may be in failure or have measurement errors.
(2) During the operation of the power distribution network, faults which may occur to the system, such as short circuit and broken line, failure of secondary equipment and the like.
(3) And communication faults caused by electromagnetic interference in the working process of the power distribution network.
For these unfavorable data, if no preprocessing is performed, the scene generation result will be affected to some extent, so the present invention performs the following brief processing:
(1) For the discrete unfavorable data, according to the characteristic of smooth curve of the power system, taking the average value of the previous data and the next data of the point as the new data of the point.
(2) For continuous bad data, the curves of adjacent periods are selected to replace the continuous bad data set.
The processed data can reduce unnecessary errors, improve the accuracy of operation, achieve better scene generation effect, and output the preprocessed data to the step 3.
And 2, step: establishing optimized dispatching model of power distribution network considering multiple novel loads
In order to improve the clean energy consumption level of the power distribution network and reduce the load peak-valley difference, the optimization scheduling model considers the uncertainty of distributed energy power generation and electric vehicle load.
1. Objective function
The objective function is that the dispatching energy consumption F of the power distribution network is minimum, including the total electric quantity F absorbed from the main network Buy Distributed power supply electricity abandoning quantity F Cur Load peak valley difference penalty energy consumption F LPV 。
minF＝F Buy +F Cur +F LPV (3)
P i Buy (t)＝P sum (t)-P i DG (t) (5)
In the formula, a Buy Is the unit amount of electricity absorbed from the main network; p i Buy (t) is the active power input from the main network to the node i in the period t; a is Cur Discarding the electric quantity for a unit of the distributed power supply; p i DG* (t)、P i DG (t) respectively representing a theoretical value of active power of the node i and an actual value of the active power after the electricity is abandoned in a period of t; p is sum (t) is the total load of the distribution network in the period of t; p is i EV (t)、P i spare (t)、P i involve (t) the charging efficiency, the standby load power and the required load power supplied to the user side of the electric vehicle of the node i in the time period t are respectively; a is LPV Energy consumption is penalized in units of load peak-to-valley difference.
2. Constraint conditions
(1) Constraint of power balance
P i (t)＝P i Buy (t)+P i DG (t)+P i ESS-dis (t)-P i ESS-cha (t)-P i sum (t) (9)
In the formula, P i (t)、Q i (t) inputting active power and reactive power to the node i within a time period t respectively; p i ESS-cha (t)、P i ESS-dis (t)、
Respectively charging and discharging active power and reactive power of the energy storage system in a t time period;
reactive power purchased from the main network within the time period t;
the reactive power output by the distributed power supply in the t time period;
the reactive power required by the total load of the power distribution network in the period t; p i sum And (t) the active power required by the total load of the power distribution network in the period of t.
(2) Operating constraints
V i min ≤V i (t)≤V i max (11)
In the formula, V i max 、V i min The maximum voltage and the minimum voltage of a node i allowed by system operation are respectively set;
the maximum value and the minimum value of the transmission active power allowed by the line ij in the system are respectively; v i (t) is the voltage of a node i where the system operates in a period of t;
the active power transmitted by the line ij in the system in the period t.
(3) Distributed power supply output constraints
0≤P i DG (t)≤P i DG-N (14)
In the formula,
the distributed power factor angle for node i is usually a fixed value; p i DG-N Is a rated value of the distributed power supply output.
(4) Energy storage system restraint
E i (t)＝E i (t-1)+ΔE (16)
P i cha-min A cha ≤P i cha (t)≤P i cha-max A cha (17)
P i dis-min A dis ≤P i dis (t)≤P i dis-max A dis (18)
0≤A cha +A dis ≤1 (19)
In the formula, delta E is the variable quantity of charging and discharging energy of the energy storage system; eta cha 、η dis Respectively are the charge and discharge efficiency coefficients of the energy storage system; a. The cha 、A dis Respectively setting the working state of the energy storage system of the node i to be a fixed value of 0 or 1; p i cha-max 、 P i cha-min 、P i dis-max 、P i dis-min The maximum value and the minimum value of the charging and discharging power of the energy storage system of the node i are respectively;
charging energy of the energy storage system for the node i;
storing the discharge energy of the energy storage system for the node i; e i And (t) is the energy of the node i energy storage system in the t period.
(5) Load side restraint
P i sum-min ≤P i sum (t)≤P i sum-max (20)
In the formula, P i sum-max 、P i sum-min Respectively the maximum value and the minimum value of the allowed bearing load of the system; p is i sum And (t) the load capacity allowed by the system.
After the dispatching model is analyzed, the peak-valley difference of the load can be reduced to a certain degree by the collaborative optimization operation of the distributed power supply and the novel load. The energy storage system discharges at the load peak and when the distributed power supply is insufficient in output, and charges at the load valley and when the distributed power supply is excessive in output, so that the peak clipping and valley filling effects are achieved, the electricity discard quantity of the distributed power supply can be reduced to a certain extent, large data required by typical scene generation can be obtained by using the model, and the model is output to the step 3.
And 3, step 3: power distribution network optimization scheduling analysis based on improved particle swarm optimization
(1) Description of particle swarm optimization algorithm
The particle swarm algorithm is a heuristic algorithm for initially simulating the activities of the bird swarm, and has the advantages of good adaptability, strong robustness, simplicity, flexibility and the like. Therefore, the invention aims to analyze the optimized dispatching of the power distribution network by adopting a particle swarm optimization algorithm.
The basic idea of the particle swarm optimization algorithm is as follows: first, assume that within the M-dimensional search space, there are N random particles flying at different velocities. Wherein the current position vector of the particle i is x i ＝(x i1 ,x i2 ,...,x iM ) The current velocity vector is v i ＝(v i1 ,v i2 ,...,v iM ) Calculating the target function value of each particle, namely the fitness value, by taking the target function as the fitness function, wherein the optimal position vector experienced by the particle i is p besti ＝(p besti1 ，p besti2 ，...，p bestiM ) Corresponding optimal fitness value of f besti In each iteration, the position and speed of the particle are in a constantly updated and changed process, and the update formula is as follows:
v im (t+1)＝w·v im (t)+c 1 r 1 [p best-im (t)-x im (t)]+c 2 r 2 [g best-m (t)-x im (t)] (21)
x im (t+1)＝x im (t)+v im (t+1) (22)
in the formula, v im (t) is the m-dimensional velocity component for particle i when iterated t times; w is the inertial weight; c. C 1 、c 2 Is a learning factor; r is 1 、r 2 Is a random number, and has a value range of [0,1 ]]；x im (t) is the m-dimensional position component for particle i when iterated t times; p is a radical of formula best-im (t) is the mth dimension optimal position component when the particle i iterates t times; g best-m (t) the mth-dimension optimal position component when the entire particle group is iterated t times.
(2) Deficiencies and improvements in particle swarm optimization algorithms
The particle swarm optimization algorithm continuously updates the state of the particles through iteration to further obtain the optimal solution, so that the solution is convenient to understand and operate, and the particle swarm optimization algorithm has better performance. However, the most prominent disadvantage of the current particle swarm optimization is that the current particle swarm optimization is easy to fall into the problem of local optimization. Many scholars improve the problem, and can be roughly divided into two improvement directions: improving parameters; in combination with other algorithms. The invention adopts an improved particle swarm optimization algorithm for improving parameters.
The inertia weight w plays an important role in the particle swarm optimization algorithm, and ensures the inertia of the particles in the motion in space, so that the particles can be searched in the next step, the particles are prevented from being stagnated in the space prematurely, and the possibility of falling into local optimum is avoided. Therefore, the inertia weight w is reasonably selected, which is related to the operation and analysis capability of the algorithm. Through research and analysis, in the operation process, the weight coefficient w should keep dynamic change: in the initial stage, in order to ensure the operation efficiency and approach the optimal state of the particles as soon as possible, the weight coefficient w should be larger; along with the continuous operation, in the later stage, in order to ensure the operation precision of the algorithm and avoid missing the optimal solution, the weight coefficient w should be in a continuously decreasing trend, which is more favorable for obtaining the optimal solution. Therefore, the dynamic change formula of the weight coefficient w is as follows:
in the formula, w max 、w min The maximum value and the minimum value of w are respectively; t, T max Respectively the current iteration number and the maximum iteration number.
(3) Step for improving particle swarm optimization algorithm
After improvement, the particle swarm optimization algorithm comprises the following steps:
step one, setting basic parameters and defining a target function.
And step two, initializing the positions and the speeds of all the particles in the specified search space.
Step three, calculating the fitness value of each particle through the objective function to obtain the fitness value of each particle in the current iteration,optimal position p of each particle besti And an optimal fitness value of f besti Selecting the best individual optimal fitness value as a global optimal fitness value f gbest Accordingly, a global optimum position g is selected best 。
And step four, iteratively updating the position vector and the velocity vector of the particle by formulas (21) and (22).
Step five, in the subsequent iteration process, the updated particle fitness value is compared with the optimal value which is obtained previously, and if the new fitness value is superior to the original value, the updated fitness value and the corresponding optimal position are reserved; otherwise, keeping the optimal value and the optimal position of the original individual unchanged.
Step six, on the basis of the step five, the updated global optimal fitness value is compared with the original value, and if the new global fitness value is superior to the original value, the updated global optimal value and the corresponding global optimal position are reserved; otherwise, the original global optimum value and the optimum position are kept unchanged.
And seventhly, judging whether the currently obtained optimal position of the particle or the global optimal fitness value meets the range or not, or whether the current iteration times exceed the maximum iteration times or not. If yes, finishing the operation and outputting the current result; otherwise, returning to the fourth step and continuing to operate.
The flow chart of the improved particle swarm optimization algorithm is shown in FIG. 3:
(3) Solving optimized scheduling model based on improved particle swarm optimization
Step one, setting parameters of an objective function: the scheduling period is one year; determining the specific electric quantity a absorbed from the main network Buy Electric unit electric quantity a of electricity abandoned by distributed power supply Cur Unit punished energy consumption a of load peak-valley difference LPV 。
Step two, determining a population size N =100, a maximum iteration number T =200 and a learning factor c in a particle swarm algorithm 1 ＝c 2 =2, particle flight velocity range [ -v [ ] m ,v m ]。
Step three, generating particles at random initially: section in t periodActual value P of active power of distributed power supply after power abandonment of point i i DG (t) and node i electric vehicle charging power P i EV (t) of (d). The position of each particle is recorded.
Step four, iteratively updating the speed and the position of the particles if v ij (t)≥v m Then order v ij (t)＝v m (ii) a If v is ij (t)≤-v m Then order v ij (t)＝-v m And continuing the iteration.
And fifthly, generating corresponding fitness values by utilizing the particles obtained through iteration of each time, and comparing the fitness values. And selecting and reserving an optimal fitness value (namely the minimum value of the objective function) and a corresponding particle individual optimal position, and then determining a global optimal position, namely the output value of the distributed power supply and the required value of the new load under the condition of minimum total energy consumption.
And step six, judging whether the currently obtained output value and the load required value of the distributed power supply meet the limited range or not, or whether the current iteration times exceed the maximum iteration times or not. If yes, outputting the output value and the load demand value of the current distributed power supply, and finishing the operation; otherwise, returning to the step four and continuing to operate.
And (3) aiming at the preprocessed data sets (1) and (2) obtained in the step (1) and the optimized scheduling model established by the formulas (3) to (20) in the step (2), obtaining the output amount of the distributed power supply and the demand data of the novel load in one year after optimization by applying the improved particle swarm algorithm in the step (3), and outputting the output amount and the demand data to the step (4).
And 4, step 4: scene clustering based on fast search density clustering method
With the development of the information society, data in various fields show a trend of rapid expansion and growth. The data mining technology can be used for mining hidden valuable information in big data, so that people can pay more attention to the hidden valuable information. The clustering analysis is an important branch of data mining, can describe common points and differences among objects, and is helpful for people to better master the development rules of the objects.
The traditional clustering method, such as the k-means method, needs to set the number of clustering centers and the initial clustering centers in advance, and the accuracy of the clustering result is influenced by the selection of the initial clustering centers. The fast search density clustering method can avoid the problems to a certain extent, and iterative solution is not needed, so the method is adopted for analysis.
The data set to be clustered is S = (x) ij ) N×T The main contents of the algorithm are as follows:
(1) Calculating local density
In order to ensure that the algorithm has good clustering performance, solving the local density rho of the data point i by adopting a Gaussian function method i ：
In the formula, x it 、x jt Respectively a data point i and a data point j at the time t; d ij Is the distance between any two data points i and j; d c The truncation distance is typically selected empirically such that the distance from each data point to other data points does not exceed d c The number of the data points is about 1% -2% of the total number of the data points, but the accuracy of the algorithm cannot be guaranteed only according to experience, so that the method is selected according to the following method:
in the formula, H is data set information entropy. The smaller the entropy of the data set information, the larger the local density difference. Thus, d is determined by solving for the minimum of the entropy of the information c ：
d c ＝argmin(H) (27)
(2) Calculating distance
Distance mu from data point i to data point j i Comprises the following steps:
A. when data point i has the greatest local density:
μ i ＝d ij (28)
B. if not, then,
(3) Determining cluster centers
In consideration of the defect that the initial fast search density clustering algorithm needs to artificially select the clustering center, the invention adopts a method for setting a boundary value to determine the clustering center. The data points are selected such that the local density and distance are both large, thereby allowing delta i ＝ρ i μ i Delta by comparison i The values to determine the boundary values.
According to the summary of the scholars, delta i Sorting in ascending order, observing delta after comparison sorting i A change in value. Non-clustered center point delta i The values all show a linear variation trend, and delta is from the non-clustering central point to the clustering central point i The values are not changed linearly any more, and according to the principle, the clustering central points can be sequentially compared and screened out. The specific implementation process is as follows:
A. starting from the first data point, (2n + 1) data points are selected, and if all the data points are non-cluster centers, the linear relation should be satisfied
B. One data point at a time is iterated until the formula is not satisfied for the first time, i.e., the first time the center point of the cluster is included, and the iteration is stopped.
C. Will delta the m And the value is reserved as a boundary value, and the corresponding data point is a cluster center point.
D. Will leave all delta remaining i Value and delta m Compared with the value, delta 'is satisfied' i ≥δ m Or | δ' i -δ m The point, | ≦ ε (ε is a positive number approaching 0) is considered the corner center point.
(4) Clustering data points
The local density ρ of all data points i Arranging values in an ascending order, marking the obtained clustering center points, and adopting a clustering principle: classifying non-cluster central point to be greater than self rho i Value, and distance d ij In the category of the smallest cluster center point, namely:
scene clustering based on a fast search density clustering method:
step one, inputting a data set to be clustered, and dividing obtained original data into four categories: obtaining data sets in spring, summer, autumn and winter
And performing cluster analysis on each data set respectively.
Step two, determining d by the formula (24) ij Constructing an information entropy function expression (26), searching for the minimum value of the information entropy function, and determining the parameter d by the expression (27) c 。
Step three, determining the local density rho by the formula (25) i The distance mu is determined by the equations (28) and (29) i 。
And step four, determining the clustering center points of the data set by using a boundary value method, wherein only one clustering center point is determined at each time point.
Marking the class of each data, classifying the non-clustering center data by the formula (30), respectively connecting the clustering center points obtained by the eight data sets into eight curves, and outputting the result.
And (5) clustering the data input in the step (3) by the scene based on the fast search density clustering method in the step (4), and outputting the obtained clustering result to the step (5).
And 5: selection of typical scenes
Due to the fact that the distributed power supply and the novel load have time sequence characteristics, calculation amount is large when time sequence data are adopted to solve a model, and time consumption is long. Therefore, the multi-scene analysis method is adopted to reduce the complexity of the operation and improve the accuracy of the operation. Meanwhile, due to the characteristics of uncertainty and randomness of the novel energy, a typical scene needs to be selected to reflect the output of the distributed power supply and the change characteristics of the novel load.
The typical scene extraction method based on the information entropy is provided, a concept of the scene information entropy is introduced to carry out typical scene extraction, scene classification is carried out by maximizing the scene information entropy change quantity, an original scene is divided into a plurality of subclasses in a recursion mode, the average value of each subclass is taken as the typical scene of the class, and a scene set of all the subclasses is an extracted typical scene set.
Firstly, inputting power distribution density function information of a scene to be divided as an initial root node, and calculating an initial root node information entropy.
Wherein H is node information entropy, v 0 Is the initial root node, x is power, p (x) is the probability density corresponding to the power, x max Is the maximum power value.
And then selecting the segmentation position of the initial root node. For a certain division position, the information entropy after division is the sum of the information entropy of two new nodes formed after division.
H(v 0 ,x 0 )＝H(v' 0 )+H(v” 0 ) (32)
Wherein x is 0 At any one of the division positions, H (v) 0 ,x 0 ) Is the information entropy, H (v' 0 ) For left child node formed after division，H(v” 0 ) For the right child node formed after segmentation, p' (x) is the probability density function of the left child node, and p "(x) is the probability density function of the right child node.
After node division, all possible power values of the parent node are also divided into two parts, and in a certain child node, the power value of another child node is not presented any more. Thus, the probability that each node retains a possible power value is increased and the power probability density function is changed.
The node segmentation is selected based on the principle of maximizing the reduction of the entropy, i.e.
ΔH＝H(v 0 )-H(v 0 ,x 0 ) (35)
Wherein Δ H is the reduction of the information entropy after node segmentation.
The segmentation position that can maximize equation (35) is the optimal segmentation position of the current node.
After the initial node is divided and two child nodes are formed, the generated new child nodes are checked, and whether the information entropy of the check nodes is smaller than the critical value epsilon or not is checked H . Less than a critical value epsilon H The information entropy of the sub-nodes of (1) is smaller, and is called as a "leaf node". The smaller information entropy represents that the power distribution in the node is more consistent, and the power values obtained by performing power sampling in a single node converge, so that larger fluctuation can not occur. Thus, for a leaf node, the expectation of the power distribution may be taken as a typical scenario for that node.
And taking the new node which does not pass the information entropy check as a node to be rooted, and continuing node segmentation until all the nodes become leaf nodes, and stopping recursion. And calculating the typical scenes corresponding to all the leaf nodes, and outputting the typical scenes as the extracted typical scene set. The proportion of the number of scenes in each leaf node to the total scenes is used as the weight of the leaf node corresponding to the typical scenes.
The scene extraction step based on information entropy is shown in fig. 4.
In order to keep the algorithm content names consistent, the initial nodes are input as leaf nodes and then modified into root nodes.
The method comprises the following steps:
the method comprises the following steps: and (5) inputting the information of the clustered data sets of spring, summer, autumn and winter in the step 4 as initial leaf nodes, and calculating the information entropy of the initial leaf nodes according to the formula (31).
Step two: judging whether all the nodes are all inseparable, if so, outputting all the leaf nodes as typical scenes, acquiring the number of hours correspondingly contained in the scenes, and calculating the probability; if not, taking the leaf node which does not meet the requirement as a root node, and continuing the following steps.
Step three: and (4) dividing the root node according to the information entropy reduction maximization principle of the formula (35) and the formulas (32) to (34) to determine the division position.
Step four: and (4) judging the new leaf nodes obtained by segmentation by using the second step, and repeating iteration until all the nodes meet the termination condition.
And (5) on the basis of the clustering data set obtained in the step (4), and through the scene generation method in the step (5), finally obtaining a typical operation scene of the complex power distribution network planning, and laying a foundation for realizing the complex power distribution network planning.
The invention also provides a distributed power supply and novel load typical scene generation system, which comprises:
the data acquisition module is used for acquiring the output quantity of the distributed power supply and the demand quantity of the novel load; the new load includes an electric vehicle.
And the model establishing module is used for establishing a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target. The model building module specifically comprises:
the target function determining unit is used for determining a target function by taking the minimum total electric quantity absorbed by the main network, the minimum electric quantity abandoned by the distributed power supply and the load peak-valley difference punishment energy consumption as targets;
a constraint condition determining unit for determining a constraint condition; the constraint conditions comprise power balance constraint, system operation constraint, distributed power supply output constraint, energy storage system constraint and load side constraint;
the model building unit is used for determining the objective function and the constraint condition as a power distribution network optimization model;
wherein,
the objective function is determined according to the following formula:
minF＝F Buy +F Cur +F LPV
P i Buy (t)＝P sum (t)-P i DG (t)
wherein F is an objective function; f Buy Total amount of power absorbed from the main network; f Cur Discarding the electric quantity for the distributed power supply; f LPV Penalizing energy consumption for load peak-to-valley difference; a is a Buy The unit amount of electricity absorbed from the main network; t is the total time period; n is the total number of nodes; p is i Buy (t) is the active power input from the main network to the node i in the period t; a is a Cur Discarding the electric quantity for a unit of the distributed power supply; p i DG* (t) is a theoretical value of the active power of the node i in a period of t; p is i DG (t) is an actual value of active power after the node i abandons power in the t time period; p sum (t) is the total load of the distribution network in the period t; p i EV (t) is the electric vehicle charging power of the node i in the t time period; p is i spare (t) is the backup load power of node i in the time period t; p i involve (t) supplying the required load power of the user side to the node i in the period t; a is a LPV Punishment of energy consumption for load peak-valley difference;
the maximum value of the total load of the power distribution network in the t period;
is the minimum value of the total load of the distribution network in the period t.
And the optimization module is used for optimizing the output capacity of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output capacity of the distributed power supply and the optimized demand of the novel load. The optimization module specifically comprises:
the optimization unit is used for optimizing the output force of the distributed power supply and the demand of the novel load by adopting an improved particle swarm algorithm according to the power distribution network optimization model to obtain the optimized output force of the distributed power supply and the optimized demand of the novel load;
the improved particle swarm optimization is to calculate the inertia weight in the speed updating formula of the particle swarm optimization by adopting the following formula:
in the formula, w max 、w min The maximum value and the minimum value of the inertia weight w are respectively; d. d max Respectively the current iteration number and the maximum iteration number.
And the clustering module is used for clustering the optimized output capacity of the distributed power supply and the optimized novel load demand to obtain a clustered data set of the output capacity of the distributed power supply and the novel load.
The clustering module specifically comprises:
and the clustering unit is used for clustering the optimized distributed power output amount and the optimized novel load demand amount by adopting a fast search density peak value clustering method to obtain a distributed power output and novel load clustering data set.
And the typical scene generation module is used for carrying out scene classification by adopting a method of maximizing scene information entropy change according to the distributed power output and the novel load clustering data set so as to generate various typical scenes.
The typical scene generation module specifically includes:
the information entropy calculation unit is used for taking the data information in the distributed power supply output and novel load clustering data set as leaf nodes and calculating the information entropy of the leaf nodes;
the judging unit is used for judging whether all leaf nodes are all inseparable or not; if yes, executing a typical scene generating unit; if not, executing the segmentation unit;
a typical scene generation unit for outputting all leaf nodes as typical scenes;
and the segmentation unit is used for taking the divisible leaf nodes as root nodes, taking the maximum information entropy reduction as a target, carrying out segmentation processing on the root nodes, obtaining the segmented leaf nodes and then executing the judgment unit.
For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points can be referred to the description of the method part.
The invention obtains the optimized scheduling scheme of the multiple novel loads by taking the minimum total energy consumption, the power balance and the safe operation as constraint conditions. Meanwhile, clustering analysis is carried out on the obtained distributed power supply output and the novel load, big data clustering is achieved, and in order to achieve multi-scene planning, a planning typical operation scene of a complex power distribution network is provided.
The invention adopts an IEEE33 node system for example analysis, and the grid structure is shown in figure 5.
The access point conditions are shown in table 1.
Table 1 access point location
| Distributed |
10，17，24，32 |
| Multiple |
9，10，11，12，13，14，19，20，21 |
| |
2，6，16，29 |
The model parameters are shown in table 2.
TABLE 2 model parameters
After the scheduling model is optimized, the annual energy consumption of the power distribution network is 105269650MWh. Before optimization, the annual energy consumption of the power distribution network is 119130160MWh, and through comparison, the model is effective.
Based on a fast search density clustering method, the optimized distributed power supply output and novel load demand are divided into four seasons of spring, summer, autumn and winter, clustering analysis is carried out, a clustering center is extracted, and a curve drawing result is shown in fig. 6. Fig. 6 (a) is a diagram of a spring distributed power supply and a novel load clustering result, fig. 6 (b) is a diagram of a summer distributed power supply and a novel load clustering result, fig. 6 (c) is a diagram of an autumn distributed power supply and a novel load clustering result, and fig. 6 (d) is a diagram of a winter distributed power supply and a novel load clustering result.
Typical scene generation is shown in tables 3 and 4.
TABLE 3 novel load Multi-Scenario results
TABLE 4 distributed Power Multi-Scenario results
The principle and the embodiment of the present invention are explained by applying specific examples, and the above description of the embodiments is only used to help understanding the method and the core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In summary, this summary should not be construed to limit the present invention.
Claims (8)
1. A typical scene generation method, comprising:
acquiring the output power of the distributed power supply and the demand of a novel load; the novel load comprises an electric automobile;
establishing a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target;
optimizing the output of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output of the distributed power supply and the optimized demand of the novel load;
clustering the optimized distributed power output and the optimized novel load demand to obtain a distributed power output and novel load clustering data set;
according to the distributed power supply output and the novel load clustering data set, carrying out scene classification by adopting a method of maximizing scene information entropy change amount to generate various typical scenes;
the method for establishing the power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target specifically comprises the following steps:
determining an objective function by taking the minimum total electric quantity absorbed by a main network, the minimum electric quantity abandoned by a distributed power supply and the minimum load peak-valley difference punishment energy consumption as targets;
determining a constraint condition; the constraint conditions comprise power balance constraint, system operation constraint, distributed power supply output constraint, energy storage system constraint and load side constraint;
determining the objective function and the constraint condition as a power distribution network optimization model;
wherein,
the objective function is determined according to the following formula:
minF＝F Buy +F Cur +F LPV
P i Buy (t)＝P sum (t)-P i DG (t)
wherein F is an objective function; f Buy Total power absorbed from the main network; f Cur Discarding the electric quantity for the distributed power supply; f LPV Penalizing energy consumption for load peak-to-valley difference; a is a Buy The unit amount of electricity absorbed from the main network; t is the total time period; n is the total number of nodes; p i Buy (t) is the active power input from the main network to the node i in the period t;a Cur the unit electricity abandonment quantity is the unit electricity abandonment quantity of the distributed power supply; p is i DG* (t) is a theoretical value of the active power of the node i in a period of t; p i DG (t) is an actual value of active power of the node i after power is abandoned in the period of t; p sum (t) is the total load of the distribution network in the period t; p i EV (t) is the electric vehicle charging power of the node i in the time period t; p i spare (t) is the standby load power of node i in the time period t; p i involve (t) supplying power of a user side demand load to a node i in a period of t; a is LPV Punishment of energy consumption for load peak-valley difference;
the maximum value of the total load of the power distribution network in the t period is obtained;
is the minimum value of the total load of the power distribution network in the period t.
2. The typical scene generation method according to claim 1, wherein the optimizing the output of the distributed power source and the demand of the new load according to the power distribution network optimization model to obtain the optimized output of the distributed power source and the optimized demand of the new load specifically includes:
according to the power distribution network optimization model, optimizing the output of the distributed power supply and the demand of the novel load by adopting an improved particle swarm algorithm to obtain the optimized output of the distributed power supply and the optimized demand of the novel load;
the improved particle swarm optimization is to calculate the inertia weight in the speed updating formula of the particle swarm optimization by adopting the following formula:
in the formula, w max 、w min The maximum value and the minimum value of the inertia weight w are respectively; d. d max Respectively the current iteration number and the maximum iteration number.
3. A typical scenario generation method according to claim 2, wherein the clustering processing is performed on the optimized distributed power output amount and the optimized new load demand amount to obtain a distributed power output and new load cluster data set, and specifically includes:
and clustering the optimized distributed power output and the optimized novel load demand by adopting a fast search density peak value clustering method to obtain a distributed power output and novel load clustering data set.
4. A typical scene generating method according to claim 3, wherein the generating of a plurality of typical scenes by performing scene classification by using a method of maximizing a scene information entropy change amount according to the distributed power supply output and the novel load clustering dataset specifically includes:
taking data information in the distributed power supply output and novel load clustering data set as leaf nodes, and calculating information entropies of the leaf nodes;
judging whether all leaf nodes are all inseparable; if yes, outputting all leaf nodes as typical scenes; if not, the leaf nodes which can be divided are used as root nodes, the information entropy reduction is maximized as a target, the root nodes are divided, and the leaf nodes after being divided are obtained, and then the step of 'judging whether all the leaf nodes are all non-divisible' is returned.
5. An exemplary scene generation system, comprising:
the data acquisition module is used for acquiring the output quantity of the distributed power supply and the demand quantity of the novel load; the novel load comprises an electric automobile;
the model building module is used for building a power distribution network optimization model by taking the minimum scheduling energy consumption of the power distribution network as a target;
the optimization module is used for optimizing the output of the distributed power supply and the demand of the novel load according to the power distribution network optimization model to obtain the optimized output of the distributed power supply and the optimized demand of the novel load;
the clustering module is used for clustering the optimized distributed power output amount and the optimized novel load demand amount to obtain a distributed power output amount and a novel load clustering data set;
the typical scene generation module is used for carrying out scene classification by adopting a method of maximizing scene information entropy change according to the distributed power supply output and the novel load clustering data set to generate various typical scenes;
the model building module specifically comprises:
the target function determining unit is used for determining a target function by taking the minimum total electric quantity absorbed by the main network, the minimum electric quantity abandoned by the distributed power supply and the load peak-valley difference punishment energy consumption as targets;
a constraint condition determining unit for determining a constraint condition; the constraint conditions comprise power balance constraint, system operation constraint, distributed power supply output constraint, energy storage system constraint and load side constraint;
the model establishing unit is used for determining the objective function and the constraint condition as a power distribution network optimization model;
wherein,
the objective function is determined according to the following formula:
minF＝F Buy +F Cur +F LPV
P i Buy (t)＝P sum (t)-P i DG (t)
in the formula, F is an objective function; f Buy Total power absorbed from the main network; f Cur Discarding the electric quantity for the distributed power supply; f LPV Penalizing energy consumption for load peak-to-valley difference; a is Buy The unit amount of electricity absorbed from the main network; t is the total time period; n is the total number of nodes; p i Buy (t) is the active power input from the main network to the node i in the period t; a is Cur Discarding the electric quantity for a unit of the distributed power supply; p i DG* (t) is the theoretical value of the active power of the node i in the period of t; p i DG (t) is an actual value of active power of the node i after power is abandoned in the period of t; p sum (t) is the total load of the distribution network in the period t; p i EV (t) is the electric vehicle charging power of the node i in the time period t; p is i spare (t) is the power of the backup load of node i during the period of t; p is i involve (t) supplying the required load power of the user side to the node i in the period t; a is a LPV Punishment of energy consumption for load peak-valley difference;
the maximum value of the total load of the power distribution network in the t period is obtained;
is the minimum value of the total load of the distribution network in the period t.
6. The typical scene generation system according to claim 5, wherein the optimization module specifically includes:
the optimization unit is used for optimizing the output of the distributed power supply and the demand of the novel load by adopting an improved particle swarm algorithm according to the power distribution network optimization model to obtain the optimized output of the distributed power supply and the optimized demand of the novel load;
the improved particle swarm optimization is to calculate the inertia weight in the speed updating formula of the particle swarm optimization by adopting the following formula:
in the formula, w max 、w min The maximum value and the minimum value of the inertia weight w are respectively; d. d max Respectively the current iteration number and the maximum iteration number.
7. The typical scene generation system according to claim 6, wherein the clustering module specifically includes:
and the clustering unit is used for clustering the optimized distributed power output amount and the optimized novel load demand amount by adopting a fast search density peak value clustering method to obtain a distributed power output amount and a novel load clustering data set.
8. The typical scene generation system according to claim 7, wherein the typical scene generation module specifically includes:
the information entropy calculation unit is used for taking the data information in the distributed power supply output and novel load clustering data set as leaf nodes and calculating the information entropy of the leaf nodes;
the judging unit is used for judging whether all leaf nodes are all inseparable or not; if yes, executing a typical scene generating unit; if not, executing the segmentation unit;
a typical scene generation unit for outputting all leaf nodes as typical scenes;
and the segmentation unit is used for taking the divisible leaf nodes as root nodes, taking the maximum information entropy reduction as a target, carrying out segmentation processing on the root nodes, obtaining the segmented leaf nodes, and then executing the judgment unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011298627.1A CN112467803B (en) | 2020-11-19 | 2020-11-19 | Distributed power supply and novel load typical scene generation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011298627.1A CN112467803B (en) | 2020-11-19 | 2020-11-19 | Distributed power supply and novel load typical scene generation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112467803A CN112467803A (en) | 2021-03-09 |
| CN112467803B true CN112467803B (en) | 2023-01-10 |
Family
ID=74836168
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011298627.1A Active CN112467803B (en) | 2020-11-19 | 2020-11-19 | Distributed power supply and novel load typical scene generation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112467803B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109508823A (en) * | 2018-11-06 | 2019-03-22 | 西安理工大学 | A kind of Distributed Generation in Distribution System planing method of method based on scene analysis |
| CN109524993A (en) * | 2018-12-19 | 2019-03-26 | 中国农业大学 | The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling |
| CN110943475A (en) * | 2019-11-08 | 2020-03-31 | 广东电网有限责任公司 | Distribution network collaborative optimization scheduling method considering distributed photovoltaic and electric automobile |
| CN111463836A (en) * | 2020-05-13 | 2020-07-28 | 陕西燃气集团新能源发展股份有限公司 | Optimized scheduling method for comprehensive energy system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9373960B2 (en) * | 2013-03-13 | 2016-06-21 | Oracle International Corporation | Computerized system and method for distributed energy resource scheduling |
-
2020
- 2020-11-19 CN CN202011298627.1A patent/CN112467803B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109508823A (en) * | 2018-11-06 | 2019-03-22 | 西安理工大学 | A kind of Distributed Generation in Distribution System planing method of method based on scene analysis |
| CN109524993A (en) * | 2018-12-19 | 2019-03-26 | 中国农业大学 | The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling |
| CN110943475A (en) * | 2019-11-08 | 2020-03-31 | 广东电网有限责任公司 | Distribution network collaborative optimization scheduling method considering distributed photovoltaic and electric automobile |
| CN111463836A (en) * | 2020-05-13 | 2020-07-28 | 陕西燃气集团新能源发展股份有限公司 | Optimized scheduling method for comprehensive energy system |
Non-Patent Citations (1)
| Title |
|---|
| 考虑DG与负荷不确定性的主动配电网重构;江泽昌 等;《电测与仪表》;20191010;第56卷(第19期);摘要，正文第1-5节，图1-3 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112467803A (en) | 2021-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108320111B (en) | A kind of ultra-large GROUP OF HYDROPOWER STATIONS practical dispatching method in short term | |
| Ghadimi et al. | PSO based fuzzy stochastic long-term model for deployment of distributed energy resources in distribution systems with several objectives | |
| CN112131733B (en) | Distributed power supply planning method considering influence of charging load of electric automobile | |
| CN108764634A (en) | A kind of electric automobile charging station dynamic programming method for considering charge requirement and increasing | |
| CN108306303A (en) | Voltage stability evaluation method considering load increase and new energy output randomness | |
| CN107528350B (en) | A kind of wind power output typical scene generation method adapting to long -- term generation expansion planning | |
| CN108649605B (en) | DRE grid-connected access capacity planning method based on double-layer scene interval power flow | |
| CN110633854A (en) | Full life cycle optimization planning method considering energy storage battery multiple segmented services | |
| CN112238781B (en) | Electric automobile ordered charging control method based on layered architecture | |
| CN111682536A (en) | Random-robust optimization operation method for virtual power plant participating in day-ahead double market | |
| CN111401755A (en) | Multi-new-energy output scene generation method, device and system based on Markov chain | |
| CN113783224A (en) | Power distribution network double-layer optimization planning method considering operation of various distributed energy sources | |
| CN114336785B (en) | Distributed power supply group control and group dispatching control method and device based on grid clustering | |
| CN115173465A (en) | Wind, light, water, fire and storage integrated coupling mechanism analysis method based on Copula theory | |
| CN110707711B (en) | User side comprehensive energy system hierarchical regulation and control method and system | |
| CN116523240A (en) | Micro-grid evaluation feedback scheduling method based on MCMC method prediction | |
| CN112257897A (en) | Electric vehicle charging optimization method and system based on improved multi-target particle swarm | |
| CN113657671A (en) | Flight delay prediction method based on ensemble learning | |
| CN110570091A (en) | Load identification method based on improved F-score feature selection and particle swarm BP neural network | |
| CN114462854A (en) | Hierarchical scheduling method and system containing new energy and electric vehicle grid connection | |
| CN112510690B (en) | Optimal scheduling method and system considering wind-fire-storage combination and demand response reward and punishment | |
| CN112467803B (en) | Distributed power supply and novel load typical scene generation method and system | |
| Pandya et al. | Stochastic wind-thermal power plants integrated multi-objective optimal power flow | |
| CN116263922A (en) | Micro-energy network planning method considering carbon-green combined trading market | |
| CN115021269A (en) | Two-stage optimal power flow solving method based on data driving |
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 |