CN111724277A - New energy and multi-element load value matching method and system - Google Patents
New energy and multi-element load value matching method and system Download PDFInfo
- Publication number
- CN111724277A CN111724277A CN202010531489.0A CN202010531489A CN111724277A CN 111724277 A CN111724277 A CN 111724277A CN 202010531489 A CN202010531489 A CN 202010531489A CN 111724277 A CN111724277 A CN 111724277A
- Authority
- CN
- China
- Prior art keywords
- load
- data
- new energy
- region
- preset time
- 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 32
- 238000012549 training Methods 0.000 claims abstract description 56
- 238000010248 power generation Methods 0.000 claims abstract description 40
- 238000009826 distribution Methods 0.000 claims abstract description 27
- 238000007781 pre-processing Methods 0.000 claims abstract description 19
- 238000013136 deep learning model Methods 0.000 claims abstract description 12
- 150000001875 compounds Chemical class 0.000 claims description 13
- 239000013598 vector Substances 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 7
- 230000029087 digestion Effects 0.000 claims description 6
- 230000009467 reduction Effects 0.000 abstract description 9
- 238000004458 analytical method Methods 0.000 abstract description 5
- 238000004134 energy conservation Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 abstract description 3
- 230000002354 daily effect Effects 0.000 description 13
- 230000006870 function Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 11
- 238000004590 computer program Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- DDNCQMVWWZOMLN-IRLDBZIGSA-N Vinpocetine Chemical compound C1=CC=C2C(CCN3CCC4)=C5[C@@H]3[C@]4(CC)C=C(C(=O)OCC)N5C2=C1 DDNCQMVWWZOMLN-IRLDBZIGSA-N 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 5
- 229960000744 vinpocetine Drugs 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 238000005265 energy consumption Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000007619 statistical method Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 241000218176 Corydalis Species 0.000 description 2
- 235000008331 Pinus X rigitaeda Nutrition 0.000 description 2
- 235000011613 Pinus brutia Nutrition 0.000 description 2
- 241000018646 Pinus brutia Species 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 230000010365 information processing Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000002329 infrared spectrum Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012800 visualization Methods 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
- 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
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a new energy and multi-element load value matching method and system, which are characterized in that the new energy power generation power data and load data in a first area in a preset time period are obtained, and the new energy power generation power data and load data in a plurality of second areas in the preset time period are obtained for preprocessing; generating a generation scene about new energy and load through a variational self-encoder to obtain scene generation data, and performing model training by using a deep learning model to obtain training generation new energy power data and load data in each region; and calculating the matching degree of the new energy and the load, and selecting a near-consumption capacity scheme of the first region with the maximum matching degree and the corresponding second region. The method has good source load probability distribution rule analysis, can improve the nearby consumption capacity of new energy, achieves the effects of energy conservation and emission reduction, and simultaneously improves the utilization rate of the new energy.
Description
Technical Field
The invention relates to the technical field of new energy power generation and power systems, in particular to a new energy and multi-load value matching method and system.
Background
With the rapid development of new energy industry in China, wind power and photovoltaic installation machines become the first in the world. In order to actively promote the construction of the electric power market, the trading mechanism and the trading strategy of the electric power market are gradually improved. But the energy problem is increasingly prominent, and the trend of rapidly developing new energy is more and more rapid along with the increasingly reduced energy supply and the continuously improved energy conservation and emission reduction constraints. In the face of the situation that the power abandonment rate of new energy in some areas is continuously increased, in order to improve the situation, the potential of new energy consumption is excavated in the power system, and the new energy needs to be developed to participate in market trading, so that multi-dimensional collaborative consumption of the new energy is realized. By reasonably standardizing the electric power market transaction, deepening the optimal configuration of electric power resources, actively promoting the electric power market construction, gradually perfecting the electric power market transaction mechanism and the transaction strategy, and further improving the new energy consumption capability.
In the face of uncertainty of load data, characteristics of the load data need to be dug urgently, and how to effectively perform probability modeling is not slow. Meanwhile, in the face of the current situation of insufficient new energy consumption, how to efficiently, accurately and automatically realize the consumption of new energy and realize the value matching of the new energy and the multiple loads becomes a very urgent task, and under the promotion of the strong demand, the related technologies and products are necessarily perfected day by day to realize rapid growth, thereby having great significance for energy conservation and emission reduction of the whole country and region.
In the prior art, patent of invention CN110571862A patent of an analysis method and a system for time sequence matching degree of a photovoltaic power station and a power load has two defects: (1) the system only collects historical load data and simple calculation data, and has the defects that if the data are abnormal, the data cannot be collected and calculated, the characteristics of the data cannot be learned, the characteristics of the data cannot be deeply mined only by achieving simple collection and calculation, and the probability distribution rule among the data cannot be analyzed. (2) The system is finished only by calculating the matching degree of each time period, and the next application description is not carried out, so that the system is difficult to apply to actual application.
Therefore, in order to solve the actual situations of the new energy field outbound force and the multiple loads, a value matching system or method of the new energy field outbound force and the multiple loads is urgently needed to be designed.
Disclosure of Invention
In view of the above problems, the present invention provides a new energy and multi-load value matching system and method that at least solve some of the above technical problems, and can realize the selection of an optimal new energy and load nearby consumption scheme, thereby achieving the purpose of energy saving and emission reduction.
In a first aspect, an embodiment of the present invention provides a method for matching new energy with a multivariate load value, including the following steps:
s10, acquiring new energy power generation data and load data in a first area in a preset time period, acquiring multiple second area new energy power generation data and load data in the preset time period, and preprocessing; the second area is an area with the potential of absorbing the new energy of the first area;
s20, generating a scene about new energy and load through a variational self-encoder by the preprocessed new energy power generation power data and load data in the first area and the new energy power generation power data and load data in the second areas to obtain scene generation data, and performing model training by using a deep learning model to obtain new energy power generation power data and load data generated by training in the first area and new energy power generation power data and load data generated by training in the second areas;
and S30, generating new energy output data and load data according to the training of each region, calculating the matching degree of the new energy and the load, and selecting a near digestion capacity scheme between the two regions with the maximum matching degree.
In one embodiment, the step S20 includes:
1) and (3) data encoding: reducing the dimensions of the new energy output data without labels and the load data into a group of characteristic vectors through an encoder, and then performing a fitting process of the model;
2) and (3) data decoding: restoring a group of feature vectors in the encoder into input data through a decoder;
3) gradient back propagation: adjusting the network weight through gradient inverse descent, and learning the distribution rule of historical data through a decoder after repeated training;
4) and (3) data generation: and generating a probability model through the VAE, generating a plurality of groups of new energy and load scene sets by an output layer, and obtaining new energy output data and load data generated by training of each region.
In one embodiment, the step S30 includes:
1) load data generated by training of each region are subtracted from new energy output data of the region respectively to obtain generalized average load of each region in a preset time period, and the formula is as follows:
(1) in the formula, PL(t) represents training generation load data in a certain area of the respective areas; pW(t) representing new energy output data in a certain area in each area; t represents the number of period time segments;
2) and superposing the generalized average load of all the preset time periods of the regions:
(2) in the formula, N represents the number of the overlapping areas; pLj(t) represents the preset time period average load of the jth superimposed region;
3) respectively calculating the maximum load P of each region in a preset time periodmaxMinimum load PminPeak to valley difference PvPeak to valley ratio RvAnd a fluctuation rate f; the formula is as follows:
Pv＝Pmax-Pmin(3)
P′v＝P′max-P′min(4)
(4) of formula (II) to (III)'vA peak-to-valley difference representing the generalized average load for each region; p'maxRepresents the maximum load among the generalized average loads of the respective regions; p'minRepresents the minimum load among the generalized average loads of the respective regions;
Rv＝Pv/Pmax(5)
R′v＝P′v/P′max(6)
(6) in the formula (II), R'vA peak-to-valley difference rate representing the generalized average load for each region;
(7) in the formula (I), the compound is shown in the specification,
the average value of the loads representing the preset time period of each area; s represents a standard value of the average load of each area in a preset time period;
4) calculating the slope association degree gamma of the curve of the average load and the generalized average load of each region in a preset time period, and calculating the load matching degree rho of new energy among the regions;
(8) in the formula (I), the compound is shown in the specification,
the expression is the generalized average load when the period is t +1 minus the generalized average load difference when the period is t; delta PL(t) represents the average load at period t +1 minus the average load difference at period t;
and
the generalized average load and the variance of the average load, respectively;
(9) in the formula, ρ represents a new energy load matching degree between the first region and the corresponding second region.
In a second aspect, an embodiment of the present invention further provides a system for matching new energy with multiple load values, including:
the system comprises an acquisition preprocessing module, a load preprocessing module and a control module, wherein the acquisition preprocessing module is used for acquiring new energy power generation power data and load data in a first region in a preset time period, acquiring a plurality of second region new energy power generation power data and load data in the preset time period and preprocessing the multiple second region new energy power generation power data and load data; the second area is an area with the potential of absorbing the new energy of the first area;
the scene generation module is used for generating a generation scene about new energy and load through a variational self-coder for the preprocessed new energy power generation power data and load data in the first area and the new energy power generation power data and load data in the second areas to obtain scene generation data, and then performing model training by using a deep learning model to obtain training generation new energy power data and load data in the first area and training generation new energy power generation power data and load data in the second areas;
and the source-load matching module is used for generating new energy output data and load data according to the training of each region, calculating the matching degree of the new energy and the load, and selecting a near digestion capacity scheme between two regions with the maximum matching degree.
In one embodiment, the scene generation module includes:
a data encoding unit: reducing the dimensions of the new energy output data without labels and the load data into a group of characteristic vectors through an encoder, and then performing a fitting process of the model;
a data decoding unit: restoring a group of feature vectors in the encoder into input data through a decoder;
gradient counter-propagating unit: adjusting the network weight through gradient inverse descent, and learning the distribution rule of historical data through a decoder after repeated training;
a data generation unit: and generating a probability model through the VAE, generating a plurality of groups of new energy and load scene sets by an output layer, and obtaining new energy output data and load data generated by training of each region.
In one embodiment, the source-to-load matching module is specifically configured to:
1) load data generated by training of each region are subtracted from new energy output data of the region respectively to obtain generalized average load of each region in a preset time period, and the formula is as follows:
(1) in the formula (I), the compound is shown in the specification,PL(t) represents training generation load data in a certain area of the respective areas; pW(t) representing new energy output data in a certain area in each area; t represents the number of period time segments;
2) and superposing the generalized average load of all the preset time periods of the regions:
(2) in the formula, N represents the number of the overlapping areas; pLj(t) represents the preset time period average load of the jth superimposed region;
3) respectively calculating the maximum load P of each region in a preset time periodmaxMinimum load PminPeak to valley difference PvPeak to valley ratio RvAnd a fluctuation rate f; the formula is as follows:
Pv＝Pmax-Pmin(3)
P′v＝P′max-P′min(4)
(4) of formula (II) to (III)'vA peak-to-valley difference representing the generalized average load for each region; p'maxRepresents the maximum load among the generalized average loads of the respective regions; p'minRepresents the minimum load among the generalized average loads of the respective regions;
Rv＝Pv/Pmax(5)
R′v＝P′v/P′max(6)
(6) in the formula (II), R'vA peak-to-valley difference rate representing the generalized average load for each region;
(7) in the formula (I), the compound is shown in the specification,
the average value of the loads representing the preset time period of each area; s represents a standard value of the average load of each area in a preset time period;
4) calculating the slope association degree gamma of the curve of the average load and the generalized average load of each region in a preset time period, and calculating the load matching degree rho of new energy among the regions;
(8) in the formula (I), the compound is shown in the specification,
the expression is the generalized average load when the period is t +1 minus the generalized average load difference when the period is t; delta PL(t) represents the average load at period t +1 minus the average load difference at period t;
and
the generalized average load and the variance of the average load, respectively;
(9) in the formula, ρ represents a new energy load matching degree between the first region and the corresponding second region.
The technical scheme provided by the embodiment of the invention has the beneficial effects that at least:
the embodiment of the invention provides a new energy and multi-element load value matching method and a system, wherein the method comprises the following steps:
1) the method is characterized in that a probability model is constructed based on a variational self-encoder, multiple groups of scene data are generated through the model aiming at the uncertainty and volatility conditions of new energy and load data, and the characteristic expression and probability distribution conditions of the data are analyzed.
2) Different from the mode of data processing in the closest technical scheme, the method also constructs a probability generation model based on the VAE, and can mine the characteristics of historical load data and discover the probability distribution rule among data. Because the invention combines the deep learning model, the constructed probability model is not only simple, but also does not need to carry out scene reduction, thereby greatly improving the efficiency of the probability model.
3) Compared with the closest technical scheme that only data collection and calculation of the matching degree of the new energy and the load are carried out, the matching degree of the new energy and the load respectively calculates various indexes of the load, analyzes the indexes, and recommends the optimal nearby consumption scheme of the new energy and the load, so that the effects of energy conservation and emission reduction are achieved, and meanwhile, the utilization rate of the new energy is improved.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
fig. 1 is a flowchart of a new energy and multi-load value matching method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a specific implementation of step S20 according to an embodiment of the present invention;
fig. 3 is a block diagram of a new energy and multi-load value matching system according to an embodiment of the present invention.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
In order to express the technical scheme of the invention more clearly, some related terms are explained first:
1) a neural network: is a mathematical model for information processing using structures similar to brain neurosynaptic connections. It is also often directly referred to in engineering and academia as simply "neural networks" or neural-like networks.
2) An auto-encoder: the artificial neural network is used in semi-supervised learning and unsupervised learning, and has the function of performing characterization learning on input information by taking the input information as a learning target.
3) Variational self-encoder: the deep learning generation model proposed by Max Welling 2013 can be applied to data dimension reduction, noise reduction and the like. The training is carried out by a random gradient descent method through construction by using a neural network.
4) Probability distribution: the probability law is used for expressing the value of a random variable. The probability of an event indicates the degree of likelihood that a certain result will occur in a trial. To fully understand a trial, it is necessary to know all possible outcomes of the trial and the probability of each possible outcome occurring, i.e., the probability distribution of the random trial.
5) Normal distribution: the term "normal distribution", also known as Gaussian distribution, was originally derived from a. zenchier in an asymptotic formula for the binomial distribution. C.f. gaussian derives it from another angle when studying the measurement error. P.s. laplace and gaussian investigated its properties. Is a very important probability distribution in the fields of mathematics, physics, engineering and the like.
6) Matching degree: generally refers to a datametric measure of the similarity of the infrared spectrum of a substance compared to a standard spectrum of the pure substance. Which refers to the degree of matching of new energy with load.
7) Slope degree of correlation: the change trend of the daily average load curve can be described by the change of the slope of the daily average load curve, and the slope correlation degree is a correlation analysis method defined by taking the slope as a criterion.
8) Maximum load: maximum value of load data during a day.
9) Minimum load: minimum value of load data in one day.
10) Load factor: the ratio of the maximum value to the minimum value of the load data during a day.
11) Peak-to-valley difference: the difference between the maximum and minimum of the load data during a day.
12) Peak-to-valley difference rate: peak to valley difference to maximum ratio of load data over the day.
13) Fluctuation ratio: mean to standard deviation ratio of load data during the day.
14) And (3) a back propagation algorithm: the method is a learning algorithm of a multilayer neuron network and is established on the basis of a gradient descent method. The input-output relationship of the BP network is substantially a mapping relationship: an n-input m-output BP neural network performs the function of continuous mapping from n-dimensional euclidean space to a finite field in m-dimensional euclidean space, which is highly non-linear. Its information processing ability comes from multiple composition of simple non-linear function, so it has strong function reproduction ability.
15) The forward propagation algorithm: that is, the output of the previous layer is taken as the input of the next layer, and then the output of the next layer is calculated, and so on, that is, the forward propagation algorithm of DNN.
16) And (3) variable deduction: there are two parts of input: the data x, the model p (z, x), requires the inference of the posterior probability p (z | x), but cannot be directly solved by constructing an approximate distribution q (z; v) of the posterior probability p (z | x) and then continuously narrowing the distance between q and p until convergence.
17) Loss function: is a function that maps the value of a random event or its associated random variable to a non-negative real number to represent the "risk" or "loss" of the random event. In application, the loss function is usually associated with the optimization problem as a learning criterion, i.e. the model is solved and evaluated by minimizing the loss function. For example, in statistics and machine learning, for parameter estimation of models.
Referring to fig. 1, a method for matching a new energy with a multi-element load value according to an embodiment of the present invention includes:
s10, acquiring new energy power generation data and load data in a first area in a preset time period, acquiring multiple second area new energy power generation data and load data in the preset time period, and preprocessing; the second area is an area with the potential of absorbing the new energy of the first area;
s20, generating a scene about new energy and load through a variational self-encoder by the preprocessed new energy power generation power data and load data in the first area and the new energy power generation power data and load data in the second areas to obtain scene generation data, and performing model training by using a deep learning model to obtain new energy power generation power data and load data generated by training in the first area and new energy power generation power data and load data generated by training in the second areas;
and S30, generating new energy output data and load data according to the training of each region, calculating the matching degree of the new energy and the load, and selecting a near digestion capacity scheme between the two regions with the maximum matching degree.
In step S10, for example, the original data of the new energy (wind power, solar power) and the load added respectively are obtained. The preset time period may be, for example, days, weeks, months, etc. When the preset time period is day, the acquired data comprises energy output data and load data of each time period in the day;
and respectively storing the obtained data into a database, wherein abnormal data are found from the original data through data analysis (for example, a visualization tool can be used), then the abnormal data are cleaned through Python, and statistical analysis is performed again (for example, the maximum and minimum values, the median, the mode, the mean, the variance and the like of Excel and a scatter diagram) so as to obtain a new energy station output data statistical list and a load data statistical list.
For example, in jilin province, the city of vinpocetine is a wind power generation area (first area), and the peripheral areas are, for example: white city, Yanbian, pine, Jilin, Baishan, Tonghua 6 cities (multiple second areas, also can have new energy to generate electricity); by the method, an optimal nearby digestion scheme can be selected from six schemes of Changchun and adjacent areas. The six schemes are as follows: vinblastine-white city (scheme one), vinblastine-rhizoma corydalis (scheme two), vinblastine-pinogen (scheme three), vinblastine-Jilin (scheme four), vinblastine-white mountain (scheme five) and vinblastine-Tonghua (scheme six).
The steps S10-S30 well complete the statistical analysis of the source load data, and can be applied to the power system; the method has the advantages that the source load probability distribution rule analysis is good, the optimal nearby consumption scheme is selected, and the nearby consumption capability of the new energy can be improved.
The above steps are described in detail below:
referring to fig. 2, the specific implementation steps of step S20 are as follows:
1) data encoding, namely, reducing the dimension of label-free data (such as historical power consumption data and load data) into a group of eigenvectors through an encoder, and then performing a model fitting process;
2) data decoding, namely restoring a group of feature vectors in the encoder into input data through a decoder;
3) the gradient backward propagation is realized, the network weight is adjusted through the gradient backward descent, and after repeated training, the distribution rule of historical data can be learned through a decoder;
4) generating data, namely generating a probability model through VAE, generating a plurality of groups of new energy and load scene sets through an output layer, and well reflecting the probability distribution condition of the new energy and the load according to the data of the new energy and load scene sets;
5) the network structures in the encoder and the decoder in the variational self-encoder (VAE) are designed according to the characteristics of the new energy output data and the load data. Four convolutional layers are designed on the encoder, and repeated experiments prove that the effect of fitting the probability distribution of the actually measured load data is best when the dimension of the scene hidden vector generated by the new energy is 6. The input in the decoder is then also set to 6, which contains two fully connected layers. In order to correspond to the encoders one by one, the embodiment of the invention also designs four deconvolution layers, and the parameters of the deconvolution layers and the encoders are the same;
6) and generating data obtained by training the variational self-coder model to be used as data for carrying out value matching calculation between source loads.
In the embodiment, a probability model is constructed based on a variational self-encoder, multiple groups of scene data are generated through the probability model according to the uncertainty and volatility conditions of new energy and load data, and the characteristic expression and probability distribution conditions of the data are analyzed. In addition, a probability generation model based on VAE is constructed, the characteristics of historical load data can be mined through the model, and the probability distribution rule among the data can be found. Because the invention combines the deep learning model, the constructed probability model is not only simple, but also does not need to carry out scene reduction, thereby greatly improving the efficiency of the probability model.
In one embodiment, the step S30 implements the specific steps of:
1) and (3) generating load data generated by training each region by using the generated data obtained by deep learning model training in scene generation, and respectively subtracting the new energy output data of the region to obtain the generalized average load of each region in a preset time period.
The formula is as follows:
(1) in the formula, PL(t) represents training generation load data in a certain area of the respective areas; pW(t) representing new energy output data in a certain area in each area; t represents the number of period time segments;
2) and superposing the generalized average load of all the preset time periods of the regions:
for example, the generalized daily average load is obtained by superposing the wind power nearby region, that is, the load of each time period in the vinpocetine region is superposed on the time period load of the nearby region, so that a new generalized daily average load formula is obtained as follows:
(2) in the formula, N represents the number of the overlapping areas; pLj(t) represents the preset time period average load of the jth superimposed region;
3) respectively calculating the maximum load P of each region in a preset time periodmaxMinimum load PminPeak to valley difference PvPeak to valley ratio RvAnd a fluctuation rate f; the formula is as follows:
Pv＝Pmax-Pmin(3)
P′v＝P′max-P′min(4)
(4) of formula (II) to (III)'vA peak-to-valley difference representing the generalized average load for each region; p'maxRepresents the maximum load among the generalized average loads of the respective regions; p'minRepresents the minimum load among the generalized average loads of the respective regions; (ii) a
Rv＝Pv/Pmax(5)
R′v＝P′v/P′max(6)
(6) In the formula (II), R'vA peak-to-valley difference rate representing the generalized average load for each region;
(7) in the formula (I), the compound is shown in the specification,
the average value of the loads representing the preset time period of each area; s represents a standard value of the average load of each area in a preset time period;
4) calculating the slope association degree gamma of the curve of the average load and the generalized average load of each region in a preset time period, and calculating the load matching degree rho of new energy among the regions;
(8) in the formula (I), the compound is shown in the specification,
the expression is the generalized average load when the period is t +1 minus the generalized average load difference when the period is t; delta PL(t) represents the average load at period t +1 minus the average load difference at period t;
and
the generalized average load and the variance of the average load, respectively;
(9) in the formula, ρ represents a new energy load matching degree between the first region and the corresponding second region.
5) And calculating the matching degree of the new energy and the load of each scheme, comparing the sizes of the new energy and the load, and selecting the scheme with the maximum matching degree, so that the optimal nearby consumption scheme can be recommended.
The technical solution of the invention is illustrated by a simple embodiment:
for example, in jilin province, the city of vinpocetine is a wind power generation area (first area), and the peripheral areas are, for example: white city, Yanbian, pine plain, Jilin, Baishan, leading to 6 cities (a plurality of second regions also have the function of new energy power generation, but are not enough to be self-sufficient); then there are six alternative schemes, respectively: vinblastine-white city (scheme one), vinblastine-rhizoma corydalis (scheme two), vinblastine-pinogen (scheme three), vinblastine-Jilin (scheme four), vinblastine-white mountain (scheme five) and vinblastine-Tonghua (scheme six).
The first step is as follows: acquiring wind power daily generation power data and daily load data of the Changchun city; acquiring daily generated power data and daily load data of 6 cities, wherein the data are daily average data; then preprocessing is performed, such as data cleansing, deleting partially invalid data, and so forth.
The second step is that: generating a generation scene about new energy and load by the daily generated power data and daily load data of the Changchun city and 6 cities obtained in the first step through a variational self-coder to obtain scene generation data, and performing model training by using a deep learning model to obtain the daily generated power data and daily load data generated by training each city;
the third step: 1) load data generated by training of each region are subtracted from new energy output data of the region respectively to obtain generalized average load of each region in a preset time period;
2) superimposing the above generalized average load over all days
3) The maximum load, the minimum load, the peak-valley difference rate and the fluctuation rate of the Changchun city and 6 cities in one day are respectively calculated.
4) Calculating the slope degree of association of the average load and the generalized average load of the vinpocetine and the 6 cities every day, and respectively calculating the load matching degree of the new energy resources of the vinpocetine and the 6 cities according to the slope degree of association of the curves;
5) and (3) according to the matching degree of the wind power and the load of the 6 schemes, comparing the matching degrees, and selecting the scheme with the maximum matching degree, so that the optimal nearby consumption scheme can be recommended.
Based on the same inventive concept, the embodiment of the invention also provides a new energy and multi-element load value matching system, and as the principle of the problem solved by the system is similar to a new energy and multi-element load value matching method, the implementation of the system can refer to the implementation of the method, and repeated parts are not repeated.
In a second aspect, an embodiment of the present invention further provides a system for matching new energy with multiple load values, as shown in fig. 3, including:
the acquiring and preprocessing module 31 is configured to acquire new energy power generation data and load data in a first area in a preset time period, acquire multiple new energy power generation data and load data in a second area in the preset time period, and perform preprocessing; the second area is an area with the potential of absorbing the new energy of the first area;
the scene generation module 32 is configured to generate a generation scene about new energy and loads through a variational self-encoder for the preprocessed new energy power generation power data and load data in the first region and the new energy power generation power data and load data in the plurality of second regions to obtain scene generation data, and then perform model training by using a deep learning model to obtain training generation new energy power data and load data in the first region and training generation new energy power data and load data in the plurality of second regions;
and the source-load matching module 33 is configured to generate new energy output data and load data according to training of each region, calculate a new energy and load matching degree, and select a near-tolerance scheme between two regions with the largest matching degree.
In one embodiment, the scene generation module 32 includes:
data encoding section 321: reducing the dimensions of the new energy output data without labels and the load data into a group of characteristic vectors through an encoder, and then performing a fitting process of the model;
the data decoding unit 322: restoring a group of feature vectors in the encoder into input data through a decoder;
gradient backpropagation unit 323: adjusting the network weight through gradient inverse descent, and learning the distribution rule of historical data through a decoder after repeated training;
the data generation unit 324: and generating a probability model through the VAE, generating a plurality of groups of new energy and load scene sets by an output layer, and obtaining new energy output data and load data generated by training of each region.
In an embodiment, the source-load matching module 33 is specifically configured to:
1) load data generated by training of each region are subtracted from new energy output data of the region respectively to obtain generalized average load of each region in a preset time period, and the formula is as follows:
(1) in the formula, PL(t) represents training generation load data in a certain area of the respective areas; pW(t) representing new energy output data in a certain area in each area; t represents the number of period time segments;
2) and superposing the generalized average load of all the preset time periods of the regions:
(2) in the formula, N represents the number of the overlapping areas; pLj(t) represents the preset time period average load of the jth superimposed region;
3) respectively calculating the maximum load P of each region in a preset time periodmaxMinimum load PminPeak to valley difference PvPeak to valley ratio RvAnd a fluctuation rate f; the formula is as follows:
Pv＝Pmax-Pmin(3)
P′v＝P′max-P′min(4)
(4) of formula (II) to (III)'vA peak-to-valley difference representing the generalized average load for each region; p'maxRepresents the maximum load among the generalized average loads of the respective regions; p'minRepresents the minimum load among the generalized average loads of the respective regions;
Rv＝Pv/Pmax(5)
R′v＝P′v/P′max(6)
(6) in the formula (II), R'vA peak-to-valley difference rate representing the generalized average load for each region;
(7) in the formula (I), the compound is shown in the specification,
the average value of the loads representing the preset time period of each area; s represents a standard value of the average load of each area in a preset time period;
4) calculating the slope association degree gamma of the curve of the average load and the generalized average load of each region in a preset time period, and calculating the load matching degree rho of new energy among the regions;
(8) in the formula (I), the compound is shown in the specification,
the expression is the generalized average load when the period is t +1 minus the generalized average load difference when the period is t; delta PL(t) represents the average load at period t +1 minus the average load difference at period t;
and
the generalized average load and the variance of the average load, respectively;
(9) in the formula, ρ represents a new energy load matching degree between the first region and the corresponding second region.
The system comprises four templates which are respectively an acquisition preprocessing module and a statistical preprocessing module, wherein the acquisition preprocessing module is used for adding, processing and counting new energy station output data; and is used for adding, processing and counting the multivariate load data; the scene generation module generates a plurality of groups of scenes by using the variational self-encoder, performs probability distribution rule analysis on the generated data and the historical data and provides data for the source load matching module; and the source-load matching module is used for recommending an optimal nearby consumption scheme by calculating the matching degree of the new energy and the load and improving the matching degree.
The overall scheme of the system is as follows: firstly, load data of each time period of each region is obtained through an acquisition preprocessing module, and the data are stored in a database; and then, a scene generation module utilizes a deep learning model VAE to construct a probability distribution model, the probability distribution rule of the source load data is analyzed, and the generated data obtained by training the generation model is used as the data of the source load matching module. And finally, calculating the matching degree of the new energy data and the load data in a source-load matching module, and recommending an optimal nearby consumption scheme according to the matching degree.
The system well completes the statistical analysis of the source load data and can be applied to a power system; the system has good source load probability distribution rule analysis and can improve the near-by consumption capability of new energy.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (6)
1. A new energy and multi-element load value matching method is characterized by comprising the following steps:
s10, acquiring new energy power generation data and load data in a first area in a preset time period, acquiring multiple second area new energy power generation data and load data in the preset time period, and preprocessing; the second area is an area with the potential of absorbing the new energy of the first area;
s20, generating a scene about new energy and load through a variational self-encoder by the preprocessed new energy power generation power data and load data in the first area and the new energy power generation power data and load data in the second areas to obtain scene generation data, and performing model training by using a deep learning model to obtain new energy power generation power data and load data generated by training in the first area and new energy power generation power data and load data generated by training in the second areas;
and S30, generating new energy output data and load data according to the training of each region, calculating the matching degree of the new energy and the load, and selecting a near digestion capacity scheme between the two regions with the maximum matching degree.
2. The method according to claim 1, wherein the step S20 includes:
1) and (3) data encoding: reducing the dimensions of the new energy output data without labels and the load data into a group of characteristic vectors through an encoder, and then performing a fitting process of the model;
2) and (3) data decoding: restoring a group of feature vectors in the encoder into input data through a decoder;
3) gradient back propagation: adjusting the network weight through gradient inverse descent, and learning the distribution rule of historical data through a decoder after repeated training;
4) and (3) data generation: and generating a probability model through the VAE, generating a plurality of groups of new energy and load scene sets by an output layer, and obtaining new energy output data and load data generated by training of each region.
3. The method according to claim 2, wherein the step S30 includes:
1) load data generated by training of each region are subtracted from new energy output data of the region respectively to obtain generalized average load of each region in a preset time period, and the formula is as follows:
(1) in the formula, PL(t) represents training generation load data in a certain area of the respective areas; pW(t) representing new energy output data in a certain area in each area; t represents the number of period time segments;
2) and superposing the generalized average load of all the preset time periods of the regions:
(2) in the formula, N represents the number of the overlapping areas; pLj(t) represents the preset time period average load of the jth superimposed region;
3) respectively calculating the maximum load P of each region in a preset time periodmaxMinimum load PminPeak to valley difference PvPeak to valley ratio RvAnd a fluctuation rate f; the formula is as follows:
Pv＝Pmax-Pmin(3)
P′v＝P′max-P′min(4)
(4) of formula (II) to (III)'vA peak-to-valley difference representing the generalized average load for each region; p'maxRepresents the maximum load among the generalized average loads of the respective regions; p'minRepresents the minimum load among the generalized average loads of the respective regions;
Rv＝Pv/Pmax(5)
R′v＝P′v/P′max(6)
(6) in the formula (II), R'vA peak-to-valley difference rate representing the generalized average load for each region;
(7) in the formula (I), the compound is shown in the specification,
the average value of the loads representing the preset time period of each area; s represents a standard value of the average load of each area in a preset time period;
4) calculating the slope association degree gamma of the curve of the average load and the generalized average load of each region in a preset time period, and calculating the load matching degree rho of new energy among the regions;
(8) in the formula (I), the compound is shown in the specification,
the expression is the generalized average load when the period is t +1 minus the generalized average load difference when the period is t; delta PL(t) represents the average load at period t +1 minus the average load difference at period t;
and
the generalized average load and the variance of the average load, respectively;
(9) in the formula, ρ represents a new energy load matching degree between the first region and the corresponding second region.
4. A new energy and multi-element load value matching system is characterized by comprising:
the system comprises an acquisition preprocessing module, a load preprocessing module and a control module, wherein the acquisition preprocessing module is used for acquiring new energy power generation power data and load data in a first region in a preset time period, acquiring a plurality of second region new energy power generation power data and load data in the preset time period and preprocessing the multiple second region new energy power generation power data and load data; the second area is an area with the potential of absorbing the new energy of the first area;
the scene generation module is used for generating a generation scene about new energy and load through a variational self-coder for the preprocessed new energy power generation power data and load data in the first area and the new energy power generation power data and load data in the second areas to obtain scene generation data, and then performing model training by using a deep learning model to obtain training generation new energy power data and load data in the first area and training generation new energy power generation power data and load data in the second areas;
and the source-load matching module is used for generating new energy output data and load data according to the training of each region, calculating the matching degree of the new energy and the load, and selecting a near digestion capacity scheme between two regions with the maximum matching degree.
5. The system according to claim 4, wherein the scene generation module comprises:
a data encoding unit: reducing the dimensions of the new energy output data without labels and the load data into a group of characteristic vectors through an encoder, and then performing a fitting process of the model;
a data decoding unit: restoring a group of feature vectors in the encoder into input data through a decoder;
gradient counter-propagating unit: adjusting the network weight through gradient inverse descent, and learning the distribution rule of historical data through a decoder after repeated training;
a data generation unit: and generating a probability model through the VAE, generating a plurality of groups of new energy and load scene sets by an output layer, and obtaining new energy output data and load data generated by training of each region.
6. The system according to claim 5, wherein the source-load matching module is specifically configured to:
1) load data generated by training of each region are subtracted from new energy output data of the region respectively to obtain generalized average load of each region in a preset time period, and the formula is as follows:
(1) in the formula, PL(t) represents training generation load data in a certain area of the respective areas; pW(t) representing new energy output data in a certain area in each area; t represents the number of period time segments;
2) and superposing the generalized average load of all the preset time periods of the regions:
(2) in the formula, N represents the number of the overlapping areas; pLj(t) represents the preset time period average load of the jth superimposed region;
3) respectively calculating the maximum load P of each region in a preset time periodmaxMinimum load PminPeak to valley difference PvPeak to valley ratio RvAnd a fluctuation rate f; the formula is as follows:
Pv＝Pmax-Pmin(3)
P′v＝P′max-P′min(4)
(4) of formula (II) to (III)'vA peak-to-valley difference representing the generalized average load for each region; p'maxRepresents the maximum load among the generalized average loads of the respective regions; p'minRepresents the minimum load among the generalized average loads of the respective regions;
Rv＝Pv/Pmax(5)
R′v＝P′v/P′max(6)
(6) in the formula (II), R'vA peak-to-valley difference rate representing the generalized average load for each region;
(7) in the formula (I), the compound is shown in the specification,
the average value of the loads representing the preset time period of each area; s represents a standard value of the average load of each area in a preset time period;
4) calculating the slope association degree gamma of the curve of the average load and the generalized average load of each region in a preset time period, and calculating the load matching degree rho of new energy among the regions;
(8) in the formula (I), the compound is shown in the specification,
the expression is the generalized average load when the period is t +1 minus the generalized average load difference when the period is t; delta PL(t) represents the average load at period t +1 minus the average load difference at period t;
and
the generalized average load and the variance of the average load, respectively;
(9) in the formula, ρ represents a new energy load matching degree between the first region and the corresponding second region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010531489.0A CN111724277A (en) | 2020-06-11 | 2020-06-11 | New energy and multi-element load value matching method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010531489.0A CN111724277A (en) | 2020-06-11 | 2020-06-11 | New energy and multi-element load value matching method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111724277A true CN111724277A (en) | 2020-09-29 |
Family
ID=72567999
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010531489.0A Pending CN111724277A (en) | 2020-06-11 | 2020-06-11 | New energy and multi-element load value matching method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111724277A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113377822A (en) * | 2021-06-21 | 2021-09-10 | 国网宁夏电力有限公司电力科学研究院 | Data processing system based on energy big data |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555628A (en) * | 2019-09-11 | 2019-12-10 | 国网能源研究院有限公司 | comprehensive evaluation method for matching degree of new energy power generation output and load characteristics |
-
2020
- 2020-06-11 CN CN202010531489.0A patent/CN111724277A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555628A (en) * | 2019-09-11 | 2019-12-10 | 国网能源研究院有限公司 | comprehensive evaluation method for matching degree of new energy power generation output and load characteristics |
Non-Patent Citations (3)
| Title |
|---|
| 刘文霞等: "基于风电-负荷匹配度提高风电消纳能力", 《智慧电力》 * |
| 杨秀璋等: "《Python 网络数据爬取及分析从入门到精通（分析篇）》", 30 June 2018, 北京航空航天大学出版社 * |
| 王守相等: ""风电和光伏随机场景生成的条件变分自动编码器方法"", 《电网技术》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113377822A (en) * | 2021-06-21 | 2021-09-10 | 国网宁夏电力有限公司电力科学研究院 | Data processing system based on energy big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Musbah et al. | Energy management using multi-criteria decision making and machine learning classification algorithms for intelligent system | |
| CN108599172B (en) | Transmission and distribution network global load flow calculation method based on artificial neural network | |
| CN109858798B (en) | Power grid investment decision modeling method and device for correlating transformation measures with voltage indexes | |
| CN110212551B (en) | Micro-grid reactive power automatic control method based on convolutional neural network | |
| Long et al. | A combination interval prediction model based on biased convex cost function and auto-encoder in solar power prediction | |
| CN116843083A (en) | Carbon emission prediction system and method based on hybrid neural network model | |
| Guo et al. | Scheme for automatic differentiation of complex loss functions with applications in quantum physics | |
| Niu et al. | Short term load forecasting model using support vector machine based on artificial neural network | |
| CN116207739A (en) | Optimal scheduling method and device for power distribution network, computer equipment and storage medium | |
| Huo et al. | Integrating learning and explicit model predictive control for unit commitment in microgrids | |
| CN115795992A (en) | Park energy Internet online scheduling method based on virtual deduction of operation situation | |
| CN111724277A (en) | New energy and multi-element load value matching method and system | |
| Kronqvist et al. | Alternating mixed-integer programming and neural network training for approximating stochastic two-stage problems | |
| CN115528750B (en) | Power grid safety and stability oriented data model hybrid drive unit combination method | |
| CN116911538A (en) | Reservoir group multi-target scheduling multi-source information joint search method, system and equipment | |
| CN114021437B (en) | Wind power photovoltaic active scene generation method and device, electronic equipment and storage medium | |
| Wang et al. | A possibilistic approach to the modeling and resolution of uncertain closed-loop logistics | |
| CN114692994A (en) | Balanced optimization device based on electric power green certificate market combined clearing model | |
| Pei et al. | A multi-objective hybrid differential optimization algorithm for flow-shop scheduling problem | |
| CN112508239A (en) | Energy storage output prediction method based on VAE-CGAN | |
| Sun et al. | Imitation learning‐based online optimal scheduling for microgrids: An approach integrating input clustering and output classification | |
| CN109816247B (en) | Power grid investment decision modeling method and device for associating transformation measures with line heavy load | |
| CN114004081B (en) | Affine algorithm-based power distribution network load supply capacity calculation method | |
| Yang et al. | Electricity load forecasting based on long and short-term memory neural network | |
| Li et al. | Data‐driven robust look‐ahead power dispatch with preidentification of the worst‐case scenario using combined multilayer perceptron |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200929 |