CN116826704A - Regulation and control method for participation of flexible load cluster in demand response - Google Patents
Regulation and control method for participation of flexible load cluster in demand response Download PDFInfo
- Publication number
- CN116826704A CN116826704A CN202310562762.XA CN202310562762A CN116826704A CN 116826704 A CN116826704 A CN 116826704A CN 202310562762 A CN202310562762 A CN 202310562762A CN 116826704 A CN116826704 A CN 116826704A
- Authority
- CN
- China
- Prior art keywords
- flexible load
- regulation
- data
- control
- flexible
- 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
- 230000004044 response Effects 0.000 title claims abstract description 50
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000001105 regulatory effect Effects 0.000 claims description 21
- 238000010606 normalization Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000001276 controlling effect Effects 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000012163 sequencing technique Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 230000005611 electricity Effects 0.000 claims description 6
- 230000007246 mechanism Effects 0.000 claims description 6
- 238000013486 operation strategy Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 238000010223 real-time analysis Methods 0.000 claims description 3
- 230000003213 activating effect Effects 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 claims description 2
- 230000000306 recurrent effect Effects 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06N3/0985—Hyperparameter optimisation; Meta-learning; Learning-to-learn
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/36—Arrangements for transfer of electric power between ac networks via a high-tension dc link
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Power Engineering (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a regulation and control method for participation of a flexible load cluster in demand response. Taking into consideration factors such as temperature, humidity, wind speed, date type (working day and non-working day), a short-term flexible load prediction flow based on a GRU model is designed, a predicted resident flexible load curve is input into a demand response system, and optimal scheduling of resident flexible load clusters is realized according to corresponding regulation and control demands of demands. The method provided by the invention can assist resident users and related management departments to formulate flexible load adjustment strategies, and improves the utilization efficiency of flexible resources.
Description
Technical Field
The invention relates to the field of power demand side response, in particular to a regulation and control method for participation of a flexible load cluster in demand response.
Background
With the development of social economy, the electric load is continuously increasing. The cost for building peak-shaving power plants and newly building installed capacity to meet short-term power demand in peak load period is high, and reasonable utilization of social resources is not facilitated. The electric power demand response technology is continuously and deeply researched, and the automatic response technology and the regulating method of the demand side load resource are increasingly paid attention to. In order to efficiently utilize flexible load resources such as air conditioner load and interruptible load and exert the application thereof in demand response, the invention provides a regulation and control method for the participation of a flexible load cluster in demand response, which can assist resident users and related management departments to formulate a flexible load regulation strategy and improve the utilization efficiency of flexible resources.
Disclosure of Invention
In order to improve the utilization efficiency of flexible resources and assist resident users and related management departments to formulate a flexible load regulation strategy, the invention provides a regulation and control method for participation of flexible load clusters in demand response, which can utilize flexible loads to predict the loads of the flexible load clusters, input a predicted flexible load curve into a demand side response system, formulate an optimized scheduling strategy, enhance flexible resource regulation and control, optimize the load curve and realize peak clipping and valley filling.
In order to achieve the purpose of the invention, the technical scheme adopted by the invention is as follows: a regulation and control method for a flexible load cluster to participate in demand response comprises the following steps:
step 1, data selection
Step 2, data preprocessing
Step 3, model selection
Step 4, super parameter selection
Step 5, model training and prediction
Step 6, formulating an optimized scheduling strategy
According to the regulation and control method for the flexible load cluster participation demand response, according to the historical data of flexible loads such as air conditioner loads and interruptible loads, short-term prediction is carried out on future flexible loads by considering factors such as temperature, meteorological factors and date types, so that a flexible load prediction curve is obtained. And formulating an optimized scheduling strategy according to the flexible load prediction curve predicted in a short period and the operation strategy. Meteorological factors include historical maximum temperature, minimum temperature, average temperature, wind speed; date type factors include weekdays, weekends, and holidays. According to the influence factors, the future load is predicted in a short period, a flexible load prediction curve is obtained, and a strategy corresponding to resident load participation needs is formulated, wherein the method mainly comprises the following steps:
further, in step 1, data of influence factors related to the flexible load are obtained from a weather forecast official network, a power dispatching mechanism or a supervision mechanism, and the like, and are used as experimental data to test the prediction effect of different models. Including various types of detailed data such as flexible load, temperature, wind speed, date type, etc.
Further, in step 2, in order to obtain meaningful data features, all the features mentioned above are visualized and the data distribution is observed. All the data features can be divided into continuous variable features and discrete variable features, and corresponding data features are selected as main features of the input model according to actual conditions.
Further, the data related to the selected flexible load prediction is preprocessed, and the preprocessing method is normalization. The normalized transformation formula is:
wherein: x is x i ' and x i The i-th value and the x-th value before and after normalization respectively max And x min The maximum and minimum of the sequence, respectively. Such normalization can normalize all values to a range of 0 to 1.
Further, in step 3, gating the cyclic units (Gate Recurrent Unit, GRU) as one of the cyclic neural networks, each of which adaptively captures the dependency of different time scales. Activation of GRU at time tIs previously activated->And candidate activation->Linear interpolation between:
wherein, the door is updatedDetermining how much the unit updates its activation or content. The update gate is calculated by equation (3).
The process of taking a linear sum between the existing state and the newly calculated state is similar to an LSTM cell. However, the GRU does not have any mechanism to control the extent of exposure of its state, but exposes the entire state each time.
Candidate activating factorsThe calculation method of (2) is similar to that of the traditional circulation unit, as shown in the formula (4)
Wherein r is t Is a set of reset gates, as is the logical operator, representing an exclusive OR operation. When closed (r t j Near 0), the reset gate effectively uses the cell gate and reads the first symbol of the input sequence while allowing the state previously calculated to be forgotten.
Similar to the update gate calculation method, reset gate r t j The calculation of (2) is shown in the formula (5).
r t j ＝σ(W r x t +U r h t-1 ) j (5)
Further, in step 4, parameters of the GRU model are set, and optimal parameters are selected so that the GRU model is in optimal super parameters.
Further, in step 5, training the model, selecting the first 90% of flexible load data for training, and allowing the model to learn rule parameters; the remaining 10% of the data was used as a test set to evaluate the model. And inputting the selected characteristics into the model for prediction to obtain a flexible load prediction curve.
Further, in step 6, according to the flexible load prediction curve, judging whether each adjustable flexible load unit can meet the target regulation and control amount; if the target regulation and control quantity is met, selecting an adjustable flexible load unit with the regulation and control potential closest to the target regulation and control quantity for regulation and control; and if the target regulation and control quantity is not met, sequencing the adjustable flexible load units according to the regulation and control potential, and sequentially selecting the flexible load unit with the largest regulation and control potential to participate in control until the regulation and control quantity reaches the target regulation and control quantity.
Meanwhile, the invention provides a flexible load cluster participation demand response auxiliary system, which comprises: the terminal response layer comprises a flexible load unit, the intermediate control layer comprises a distributed controller and is used for carrying out short-term prediction on future loads according to historical data of the flexible loads, temperature, weather date types and the like to obtain flexible load prediction curves, the flexible load prediction curves are uploaded to the system architecture layer, the system architecture layer comprises an electric power service cloud platform, the electric power service cloud platform formulates an optimal scheduling strategy according to the flexible load prediction curves uploaded by the distributed controller and an operation strategy and sends the optimal scheduling strategy to the intermediate control layer, and the intermediate control layer schedules the flexible load unit according to the optimal scheduling strategy. The terminal response layer is also used for uploading the unit work information to the middle control layer in real time, the distributed controller is also used for storing the unit work information uploaded by the terminal response layer, and sequencing the flexible load units after real-time analysis and calculation of the unit demand response capacity to obtain a time sequence of the power load.
The regulation and control method for the flexible load cluster to participate in the demand response has the beneficial effects that: 1. aiming at the problem that the user preference of a single-machine group system is difficult to consider in the regulation and control of the flexible load cluster participation demand response, flexible and accurate regulation and control of the flexible load cluster in a transformer area cannot be realized, the embodiment obtains a flexible load prediction curve through a short-term load prediction technology, and realizes the optimization division and accurate regulation and control of a large-scale flexible load cluster by utilizing a three-level layered auxiliary regulation and control system; 2. by utilizing the variable power property of the flexible load and combining the flexible load cluster participation power system strategy based on short-time power prediction, under the condition of considering personal electricity preference of users and peak clipping requirements of a power system, the flexible load cluster in each platform area is effectively optimally scheduled, the utilization rate of the flexible load cluster is improved, the peak-valley difference of the power grid load is greatly reduced, and the power grid load and the power system are enabled to run more stably.
Drawings
FIG. 1 is a flow chart of a flexible load cluster participation demand response provided by an embodiment of the invention;
FIG. 2 is a schematic diagram of a flexible load cluster participation demand side response auxiliary system according to an embodiment of the present invention;
FIG. 3 is a flowchart of optimizing demand response scheduling according to an embodiment of the present invention.
Detailed Description
The technical solutions of the inventive embodiments of the present invention will be explained and illustrated below with reference to the drawings of the inventive embodiments of the present invention, but the following embodiments are only preferred embodiments of the inventive embodiments of the present invention, not all. Based on the examples in the implementation manner, other examples obtained by a person skilled in the art without making any inventive effort fall within the scope of protection created by the present invention.
Referring to fig. 1-3, the regulation method for the flexible load cluster participation demand response comprises the following steps:
1. data selection
2. Data preprocessing
3. Model selection
4. Super parameter selection
5. Model training and prediction
6. Formulating an optimized scheduling policy
Examples
According to the method, the flexible load clusters of Zhejiang are regulated and controlled to participate in demand response. The method is divided into two parts, firstly, flexible load clusters are predicted, and then flexible load participation demand response strategies are formulated.
1. Prediction of flexible load clusters
(1) Data selection
And (3) selecting related data sets such as flexible loads and weather of 7 months and 8 months of 2022 of Zhejiang as experimental data, and considering core characteristics such as temperature, weather, date type and the like in a characteristic range. Flexible load data during days 7, 1 and 8, 31 of Zhejiang 2022 are selected.
(2) Data preprocessing
Identifying abnormal data, processing the abnormal data, and replacing the missing data and the abnormal data with the data of the next time point; and normalizing the data, and limiting the numerical value of the flexible load data between 0 and 1 by adopting Min-Max normalization, so as to avoid the local optimum of the sinking operation, wherein the normalization of the data is shown in a formula (1).
(3) Model selection
Six different models are selected for comparison, and the influence of the different prediction models on the prediction effect is explored. In this experiment, load prediction was performed using a GRU model with flexible load data of 7 to 9 months of 2022, which was the prediction target.
(4) Super parameter setting
Table 1 is the settings of all the hyper-parameters for the different models.
TABLE 1GRU super parameter settings
(5) Model training and load prediction
The model selects flexible load characteristic codes and extracts for 48 hours; the electricity price for the next 24 hours is selected as the predicted value. Table 2 shows the predicted performance index of the GRU model,
TABLE 2 comparison of predictive performance of different models under Spain-E dataset
2. Formulating resident flexible load participation demand response strategy
The flexible load clusters receive the dispatching tasks of the power company, and the distributed flexible load clusters are reasonably aggregated according to the economy and the load capacity of the flexible load clusters. In the prior art, the regulation and control requirements are mainly provided by the flexible load clusters according to the economic indexes and the load capacity of the flexible load clusters. And the power company dispatching center makes dispatching plans such as the participated peak clipping time length, the peak clipping number and the like according to the load prediction result. In practical application, the economy of the flexible load cluster is taken as a starting point, and the large-area flexible load is integrally and intensively regulated.
Referring to fig. 1, the embodiment of the invention provides a flexible load cluster participation demand response regulation and control method, a flexible load prediction curve is obtained through a short-term load prediction technology, an active load regulation control strategy is provided, the running stability of an electric power system is improved, and peak-to-valley load difference is effectively reduced. Specifically, the flexible load cluster participation demand response regulation and control method comprises the following steps:
and formulating an optimized scheduling strategy according to the flexible load prediction curve predicted in a short period and the operation strategy. The operation strategy comprises the steps of combining the power grid load capacity of the transformer area, personal electricity preference of a user, peak clipping requirement of the power grid load, meeting the target regulation and control quantity and achieving the peak clipping effect. The specific content of the optimized scheduling strategy comprises the following steps: judging whether each adjustable flexible load unit can meet the target regulation and control amount according to the flexible load prediction curve; if the target regulation and control quantity is met, selecting an adjustable flexible load unit with the regulation and control potential closest to the target regulation and control quantity for regulation and control; and if the target regulation and control quantity is not met, sequencing the adjustable flexible load units according to the regulation and control potential, and sequentially selecting the flexible load unit with the largest regulation and control potential to participate in control until the regulation and control quantity reaches the target regulation and control quantity.
Referring to fig. 2, the embodiment of the invention further provides a flexible load cluster participation demand side response auxiliary system, which comprises a terminal response layer, an intermediate control layer and a system architecture layer.
The terminal response layer comprises a flexible load and is used for uploading unit working information to the intermediate control layer in real time, wherein the unit working information comprises real-time power of a compressor, working frequency, user set temperature, required response control times, required response control variables, indoor real-time temperature, outdoor real-time temperature, a flexible load working state zone bit and the like.
The intermediate control layer comprises a distributed controller, is used for storing unit work information uploaded by the terminal response layer, and sequencing flexible load units after real-time analysis and calculation of unit demand response capacity to obtain a time sequence of electricity utilization loads, coupling the time sequence of the electricity utilization loads with meteorological factors according to historical data of resident flexible loads, carrying out short-term prediction on future loads to obtain flexible load prediction curves, and uploading platform area data to the system architecture layer, wherein the platform area data comprises flexible load prediction curves, is ready to respond to power system scheduling instructions at any time, and feeds back response results to the system architecture layer in real time. The system architecture layer comprises an electric power service cloud platform, wherein the electric power service cloud platform formulates an optimal scheduling strategy according to platform region data and an operation strategy uploaded by the distributed controller, and transmits an electric power system scheduling instruction to the intermediate control layer, and the intermediate control layer schedules the flexible load unit according to the optimal scheduling strategy.
Referring to fig. 3, a specific operation flow of the flexible load cluster participating in the demand side response is as follows:
s1: inputting target regulation and control quantity.
S2: it is determined whether to enter a controlled period. In the case of not entering the controlled period, then executing
S3: the normal control mode is performed. In case of entering a controlled period, then executing
S4: and judging whether each adjustable flexible load unit can meet the regulation and control requirements according to the flexible load prediction curve. In the case that not every adjustable flexible load unit can meet the regulation and control requirements, then
S5: and sequencing the flexible load units according to the mode of regulating and controlling potential from small to large.
S6: and sequentially selecting the flexible load units with the maximum regulating potential to participate in control until the regulating quantity reaches the target regulating quantity.
S7: it is determined whether the next controlled period is reached. In the event that the next controlled period is reached, execution returns to S4. In the case where the next controlled period is not reached, S3 is performed.
Under the condition that each adjustable flexible load unit can meet the regulation and control requirement, executing S8: and sequentially selecting the adjustable flexible load units with maximum regulating potential and not reaching the upper limit of the controlled times to perform out-of-line regulation until the regulating quantity reaches the target regulating quantity. After execution is completed S8, S7 is executed.
The detailed mathematical model of the regulation strategy is as follows:
setting a target regulation and control quantity G, adjusting the number n of flexible load units and adjusting the load regulation and control potential [ G ] 1 ,g 2 ,g 3 ......g n ]. When the target modulation amount is less than the modulation potential of the single tunable flexible load unit: g is less than or equal to max [ G ] n ]n＝1,2,3……n。
Selecting a controlled flexible load cell g n The method meets the following conditions: min [ g ] n -G]Gtoreq 0 n=1, 2,3 … … n. At the moment, any controlled flexible load unit which accords with regulation and control is selected.
When the target regulatory quantity is greater than the regulatory potential of a single tunable flexible load unit: g is greater than or equal to max [ G ] n ]n＝1,2,3……n。
To the controlled unit [ g ] 1 ,g 2 ,g 3 ......g n ]Ordered from big to small: to obtain [ p ] 1 ,p 2 ,p 3 ......p n ]。
The adjustable flexible load units with the maximum regulating potential are sequentially selected to participate in control until the regulating quantity reaches the target regulating quantity, so that the peak clipping purpose is achieved by a small amount of effective adjustable flexible load units:
at this time, the controlled flexible load unit is [ p ] 1 ,p 2 ,p 3 ......p i ]。
While the applicant has described and illustrated the embodiments of the present invention in detail with reference to the drawings, it should be understood by those skilled in the art that the above embodiments are only preferred embodiments of the present invention, and the detailed description is only for the purpose of helping the reader to better understand the spirit of the present invention, and not to limit the scope of the present invention, but any improvements or modifications based on the spirit of the present invention should fall within the scope of the present invention.
Claims (7)
1. A method for regulating and controlling participation demand response of a flexible load cluster, the method comprising the steps of:
step 1, selecting data;
step 2, data preprocessing;
step 3, selecting a model;
step 4, super parameter selection;
step 5, model training and prediction;
and 6, formulating an optimized scheduling strategy.
2. The method for regulating and controlling the flexible load cluster to participate in demand response according to claim 1, wherein in step 1, data of influence factors related to the flexible load are obtained from a weather forecast official network, a power dispatching mechanism or a supervision mechanism, etc., and are used as experimental data to test the prediction effect of different models, including detailed data of various types such as flexible load, temperature, wind speed, date type, etc.
3. The method according to claim 1, wherein in step 2, in order to obtain meaningful data features, all the features are visualized and the data distribution is observed, all the data features are divided into continuous variable features and discrete variable features, the corresponding data features are selected as the main features of the input model according to the actual situation,
preprocessing the data related to the selected flexible load prediction, wherein the preprocessing method is normalization, and the normalization transformation formula is as follows:
wherein: x is x i ' and x i The i-th value and the x-th value before and after normalization respectively max And x min Such normalization can normalize all values to a range of 0 to 1, respectively the maximum and minimum of the sequence.
4. The flexible load cluster participation demand response regulation method according to claim 1, wherein in the step 3, a gating loop unit (Gate Recurrent Unit, GRU) is used as one of the loop neural networks, each loop unit thereof adaptively captures the dependency of different time scales,
activation of GRU at time tIs previously activated->And candidate activation->Linear interpolation between:
wherein, the door is updatedDetermining how much the unit updates its activation or content, the update gate is calculated by equation (3),
the process of taking a linear sum between the existing state and the newly calculated state is similar to an LSTM cell, however, the GRU has no mechanism to control the exposure of its state, but exposes the entire state each time,
candidate activating factorsThe calculation method of (2) is similar to that of the traditional circulation unit, as shown in the formula (4)
Wherein r is t Is a set of reset gates, the logical operator, represents an exclusive OR operation, when turned off (r t j Near 0), the reset gate effectively uses the cell gate and reads the first symbol of the input sequence, while allowing the state previously calculated to be forgotten,
reset gate r t j The calculation of (2) is shown in the formula (5).
r t j ＝σ(W r x t +U r h t-1 ) j (5)
5. The method for regulating and controlling the participation demand response of the flexible load cluster according to claim 1, wherein in the step 4, parameters of the GRU model are set, and optimal parameters are selected so that the GRU model is in optimal super parameters.
6. The method for regulating and controlling the flexible load clusters to participate in demand response according to claim 1, wherein in the step 5, the model is trained, and the first 90% of flexible load data is selected for training, so that the model learns rule parameters; and taking the rest 10% of data as a test set, evaluating the model, and inputting the selected characteristics into the model for prediction to obtain a flexible load prediction curve.
7. The method for regulating and controlling the flexible load clusters to participate in demand response according to claim 1, wherein in the step 6, whether each adjustable flexible load unit can meet the target regulating and controlling amount is judged according to a flexible load prediction curve; if the target regulation and control quantity is met, selecting an adjustable flexible load unit with the regulation and control potential closest to the target regulation and control quantity for regulation and control; if the target regulation and control quantity is not met, sequencing the adjustable flexible load units according to the regulation and control potential, and sequentially selecting the flexible load unit with the largest regulation and control potential to participate in control until the regulation and control quantity reaches the target regulation and control quantity;
meanwhile, the invention provides a flexible load cluster participation demand response auxiliary system, which comprises: the terminal response layer comprises a flexible load unit, the intermediate control layer and a system architecture layer, the intermediate control layer comprises a distributed controller and is used for carrying out short-term prediction on future loads according to historical data of the flexible loads, temperature, weather date types and the like, obtaining flexible load prediction curves and uploading the flexible load prediction curves to the system architecture layer, the system architecture layer comprises an electric power service cloud platform, the electric power service cloud platform formulates an optimal scheduling strategy according to the flexible load prediction curves and operation strategies uploaded by the distributed controller and issues the optimal scheduling strategy to the intermediate control layer, the intermediate control layer is used for scheduling the flexible load unit according to the optimal scheduling strategy, the terminal response layer is also used for uploading unit work information to the intermediate control layer, the distributed controller is also used for storing the unit work information uploaded by the terminal response layer, and sequencing the flexible load units after carrying out real-time analysis and calculation on unit demand response capacity, so as to obtain a time sequence of electricity loads.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310562762.XA CN116826704A (en) | 2023-05-18 | 2023-05-18 | Regulation and control method for participation of flexible load cluster in demand response |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310562762.XA CN116826704A (en) | 2023-05-18 | 2023-05-18 | Regulation and control method for participation of flexible load cluster in demand response |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116826704A true CN116826704A (en) | 2023-09-29 |
Family
ID=88126579
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310562762.XA Pending CN116826704A (en) | 2023-05-18 | 2023-05-18 | Regulation and control method for participation of flexible load cluster in demand response |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116826704A (en) |
-
2023
- 2023-05-18 CN CN202310562762.XA patent/CN116826704A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107292766B (en) | Wind power consumption-oriented power system peak regulation means economical evaluation method and system | |
| CN109886567B (en) | Short-term load prediction method considering somatosensory temperature and radiation intensity | |
| CN104037761B (en) | AGC power multi-objective random optimization distribution method | |
| CN112686571B (en) | Comprehensive intelligent energy optimization scheduling method and system based on dynamic adaptive modeling | |
| Ju et al. | Hierarchical control of air-conditioning loads for flexible demand response in the short term | |
| CN113158450A (en) | Building energy management system-based economic scheduling method and system | |
| CN112001523A (en) | Comprehensive energy multi-scale optimization scheduling control method and system considering multiple energy storages | |
| CN114322208A (en) | Intelligent park air conditioner load regulation and control method and system based on deep reinforcement learning | |
| CN116488223A (en) | Household light-storage-flexible double-layer multi-time scale control method, device and medium | |
| CN116976513A (en) | Method for optimizing aggregators under different types of load participation conditions | |
| CN116960985A (en) | Method for evaluating demand response potential of virtual power plant of power system | |
| CN115115145B (en) | Demand response scheduling method and system for distributed photovoltaic intelligent residence | |
| CN116826704A (en) | Regulation and control method for participation of flexible load cluster in demand response | |
| Molina et al. | Approach to multivariable predictive control applications in residential HVAC direct load control | |
| CN113609778B (en) | Multi-objective optimization method and system for comprehensive energy system | |
| CN113487131B (en) | Station network interaction method based on charging demand and energy storage adjustment | |
| CN109767037A (en) | A kind of prediction technique of summer Daily treatment cost | |
| Pang et al. | A bi-objective low-carbon economic scheduling method for cogeneration system considering carbon capture and demand response | |
| CN114372645A (en) | Energy supply system optimization method and system based on multi-agent reinforcement learning | |
| CN115051374A (en) | Control method and device for electric heating equipment to participate in electric power peak regulation and storage medium | |
| Mashapu et al. | The Need for Artificial Intelligence for Energy-Efficiency Management: A Review | |
| CN113904331B (en) | Auxiliary regulation and control method, device and system for variable-frequency air conditioner cluster participation power system | |
| CN117559464B (en) | Power demand response regulation and control method, system, electronic equipment and medium | |
| CN117578534B (en) | Scheduling method, device, equipment and storage medium of photovoltaic energy storage system | |
| CN115840986B (en) | Energy management method based on stochastic model predictive control |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |