CN110991731A - Electric power real-time market constraint self-identification clearing method and system based on deep learning - Google Patents
Electric power real-time market constraint self-identification clearing method and system based on deep learning Download PDFInfo
- Publication number
- CN110991731A CN110991731A CN201911194707.XA CN201911194707A CN110991731A CN 110991731 A CN110991731 A CN 110991731A CN 201911194707 A CN201911194707 A CN 201911194707A CN 110991731 A CN110991731 A CN 110991731A
- Authority
- CN
- China
- Prior art keywords
- constraint
- clearing
- model
- power
- output
- 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 48
- 238000013135 deep learning Methods 0.000 title claims abstract description 29
- 239000013598 vector Substances 0.000 claims abstract description 104
- 238000012545 processing Methods 0.000 claims abstract description 55
- 238000004458 analytical method Methods 0.000 claims abstract description 45
- 230000003068 static effect Effects 0.000 claims abstract description 42
- 238000010606 normalization Methods 0.000 claims abstract description 35
- 230000009471 action Effects 0.000 claims abstract description 34
- 230000005540 biological transmission Effects 0.000 claims abstract description 28
- 238000013528 artificial neural network Methods 0.000 claims abstract description 27
- 238000003062 neural network model Methods 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 22
- 230000006870 function Effects 0.000 claims abstract description 18
- 230000004913 activation Effects 0.000 claims abstract description 16
- 230000009194 climbing Effects 0.000 claims description 11
- 238000002347 injection Methods 0.000 claims description 9
- 239000007924 injection Substances 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 6
- 230000035945 sensitivity Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 210000002569 neuron Anatomy 0.000 claims description 5
- 238000009877 rendering Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 230000005611 electricity Effects 0.000 description 9
- 238000010248 power generation Methods 0.000 description 8
- 238000010926 purge Methods 0.000 description 6
- 238000000605 extraction Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 239000000243 solution Substances 0.000 description 4
- 230000003466 anti-cipated effect Effects 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 239000000446 fuel Substances 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000007781 pre-processing Methods 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
- 230000001502 supplementing effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000003442 weekly effect Effects 0.000 description 1
Images
Classifications
-
- 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
- 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
- G06N3/08—Learning methods
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0206—Price or cost determination based on market factors
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Marketing (AREA)
- Finance (AREA)
- Entrepreneurship & Innovation (AREA)
- Accounting & Taxation (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Tourism & Hospitality (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Game Theory and Decision Science (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Primary Health Care (AREA)
- Quality & Reliability (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Operations Research (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a deep learning-based electric power real-time market constraint self-identification clearing method and system, wherein the method comprises the following steps: after the system characteristic vector of the power system is subjected to standardization processing, iterative training is carried out through an activation model to obtain a deep neural network model; performing inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit, and performing first static security analysis and line transmission capacity constraint processing on the generator optimal output to obtain an action constraint set; and calculating the clearing result of the current system according to the clearing model, the function constraint set and the system constraint, performing second static security analysis on the clearing result, and taking the clearing result as the optimal clearing result when all parameters of the current system meet given requirements. The SCED function constraint high-efficiency identification method can realize SCED function constraint high-efficiency identification, reduce function constraint search times and improve market clearing efficiency.
Description
Technical Field
The invention relates to the technical field of power markets, in particular to a deep learning-based power real-time market constraint self-identification clearing method and system.
Background
The electric power system is an electric energy production and consumption system which consists of links of power generation, power transmission, power transformation, power distribution, power utilization and the like. The function of the device is to convert the primary energy of the nature into electric energy through a power generation device, and then supply the electric energy to each user through power transmission, power transformation and power distribution. In order to realize the function, the power system is also provided with corresponding information and control systems at each link and different levels, and the production process of the electric energy is measured, regulated, controlled, protected, communicated and scheduled so as to ensure that users obtain safe, economic and high-quality electric energy. The whole formed by the substation and the transmission and distribution lines of various voltages in the power system is called as a power grid.
The EMS energy management system is a general name of a modern power grid dispatching automation system, and the main function of the EMS energy management system consists of a basic function and an application function. The basic functions comprise a computer, an operation system and an EMS support system; the application functions include data acquisition and monitoring (SCADA), Automatic Generation Control (AGC) and planning, and network application analysis. The economic dispatching of the power system is the main content of an Energy Management System (EMS), and is equivalent to a power generation plan in the concept category under some specific environments, wherein the power generation plan comprises a unit combination, a water-fire-electricity plan, an exchange plan, a maintenance plan, a fuel plan and the like; according to the period, the system comprises an ultra-short period plan, namely Automatic Generation Control (AGC), and a short period generation plan, namely a daily or weekly plan; a middle-term power generation plan, namely a monthly to yearly plan and correction; long-term planning, i.e., planning for years to decades, includes power supply development planning, network development planning, and the like.
The reasonable and optimized configuration of energy resources is guided by using the electricity price, the high-proportion new energy consumption of the power system is promoted, and the node electricity price is generally adopted as a market pricing mechanism at present. Market clearing and node electricity price formulation need to solve a Security-constrained eco-not dispatch (SCED) model. However, since the SCED model contains N-1 massive security constraints. Although the SCED model is a linear optimization model, even if the existing commercial linear solvers such as CPLEX and GUROBI are adopted, the SCED model of the large-scale practical system is still difficult to solve effectively. However, although the scale of the constraints in the SCED model is huge, the proportion of the real functioning constraints is small, and therefore, a scientific and effective efficient method for identifying the functioning constraints of the SCED is urgently needed in the actual industry. In view of the above, it is urgently needed to provide a method capable of mining a nonlinear relationship between a system operating condition and an action constraint, implementing efficient identification of SCED action constraint, reducing the number of times of action constraint search, and improving market clearing efficiency.
Disclosure of Invention
In order to solve the above problems, the present invention aims to provide a deep learning-based real-time electric power market constraint self-identification clearing method and system, which can realize efficient identification of SCED functioning constraints, reduce the number of functioning constraint lookups, and improve the market clearing efficiency.
A deep learning based power real-time market constraint self-identification clearing method, the method comprising:
acquiring power information of each node area in a power system, standardizing the extracted system characteristic vector of the system, and performing iterative training through an activation model to obtain a deep neural network model, wherein the system characteristic vector comprises an input characteristic vector and an output characteristic vector;
performing inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit, and performing first static safety analysis and line transmission capacity constraint processing on the generator optimal output to obtain an action constraint set;
and calculating the clearing result of the current system according to the clearing model, the action constraint set and the system constraint, performing second static security analysis on the clearing result, and taking the clearing result as the optimal clearing result when all parameters of the current system meet given requirements.
According to the electric power real-time market constraint self-identification clearing method based on deep learning, system feature extraction is carried out on electric power information in an electric power system, so that system feature vectors capable of effectively reflecting new energy, load fluctuation and electricity price fluctuation are obtained; the method comprises the steps that through standardization processing and iterative training of system feature vectors, a well-trained deep neural network model is obtained, and necessary conditions are provided for subsequent action constraint through standard optimal output and actual optimal output of a generator; performing inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit; the optimal output of the generator is subjected to first static safety analysis and line transmission capacity constraint processing to obtain an acting constraint set, so that the high-efficiency identification of the acting constraint is improved, and the searching times of the acting constraint are reduced; and calculating the clearing result of the current system according to the clearing model, the action constraint set and the system constraint, and performing second static security analysis on the clearing result so as to take the clearing result as the optimal clearing result when all parameters of the current system meet given requirements. The invention provides technical support for the deep learning technology to effectively identify the functional constraint in the definition model from two aspects of feature vector and deep neural network result processing, excavates the nonlinear relation between the system operation condition and the functional constraint, realizes the efficient identification of the functional constraint in the definition model, reduces the searching times of the functional constraint, improves the definition efficiency of the market, and meets the actual application requirements.
In addition, according to the above-mentioned electric real-time market constraint self-identification clearing method based on deep learning of the present invention, the following additional technical features may also be provided:
further, after the step of performing the second static security analysis on the clearing result, the method further includes:
and when any parameter of the current system does not meet the given requirement, converting the out-of-limit line which does not meet the given requirement into a newly added constraint, and recalculating the clearing result of the current system according to the clearing model.
Further, the method for obtaining the power information of each node area in the power system, after the extracted system feature vector of the system is subjected to standardization processing, iterative training is carried out through an activation model to obtain the deep neural network model comprises the following steps:
extracting node injection active power, node injection reactive power and power plant quotation secondary curve coefficients of each node area in the power system as input characteristic vectors, and taking the output of a corresponding generator as output characteristic vectors;
respectively carrying out standardization processing on the input feature vector and the output feature vector through a standardization model to obtain a corresponding standard input feature vector and a corresponding standard output feature vector;
inputting the standard input characteristic vector and the standard output characteristic vector into the activation model for iterative training to obtain the deep neural network model.
Further, the deep neural network model is:
where ω includes the weight of each connection of the neural network and the deviation value on each neuron,
the gradient of ω is α is the learning rate constant, β is the weight coefficient, and ε is the adjustment value.
Further, the method for obtaining the optimal output of the generator of the current unit, and performing the first static safety analysis and the line transmission capacity constraint processing on the optimal output of the generator to obtain the action constraint set comprises the following steps:
inputting the standard input characteristic vector and the standard output characteristic vector into a deep neural network for calculation so as to obtain the standardized optimal output of the generator;
performing inverse normalization processing on the normalized optimal output of the generator through an inverse normalization model to obtain the optimal output of the generator of the current unit;
and carrying out first static safety analysis on the optimal output of the generator, and adding the out-of-limit line into a line transmission capacity constraint to obtain an action constraint set.
Further, the method for calculating the clearing result of the current system according to the clearing model, the action constraint set and the system constraint and performing the second static security analysis on the clearing result comprises the following steps:
constraining a clearing model for calculating a clearing result of the current system according to the acting constraint set, the system balance constraint, the unit output upper and lower limit constraint, the unit climbing constraint and the power grid safety constraint;
and calculating the clearing result of the current system through the clearing model after constraint, and performing second static security analysis on the clearing result to judge whether each parameter of the current system meets the given requirement.
Further, the output model is as follows:
Pi,min≤Pi,t≤Pi,max
-Rd≤Pi,t-Pi,t-1≤Ru
wherein, Pi,tThe output of the unit i at time t, ai、bi、ciIs the running cost coefficient, L, of the unit id,tFor node d load demand at time t, Pi,minAnd Pi,maxRespectively the upper and lower limits of the unit output, RdAnd RuRespectively the upper and lower climbing speed limit values of the unit,
for the transmission active power of the branch (i, j) during the t-th time period under the c-th anticipated fault,
and
corresponding to the upper and lower power limits of the branch (i, j) under the expected failure of the c-th branch,
the sensitivity coefficients K, C, M of the injected power for node i to branch (i, j) are branch set, fault set, and node set, respectively.
Another embodiment of the invention provides a deep learning-based real-time electric power market restriction self-recognition clearing system, which solves the problems that the existing SCED has low efficiency of acting restriction recognition and more times of acting restriction search, thereby causing low market clearing efficiency.
The electric power real-time market constraint self-identification system based on deep learning comprises the following components:
the acquisition module is used for acquiring power information of each node area in a power system, carrying out standardization processing on the extracted system characteristic vector of the system, and then carrying out iterative training through an activation model to obtain a deep neural network model, wherein the system characteristic vector comprises an input characteristic vector and an output characteristic vector;
the processing module is used for carrying out inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model so as to obtain the generator optimal output of the current unit, and carrying out first static safety analysis and line transmission capacity constraint processing on the generator optimal output so as to obtain an action constraint set;
and the clearing module is used for calculating a clearing result of the current system according to the clearing model, the play function constraint set and the system constraint, performing second static security analysis on the clearing result, and taking the clearing result as an optimal clearing result when all parameters of the current system meet given requirements.
Another embodiment of the invention also proposes a storage medium on which a computer program is stored which, when being executed by a processor, carries out the steps of the above-mentioned method.
Another embodiment of the present invention also proposes a rendering device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, which when executing the program implements the above method.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
FIG. 1 is a flow chart of a deep learning-based real-time power market restriction self-recognition clearing method according to a first embodiment of the present invention;
FIG. 2 is a detailed flowchart of step S101 in FIG. 1;
FIG. 3 is a detailed flowchart of step S102 in FIG. 1;
FIG. 4 is a flow diagram of functional constraint recognition of a deep neural network;
FIG. 5 is a detailed flowchart of step S103 in FIG. 1;
FIG. 6 is a detailed flow chart of real-time market clearing of electricity;
fig. 7 is a block diagram of a deep learning-based real-time power market restriction self-recognition system according to a second embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. 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.
Referring to fig. 1, a method for self-identifying and clearing electric real-time market constraints based on deep learning according to a first embodiment of the present invention includes steps S101 to S103:
step S101, acquiring power information of each node area in a power system, performing standardization processing on the extracted system characteristic vector of the system, and performing iterative training through an activation model to obtain a deep neural network model, wherein the system characteristic vector comprises an input characteristic vector and an output characteristic vector.
The method is used for guiding reasonable and optimized configuration of energy resources by using electricity prices and promoting high-proportion new energy consumption of the power system, so that a Security-constrained economic dispatch (SCED) model is applied. However, since the SCED model includes N-1 massive security constraints, even though the SCED model is a linear optimization model, it is still difficult to effectively solve the SCED model of the large-scale practical system by using the existing linear solvers such as CPLEX, GUROBI, and the like. It can be understood that although the scale of the constraint in the existing SCED model is huge, the proportion of the real functioning constraint is very small, so that it is difficult to quickly and effectively identify all functioning constraint sets of the SCED, and thus the obtained functioning constraint has low recognition degree and high search times, thereby increasing the system operation cost and reducing the market clearing efficiency. In the embodiment of the invention, in order to solve the problem, a technical guarantee is provided for effectively identifying the functional constraint in the definition model by the deep learning technology from two aspects of feature vector and deep neural network result processing, the nonlinear relation between the system operation condition and the functional constraint is excavated, the functional constraint in the definition model is efficiently identified, the search times of the functional constraint are reduced, the clearing efficiency of the market is improved, and the actual application requirement is met.
In the embodiment, system characteristic extraction is carried out on the electric power information in the electric power system to obtain a system characteristic vector which can effectively reflect new energy, load fluctuation and electricity price fluctuation; by carrying out standardization processing and iterative training on the system characteristic vector, a deep neural network model with complete training is obtained, and necessary conditions are provided for obtaining functional constraints through standard optimal output and actual optimal output of the generator subsequently.
Referring to fig. 2, the method for obtaining the deep neural network model by obtaining the power information of each node area in the power system, performing normalization processing on the extracted system feature vector of the system, and performing iterative training through the activation model includes the following steps:
step S1011, extracting node injection active power, node injection reactive power and power plant quotation secondary curve coefficients of each node area in the power system as input characteristic vectors, and taking the output of the corresponding generator as output characteristic vectors.
As described above, in consideration of frequent changes of new energy, load and power plant quotation in the power system, the nodes in each node area in the power system are selected to inject active power PiNode injection reactive power QiAnd the power plant quoted secondary curve coefficient ai,bi,ciAs input characteristic vector, i.e. X ═ Pi,Qi,ai,bi,ci]. In this embodiment, reactive power Q is injected into the nodeiThe fluctuation of new forms of energy and load can be effectively responded to, and the dimensionality is only the number of system nodes. In addition, because the power generation output can effectively reflect the power price fluctuation situation and the dimension of the power generation output is only linearly related to the number of the system power plants, the power generation output is taken as an output characteristic vector, namely Y is PG. It can be appreciated that by injecting a dimensionally simple node into the reactive power QiOutput of generator and power plant quotation secondary curve coefficient ai,bi,ciAs a system feature vector, the sensitivity, reliability, unity and representativeness of the system feature vector are improved.
Step S1012, respectively normalizing the input feature vector and the output feature vector by a normalization model to obtain a corresponding standard input feature vector and a corresponding standard output feature vector.
As described above, the input feature vector and the output feature vector are normalized by the normalization model, respectively, to obtain normalized feature vectors of the current system. The input feature vector and the output feature vector can be processed by effective abnormal values, so that the reliability of obtaining the standard input feature vector and the standard output feature vector is improved.
Specifically, the expression of the standardized model is as follows:
where V is the system feature vector to be normalized, VmeanAnd vstdMean and variance, respectively, of vector V.
And S1013, inputting the standard input characteristic vector and the standard output characteristic vector into an activation model for iterative training to obtain a deep neural network model.
Specifically, the activation model is:
where x is an input variable.
The deep neural network model is as follows:
where ω includes the weight of each connection of the neural network and the deviation value on each neuron,
a gradient of ω, α a learning rate constant (typically 0.8), β a weight coefficient (typically 0.999), ε an adjustment value (typically 10)-8)。
Specifically, in this embodiment, ω is composed of a weight w of each connection of the neural network and a bias value b on each neuron, and iterative training is performed by inputting the standard input feature vector and the standard output feature vector into the activation model until the update amount of w and b is less than 0.01%.
Step S102, performing inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit, and performing first static security analysis and line transmission capacity constraint processing on the generator optimal output to obtain an action constraint set.
As described above, the denormalization processing is performed on the generator optimal output result predicted by the deep neural network through the denormalization model to obtain the generator optimal output of the current unit; the optimal output of the generator is subjected to first static safety analysis and line transmission capacity constraint processing to obtain an action constraint set, so that the efficient identification of the action constraint is improved, and the search times of the action constraint are reduced.
Referring to fig. 3, the method for obtaining the optimal output of the generator of the current unit, and performing the first static safety analysis and the line transmission capacity constraint processing on the optimal output of the generator to obtain the acting constraint set includes the following steps:
and step S1021, inputting the standard input characteristic vector and the standard output characteristic vector into a deep neural network for calculation so as to obtain the standardized optimal output of the generator.
And step S1022, performing inverse normalization processing on the normalized optimal output of the generator through an inverse normalization model to obtain the optimal output of the generator of the current unit.
And S1023, performing first static safety analysis on the optimal output of the generator, and adding an out-of-limit line into a line transmission capacity constraint to obtain an action constraint set.
As described above, in order to improve the reliability of system market clearing, the optimal output of the generator needs to be standardized, and in this embodiment, the standard input feature vector and the standard output feature vector are input into the deep neural network model again to obtain the standardized optimal output of the generator; then, the standardized optimal output of the generator is input into and output from the anti-standardization model to perform anti-normalization processing on the standardized optimal output of the generator, so that the optimal output of the generator of the current unit is obtained; performing N-1 first static safety analysis on the optimal output of the generator, and adding an out-of-limit line into a line transmission capacity constraint to obtain an action constraint set K(1)、C(1)In which K is(1)Is an initial set of branches, C(1)Is the initial set of incidents. Thereby to obtainThe efficient identification of the action constraint is improved, and the search times of the action constraint are reduced.
The safety analysis is to study the safety condition and safety margin of the network after the operation elements quit operation due to faults according to the principle of N-1 for the network in operation or the network in a certain research state. Specifically, the method for adding the out-of-limit line into the line transmission capacity constraint includes: the limit for a line is 100MW, and if our purge results (generator output) are such that the line flow will reach 120MW, exceeding 20MW, the constraint of "line power < > 100MW" will be added to the purge calculation.
Specifically, the expression of the denormalization model is as follows:
where V' is the normalized system feature vector, VmeanAnd vstdMean and variance of the vector V' respectively.
Referring to fig. 4, as an embodiment, the feature vectors are input: x ═ Pi,Qi,ai,bi,ci]And outputting the feature vector: y ═ PGAs a system characteristic vector, obtaining a standard characteristic vector after data preprocessing through a standardized model, and finally obtaining a well-trained deep neural network through an activated model, thereby realizing the off-line training of the neural network; and then, the normalized input feature vector is subjected to a deep neural network model to predict the optimal output of the generator, then, the inverse normalization processing is carried out through an inverse normalization model to obtain the optimal output of the current generator, and finally, an action constraint set is obtained through N-1 static safety analysis, so that the action constraint pre-identification online application is realized.
And S103, calculating the clearing result of the current system according to the clearing model, the function constraint set and the system constraint, performing second static security analysis on the clearing result, and taking the clearing result as the optimal clearing result when all parameters of the current system meet given requirements.
As described above, in order to ensure the accuracy and optimality of the clearing result, the second static security analysis needs to be performed on the clearing result of the current system calculated according to the clearing model, the functional constraint set and the system constraint, so that each parameter of the current system meets the given requirement.
Referring to fig. 5, the method for calculating the clearing result of the current system according to the clearing model, the functional constraint set and the system constraint, and performing the second static security analysis on the clearing result includes the following steps:
and step S1031, constraining a clearing model for calculating a clearing result of the current system according to the action constraint set, the system balance constraint, the unit output upper and lower limit constraint, the unit climbing constraint and the power grid safety constraint.
And step S1032, calculating the clearing result of the current system through the cleared model after the restriction, and performing second static security analysis on the clearing result to judge whether each parameter of the current system meets the given requirement.
Specifically, the output model is as follows:
Pi,min≤Pi,t≤Pi,max(7)
-Rd≤Pi,t-Pi,t-1≤Ru(8)
wherein, Pi,tAs a uniti force at time t, ai、bi、ciIs the running cost coefficient, L, of the unit id,tFor node d load demand at time t, Pi,minAnd Pi,maxRespectively the upper and lower limits of the unit output, RdAnd RuRespectively the upper and lower climbing speed limit values of the unit,
for the transmission active power of the branch (i, j) during the t-th time period under the c-th anticipated fault,
and
corresponding to the upper and lower power limits of the branch (i, j) under the expected failure of the c-th branch,
the sensitivity coefficient of the injected power for node i to branch (i, j).
It will be appreciated that the model (5) is for the purpose of minimising the resulting unit operating costs; the model (6) is a system balance constraint for clearing the model so as to balance the power supply and demand and ensure the frequency and the voltage quality of a power grid; the model (7) is a unit output upper and lower limit constraint of a clear model so as to ensure that the unit output has the highest economical efficiency; the model (8) is a unit climbing constraint of a clear model so as to constrain the output inertia of the unit; the model (9) is used for clearing the power grid safety constraint in the model, namely the N-1 safety constraint, and indicates that the branch power can be maintained in a given range under the forecast accident set C. In the model, the model is divided into a plurality of models,
for the transmission of active power of the branch (i, j) during the t-th period under the c-th expected failure,
and
corresponding to the upper and lower power limits of the branch (i, j) under the expected failure of the c-th branch,
the sensitivity coefficient of the injected power for node i to branch (i, j).
It should be noted that, in other embodiments of the present invention, when any parameter of the current system does not meet the given requirement, the out-of-limit line that does not meet the given requirement is converted into a newly added constraint, and the clearing result of the current system is recalculated according to the clearing model.
In this embodiment, N-1 static security analysis is performed on the pre-clearing result, that is, after one line is arbitrarily disconnected in all N lines, whether each index of the system meets a given requirement is determined. When any line is disconnected, an out-of-limit line still exists, and the network does not meet the N-1 test. Converting the out-of-limit line into a newly added constraint, supplementing the newly added constraint into a clear model, and solving the clear again; and if the out-of-limit line does not exist, the current clearing result is the optimal clearing result of the electric power real-time market.
Referring to fig. 6, in a most specific embodiment, the identification of the active constraint is implemented according to the system operating condition and the active constraint set, and then whether the current system has a new constraint is determined through the active constraint set, the release model and the static analysis.
According to the electric power real-time market constraint self-identification clearing method based on deep learning, system feature extraction is carried out on electric power information in an electric power system, so that system feature vectors capable of effectively reflecting new energy, load fluctuation and electricity price fluctuation are obtained; the method comprises the steps that through standardization processing and iterative training of system feature vectors, a well-trained deep neural network model is obtained, and necessary conditions are provided for subsequent action constraint through standard optimal output and actual optimal output of a generator; performing inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit; the optimal output of the generator is subjected to first static safety analysis and line transmission capacity constraint processing to obtain an acting constraint set, so that the high-efficiency identification of the acting constraint is improved, and the searching times of the acting constraint are reduced; and calculating the clearing result of the current system according to the clearing model, the action constraint set and the system constraint, and performing second static security analysis on the clearing result so as to take the clearing result as the optimal clearing result when all parameters of the current system meet given requirements. The invention provides technical support for the deep learning technology to effectively identify the functional constraint in the definition model from two aspects of feature vector and deep neural network result processing, excavates the nonlinear relation between the system operation condition and the functional constraint, realizes the efficient identification of the functional constraint in the definition model, reduces the searching times of the functional constraint, improves the definition efficiency of the market, and meets the actual application requirements.
Referring to fig. 7, based on the same inventive concept, a deep learning-based real-time market constraint self-identification system according to a second embodiment of the present invention includes:
the obtaining module 10 is configured to obtain power information of each node area in the power system, perform normalization processing on the extracted system feature vector of the system, and perform iterative training through an activation model to obtain a deep neural network model, where the system feature vector includes an input feature vector and an output feature vector.
Further, the obtaining module 10 includes:
and the extraction unit is used for extracting the node injection active power, the node injection reactive power and the power plant quotation secondary curve coefficient of each node area in the power system as input characteristic vectors and taking the output of the corresponding generator as output characteristic vectors.
The first processing unit is used for respectively carrying out standardization processing on the input characteristic vector and the output characteristic vector through a standardization model so as to obtain a corresponding standard input characteristic vector and a corresponding standard output characteristic vector.
And the training unit is used for inputting the standard input characteristic vector and the standard output characteristic vector into the activation model for iterative training to obtain the deep neural network model.
Further, the deep neural network model is:
where ω includes the weight of each connection of the neural network and the deviation value on each neuron,
the gradient of ω is α is the learning rate constant, β is the weight coefficient, and ε is the adjustment value.
The processing module 20 is configured to perform inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit, and perform first static security analysis and line transmission capacity constraint processing on the generator optimal output to obtain an action constraint set.
Further, the processing module 20 includes:
and the calculating unit is used for inputting the standard input characteristic vector and the standard output characteristic vector into the deep neural network for calculation so as to obtain the standardized optimal output of the generator.
And the second processing unit is used for carrying out inverse normalization processing on the normalized optimal output of the generator through an inverse normalization model so as to obtain the optimal output of the generator of the current unit.
And the analysis unit is used for carrying out first static safety analysis on the optimal output of the generator and adding the out-of-limit line into a line transmission capacity constraint to obtain an action constraint set.
And the clearing module 30 is used for calculating a clearing result of the current system according to the clearing model, the function constraint set and the system constraint, performing second static security analysis on the clearing result, and taking the clearing result as an optimal clearing result when all parameters of the current system meet given requirements. The system constraints comprise system balance constraints, unit output upper and lower limit constraints, unit climbing constraints and power grid safety constraints.
Further, the purge module 30 includes:
the constraint unit is used for constraining a clearing model used for calculating a clearing result of the current system through the acting constraint set, the system balance constraint, the upper and lower limit constraints of the unit output force, the unit climbing constraint and the power grid safety constraint;
and the judging unit is used for calculating the clearing result of the current system through the cleared model after constraint and performing second static security analysis on the clearing result so as to judge whether all parameters of the current system meet given requirements.
Further, the output model is as follows:
Pi,min≤Pi,t≤Pi,max
-Rd≤Pi,t-Pi,t-1≤Ru
wherein, Pi,tThe output of the unit i at time t, ai、bi、ciIs the running cost coefficient, L, of the unit id,tFor node d load demand at time t, Pi,minAnd Pi,maxRespectively output to the unitUpper and lower limits, RdAnd RuRespectively the upper and lower climbing speed limit values of the unit,
for the transmission active power of the branch (i, j) during the t-th time period under the c-th anticipated fault,
and
corresponding to the upper and lower power limits of the branch (i, j) under the expected failure of the c-th branch,
the sensitivity coefficient of the injected power for node i to branch (i, j).
It should be further noted that, the purge module 30 is further configured to, when any parameter of the current system does not meet the given requirement, convert the out-of-limit line that does not meet the given requirement into a newly added constraint, and recalculate the purge result of the current system according to the purge model.
According to the deep learning-based real-time electric power market constraint self-identification clear system provided by the invention, system feature extraction is carried out on electric power information in an electric power system to obtain a system feature vector capable of effectively reflecting new energy, load fluctuation and electricity price fluctuation; the method comprises the steps that through standardization processing and iterative training of system feature vectors, a well-trained deep neural network model is obtained, and necessary conditions are provided for subsequent action constraint through standard optimal output and actual optimal output of a generator; performing inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit; the optimal output of the generator is subjected to first static safety analysis and line transmission capacity constraint processing to obtain an acting constraint set, so that the high-efficiency identification of the acting constraint is improved, and the searching times of the acting constraint are reduced; and calculating the clearing result of the current system according to the clearing model, the action constraint set and the system constraint, and performing second static security analysis on the clearing result so as to take the clearing result as the optimal clearing result when all parameters of the current system meet given requirements. The invention provides technical support for the deep learning technology to effectively identify the functional constraint in the definition model from two aspects of feature vector and deep neural network result processing, excavates the nonlinear relation between the system operation condition and the functional constraint, realizes the efficient identification of the functional constraint in the definition model, reduces the searching times of the functional constraint, improves the definition efficiency of the market, and meets the actual application requirements.
The technical features and technical effects of the deep learning-based real-time power market constraint self-identification clear system provided by the embodiment of the invention are the same as those of the method provided by the embodiment of the invention, and are not repeated herein.
Furthermore, an embodiment of the present invention also proposes a storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the above-mentioned method.
Furthermore, an embodiment of the present invention further provides a rendering device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method when executing the program.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Further, the computer readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, a dedicated integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the exemplary embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (10)
1. A deep learning-based electric power real-time market constraint self-identification clearing method is characterized by comprising the following steps:
acquiring power information of each node area in a power system, standardizing the extracted system characteristic vector of the system, and performing iterative training through an activation model to obtain a deep neural network model, wherein the system characteristic vector comprises an input characteristic vector and an output characteristic vector;
performing inverse normalization processing on the generator optimal output result predicted by the deep neural network through an inverse normalization model to obtain the generator optimal output of the current unit, and performing first static security analysis and line transmission capacity constraint processing on the generator optimal output to obtain an action constraint set;
and calculating the clearing result of the current system according to the clearing model, the function constraint set and the system constraint, performing second static security analysis on the clearing result, and taking the clearing result as the optimal clearing result when all parameters of the current system meet given requirements.
2. The deep learning-based power real-time market constraint self-identification clearing method according to claim 1, wherein after the step of performing the second static security analysis on the clearing result, the method further comprises:
and when any parameter of the current system does not meet the given requirement, converting the out-of-limit line which does not meet the given requirement into a newly added constraint, and recalculating the clearing result of the current system according to the clearing model.
3. The deep learning-based real-time power market constraint self-identification method according to claim 1, wherein the method for obtaining the deep neural network model by obtaining power information of each node area in the power system, normalizing the extracted system feature vectors of the system, and performing iterative training through an activation model comprises the following steps:
extracting node injection active power, node injection reactive power and power plant quotation secondary curve coefficients of each node area in the power system as input characteristic vectors, and taking the output of a corresponding generator as output characteristic vectors;
respectively carrying out standardization processing on the input feature vector and the output feature vector through a standardization model to obtain a corresponding standard input feature vector and a corresponding standard output feature vector;
inputting the standard input characteristic vector and the standard output characteristic vector into the activation model for iterative training to obtain a deep neural network model.
4. The deep learning-based power real-time market constraint self-identification method according to claim 3, wherein the deep neural network model is as follows:
wherein ω includes a weight value for each connection of the neural network and a deviation value on each neuron,
the gradient of ω is α is the learning rate constant, β is the weight coefficient, and ε is the adjustment value.
5. The deep learning-based real-time market constraint self-identification method for electric power as claimed in claim 3, wherein the method for obtaining the optimal output of the generator of the current unit and performing the first static safety analysis and the line transmission capacity constraint processing on the optimal output of the generator to obtain the action constraint set comprises:
inputting the standard input characteristic vector and the standard output characteristic vector into a deep neural network for calculation so as to obtain the standardized optimal output of the generator;
performing inverse normalization processing on the normalized optimal output of the generator through an inverse normalization model to obtain the optimal output of the generator of the current unit;
and carrying out first static safety analysis on the optimal output of the generator, and adding the out-of-limit line into a line transmission capacity constraint to obtain an action constraint set.
6. The deep learning-based power real-time market constraint self-identification clearing method according to claim 1, wherein the system constraints comprise system balance constraints, unit output upper and lower limit constraints, unit climbing constraints and grid safety constraints, the method for calculating the clearing result of the current system according to the clearing model, the action constraint set and the system constraints and performing a second static safety analysis on the clearing result comprises:
constraining a clearing model for calculating a clearing result of the current system according to the acting constraint set, the system balance constraint, the unit output upper and lower limit constraint, the unit climbing constraint and the power grid safety constraint;
and calculating the clearing result of the current system through the clearing model after constraint, and performing second static security analysis on the clearing result to judge whether each parameter of the current system meets the given requirement.
7. The deep learning-based power real-time market constraint self-identification clearing method according to claim 6, wherein the clearing model is as follows:
Pi,min≤Pi,t≤Pi,max
-Rd≤Pi,t-Pi,t-1≤Ru
wherein, Pi,tThe output of the unit i at time t, ai、bi、ciIs the running cost coefficient, L, of the unit id,tFor node d load demand at time t, Pi,minAnd Pi,maxRespectively the upper and lower limits of the unit output, RdAnd RuRespectively the upper and lower climbing speed limit values of the unit,
for the transmission active power of the branch (i, j) during the t-th time period under the c-th expected failure,
and
corresponding to the upper and lower power limits of the branch (i, j) under the expected failure of the c-th branch,
the sensitivity coefficient of the injected power for node i to branch (i, j).
8. A deep learning based real-time market-constrained self-identification system for power, the system comprising:
the acquisition module is used for acquiring power information of each node area in a power system, carrying out standardization processing on the extracted system characteristic vector of the system, and then carrying out iterative training through an activation model to obtain a deep neural network model, wherein the system characteristic vector comprises an input characteristic vector and an output characteristic vector;
the processing module is used for carrying out inverse normalization processing on the optimal output result of the generator predicted by the deep neural network through an inverse normalization model so as to obtain the optimal output of the generator of the current unit, and carrying out first static safety analysis and line transmission capacity constraint processing on the optimal output of the generator so as to obtain an action constraint set;
and the clearing module is used for calculating a clearing result of the current system according to the clearing model, the function constraint set and the system constraint, performing second static security analysis on the clearing result, and taking the clearing result as an optimal clearing result when all parameters of the current system meet given requirements.
9. A storage medium on which a computer program is stored, which program, when being executed by a processor, carries out the method according to any one of claims 1 to 7.
10. A rendering device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 1 to 7 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911194707.XA CN110991731A (en) | 2019-11-28 | 2019-11-28 | Electric power real-time market constraint self-identification clearing method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911194707.XA CN110991731A (en) | 2019-11-28 | 2019-11-28 | Electric power real-time market constraint self-identification clearing method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110991731A true CN110991731A (en) | 2020-04-10 |
Family
ID=70087891
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911194707.XA Pending CN110991731A (en) | 2019-11-28 | 2019-11-28 | Electric power real-time market constraint self-identification clearing method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110991731A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108921601A (en) * | 2018-06-20 | 2018-11-30 | 浙江大学 | A kind of power spot market auxiliary method of commerce neural network based |
| CN109546646A (en) * | 2018-11-19 | 2019-03-29 | 南方电网科学研究院有限责任公司 | A kind of region power spot market distributing goes out clearing method, device, equipment and medium |
| CN109784692A (en) * | 2018-12-29 | 2019-05-21 | 重庆大学 | A kind of fast and safely constraint economic load dispatching method based on deep learning |
-
2019
- 2019-11-28 CN CN201911194707.XA patent/CN110991731A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108921601A (en) * | 2018-06-20 | 2018-11-30 | 浙江大学 | A kind of power spot market auxiliary method of commerce neural network based |
| CN109546646A (en) * | 2018-11-19 | 2019-03-29 | 南方电网科学研究院有限责任公司 | A kind of region power spot market distributing goes out clearing method, device, equipment and medium |
| CN109784692A (en) * | 2018-12-29 | 2019-05-21 | 重庆大学 | A kind of fast and safely constraint economic load dispatching method based on deep learning |
Non-Patent Citations (4)
| Title |
|---|
| RAMTIN MADANI ET AL.: "Constraint Screening for Security Analysis of Power Networks", 《IEEE TRANSACTIONS ON POWER SYSTEMS》 * |
| 丁涛 等: "考虑不确定功率注入的安全约束经济调度模型非有效约束识别方法", 《中国电机工程学报》 * |
| 张昆 等: "南方区域电力现货市场技术支持系统：出清功能部分", 《南方电网技术》 * |
| 李洪美 等: "考虑风电接入不确定性的SCED模型和区间优化算法", 《电器与能效管理技术》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yang et al. | Intelligent data-driven decision-making method for dynamic multisequence: An E-seq2seq-based SCUC expert system | |
| Wang et al. | An approach for integrated generation and transmission maintenance scheduling considering N-1 contingencies | |
| CN112085285B (en) | Bus load prediction method, device, computer equipment and storage medium | |
| CN114021425B (en) | Power system operation data modeling and feature selection method and device, electronic equipment and storage medium | |
| Wang et al. | Analysis of network loss energy measurement based on machine learning | |
| CN110086166B (en) | Method and system for representing stable operation quota of power grid | |
| CN110991731A (en) | Electric power real-time market constraint self-identification clearing method and system based on deep learning | |
| CN115660893A (en) | Transformer substation bus load prediction method based on load characteristics | |
| CN114759579A (en) | Power grid active power optimization control system, method and medium based on data driving | |
| Liu et al. | Deep reinforcement learning for real-time economic energy management of microgrid system considering uncertainties | |
| CN113808071A (en) | Non-invasive load monitoring method and system based on deep learning | |
| Kong et al. | Running state prediction and evaluation of power transformers | |
| CN105389641A (en) | Short-term data fusion-based power grid project optimum decision method | |
| Li et al. | Power system data-driven dispatch using improved scenario generation considering time-series correlations | |
| CN113139271A (en) | Aggregation resource modeling method and device, electronic device and storage medium | |
| Qin et al. | Risk assessment technology on power grid-communication networks considering spatiotemporal correlation | |
| CN113434667B (en) | Text classification method based on distribution network automation terminal text classification model | |
| CN116340765B (en) | Electricity larceny user prediction method and device, storage medium and electronic equipment | |
| CN117993868B (en) | Dual-mode attention-based power grid engineering project audit early warning method and system | |
| Abusida et al. | A New Approach to Load Shedding Prediction in GECOL Using Deep Learning Neural Network | |
| Huang et al. | A simplification method for robust optimization of power system based on LMP | |
| CN117833198A (en) | Multi-level load prediction method of novel power distribution system | |
| Ghaljehei et al. | Effect of optimal generation scheduling of compressed air energy storage and wind power generation on economic and technical issues | |
| CN117977556A (en) | Power demand prediction and energy storage optimization method, device, medium and equipment | |
| Cong et al. | A Fast Calculation Method of SCUC Based on Deep Learning |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200410 |
|
| RJ01 | Rejection of invention patent application after publication |