CN117293923A - Method, device, equipment and storage medium for generating day-ahead scheduling plan of power grid - Google Patents
Method, device, equipment and storage medium for generating day-ahead scheduling plan of power grid Download PDFInfo
- Publication number
- CN117293923A CN117293923A CN202311250242.1A CN202311250242A CN117293923A CN 117293923 A CN117293923 A CN 117293923A CN 202311250242 A CN202311250242 A CN 202311250242A CN 117293923 A CN117293923 A CN 117293923A
- Authority
- CN
- China
- Prior art keywords
- power grid
- target
- state data
- scheduling
- constraint
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012549 training Methods 0.000 claims abstract description 36
- 238000012545 processing Methods 0.000 claims abstract description 34
- 238000005457 optimization Methods 0.000 claims abstract description 23
- 239000011159 matrix material Substances 0.000 claims description 25
- 238000004590 computer program Methods 0.000 claims description 16
- 238000013528 artificial neural network Methods 0.000 claims description 12
- 230000003042 antagnostic effect Effects 0.000 claims description 8
- 230000009194 climbing Effects 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000005265 energy consumption Methods 0.000 abstract description 12
- 238000004891 communication Methods 0.000 description 8
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000009471 action Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000010248 power generation Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 239000000243 solution Substances 0.000 description 4
- 238000011217 control strategy Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000002787 reinforcement Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000002347 injection Methods 0.000 description 2
- 239000007924 injection Substances 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000004883 computer application Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000010977 unit operation Methods 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0475—Generative networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/094—Adversarial learning
-
- 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
- G06Q10/06313—Resource planning in a project environment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—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
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/10—Power transmission or distribution systems management focussing at grid-level, e.g. load flow analysis, node profile computation, meshed network optimisation, active network management or spinning reserve management
-
- 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]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Marketing (AREA)
- Mathematical Physics (AREA)
- Tourism & Hospitality (AREA)
- General Engineering & Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Educational Administration (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Power Engineering (AREA)
- Biodiversity & Conservation Biology (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a method, a device, equipment and a storage medium for generating a day-ahead scheduling plan of a power grid. The method for generating the day-ahead scheduling plan of the power grid comprises the following steps: acquiring original state data of a target power grid at a target moment; processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of the thermal power unit in the target power grid; predicting the input original state data and the target scheduling instruction through a virtual power grid model after training to obtain target state data of the target power grid in a first preset time period in the future; and obtaining a daily schedule of the target power grid based on the target state data. Based on the technical scheme of the embodiment of the invention, the daily scheduling plan of the target power grid with low energy consumption can be automatically determined, the generation efficiency of the daily scheduling plan of the power grid is improved, and the scheduling cost of the power grid is reduced.
Description
Technical Field
The present invention relates to the field of computer application technologies, and in particular, to a method, an apparatus, a device, and a storage medium for generating a day-ahead scheduling plan of a power grid.
Background
The current proportion of new energy power generation and new energy automobiles is gradually increased, and the uncertainty of the output of new energy equipment and the load of electric equipment is increased, so that the solving scale and difficulty of new energy power grid dispatching are increased.
At present, a linear programming method is generally used for solving a day-ahead scheduling plan of the thermal power generating unit. However, the linear programming method requires that the parameters in the problem are assumed to be deterministic, and for large-scale solving of the problem, the use of linear programming for solving requires high computational resources and time. Therefore, it is difficult to solve a safe and effective thermal power plant day-ahead schedule within a prescribed time, and there are often cases where the determined day-ahead schedule is too costly to schedule.
Disclosure of Invention
The invention provides a method, a device, equipment and a storage medium for generating a day-ahead scheduling plan of a power grid, which are used for solving the technical problems that the generation efficiency of the day-ahead scheduling plan is low and the scheduling cost of the generated day-ahead scheduling plan is high.
According to an aspect of the present invention, there is provided a method for generating a day-ahead schedule of a power grid, wherein the method includes:
acquiring original state data of a target power grid at a target moment, wherein the original power grid state data comprises at least one of topological structure information, new energy output, power grid node voltage, a starting and stopping state of a thermal power unit, power output of the thermal power unit, a branch current value and power grid load;
Processing the input original state data through a trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, wherein the target scheduling instruction comprises an on-off instruction and/or an output instruction corresponding to the thermal power unit, and the strategy network model is obtained by training the original strategy model based on a historical state set of the target power grid;
predicting the input original state data and the target scheduling instruction through a trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future, wherein the virtual power grid model is obtained by training an antagonistic neural network based on the historical state set and the historical scheduling set of the thermal power generating unit;
and obtaining a daily schedule of the target power grid based on the target state data.
According to another aspect of the present invention, there is provided a day-ahead schedule generation apparatus for an electric network, wherein the apparatus includes:
the state data acquisition module is used for acquiring original state data of a target power grid at a target moment, wherein the original power grid state data comprises at least one of topological structure information, new energy output, power grid node voltage, a starting and stopping state of a thermal power unit, the output of the thermal power unit, a branch current value and a power grid load;
The scheduling instruction determining module is used for processing the input original state data through a trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, wherein the target scheduling instruction comprises an on-off instruction and/or an output instruction corresponding to the thermal power unit, and the strategy network model is obtained by training the original strategy model based on a historical state set of the target power grid;
the state data prediction module is used for predicting the input original state data and the target scheduling instruction through a trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future, wherein the virtual power grid model is obtained by training an antagonistic neural network based on the historical state set and the historical scheduling set of the thermal power generating unit;
and the scheduling plan generation module is used for obtaining a daily scheduling plan of the target power grid based on the target state data.
According to another aspect of the present invention, there is provided an electronic apparatus including:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the method of generating a day-ahead schedule for a power grid according to any of the embodiments of the present invention.
According to another aspect of the present invention, there is provided a computer readable storage medium storing computer instructions for causing a processor to implement the method for generating a daily schedule plan for a power grid according to any embodiment of the present invention when executed.
According to the technical scheme, the original state data of the target power grid at the target moment is obtained; processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, wherein the target scheduling instruction comprises a start-stop instruction and/or an output instruction corresponding to the thermal power unit; predicting the input original state data and the target scheduling instruction through a virtual power grid model after training to obtain target state data of the target power grid in a first preset time period in the future; and obtaining a day-ahead dispatch plan of the target power grid based on the target state data, so that the day-ahead dispatch plan of the target power grid with low energy consumption is automatically determined, the generation efficiency of the day-ahead dispatch plan of the power grid is improved, and the dispatch cost of the power grid is reduced.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the invention or to delineate the scope of the invention. Other features of the present invention will become apparent from the description that follows.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for generating a day-ahead schedule of a power grid according to a first embodiment of the present invention;
fig. 2 is a flowchart of a method for generating a day-ahead schedule of a power grid according to a second embodiment of the present invention;
FIG. 3 is an overall flowchart of a method for generating a day-ahead schedule of a power grid according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a day-ahead schedule generating apparatus of a power grid according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device implementing a method for generating a daily schedule of a power grid according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
Fig. 1 is a flowchart of a method for generating a daily schedule of a power grid according to an embodiment of the present invention, where the method may be applied to a case of solving a daily schedule, and the method may be performed by a daily schedule generating device of the power grid, where the daily schedule generating device of the power grid may be implemented in the form of hardware and/or software, and where the daily schedule generating device of the power grid may be configured in computer software. As shown in fig. 1, the method includes:
s110, acquiring original state data of a target power grid at a target moment.
The target power grid may be understood as a power grid from which the daily schedule is to be generated.
The target time may be understood as a starting time of the day-ahead schedule. In the embodiment of the present invention, the target time may be preset according to a scene requirement, which is not specifically limited herein. Alternatively, the target time may be 0 point of the day.
The raw state data may be understood as state data of the target grid at a target time. In the embodiment of the present invention, the raw state data may be preset according to a scene requirement, which is not specifically limited herein. Optionally, the raw data may include at least one of topology information, new energy output, grid node voltage, thermal power unit startup and shutdown state, thermal power unit output, branch current value, and grid load.
S120, the input original state data is processed through the trained optimization strategy model, and a target scheduling instruction of the thermal power unit in the target power grid is obtained.
The optimization strategy model can be understood as a model for determining a scheduling instruction of a thermal power unit in the power grid according to the state data of the power grid. Specifically, training an original strategy model based on the historical state set of the target power grid to obtain the strategy network model. Optionally, the set of historical states includes historical state data.
The thermal power generating unit can be understood as a unit for generating electricity in the target power grid.
The target scheduling instruction may be understood as an instruction to schedule the thermal power generating unit. Optionally, the target scheduling instruction may include an on/off instruction and/or an output instruction corresponding to the thermal power generating unit.
S130, predicting the input original state data and the target scheduling instruction through the trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future.
The virtual power grid model can be understood as a model for predicting state data of a future time period according to current state data of a power grid and scheduling instructions. Specifically, training the antagonistic neural network based on the historical state set and the historical scheduling set of the thermal power generating unit to obtain the virtual power grid model. Wherein the historical schedule set includes historical schedule data.
The first preset time period may be understood as a predicted time period of the virtual grid model. In the embodiment of the present invention, the first preset time period may be preset according to a scene requirement, which is not specifically limited herein. Alternatively, the first preset period of time may be 10min, 15min, 30min, or the like.
The target state data can be understood as the original state data predicted by the virtual power grid model and the future state data of the target power grid corresponding to the target scheduling instruction. Optionally, the target state data may include at least one of topology information, new energy output, power grid node voltage, thermal power unit startup and shutdown state, thermal power unit output, branch current value, and power grid load.
And S140, obtaining a daily scheduling plan of the target power grid based on the target state data.
The future scheduling plan of the target power grid can be understood as the future scheduling plan of the target power grid. Alternatively, the day-ahead schedule may be a schedule of the next day of the target grid.
Optionally, the obtaining the daily schedule of the target power grid based on the target state data includes:
Returning to execute the operation of processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of the thermal power unit in the target power grid, predicting the input original state data and the target scheduling instruction through the trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future,
and under the condition that the number of times of returning to executing operation reaches an execution number threshold, taking a plurality of target scheduling instructions and a plurality of target state data as the daily schedule of the target power grid.
The execution times threshold may be understood as a threshold for ending the return execution operation. In the embodiment of the present invention, the execution frequency threshold may be determined according to the first preset time period and a day-ahead time period corresponding to the day-ahead scheduling plan. For example, specifically, in the case where the first preset period is 15min and the day-ahead period is 24h, the execution number threshold may be 96.
Optionally, the historical state set includes historical state data, the historical scheduling set includes historical scheduling data, and before the original state data of the target power grid at the target moment is obtained, the method further includes:
Acquiring the historical state data of the target power grid, processing the historical state data to obtain a first matrix, acquiring the historical scheduling data of the thermal power generating unit, and processing the historical scheduling data to obtain a second matrix;
training the countermeasure neural network based on the first matrix and the second matrix to obtain the trained virtual power grid model, wherein the countermeasure training model comprises a generator and/or an evaluator.
The historical state data can be understood as historical state data corresponding to the target power grid. Optionally, the historical state data includes at least one of topology information, new energy output, power grid node voltage, thermal power unit startup and shutdown state, thermal power unit output, branch current value and power grid load. The first matrix may be understood as a matrix corresponding to the historical state data.
Specifically, the topology information of the target power grid and the branch current value are represented together by using a directed graph, and then the directed graph is converted into an adjacency matrix. The nodes of the target power grid have voltage attributes, and a matrix can be created, wherein each column contains node voltage values corresponding to the node voltages. The grid load is represented using an undirected graph, which is then converted into an adjacency matrix. And splicing the data of the matrixes, and carrying out normalization processing to obtain a first matrix corresponding to the historical state data.
The historical scheduling data can be understood as historical scheduling data corresponding to the target power grid. The second matrix may be understood as a matrix corresponding to the historical schedule data.
Optionally, the obtaining the historical scheduling data of the thermal power generating unit, and processing the historical scheduling data to obtain a second matrix, includes:
determining a corresponding coding value of each thermal power generating unit based on the actual output, the maximum output and the minimum output corresponding to each thermal power generating unit in the historical scheduling data;
and determining a second matrix corresponding to the historical scheduling data based on the code value.
Specifically, the maximum output of the thermal power unit is represented by Pmax, the minimum output is represented by Pmin, and then the encoding value of the thermal power unit is (actual output-Pmin)/(Pmax-Pmin). The actual output force of the thermal power generating unit is expressed by using a decimal between 0 and 1, if the coding value is 0, the thermal power generating unit is stopped, and if the coding value is 1, the output force of the thermal power generating unit is maximum. If the number of the thermal power generating units is N, the scheduling data of the thermal power generating units are represented by using N-dimensional vectors so as to obtain a second matrix.
Specifically, after the power grid state data and the thermal power generating unit dispatching data are subjected to data, training is performed on the antagonistic neural network based on the first matrix and the second matrix so as to obtain a virtual power grid model. Wherein the antagonistic neural network comprises a generator and an evaluator. The goal of the generator is to generate false grid state data so that it more closely approximates the true grid state data, and the goal of the evaluator is to maximize the probability that the true grid state data is judged to be true and the generated grid state data is judged to be false. In this way, the generator and evaluator game with each other, improving performance. Finally, a trained generator model is used for generating a power grid state which is more similar to the real power grid state, so that the starting/stopping state and the output power of the thermal power unit are adjusted under any power grid state, the next power grid state can be accurately predicted, and an accurate power grid virtual model is trained.
According to the technical scheme, the original state data of the target power grid at the target moment is obtained; processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, wherein the target scheduling instruction comprises a start-stop instruction and/or an output instruction corresponding to the thermal power unit; predicting the input original state data and the target scheduling instruction through a virtual power grid model after training to obtain target state data of the target power grid in a first preset time period in the future; and obtaining a day-ahead dispatch plan of the target power grid based on the target state data, so that the day-ahead dispatch plan of the target power grid with low energy consumption is automatically determined, the generation efficiency of the day-ahead dispatch plan of the power grid is improved, and the dispatch cost of the power grid is reduced.
Example two
Fig. 2 is a flowchart of a method for generating a day-ahead schedule of a power grid according to a second embodiment of the present invention, where the method is to add the original state data of the target power grid at the target time in the foregoing embodiment. As shown in fig. 2, the method includes:
S210, processing the historical state data of the input sample time through the original strategy model to obtain a reference scheduling instruction.
Wherein the original policy model may be understood as a non-optimized policy model.
In the embodiment of the present invention, the sample time may be the same as or different from the target time.
The reference scheduling instruction can be understood as a scheduling instruction corresponding to the historical state data of the sample time determined by the original policy model.
S220, predicting the input reference scheduling instruction and the historical state data through the virtual power grid model after training, and obtaining the reference state data of the target power grid in a second preset time period in the future.
In the embodiment of the present invention, the second preset time period may be the same as or different from the first preset time period. Alternatively, the second preset time period may be 10min, 15min, 30min, or the like.
The reference state data may be understood as future state data of the target power grid corresponding to the reference scheduling instruction and the reference state data predicted by the virtual power grid model. Optionally, the reference state data may include at least one of topology information, new energy output, power grid node voltage, thermal power unit startup and shutdown state, thermal power unit output, branch current value, and power grid load.
S230, determining the reference scheduling instruction and the reward data corresponding to the reference state data, and updating the model parameters of the original strategy model based on the historical state data, the reference scheduling instruction, the reward data and the reference state data to obtain the optimized strategy model.
The reward data may be understood as a reward value corresponding to the current reference scheduling instruction and the current reference state data.
The optimization strategy model can be understood as an optimized strategy model.
Optionally, the determining the reward data corresponding to the reference scheduling instruction and the reference state data includes:
determining a scheduling constraint of the thermal power generating unit and a state constraint of the target power grid, determining the reward data based on the reference scheduling instruction, the reference state data, the scheduling constraint and the state constraint,
the reward data comprises a positive reward value and/or a negative reward value, the scheduling constraint comprises at least one of a load balance constraint, a climbing constraint of a thermal power unit, an upper limit and a lower limit of the output force of the thermal power unit, a system-level balance constraint, a power transmission and transformation load capacity constraint and a bus voltage constraint, and the state constraint comprises at least one of a standby balance, a start-stop frequency constraint of the thermal power unit, a power grid safety constraint and a bus voltage constraint.
In the power system provided by the embodiment of the invention, standby balance, thermal power unit start-stop frequency constraint, power grid safety constraint, bus voltage constraint and the like are important power grid state constraint. The standby balance requirement is that a certain amount of standby capacity must be reserved to cope with the conditions of sudden load or thermal power unit faults and the like; the amount of the up-and-down adjustment of the output power of the thermal power generating unit is required to be larger than a constraint value; the constraint of the start-stop times of the thermal power generating unit requires that the maximum start-stop times of the thermal power generating unit in a dispatching period is smaller than a certain value; the power grid safety constraint requires that an environment characteristic value is calculated according to the state of the power system, and the environment characteristic value is in a range; the load capacity constraint of the power transmission and transformation equipment requires that the branch current of the state of the power system is smaller than a certain value at any moment; bus voltage constraints require whether the node voltage is within a certain range at any time depending on the state of the power system. By constraining the grid state, when the grid state exceeds the operating range, the round is terminated and deduction is not continued.
In the power system provided by the embodiment of the invention, load balance constraint, climbing constraint of the thermal power unit, upper and lower limit constraint of the output of the thermal power unit, system-level balance constraint, power transmission and transformation load capacity constraint, bus voltage constraint and the like are important thermal power unit scheduling constraint. Aiming at load balance constraint, the sum of the output of thermal power units is constrained according to the load of a local power grid and the injection power of other energy sources, and the total load of the local power grid is the sum of the output of each unit of the local power grid and the injection power of other energy sources; aiming at climbing constraint of the thermal power unit and upper and lower limit constraint of output force of the thermal power unit, calculating a feasible region range of output power of the thermal power unit, and limiting the calculated output force of the thermal power unit by using the feasible region range so that the output power of the thermal power unit meets the climbing constraint; aiming at system-level balance constraint, the sum of new energy power generation and thermal power generation of the whole power grid is equal to the load of the whole power system.
Optionally, the determining the reward data based on the reference scheduling instruction, the reference status data, the scheduling constraint, and the status constraint includes:
determining a first parameter and a second parameter under the condition that the reference scheduling instruction does not exceed the scheduling constraint and the reference state data does not exceed the state constraint, wherein the first parameter is a parameter representing the sum of the total running cost and the total starting cost of the thermal power unit, and the second parameter is a parameter representing the sum of the output of the thermal power unit;
and determining the positive rewards value based on the first parameter and the second parameter, and taking the current positive rewards value as the rewards data corresponding to the current historical state data.
In the embodiment of the invention, the smaller the first parameter (i.e. the sum of the running total cost and the starting total cost of the thermal power generating unit is lower) and the smaller the second parameter (the smaller the sum of the output of the thermal power generating unit is, i.e. the higher the new energy consumption rate is), the larger the positive rewarding value is. Thus, the larger the positive prize value is indicative of lower overall running and starting costs, the higher the new energy consumption rate, the higher the current degree of optimization of the reference scheduling instructions and the reference status data, and the lower the energy consumption of the day-ahead scheduling plan.
Optionally, the determining the reward data based on the reference scheduling instruction, the reference status data, the scheduling constraint, and the status constraint further includes:
when the reference scheduling instruction exceeds the scheduling constraint and/or the reference state data exceeds the state constraint, taking the preset negative rewarding value as the rewarding data corresponding to the current historical state data;
and finishing training the original strategy model by the historical state data of the current sample time, and training the original strategy model based on the historical state data of the next sample time.
The negative reward value may be preset according to the scene requirement, and is not specifically limited herein. Alternatively, the negative prize value may be-1000.
In the embodiment of the invention, in the electric power system, a reward function is designed according to the data such as the total running cost of the thermal power unit, the starting cost of the thermal power unit, the new energy consumption rate, whether the power grid state meets the constraint condition or not and the like. Specifically, the total operation cost and the starting cost are calculated according to the active power and/or the start-stop state of the thermal power unit, the sum of the total operation cost and the starting cost is marked as C, and the sum of the output of the thermal power unit is marked as P.
According to the constraint interface, whether the current start-stop and/or output of the thermal power generating unit exceeds preset scheduling constraint or not is judged under the current power grid state, and whether the current power grid state exceeds preset state constraint or not is judged. If there is a condition that the constraint is exceeded, an error is noted.
Specifically, the calculation formula for determining the bonus data is:
wherein a represents a coefficient of influence degree of sum of thermal power unit output on rewards, C represents a first parameter, and P represents a second parameter.
According to the embodiment of the invention, the rewarding data are introduced, and the model parameters of the original strategy model are updated based on the rewarding data, so that the new energy consumption rate corresponding to the target scheduling instruction determined based on the optimization strategy model is higher, and the output of the thermal power unit is reduced, the new energy consumption rate is improved, and the total energy consumption is reduced under the condition that the target scheduling instruction meets the scheduling constraint and the target power grid state meets the state constraint.
Specifically, the updating the model parameters of the original policy model based on the historical state data, the reference scheduling instruction, the reward data and the reference state data to obtain the optimized policy model may be: reinforcement learning algorithm implementation. The scheduling strategy of the power grid is optimized by using a near-end strategy optimization (Proximal Policy Optimization, PPO) algorithm, the PPO algorithm inputs historical state data into an actor network (original strategy model) to obtain two values, one is a mean value and the other is a variance, the two values are respectively used as the mean value and the variance of Normal distribution to construct Normal distribution Normal1, a reference scheduling instruction is generated from the Normal1, the reference scheduling instruction and the historical state data are input into a virtual power grid model with training completed to obtain reward data and reference state data of the next step, and finally the quadruple data (the historical state data, the reference scheduling instruction, the reward data and the reference state data) are stored in a buffer pool. Inputting all stored state data into an action-old network and an action-new network to respectively obtain Normal distribution Normal1 and Normal2, combining all stored actions into actions, inputting the actions into the Normal distribution Normal1 and Normal2 to obtain prob1 and prob2 corresponding to each action, and dividing prob2 by prob1 to obtain a ratio. And calculating a loss value, and then, reversely transmitting and updating an actor-new network, so as to optimize the control of the thermal power unit and obtain a higher Critic score. The actor-new parameter is updated by the decision gradient. After updating several times, the parameters of the actor-new network are synchronized to the actor-old network. And calculating the loss by adopting a mean square error for the Critic network, and updating the neural network parameters by using a back propagation algorithm to obtain the optimization strategy model.
S240, acquiring original state data of the target power grid at the target moment.
S250, processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of the thermal power unit in the target power grid.
S260, predicting the input original state data and the target scheduling instruction through the trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future.
And S270, obtaining a daily scheduling plan of the target power grid based on the target state data.
According to the technical scheme, the original strategy model is used for processing the historical state data of the input sample moment to obtain a reference scheduling instruction; predicting the input reference scheduling instruction and the historical state data through the trained virtual power grid model to obtain reference state data of the target power grid in a second preset time period in the future; determining reward data corresponding to the reference scheduling instruction and the reference state data, and updating model parameters of the original strategy model based on the historical state data, the reference scheduling instruction, the reward data and the reference state data to obtain the optimization strategy model, so that the accuracy of the determined optimization strategy model is improved.
Fig. 3 is an overall flowchart of a method for generating a day-ahead schedule of a power grid according to an embodiment of the present invention. As shown in fig. 3, the overall flow of the method for generating the daily schedule of the power grid may be:
1. and (5) processing power grid state data. 2. And (5) scheduling data processing of the thermal power generating unit. 3. Virtual grid model training. 4. Grid state constraints. 5. Thermal power generating unit scheduling constraints. 6. And (5) scheduling and rewarding design of the thermal power generating unit. 7. Reinforcement learning algorithm implementation.
8. And solving a power grid day-ahead scheduling plan. Day-ahead dispatch plan solutions were performed at 0 point per day. Specifically, the original state data of the target power grid at the target moment is obtained, the original state data is encoded and then is input into an optimized Actor network (optimizing strategy model), the Actor network outputs a thermal power unit starting/stopping and power output dispatching instruction, the dispatching instruction and the power grid state are input into a trained virtual power grid model, the virtual power grid model is used for predicting power grid node voltage and/or power grid topological structure energy data (target state data) after 15 minutes, the target state data is input into the Actor network, and the thermal power unit dispatching instruction after 15 minutes is output until the thermal power unit dispatching control of the next day is calculated and used as a future dispatching plan of the target power grid.
According to the invention, a virtual power grid model is constructed through historical state data and historical scheduling data of a power grid, and on the basis, the power grid constraint conditions such as standby balance of the power grid, start-stop times constraint of the thermal power unit and bus voltage constraint are met through constraint on the power grid state and thermal power unit scheduling strategy. And then designing a reward function, guiding a regulation strategy of the thermal power unit to reduce the total cost of the unit and improving the new energy consumption rate. And then, a reinforcement learning PPO algorithm is realized, and a strategy network and an evaluation network are optimized, so that the evaluation network can evaluate the current power grid state more accurately, and the strategy network outputs a better thermal power unit control strategy. And finally, outputting a thermal power unit control strategy according to the predicted next-day power grid load and the new energy power generation amount. Under the condition of uncertainty of power generation and load of new energy equipment, the control strategy output of the thermal power unit is faster when the dispatching of the thermal power unit meets all conventional thermal power unit operation constraint conditions.
Example III
Fig. 4 is a schematic structural diagram of a device for generating a daily schedule of a power grid according to a third embodiment of the present invention. As shown in fig. 4, the apparatus includes: a status data acquisition module 310, a scheduling instruction determination module 320, and a status data prediction module 330; wherein,
The state data obtaining module 310 is configured to obtain original state data of a target power grid at a target moment, where the original power grid state data includes at least one of topology information, new energy output, power grid node voltage, a start-stop state of a thermal power unit, output of the thermal power unit, a branch current value, and a power grid load; the scheduling instruction determining module 320 is configured to process the input original state data through a trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, where the target scheduling instruction includes an on-off instruction and/or an output instruction corresponding to the thermal power unit, and the strategy network model trains the original strategy model based on a historical state set of the target power grid; the state data prediction module 330 is configured to predict, by using a trained virtual power grid model, the input original state data and the target scheduling instruction to obtain target state data of the target power grid within a first preset time period in the future, where the virtual power grid model is obtained by training an antagonistic neural network based on the historical state set and the historical scheduling set of the thermal power generating unit; the scheduling plan generating module 340 is configured to obtain a daily scheduling plan of the target power grid based on the target state data.
According to the technical scheme, the original state data of the target power grid at the target moment is obtained; processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, wherein the target scheduling instruction comprises a start-stop instruction and/or an output instruction corresponding to the thermal power unit; predicting the input original state data and the target scheduling instruction through a virtual power grid model after training to obtain target state data of the target power grid in a first preset time period in the future; and obtaining a day-ahead dispatch plan of the target power grid based on the target state data, so that the day-ahead dispatch plan of the target power grid with low energy consumption is automatically determined, the generation efficiency of the day-ahead dispatch plan of the power grid is improved, and the dispatch cost of the power grid is reduced.
Optionally, the scheduling plan generating module 340 is configured to:
returning to execute the operation of processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of the thermal power unit in the target power grid, predicting the input original state data and the target scheduling instruction through the trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future,
And under the condition that the number of times of returning to executing operation reaches an execution number threshold, taking a plurality of target scheduling instructions and a plurality of target state data as the daily schedule of the target power grid.
Optionally, the history state set includes history state data, the history scheduling set includes history scheduling data, and the device for generating a day-ahead scheduling plan of the power grid further includes a history data processing module and a virtual model training module; wherein,
the historical data processing module is used for acquiring the historical state data of the target power grid before the original state data of the target power grid at the target moment is acquired, processing the historical state data to obtain a first matrix, acquiring the historical scheduling data of the thermal power generating unit, and processing the historical scheduling data to obtain a second matrix;
the virtual model training module is configured to train the countermeasure neural network based on the first matrix and the second matrix, and obtain the trained virtual power grid model, where the countermeasure training model includes a generator and/or an evaluator.
Optionally, the device for generating the daily schedule of the power grid further comprises a reference instruction determining module, a reference state predicting module and a strategy model training module; wherein,
The reference instruction determining module is used for processing the historical state data of the input sample moment through the original strategy model before the original state data of the target power grid at the target moment is acquired, so as to obtain a reference scheduling instruction;
the reference state prediction module is used for predicting the input reference scheduling instruction and the historical state data through the virtual power grid model after training is completed, so as to obtain reference state data of the target power grid in a second preset time period in the future;
the strategy model training module is used for determining the reference scheduling instruction and the reward data corresponding to the reference state data, and updating the model parameters of the original strategy model based on the historical state data, the reference scheduling instruction, the reward data and the reference state data to obtain the optimized strategy model.
Optionally, the policy model training module includes a data constraint unit, configured to:
determining a scheduling constraint of the thermal power generating unit and a state constraint of the target power grid, determining the reward data based on the reference scheduling instruction, the reference state data, the scheduling constraint and the state constraint,
The reward data comprises a positive reward value and/or a negative reward value, the scheduling constraint comprises at least one of a load balance constraint, a climbing constraint of a thermal power unit, an upper limit and a lower limit of the output force of the thermal power unit, a system-level balance constraint, a power transmission and transformation load capacity constraint and a bus voltage constraint, and the state constraint comprises at least one of a standby balance, a start-stop frequency constraint of the thermal power unit, a power grid safety constraint and a bus voltage constraint.
Optionally, the data constraint unit is configured to:
determining a first parameter and a second parameter under the condition that the reference scheduling instruction does not exceed the scheduling constraint and the reference state data does not exceed the state constraint, wherein the first parameter is a parameter representing the sum of the total running cost and the total starting cost of the thermal power unit, and the second parameter is a parameter representing the sum of the output of the thermal power unit;
and determining the positive rewards value based on the first parameter and the second parameter, and taking the current positive rewards value as the rewards data corresponding to the current historical state data.
Optionally, the data constraint unit is further configured to:
when the reference scheduling instruction exceeds the scheduling constraint and/or the reference state data exceeds the state constraint, taking the preset negative rewarding value as the rewarding data corresponding to the current historical state data;
And finishing training the original strategy model by the historical state data of the current sample time, and training the original strategy model based on the historical state data of the next sample time.
The day-ahead schedule generating device of the power grid provided by the embodiment of the invention can execute the day-ahead schedule generating method of the power grid provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the executing method.
Example IV
Fig. 5 shows a schematic diagram of the structure of an electronic device 10 that may be used to implement an embodiment of the invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.
As shown in fig. 5, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various appropriate actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.
Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above, such as the day-ahead schedule generation method of the power grid.
In some embodiments, the day-ahead schedule generation method of the power grid may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the grid day-ahead schedule generation method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the day-ahead schedule generation method of the power grid in any other suitable manner (e.g., by means of firmware).
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
A computer program for carrying out methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.
The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.
The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method for generating a day-ahead schedule of a power grid, comprising:
acquiring original state data of a target power grid at a target moment, wherein the original power grid state data comprises at least one of topological structure information, new energy output, power grid node voltage, a starting and stopping state of a thermal power unit, power output of the thermal power unit, a branch current value and power grid load;
processing the input original state data through a trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, wherein the target scheduling instruction comprises an on-off instruction and/or an output instruction corresponding to the thermal power unit, and the strategy network model is obtained by training the original strategy model based on a historical state set of the target power grid;
Predicting the input original state data and the target scheduling instruction through a trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future, wherein the virtual power grid model is obtained by training an antagonistic neural network based on the historical state set and the historical scheduling set of the thermal power generating unit;
and obtaining a daily schedule of the target power grid based on the target state data.
2. The method of claim 1, wherein the deriving a daily schedule of the target grid based on the target state data comprises:
returning to execute the operation of processing the input original state data through the trained optimization strategy model to obtain a target scheduling instruction of the thermal power unit in the target power grid, predicting the input original state data and the target scheduling instruction through the trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future,
and under the condition that the number of times of returning to executing operation reaches an execution number threshold, taking a plurality of target scheduling instructions and a plurality of target state data as the daily schedule of the target power grid.
3. The method of claim 1, wherein the set of historical states includes historical state data, the set of historical schedules includes historical schedule data, and further comprising, prior to the obtaining of the raw state data of the target grid at the target time:
acquiring the historical state data of the target power grid, processing the historical state data to obtain a first matrix, acquiring the historical scheduling data of the thermal power generating unit, and processing the historical scheduling data to obtain a second matrix;
training the countermeasure neural network based on the first matrix and the second matrix to obtain the trained virtual power grid model, wherein the countermeasure training model comprises a generator and/or an evaluator.
4. The method of claim 1, further comprising, prior to said obtaining raw state data of the target grid at the target time,:
processing the historical state data of the input sample moment through the original strategy model to obtain a reference scheduling instruction;
predicting the input reference scheduling instruction and the historical state data through the trained virtual power grid model to obtain reference state data of the target power grid in a second preset time period in the future;
Determining reward data corresponding to the reference scheduling instruction and the reference state data, and updating model parameters of the original strategy model based on the historical state data, the reference scheduling instruction, the reward data and the reference state data to obtain the optimized strategy model.
5. The method of claim 4, wherein the determining the bonus data corresponding to the reference scheduling instruction and the reference status data comprises:
determining a scheduling constraint of the thermal power generating unit and a state constraint of the target power grid, determining the reward data based on the reference scheduling instruction, the reference state data, the scheduling constraint and the state constraint,
the reward data comprises a positive reward value and/or a negative reward value, the scheduling constraint comprises at least one of a load balance constraint, a climbing constraint of a thermal power unit, an upper limit and a lower limit of the output force of the thermal power unit, a system-level balance constraint, a power transmission and transformation load capacity constraint and a bus voltage constraint, and the state constraint comprises at least one of a standby balance, a start-stop frequency constraint of the thermal power unit, a power grid safety constraint and a bus voltage constraint.
6. The method of claim 5, wherein the determining the bonus data based on the reference scheduling instruction, the reference status data, the scheduling constraint, and the status constraint comprises:
determining a first parameter and a second parameter under the condition that the reference scheduling instruction does not exceed the scheduling constraint and the reference state data does not exceed the state constraint, wherein the first parameter is a parameter representing the sum of the total running cost and the total starting cost of the thermal power unit, and the second parameter is a parameter representing the sum of the output of the thermal power unit;
and determining the positive rewards value based on the first parameter and the second parameter, and taking the current positive rewards value as the rewards data corresponding to the current historical state data.
7. The method of claim 5, wherein the determining the bonus data based on the reference scheduling instruction, the reference status data, the scheduling constraint, and the status constraint further comprises:
when the reference scheduling instruction exceeds the scheduling constraint and/or the reference state data exceeds the state constraint, taking the preset negative rewarding value as the rewarding data corresponding to the current historical state data;
And finishing training the original strategy model by the historical state data of the current sample time, and training the original strategy model based on the historical state data of the next sample time.
8. A day-ahead schedule generation device for an electric network, comprising:
the state data acquisition module is used for acquiring original state data of a target power grid at a target moment, wherein the original power grid state data comprises at least one of topological structure information, new energy output, power grid node voltage, a starting and stopping state of a thermal power unit, the output of the thermal power unit, a branch current value and a power grid load;
the scheduling instruction determining module is used for processing the input original state data through a trained optimization strategy model to obtain a target scheduling instruction of a thermal power unit in the target power grid, wherein the target scheduling instruction comprises an on-off instruction and/or an output instruction corresponding to the thermal power unit, and the strategy network model is obtained by training the original strategy model based on a historical state set of the target power grid;
the state data prediction module is used for predicting the input original state data and the target scheduling instruction through a trained virtual power grid model to obtain target state data of the target power grid in a first preset time period in the future, wherein the virtual power grid model is obtained by training an antagonistic neural network based on the historical state set and the historical scheduling set of the thermal power generating unit;
And the scheduling plan generation module is used for obtaining a daily scheduling plan of the target power grid based on the target state data.
9. An electronic device, the electronic device comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the method of generating a daily schedule for a power grid as claimed in any one of claims 1 to 7.
10. A computer readable storage medium storing computer instructions for causing a processor to perform the method of generating a daily schedule for a power grid according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311250242.1A CN117293923A (en) | 2023-09-25 | 2023-09-25 | Method, device, equipment and storage medium for generating day-ahead scheduling plan of power grid |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311250242.1A CN117293923A (en) | 2023-09-25 | 2023-09-25 | Method, device, equipment and storage medium for generating day-ahead scheduling plan of power grid |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117293923A true CN117293923A (en) | 2023-12-26 |
Family
ID=89253077
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311250242.1A Pending CN117293923A (en) | 2023-09-25 | 2023-09-25 | Method, device, equipment and storage medium for generating day-ahead scheduling plan of power grid |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117293923A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190165580A1 (en) * | 2017-11-27 | 2019-05-30 | Ihi Inc. | System and method for optimal control of energy storage system |
| CN109919399A (en) * | 2019-04-15 | 2019-06-21 | 中国科学院电工研究所 | A kind of integrated energy system economic load dispatching method and system a few days ago |
| US20210055700A1 (en) * | 2019-08-19 | 2021-02-25 | Invenia Technical Computing Corporation | Hybrid machine learning and simulation based system for forecasting in electricity systems |
| CN115526737A (en) * | 2022-09-23 | 2022-12-27 | 国网山西省电力公司 | Power grid energy management method and system based on digital twinning and terminal equipment |
| CN116050730A (en) * | 2022-12-02 | 2023-05-02 | 中国华能集团清洁能源技术研究院有限公司 | Method and system for generating daily scheduling plan of comprehensive energy base |
| CN116258192A (en) * | 2022-11-22 | 2023-06-13 | 华能陇东能源有限责任公司 | Day-ahead scheduling plan generation model training method, plan generation method and device |
| CN116345578A (en) * | 2023-05-26 | 2023-06-27 | 南方电网数字电网研究院有限公司 | Micro-grid operation optimization scheduling method based on depth deterministic strategy gradient |
| CN116545025A (en) * | 2023-04-04 | 2023-08-04 | 广东电网有限责任公司 | Optimal scheduling method, device, equipment and storage medium for power distribution network |
-
2023
- 2023-09-25 CN CN202311250242.1A patent/CN117293923A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190165580A1 (en) * | 2017-11-27 | 2019-05-30 | Ihi Inc. | System and method for optimal control of energy storage system |
| CN109919399A (en) * | 2019-04-15 | 2019-06-21 | 中国科学院电工研究所 | A kind of integrated energy system economic load dispatching method and system a few days ago |
| US20210055700A1 (en) * | 2019-08-19 | 2021-02-25 | Invenia Technical Computing Corporation | Hybrid machine learning and simulation based system for forecasting in electricity systems |
| CN115526737A (en) * | 2022-09-23 | 2022-12-27 | 国网山西省电力公司 | Power grid energy management method and system based on digital twinning and terminal equipment |
| CN116258192A (en) * | 2022-11-22 | 2023-06-13 | 华能陇东能源有限责任公司 | Day-ahead scheduling plan generation model training method, plan generation method and device |
| CN116050730A (en) * | 2022-12-02 | 2023-05-02 | 中国华能集团清洁能源技术研究院有限公司 | Method and system for generating daily scheduling plan of comprehensive energy base |
| CN116545025A (en) * | 2023-04-04 | 2023-08-04 | 广东电网有限责任公司 | Optimal scheduling method, device, equipment and storage medium for power distribution network |
| CN116345578A (en) * | 2023-05-26 | 2023-06-27 | 南方电网数字电网研究院有限公司 | Micro-grid operation optimization scheduling method based on depth deterministic strategy gradient |
Non-Patent Citations (1)
| Title |
|---|
| 蔡秋娜 等: "考虑出力稳定性的日前调度计划模型与方法", 电气应用, vol. 2018, no. 7, 5 April 2018 (2018-04-05), pages 46 - 49 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115345380A (en) | New energy consumption electric power scheduling method based on artificial intelligence | |
| CN116931438B (en) | Method, device, equipment and medium for determining parameters of speed regulator of water turbine | |
| CN115833101B (en) | Power scheduling method, device, electronic equipment and storage medium | |
| CN117332896A (en) | New energy small time scale power prediction method and system for multilayer integrated learning | |
| CN115347621A (en) | Scheduling method and device of combined power generation system, electronic equipment and medium | |
| CN116502771A (en) | Power distribution method and system based on electric power material prediction | |
| CN116454902A (en) | Power distribution network voltage regulating method, device, equipment and storage medium based on reinforcement learning | |
| CN117293923A (en) | Method, device, equipment and storage medium for generating day-ahead scheduling plan of power grid | |
| CN114943384A (en) | Transformer substation load prediction method, device, equipment and storage medium | |
| CN114611805A (en) | Net load prediction method and device, electronic equipment and storage medium | |
| CN117094537B (en) | Power grid planning method, device, electronic equipment and storage medium | |
| CN117575175B (en) | Carbon emission evaluation method, device, electronic equipment and storage medium | |
| CN115373449B (en) | Data processing method, device, equipment and storage medium | |
| CN117130280B (en) | Pump room control method and device, electronic equipment and storage medium | |
| CN118100233A (en) | Micro-grid group energy storage capacity configuration method and device, electronic equipment and storage medium | |
| CN114626636A (en) | Power grid load prediction method and device, modeling method, computer equipment and medium | |
| CN117955150A (en) | Control method and system of wind driven generator grid-connected system | |
| CN117252612A (en) | Method and device for determining power supply and selling scheme, electronic equipment and storage medium | |
| CN115912355A (en) | Method, device, equipment and medium for dividing power supply area of transformer substation | |
| CN115425646A (en) | Power supply planning method and device for low-voltage transformer area, electronic equipment and storage medium | |
| CN115375186A (en) | Distributed renewable energy dynamic polymerization method, device, equipment and medium | |
| CN118134019A (en) | Power distribution network topology optimization method, device, equipment and storage medium | |
| CN116646991A (en) | Power generation scheduling method, device, equipment and storage medium of power system | |
| Liu et al. | The application of the improved Particle Swarm Optimization on dynamic economic dispatch of power system with wind farms | |
| CN116523249A (en) | Production line determining method, device, equipment and storage medium |
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 |