CN116362136A - Self-dispatching optimization method and system for independent energy storage system - Google Patents
Self-dispatching optimization method and system for independent energy storage system Download PDFInfo
- Publication number
- CN116362136A CN116362136A CN202310357445.4A CN202310357445A CN116362136A CN 116362136 A CN116362136 A CN 116362136A CN 202310357445 A CN202310357445 A CN 202310357445A CN 116362136 A CN116362136 A CN 116362136A
- Authority
- CN
- China
- Prior art keywords
- energy storage
- price
- time
- self
- storage system
- 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
- 238000004146 energy storage Methods 0.000 title claims abstract description 146
- 238000005457 optimization Methods 0.000 title claims abstract description 82
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000005611 electricity Effects 0.000 claims abstract description 127
- 230000006870 function Effects 0.000 claims description 12
- 238000010276 construction Methods 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000013139 quantization Methods 0.000 claims description 4
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000003062 neural network model Methods 0.000 claims description 2
- 238000004590 computer program Methods 0.000 claims 2
- 238000007599 discharging Methods 0.000 description 17
- 230000002354 daily effect Effects 0.000 description 15
- 230000008569 process Effects 0.000 description 6
- 230000008901 benefit Effects 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000005855 radiation Effects 0.000 description 4
- 238000010248 power generation Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008033 biological extinction Effects 0.000 description 1
- 230000009133 cooperative interaction Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
-
- 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
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Water Supply & Treatment (AREA)
- Geometry (AREA)
- Educational Administration (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
Abstract
The invention belongs to the technical field of energy storage system optimal scheduling, and particularly relates to a self-scheduling optimization method and system of an independent energy storage system, wherein the self-scheduling optimization method comprises the following steps: acquiring historical data of the operation of an independent energy storage system and a power grid; constructing a key feature set of electricity price prediction according to the acquired data; based on the constructed key feature set, combining a preset prediction model to respectively predict the day-ahead power price and the real-time power price, and calculating a price difference predicted value of the day-ahead power price predicted value and the real-time power price predicted value; constructing a priori distribution model of the price difference predicted value by using the conditional kernel density estimation, and calculating the confidence coefficient of the price difference predicted value; and according to the confidence coefficient of the obtained price difference predicted value, the daily gain of the independent energy storage power station is used as an optimization target to complete the self-dispatching optimization of the independent energy storage system.
Description
Technical Field
The invention belongs to the technical field of energy storage system optimal scheduling, and particularly relates to a self-scheduling optimization method and system of an independent energy storage system.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
With the rise of a novel power system taking new energy as a main body, the cooperative interaction of source-network-load-storage becomes a necessary requirement for the operation of the power system. Because of the large amount of access of renewable energy sources, the energy storage can well solve the challenges caused by the access of renewable energy sources. On one hand, the energy storage system can cut peaks and fill valleys, increase valley load to promote the absorption of renewable energy sources, and reduce peak load to delay capacity investment requirements; on the other hand, the frequency stability problem caused by wind-light output randomness and fluctuation can be solved, and the reliability of the power grid is improved. With the advancement of the power market reform, the construction and operation of energy storage are mainly realized in the market background, so that the problem about participation of energy storage in the power market is needed to be solved.
Currently in the power spot market, energy storage systems are allowed to participate in market transactions in a self-tuning mode as independent market agents. The self-dispatching mode is that the energy storage power station automatically arranges the charging and discharging plans according to the market electricity price, but needs to report to a dispatching department in advance. When the energy storage is charged, the energy storage is regarded as a power user, and the power is directly purchased from the power spot market; and when discharging, the electric power is regarded as a power generation enterprise, electricity selling is carried out in the electric power spot market, and the price of the electric power market is executed in both charging and discharging. The power spot market is divided into the day-ahead market and the real-time market. The day-ahead market is an electric energy transaction market developed on the day-ahead day of actual operation, and an energy storage power station in the day-ahead market needs to declare a charging and discharging plan of the operation day of the energy storage power station, and the day-ahead market is cleared to obtain the day-ahead electricity price; the real-time market is an electric energy trading market developed 15min before an actual running point on the day of actual running, and in the real-time market, a trading center can clear according to the latest load and new energy output prediction conditions to obtain real-time electricity price. All market agents must participate in both the day-ahead market and the real-time market according to the rules of the transaction. And the settlement in the power spot market generally adopts a dual settlement mode of the market in the day-ahead and the real-time market.
Assuming that the power market is divided into T time periods in one day, the day-ahead market of the energy storage power station declares that the charge and discharge quantity in the T time periods is
Wherein the value is negative when charged and positive when discharged; the actual charge/discharge amount is +.>
the day front electricity price and the real-time electricity price of the t period are respectively +.>
The market benefit B of the energy storage power station is:
wherein the first term represents the revenue of the energy storage power station in the market in the day before and the second term is the revenue of the energy storage power station in the real-time market. Therefore, the settlement electric quantity in the market in the day-ahead is the declared charge and discharge quantity, and the settlement electric quantity in the real-time market is the difference value between the actual charge and discharge quantity and the declared charge and discharge quantity. In this mode, the energy storage power station chooses to charge at low electricity prices and discharges at high electricity prices, thereby earning charge-discharge price differences. Meanwhile, if the charge and discharge plan of the energy storage power station declared before the day is consistent with the actual charge and discharge plan, namely +.>
Market revenue is then all from the first item +.>
The electricity price is settled until all electricity prices are used; if the charge and discharge plan declared by the energy storage power station in the future is inconsistent with the actual charge and discharge plan, both the two items have income. In this case, if the day-ahead electricity price is the same as the real-time electricity price, i.e. +.>
Then no matter how far before the day the energy storage station declares +>
The market income is all
If the day-ahead electricity price is greater than the real-time electricity price, i.e. +.>
The energy storage power station improves the declaration before the day
Then there may be multiple profitability; if the day-ahead electricity price is smaller than the real-time electricity price, i.e. +.>
The energy storage power station reduces the +.>
There may be multiple profitability.
In conclusion, the independent energy storage power station participates in electric quantity market transaction in a self-dispatching mode, a daily operation day charge and discharge plan is declared to a transaction center in advance, and the declaration plan and the actually executed charge and discharge decision directly influence the settlement of the daily market and the real-time market. Therefore, from the perspective of the energy storage power station, the day-ahead electricity price and the real-time electricity price of the operation day must be predicted in advance, and charge and discharge declaration optimization decisions must be developed based on the predicted electricity price so as to maximize self-income.
Disclosure of Invention
In order to solve the problems, the invention provides a self-dispatching optimization method and a self-dispatching optimization system of an independent energy storage system, which carry out self-dispatching optimization decision of an independent energy storage power station according to predicted electricity price, and consider the possibility of reasonably selecting charge and discharge time periods and the arbitrage space between the market in the future and the real-time market, so that the income of the independent energy storage system in the electric power market is improved.
According to some embodiments, a first aspect of the present invention provides a self-tuning optimization method for an independent energy storage system, which adopts the following technical scheme:
a self-tuning optimization method of an independent energy storage system, comprising:
acquiring historical data of the operation of an independent energy storage system and a power grid;
constructing a key feature set of electricity price prediction according to the acquired data;
based on the constructed key feature set, combining a preset prediction model to respectively predict the day-ahead power price and the real-time power price, and calculating a price difference predicted value of the day-ahead power price predicted value and the real-time power price predicted value;
constructing a priori distribution model of the price difference predicted value by using the conditional kernel density estimation, and calculating the confidence coefficient of the price difference predicted value;
and according to the confidence coefficient of the obtained price difference predicted value, the daily gain of the independent energy storage power station is used as an optimization target to complete the self-dispatching optimization of the independent energy storage system.
As a further technical definition, the acquired historical data of the independent energy storage system and grid operation includes real-time electricity price data, day-ahead electricity price data, supply side power prediction data, demand side power prediction data, and numerical weather forecast data.
As a further technical limitation, in the process of constructing the key feature set of electricity price prediction, the autocorrelation coefficient is adopted for quantization analysis, namely, the autocorrelation coefficient ρ Yt (Δd) is:
wherein (1)>
Indicating the electricity price at time t on day d, < >>
Represents the electricity price, mu, at time t on day d-delta d Yt Mean value of electricity price at time t, < >>
A variance indicating the electricity price at time t; when meteorological factors are considered, the cross-correlation coefficient is adopted for quantization analysis, and the cross-correlation coefficient rho (X, Y) is as follows: />
Wherein X is i And Y i Respectively representing the meteorological element and electricity price of the ith sample, mu X Sum mu Y Respectively represent X i And Y i Is a mean value of (c).
As a further technical limitation, the preset prediction model adopts a long-short-time memory neural network, a key feature set is regarded as input, an implicit state is regarded as electricity price, training is carried out by utilizing a training data set, and an Adam optimizer is adopted to optimize long-short-time memory neural network model parameters, so that the construction of the preset prediction model is completed.
As a further technical definition, the expression of the kernel density estimation is
n represents the number of samples, h represents window width, h determines the smoothness of the probability density function, x i Representing the ith sample, K represents the kernel function.
Further, after the day-ahead electricity price and the real-time electricity price predicted value at the time t are obtained based on the long-short-time memory neural network, the price difference predicted value y 'at the time can be obtained' t Obtaining a constructed valence difference posterior conditional probability distribution model Pr (y) based on conditional kernel density estimation t |x t ) Calculate Pr (y' t |x t ) Find y t So that Pr (y) t |x t )＝Pr(y′ t |x t ) Confidence xi t Is that
The method comprises the steps of constructing a conditional probability model Pr (y|x) based on kernel density estimation, and according to a Bayesian full probability rule, the conditional probability form of the kernel density estimation is as follows:
wherein f XY For joint probability distribution, f X For the edge density distribution, K (·) is the kernel function used, h x And h y Is the bandwidth parameter of variables x and y, controls the degree of smoothness of the output probability density,
representing a given sample set.
As a further technical limitation, pre-judging according to the obtained price difference predicted value and confidence, when the price difference threshold is met, constructing a self-dispatching optimization model by taking the daily gain of the independent energy storage power station as an optimization target and taking daily market constraint, real-time market constraint and joint constraint of daily and real-time market strategies as constraint conditions, and carrying out self-dispatching optimization of the independent energy storage system.
According to some embodiments, a second aspect of the present invention provides a self-tuning optimization system of an independent energy storage system, which adopts the following technical scheme:
a self-tuning optimization system for an independent energy storage system, comprising:
an acquisition module configured to acquire historical data of the operation of the independent energy storage system and the grid;
a construction module configured to construct a key feature set of electricity price predictions from the acquired data;
the calculation module is configured to respectively predict the day-ahead power price and the real-time power price based on the constructed key feature set and in combination with a preset prediction model, and calculate a price difference predicted value of the day-ahead power price predicted value and the real-time power price predicted value; constructing a priori distribution model of the price difference predicted value by using the conditional kernel density estimation, and calculating the confidence coefficient of the price difference predicted value;
and the scheduling module is configured to complete the self-scheduling optimization of the independent energy storage system by taking the daily gain of the independent energy storage power station as an optimization target according to the confidence level of the obtained price difference predicted value.
According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium, which adopts the following technical solutions:
a computer readable storage medium having stored thereon a program which when executed by a processor performs the steps of a self-tuning optimization method of a stand alone energy storage system according to the first aspect of the present invention.
According to some embodiments, a fourth aspect of the present invention provides an electronic device, which adopts the following technical solutions:
an electronic device comprising a memory, a processor and a program stored on the memory and executable on the processor, the processor implementing the steps in a self-tuning optimization method of a stand-alone energy storage system according to the first aspect of the invention when the program is executed.
Compared with the prior art, the invention has the beneficial effects that:
the method predicts the day-ahead electricity price and the real-time electricity price in the electric power market based on deep learning and conditional kernel density estimation, and calculates the confidence level of the corresponding electricity price prediction result; the self-dispatching optimization decision of the independent energy storage power station in the double-settlement electric power market mode is scientifically developed by energy storage operators according to electricity price prediction conditions.
According to the invention, a self-dispatching optimization decision model of the energy storage power station is established from a settlement mode of the electric power market, so that the charge-discharge price difference profit of the energy storage power station is considered, the financial arbitrage possibility of the energy storage power station between the market in the future and the real-time market is also considered, and the income of the energy storage power station in the electric power market can be better improved.
According to the method, the optimization decision judgment is carried out according to the predicted electricity price condition, the energy storage charging and discharging price difference threshold value and the strategy adjustment price difference threshold value are set, and the optimization decision flow is carried out after the price difference threshold value condition is met, so that the optimization decision efficiency of the energy storage power station is improved; the energy storage charging and discharging price difference threshold reflects the construction and operation cost of the energy storage power station, the strategy adjustment price difference threshold reflects the level of excitation of an operator of the energy storage power station on adjusting the reporting strategy of the operator, or the preference degree of risks, and the energy storage charging and discharging decision can be better developed through reasonably setting two price differences.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments and are incorporated in and constitute a part of this specification, illustrate and explain the embodiments and together with the description serve to explain the embodiments.
FIG. 1 is a flow chart of a self-tuning optimization method for an independent energy storage system according to a first embodiment of the invention;
FIG. 2 is a flow chart illustrating the operation of the independent energy storage self-tuning optimization decision method according to the first embodiment of the invention;
FIG. 3 is a schematic diagram of electricity price prediction based on deep learning and conditional kernel density estimation in accordance with a first embodiment of the present invention;
FIG. 4 is a schematic diagram of an independent energy storage self-tuning optimization decision-making in accordance with a first embodiment of the present invention;
FIG. 5 is a schematic diagram of a prediction curve of the current price and the real-time current price before a certain operation day in the first embodiment of the present invention;
FIG. 6 is a schematic diagram of an optimization decision result of an energy storage power station on a certain operation day in the first embodiment of the present invention;
fig. 7 is a block diagram of a self-tuning optimization system for an independent energy storage system in accordance with a second embodiment of the present invention.
Detailed Description
The invention will be further described with reference to the drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present invention. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
Embodiments of the invention and features of the embodiments may be combined with each other without conflict.
Term interpretation:
double settlement: market subjects participating in the power market settle accounts in both the power day-ahead market and the real-time market, respectively.
Independent energy storage: and the energy storage power station participates in the electric power market by an independent settlement main body.
Self-scheduling: and automatically deciding a charging and discharging scheme of the energy storage power station according to market conditions.
Market in the daytime: the market for electric quantity transaction carried out day before day is actually operated.
Day-ahead electricity price: the day-ahead market clearly obtains day-ahead node electricity prices.
Real-time market: and the electric quantity transaction market is developed 15min before the actual operation point on the day of the actual operation.
Real-time electricity price: and the real-time market is clear, and the obtained real-time node electricity price is obtained.
Example 1
The embodiment of the invention introduces a self-dispatching optimization method of an independent energy storage system.
A method of self-tuning optimization of an independent energy storage system as shown in fig. 1 and 2, comprising:
acquiring historical data of the operation of an independent energy storage system and a power grid;
constructing a key feature set of electricity price prediction according to the acquired data;
based on the constructed key feature set, combining a preset prediction model to respectively predict the day-ahead power price and the real-time power price, and calculating a price difference predicted value of the day-ahead power price predicted value and the real-time power price predicted value;
constructing a priori distribution model of the price difference predicted value by using the conditional kernel density estimation, and calculating the confidence coefficient of the price difference predicted value;
and according to the confidence coefficient of the obtained price difference predicted value, the daily gain of the independent energy storage power station is used as an optimization target to complete the self-dispatching optimization of the independent energy storage system.
The present embodiment is described in detail in two ways:
a day-ahead electricity price prediction and a real-time electricity price prediction based on deep learning and nuclear density estimation;
and (II) a self-dispatching optimization decision method of an independent energy storage system in a double settlement mode.
First, future electricity price prediction and real-time electricity price prediction based on deep learning and nuclear density estimation
The price of the power spot transaction is mainly influenced by two aspects: the supply amount on the supply side and the demand amount on the demand side together determine the electric power spot price. The influence factors on the supply side include: the supply condition of the conventional thermal power generating unit, the supply condition of the extra-saving incoming call and the output condition of the new energy; the influence factors on the demand side mainly include: load change conditions. The power transmission of the extra-provincial call among the influencing factors is generally carried out according to a power transmission curve established in advance, the conventional unit is adjusted according to new energy and load changes, and the uncertain factors are mainly new energy output and load changes. The most important factor affecting the output and load change of new energy is weather. Therefore, the new energy output prediction condition, the load prediction condition and the meteorological factors are comprehensively considered, and the electricity price prediction can be effectively carried out by selecting a proper prediction model.
In this embodiment, as shown in fig. 3, the specific steps of the day-ahead electricity price prediction and the real-time electricity price prediction are:
constructing a historical data set, wherein the historical data set comprises real-time electricity price data, day-ahead electricity price data, supply side predicted power data, demand side predicted power data and numerical weather forecast data;
different factors influencing the day-ahead power price and the real-time power price are analyzed by adopting a means combining correlation coefficient numerical quantification and theoretical analysis, and a key feature set facing the day-ahead power price and the real-time power price prediction is constructed;
based on the constructed key feature set, respectively establishing a day-ahead power price prediction model and a real-time power price prediction model by utilizing a long-short-period memory neural network to obtain a day-ahead power price prediction result, a real-time power price prediction result and a price difference prediction result of the day-ahead power price prediction result and the real-time power price prediction result;
and constructing a prior distribution model of the price difference between the day-ahead electricity price and the real-time electricity price by using the conditional kernel density estimation, and evaluating the confidence coefficient of the price difference prediction result based on the price difference prior distribution and the price difference prediction result.
As one or more embodiments, the real-time electricity price data, the day-ahead electricity price data, the supply side power prediction data and the demand side power prediction data are all from the electric power trading center in the process of constructing the data set. The time resolution of the data is 15 minutes, and the power prediction data of the supply side and the power prediction data of the demand side are the predicted results of the day before. The supply-side power prediction data includes: wind power prediction data, photovoltaic power prediction data, local power plant power generation prediction data, nuclear power prediction data, self-contained unit power generation prediction data and tie line power prediction data. The demand-side power prediction data is direct-tuning load prediction data. The method for acquiring the numerical weather forecast data comprises the following steps: using a global forecasting system (Global Forecasting System) to drive WRF (Weather Research and Forecasting) modes to carry out multi-layer nested downscaling, obtaining 3km multiplied by 3km grid numerical weather forecast data, carrying out weighted average on grid data covered by a region of interest to obtain numerical weather forecast results of the region, wherein the weather elements comprise: temperature, wind speed, short wave radiation, relative humidity, etc.
As one or more embodiments, during the construction of the key feature set, the electricity price is directly determined by the supply side power and the demand side power, so that supply side and demand side power predictions need to be considered in constructing the feature set. Meanwhile, the electricity price has certain periodic characteristics, the electricity price has high similarity at the same time close to a plurality of days, and the electricity price is quantitatively analyzed by adopting an autocorrelation coefficient and calculated as follows:
wherein (1)>
Indicating the electricity price at time t on day d, < >>
Represents the electricity price, mu, at time t on day d-delta d Yt Mean value of electricity price at time t, < >>
The variance of the electricity price at time t is shown.
Meanwhile, factors such as temperature, irradiance and the like influence the wind-solar new energy output and the demand side load power, so that meteorological factors can be considered when a feature set is constructed, and the cross-correlation coefficient is adopted for quantitative analysis, and the calculation formula is as follows:
wherein X is i And Y i Respectively representing the meteorological element and electricity price of the ith sample, mu X Sum mu Y Respectively represent X i And Y i Is a mean value of (c).
As shown in table 1, the cross correlation coefficient of weather and day-ahead electricity prices and the autocorrelation coefficient of day-ahead electricity prices are available, among the weather factors, the cross correlation coefficient of short wave radiation and wind speed with day-ahead electricity prices is high, and the autocorrelation coefficient of day-ahead electricity prices at the same time in the vicinity of four days is high, and the characteristic set at the time of dimensionally predicting day-ahead electricity prices includes supply side predicted power, demand side predicted power, short wave radiation, wind speed, and day-ahead electricity prices at the same time in the preceding four days. The cross correlation coefficient of weather and real-time electricity prices and the autocorrelation coefficient of real-time electricity prices as shown in table 2 can be obtained similarly to the day-ahead electricity price, so that the feature set when predicting real-time electricity prices includes the supply side predicted power, the demand side predicted power, the short wave radiation, the wind speed, and the real-time electricity price at the same time of the previous four days.
TABLE 1 day-ahead Power price auto-correlation coefficient and Meteorological factor Cross-correlation coefficient
TABLE 2 real-time Power price auto-correlation coefficient and air-phase factor Cross-correlation coefficient
As one or more embodiments, long-short-term memory neural networks (LSTM) are an improved recurrent neural network that overcomes the phenomenon of gradient extinction or gradient explosion that occurs when the time step increases in conventional recurrent neural networks, and can mine the timing dependence on longer time scales. The LSTM mainly comprises a forgetting gate, an input gate and an output gate; the forget gate controls the information to be reserved in each LSTM unit, and the calculation formula of the process is as follows: f (f) t ＝σ(w f [h t-1 ,x t ]+b f ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein h is t-1 Indicating the implicit state of the last moment; x is x t An input representing a current time; w (w) f And b f Respectively representing a weight matrix and a bias vector; sigma represents a sigmoid activation function; [ h ] t-1 ,x t ]Representation of pair h t-1 And x t Vector splicing operation is carried out; the input gate determines the information for updating the memory state, i.e./ t ＝σ(w i [h t-1 ,x t ]+b i ) The method comprises the steps of carrying out a first treatment on the surface of the The output gate outputs according to the current input and conditional determination of the previous time state, i.e. o t ＝σ(w o [h t-1 ,x t ]+b o )。
For time t, the actual input characteristic of this time is represented by input x t And the implicit state h of the last moment t-1 The calculation results show that the input characteristics are normalized to be between-1 and 1 by using the tanh function, and the calculation process is that
Memory cell c t Is based on the adjusted input features +.>
And a memory unit c at the previous time t-1 The calculation process is as follows: />
The hidden state of the final current moment can be obtained by the output gate o t And a memory unit c t Calculated, i.e. h t ＝o t *tanh(c t ) The method comprises the steps of carrying out a first treatment on the surface of the Where, represents Hadamard product.
When constructing the electricity price prediction model, the key feature set is regarded as input x t Will imply state h t And taking the model as the electricity price, training by using a training data set, and optimizing LSTM model parameters by using an Adam optimizer to construct an electricity price prediction model.
As one or more embodiments, a prediction model of the day-ahead electricity price and the real-time electricity price is respectively built based on a long-short-time memory neural network, a price difference prediction result can be obtained on the basis of predicting the day-ahead electricity price and the real-time electricity price, and the confidence level of price difference prediction provides more reference information for subsequent decisions. And constructing a posterior probability distribution model of the price difference by adopting conditional kernel density estimation, and further carrying out confidence assessment on the price difference prediction result.
The expression for the nuclear density estimation is:
where n represents the number of samples, h represents window width, h determines the smoothness of the probability density function, x i Representing the ith sample, K represents the kernel function.
For posterior probability distribution estimation of valence differences, a set of samples is given
Wherein x is i Representing factors related to price difference, where the union of the key feature set of day-ahead power price and the key feature set of real-time power price is taken, i.e. comprising supply-side and demand-side power pre-limitsThe method comprises the steps of constructing a conditional probability model Pr (y|x) based on nuclear density estimation, wherein the conditional probability form of the nuclear density estimation can be deduced according to a Bayesian full probability rule:
wherein f XY For joint probability distribution, f X For the edge density distribution, K (·) is the kernel function used, h x And h y Is the bandwidth parameter of variables x and y, which controls the degree of smoothness of the output probability density.
After the day-ahead electricity price and the real-time electricity price predicted value at the time t are obtained based on LSTM, the price difference predicted value y 'at the time can be obtained' t Meanwhile, a constructed valence difference posterior conditional probability distribution model Pr (y) is obtained based on conditional kernel density estimation t |x t ) Calculate Pr (y' t |x t ) Find y t So that Pr (y) t |x t )＝Pr(y′ t |x t ) Confidence is then
Self-dispatching optimization decision method of independent energy storage system in (II) double settlement mode
The existing self-scheduling optimization decision-making method of the energy storage power station mainly comprises two methods: firstly, according to the historical trend of electricity price, the operators charge and discharge in a fixed time period every day. Because the electricity price is continuously changed under the market condition, the low point and the high point of the daily electricity price are difficult to judge by experience of operators, and the method depending on manual experience can cause the reduction of the income of the energy storage power station. The second method is to make decisions based on the predicted day-ahead electricity prices. In this way, since only the day-ahead electricity prices are considered, and the settlement method of the double settlement market mode is omitted, the energy storage power station loses the opportunity of arbitrage in the day-ahead market and the real-time market in some cases, and the income is reduced. Therefore, the self-scheduling optimization decision of the energy storage power station in the double settlement mode must consider the reasonable selection of the charging and discharging time period and the possibility of the existence of the market-ahead and real-time market arbitrage space.
In the process of self-dispatching optimization decision of the independent energy storage system, as shown in fig. 4, firstly, pre-judging is carried out according to the predicted day-ahead electricity price and the real-time electricity price, judging contents comprise whether the peak-valley price difference of the operation day meets a charging and discharging threshold value, whether a arbitrage space exists in the day-ahead market and the real-time market, and if one of the peak-valley price difference and the charging and discharging threshold value is met, the charging and discharging optimization decision is carried out, and if both the peak-valley price and the real-time price are not met, the charging and discharging operation is not carried out on the operation day.
(1) Optimization decision making
Setting a charge-discharge valence difference threshold as delta Q (yuan/MWh), wherein delta Q can be reasonably set according to the construction and operation cost of the energy storage power station; setting a policy adjustment price difference threshold value as beta (element/MWh), wherein the value represents the aggressive level of a user for adjusting the reporting policy of the user, the smaller the value is, the more aggressive the user is to the adjustment policy, and the larger the value is, the more conservative the user is to the adjustment policy is; confidence in predicted day-ahead and real-time price difference for period t is ζ t Meaning the confidence level of the price difference.
If the peak-valley difference satisfying the predicted day-ahead electricity price is greater than the set charge-discharge price difference threshold (i.e
) Or the day-ahead and real-time tariffs meeting the t-period prediction are greater than a set strategically adjusted tariff threshold (i.e
) The method comprises the steps of carrying out a first treatment on the surface of the And entering a second step of optimization decision, and if the optimization decision is not satisfied, not performing charge and discharge operation on the operation day.
(2) Self-dispatching optimization decision of energy storage power station
Establishing an independent energy storage power station self-dispatching optimization decision model, wherein the optimization variable is the charge and discharge power of the independent energy storage power station in the daily market and the real-time market in each transaction period in the operation day, the optimization target is the maximum daily income of the operation of the independent energy storage power station, and the optimization target function is as follows:
wherein F is 1 Representing the income of an energy storage power station in the market in the day-ahead, wherein DeltaT is the time interval of each period of the electric power market;
and->
Respectively representing the charging power and the discharging power of the energy storage power station in the period t of the market before the day,/>
Predicted electricity prices for the time period t of the market in the day; f (F) 2 Represents the return of the energy storage power station in the real-time market, +.>
And->
Respectively representing the charge and discharge power of the stored energy in the real-time market t period,/>
Predicted electricity prices for the time period t of the real-time market; f (F) 3 Representing the construction and operation costs of the energy storage power station, F kc The electricity-measuring cost of the energy storage power station means the average cost consumed by the energy storage power station for generating the unit electric quantity, including construction and operation and maintenance costs; f (F) 4 Representing the power transmission and distribution fees and the additional funds fees which need to be paid by the energy storage power station, Q net The method comprises the steps that the energy storage power station is required to pay the power transmission and distribution fee and the foundation additional fee of the consumed unit electric quantity to the power grid company, and only the charge and discharge quantity difference value is required to pay the fee according to typical electric power market regulations; />
The four terms are variables to be optimized.
The constraint conditions to be met by the optimization target include:
1) Market constraints in the future:
wherein,,
and->
The state of charge variable and the state of discharge variable of the energy storage system in the period t of the market before the day are respectively represented. When the value is 1, the energy storage system is in a charge/discharge state, and when the value is 0, the energy storage system is not in a charge/discharge state. />
And->
Representing the maximum charge and discharge power, respectively, of the energy storage system.
2) Real-time market constraints:
wherein,,
and->
The charging state variable and the discharging state variable of the energy storage system in the period of the real-time market t are respectively represented. />
The state of charge is the period t of the energy storage power station; η (eta) ch For the charging efficiency of the energy storage power station, eta dis The discharge efficiency of the energy storage power station; e (E) rate The total capacity of the energy storage power station.
3) Linking constraints for day-ahead and real-time marketing strategies:
when the daily and real-time valence difference confidence is smaller than a given threshold, the daily and real-time decisions are consistent, and when the daily and real-time valence difference confidence is larger than the given threshold, different decisions can be made. The market in the future only carries out financial settlement without considering actual implementation, so that the constraint of the energy storage state of charge is not considered.
Calculation case analysis
Taking an independent energy storage power station as an example, the installed capacity of the energy storage power station is 100MW/200MWh, the charging and discharging efficiency of the energy storage power station is 90%, the initial charge state is 0, the additional cost of the power transmission and distribution cost and the foundation is 200 yuan/MWh, the charging and discharging price difference threshold is 200 yuan/MWh, and the strategy adjustment price difference threshold is 100 yuan/MWh. The predicted day-ahead and real-time electricity prices for a certain day are shown in fig. 5, and the self-scheduling optimization decision result of the energy storage power station is shown in fig. 6.
As can be seen from the electricity price curve in fig. 5, the operation day has two electricity price peaks and two electricity price valleys, the energy storage power station can perform two-charge and two-discharge operation, and the actual implementation strategy is two-charge and two-discharge, the electricity is charged at the two electricity price valleys, and the electricity is discharged at the two electricity price peaks; from the aspect of the real-time electricity price and the trend of the day-ahead electricity price, the real-time electricity price is higher than the day-ahead electricity price in most of the time period, and the price difference is larger than the strategy adjustment price difference threshold, so that in the optimization result shown in fig. 6, the day-ahead reporting strategy is adjusted, and the reporting power is reduced in the time period that the day-ahead electricity price is lower than the real-time electricity price, so that the benefits in the day-ahead and real-time markets are realized. Under the condition, the expected benefit of the operation day energy storage power station is 726871 yuan; if the set benefit of the energy storage power station is not considered, the strategy adjustment price difference threshold value can be increased to 1000 yuan/MWh, the future declaration of the energy storage power station is consistent with the real-time execution strategy, and the expected benefit of the energy storage power station is 218551 yuan.
The embodiment predicts the day-ahead electricity price and the real-time electricity price in the electric power market based on deep learning and conditional kernel density estimation, and calculates the confidence level of the corresponding electricity price prediction result; the self-dispatching optimization decision of the independent energy storage power station in the double-settlement electric power market mode is scientifically developed by energy storage operators according to the electricity price prediction condition; starting from a settlement mode of an electric power market, a self-dispatching optimization decision model of the energy storage power station is established, charge and discharge price difference profit of the energy storage power station is considered, financial arbitrage possibility between a market in the future and a real-time market of the energy storage power station is also considered, and the income of the energy storage power station in the electric power market can be better improved.
Example two
The second embodiment of the invention introduces a self-dispatching optimization system of an independent energy storage system.
A self-tuning optimization system for a self-contained energy storage system as shown in fig. 7, comprising:
an acquisition module configured to acquire historical data of the operation of the independent energy storage system and the grid;
a construction module configured to construct a key feature set of electricity price predictions from the acquired data;
the calculation module is configured to respectively predict the day-ahead power price and the real-time power price based on the constructed key feature set and in combination with a preset prediction model, and calculate a price difference predicted value of the day-ahead power price predicted value and the real-time power price predicted value; constructing a priori distribution model of the price difference predicted value by using the conditional kernel density estimation, and calculating the confidence coefficient of the price difference predicted value;
and the scheduling module is configured to complete the self-scheduling optimization of the independent energy storage system by taking the daily gain of the independent energy storage power station as an optimization target according to the confidence level of the obtained price difference predicted value.
The detailed steps are the same as those of the self-tuning optimization method of the independent energy storage system provided in the first embodiment, and are not described herein.
Example III
The third embodiment of the invention provides a computer readable storage medium.
A computer readable storage medium having stored thereon a program which when executed by a processor performs the steps of a self-tuning optimization method for a self-contained energy storage system according to an embodiment of the present invention.
The detailed steps are the same as those of the self-tuning optimization method of the independent energy storage system provided in the first embodiment, and are not described herein.
Example IV
The fourth embodiment of the invention provides electronic equipment.
An electronic device comprising a memory, a processor and a program stored on the memory and executable on the processor, the processor implementing the steps in a self-tuning optimization method for a stand-alone energy storage system according to an embodiment of the invention when the program is executed by the processor.
The detailed steps are the same as those of the self-tuning optimization method of the independent energy storage system provided in the first embodiment, and are not described herein.
The above description is only a preferred embodiment of the present embodiment, and is not intended to limit the present embodiment, and various modifications and variations can be made to the present embodiment by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present embodiment should be included in the protection scope of the present embodiment.
Claims (10)
1. A method of self-tuning optimization of an independent energy storage system, comprising:
acquiring historical data of the operation of an independent energy storage system and a power grid;
constructing a key feature set of electricity price prediction according to the acquired data;
based on the constructed key feature set, combining a preset prediction model to respectively predict the day-ahead power price and the real-time power price, and calculating a price difference predicted value of the day-ahead power price predicted value and the real-time power price predicted value;
constructing a priori distribution model of the price difference predicted value by using the conditional kernel density estimation, and calculating the confidence coefficient of the price difference predicted value;
and according to the confidence coefficient of the obtained price difference predicted value, the daily gain of the independent energy storage power station is used as an optimization target to complete the self-dispatching optimization of the independent energy storage system.
2. A method of self-tuning an independent energy storage system as defined in claim 1, wherein the acquired historical data for the independent energy storage system and grid operation includes real-time electricity rate data, day-ahead electricity rate data, supply side power forecast data, demand side power forecast data, and numerical weather forecast data.
3. A method of self-tuning optimization of an independent energy storage system as defined in claim 1, wherein in constructing the key feature set of electricity price prediction, the quantization analysis is performed using autocorrelation coefficients, i.e., autocorrelation coefficients
The method comprises the following steps:
wherein (1)>
Indicating the electricity price at time t on day d, < >>
Represents the electricity price at time t on day d- Δd,/->
Mean value of electricity price at time t, < >>
A variance indicating the electricity price at time t; when meteorological factors are considered, the cross-correlation coefficient is adopted for quantization analysis, and the cross-correlation coefficient rho (X, Y) is as follows:
wherein X is i And Y i Respectively representing the meteorological element and electricity price of the ith sample, mu X Sum mu Y Respectively represent X i And Y i Is a mean value of (c).
4. The self-tuning optimization method of an independent energy storage system according to claim 1, wherein the preset prediction model is built by adopting a long-short-time memory neural network, taking a key feature set as input, taking an implicit state as electricity price, training by using a training data set, and optimizing long-short-time memory neural network model parameters by using an Adam optimizer.
5. A method of self-tuning optimization of an independent energy storage system as defined in claim 1, wherein said expression of the core density estimate is
n represents the number of samples, h represents window width, h determines the smoothness of the probability density function, x i Representing the ith sample, K represents the kernel function.
6. The self-tuning optimization method of an independent energy storage system as claimed in claim 5, wherein the price difference predicted value y 'at time is obtained after obtaining the day front price and the real-time price predicted value at time t based on a long-short-term memory neural network' t Obtaining a constructed valence difference posterior conditional probability distribution model Pr (y) based on conditional kernel density estimation t |x t ) Calculate Pr (y' t |x t ) Find y t So that Pr (y) t |x t )＝Pr(y′ t |x t ) Confidence xi t Is that
The method comprises the steps of constructing a conditional probability model Pr (y|x) based on kernel density estimation, and according to a Bayesian full probability rule, the conditional probability form of the kernel density estimation is as follows:
wherein f XY For joint probability distribution, f X For the edge density distribution, K (·) is the kernel function used, h x And h y Is the bandwidth parameter of variables x and y, controls the degree of smoothness of the output probability density,
representing a given sample set.
7. The self-dispatching optimization method of the independent energy storage system according to claim 1, wherein the self-dispatching optimization model is built by taking daily gain of the independent energy storage power station as an optimization target and taking daily market constraint, real-time market constraint and joint constraint of daily and real-time market strategies as constraint conditions when the price difference threshold is met according to the obtained price difference predicted value and the confidence coefficient, and self-dispatching optimization of the independent energy storage system is carried out.
8. A self-tuning optimization system for an independent energy storage system, comprising:
an acquisition module configured to acquire historical data of the operation of the independent energy storage system and the grid;
a construction module configured to construct a key feature set of electricity price predictions from the acquired data;
the calculation module is configured to respectively predict the day-ahead power price and the real-time power price based on the constructed key feature set and in combination with a preset prediction model, and calculate a price difference predicted value of the day-ahead power price predicted value and the real-time power price predicted value; constructing a priori distribution model of the price difference predicted value by using the conditional kernel density estimation, and calculating the confidence coefficient of the price difference predicted value;
and the scheduling module is configured to complete the self-scheduling optimization of the independent energy storage system by taking the daily gain of the independent energy storage power station as an optimization target according to the confidence level of the obtained price difference predicted value.
9. A computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the steps of a self-tuning optimization method of an independent energy storage system according to any one of claims 1-7.
10. An electronic 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 steps of the self-tuning optimization method of the independent energy storage system of any one of claims 1-7 when the program is executed by the processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310357445.4A CN116362136A (en) | 2023-03-31 | 2023-03-31 | Self-dispatching optimization method and system for independent energy storage system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310357445.4A CN116362136A (en) | 2023-03-31 | 2023-03-31 | Self-dispatching optimization method and system for independent energy storage system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116362136A true CN116362136A (en) | 2023-06-30 |
Family
ID=86941734
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310357445.4A Pending CN116362136A (en) | 2023-03-31 | 2023-03-31 | Self-dispatching optimization method and system for independent energy storage system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116362136A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117254505A (en) * | 2023-09-22 | 2023-12-19 | 南方电网调峰调频(广东)储能科技有限公司 | Energy storage power station optimal operation mode decision method and system based on data processing |
-
2023
- 2023-03-31 CN CN202310357445.4A patent/CN116362136A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117254505A (en) * | 2023-09-22 | 2023-12-19 | 南方电网调峰调频(广东)储能科技有限公司 | Energy storage power station optimal operation mode decision method and system based on data processing |
| CN117254505B (en) * | 2023-09-22 | 2024-03-26 | 南方电网调峰调频(广东)储能科技有限公司 | Energy storage power station optimal operation mode decision method and system based on data processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bottieau et al. | Very-short-term probabilistic forecasting for a risk-aware participation in the single price imbalance settlement | |
| Lu et al. | A hybrid deep learning-based online energy management scheme for industrial microgrid | |
| Sedighizadeh et al. | A two-stage optimal energy management by using ADP and HBB-BC algorithms for microgrids with renewable energy sources and storages | |
| Wang et al. | Robust energy systems scheduling considering uncertainties and demand side emission impacts | |
| Abedi et al. | Rolling-horizon optimization integrated with recurrent neural network-driven forecasting for residential battery energy storage operations | |
| Bagheri et al. | Stochastic optimization and scenario generation for peak load shaving in Smart District microgrid: sizing and operation | |
| Giglio et al. | An efficient artificial intelligence energy management system for urban building integrating photovoltaic and storage | |
| Khalid | Wind power economic dispatch–impact of radial basis functional networks and battery energy storage | |
| CN116388293A (en) | Combined optimization scheduling method and system for new energy matched energy storage power station | |
| CN112184479A (en) | Reservoir group dispatching rule type and parameter adaptability research method for climate change | |
| Falabretti et al. | Scheduling and operation of RES-based virtual power plants with e-mobility: A novel integrated stochastic model | |
| Saini et al. | Data driven net load uncertainty quantification for cloud energy storage management in residential microgrid | |
| Henni et al. | Industrial peak shaving with battery storage using a probabilistic forecasting approach: Economic evaluation of risk attitude | |
| CN116362136A (en) | Self-dispatching optimization method and system for independent energy storage system | |
| Hosseini et al. | Data-driven multi-resolution probabilistic energy and reserve bidding of wind power | |
| CN117726143B (en) | Environment-friendly micro-grid optimal scheduling method and system based on deep reinforcement learning | |
| Gaspar et al. | Optimisation and economic feasibility of Battery Energy Storage Systems in electricity markets: The Iberian market case study | |
| CN115313361A (en) | Joint optimization control method and device for large-scale adjustable resources | |
| Kim et al. | Economical energy storage systems scheduling based on load forecasting using deep learning | |
| Alikhani et al. | Optimal implementation of consumer demand response program with consideration of uncertain generation in a microgrid | |
| Wang et al. | Optimal scheduling strategy for virtual power plants with aggregated user-side distributed energy storage and photovoltaics based on CVaR-distributionally robust optimization | |
| CN112884572A (en) | Multi-microgrid transaction optimization method and system under credit risk | |
| Wang et al. | Optimal allocation of customer energy storage based on power big data and improved LSTM load forecasting | |
| Ghanbari et al. | Modeling market trading strategies of the intermediary entity for microgrids: A reinforcement learning-based approach | |
| Vieira et al. | Two-stage robust market clearing procedure model for day-ahead energy and reserve auctions of wind–thermal systems |
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 |