CN117674098A - Multi-element load space-time probability distribution prediction method and system for different permeability - Google Patents
Multi-element load space-time probability distribution prediction method and system for different permeability Download PDFInfo
- Publication number
- CN117674098A CN117674098A CN202311621657.5A CN202311621657A CN117674098A CN 117674098 A CN117674098 A CN 117674098A CN 202311621657 A CN202311621657 A CN 202311621657A CN 117674098 A CN117674098 A CN 117674098A
- Authority
- CN
- China
- Prior art keywords
- data
- power supply
- grid
- space
- load
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000035699 permeability Effects 0.000 title claims abstract description 27
- 230000002776 aggregation Effects 0.000 claims abstract description 8
- 238000004220 aggregation Methods 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 230000001131 transforming effect Effects 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 8
- 230000001502 supplementing effect Effects 0.000 claims description 6
- 238000013135 deep learning Methods 0.000 claims description 4
- 230000004931 aggregating effect Effects 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 2
- 230000004044 response Effects 0.000 abstract description 5
- 239000011159 matrix material Substances 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 6
- 239000013598 vector Substances 0.000 description 6
- 230000003238 somatosensory effect Effects 0.000 description 4
- 210000004027 cell Anatomy 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- 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
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a multi-element load space-time probability distribution prediction method and system for different permeability, which relate to the technical field of intelligent power grids and comprise the following steps: collecting power data and related information data sets of each power supply station area, and preprocessing; constructing a space grid according to the power supply grids, transforming the preprocessed power data and related information data sets according to the structure of the space grid, and extracting the data characteristic groups of each power supply grid; adopting a conv-LSTM network to encode the data feature group sequences of all power supply grids to obtain encoded data; carrying out space aggregation on the coded data of each grid in the space grid by adopting a self-attention module to obtain aggregated data; and inputting the aggregated data into a probability prediction model to obtain the discrete load probability distribution of each power supply grid. The invention fully considers uncertainty, is more suitable for strong randomness scenes of multiple loads, and can provide more reliable schedulable capacity information for the response of the demand side.
Description
Technical Field
The invention relates to the technical field of smart grids, in particular to a multi-element load space-time probability distribution prediction method and system for different permeability.
Background
In recent years, as the permeability of new energy is continuously improved, the contradiction between electricity supply and demand is increasingly prominent. The demand side response is used as an important novel power system regulation and control means, and the contradiction between power supply and demand can be greatly relieved. The load prediction is an important component in an intelligent traction novel electric power system technology system. Under the novel power system, the power load presents a new trend: load structure diversification, load spike, load uncertainty improvement and the like. This trend of variation places higher demands on the load prediction technique.
The load prediction of the traditional power system is mainly deterministic prediction, and is a single-point expected value of renewable energy power at a certain moment, and the load prediction is mainly calculated according to the theories of regression analysis, gray systems, expert systems and the like. In recent years, with the improvement of computing power of computers, algorithms such as support vector machines, ensemble learning, deep learning and the like are widely applied to prediction of loads of power systems. However, deterministic predictions, represented by point predictions, are difficult to reflect the uncertainty and randomness of the electrical load. How to extract more reliable schedulable capacity information from the demand side response is a matter that the skilled person is urgent to solve.
Disclosure of Invention
In view of the above, the invention provides a multi-load space-time probability distribution prediction method and system for different permeabilities, which considers the influence of various factors on the load of a power system, has high precision and strong applicability, and can provide space-time load prediction with uncertain information for the response of a demand side.
In order to achieve the above object, the present invention provides the following technical solutions:
a multi-element load space-time probability distribution prediction method for different permeability comprises the following specific steps:
collecting power data and related information data sets of each power supply station area, and preprocessing the power data and the related information data sets;
constructing a space grid according to the power supply grids, transforming the preprocessed power data and related information data sets according to the structure of the space grid, and extracting the data characteristic groups of each power supply grid;
adopting a conv-LSTM network to encode the data feature group sequences of all power supply grids to obtain encoded data;
spatially aggregating the encoded data of each grid in the spatial grid by using a position-sensitive self-attention module to obtain aggregated data;
and inputting the aggregated data into a probability prediction model to obtain the discrete load probability distribution of each power supply grid.
Optionally, the power data and related information data sets of each power supply station area include historical load, time information, calendar information and weather information.
Optionally, the pretreatment is as follows: defining any piece of missing data in the power data and related information data set as i; and supplementing the missing data i according to the scene similarity of the missing data i and other data by adopting a hot card filling method.
Optionally, the step of acquiring the data feature group of each power supply grid includes:
sequencing each power supply grid according to the longitude and latitude of the center of the region to obtain the horizontal space grid position of the power supply grid;
calculating the renewable energy permeability of each power supply grid according to the power data of each power supply grid;
calculating a temperature prediction sequence of the temperature of each power supply grid according to weather information;
and establishing a data characteristic set of each power supply grid based on the renewable energy permeability, the temperature sensing prediction sequence, the power data and the related information data set of each power supply grid, wherein the data characteristic set comprises historical load, power supply grid position, renewable energy permeability at the predicted time, month and day information, predicted target time, predicted time, holiday information and predicted time temperature sensing prediction value.
Optionally, the step of obtaining the encoded data includes: and (3) encoding the data feature group sequences of the power supply grids by adopting a conv-LSTM network, and extracting load association information at different times.
Optionally, the step of acquiring the aggregate data includes: based on the coded data, load related information of different spatial positions is extracted by adopting a position sensitive self-focusing module, and the expression is as follows:
wherein m×n represents all spatial positions, softmax p Representing a softmax function operation in all spatial locations; q o Representing query terms, k p Representing a key item; v p And representing a value item, wherein T represents characteristic information of each power supply grid.
Optionally, the training steps of the probability prediction model are as follows:
constructing a probability prediction model based on a deep learning neural network of error forward propagation;
and carrying out parameter optimization on the probability prediction model according to the expected value difference value of the actual measured load and the predicted load.
A multi-element load space-time probability distribution prediction system for different permeabilities, comprising:
and a data acquisition module: the power supply station is used for collecting power data and related information data sets of each power supply station area;
and a pretreatment module: the system is used for supplementing missing data to the power data and related information data sets;
the space gridding module is used for constructing a space grid according to the power supply grids, and converting the preprocessed power data and related information data sets according to the structure of the space grid to obtain data characteristic groups of each power supply grid;
the coding module is used for coding the data feature group sequences of the power supply grids by adopting a conv-LSTM network to obtain coded data;
location-sensitive self-attention module: the method comprises the steps of performing space aggregation on coded data of each grid in space grids to obtain aggregated data;
load distribution prediction module: the method is used for obtaining the discrete load probability distribution of each power supply grid through aggregation data according to the probability prediction model.
Compared with the prior art, the multi-element load space-time probability distribution prediction method and system for different permeability are provided, various factors are fully considered, a thermal card filling method is used for filling a data set, the data integrity is ensured, the data quality is ensured, and the follow-up prediction result is more scientific and accurate. By using high-dimensional data, the influence of various factors on the power load is considered, so that the prediction model is higher in robustness and less sensitive to noise data. The influence of time sequence and space distribution relation on the result is considered by using the space grid transformation and time coding modes, so that the prediction accuracy is higher. The position-sensitive self-focusing module is adopted, so that the calculation complexity is greatly reduced, and the model operation efficiency is improved. The final output is probability distribution of the power load, uncertainty is fully considered, the method is more suitable for a strong randomness scene of a multi-element load, and more reliable schedulable capacity information can be provided for response of a demand side.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of the method of the present invention;
FIG. 2 is a schematic diagram showing the effect of the method of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The embodiment of the invention discloses a multi-element load space-time probability distribution prediction method for different permeabilities, which comprises the following specific steps as shown in figure 1:
s1, collecting power data and related information data sets of each power supply station area, and preprocessing the power data and the related information data sets;
s2, constructing a space grid according to the power supply grids, and transforming the preprocessed power data and related information data sets according to the structure of the space grid to obtain data characteristic groups of all the power supply grids;
s3, adopting a conv-LSTM network to encode the data feature group sequences of all power supply grids to obtain encoded data;
s4, spatially aggregating the coded data of each grid in the spatial grid by adopting a position-sensitive self-attention module to obtain aggregated data;
s5, outputting the discrete load probability distribution of each power supply grid according to the aggregated data.
In this embodiment, the specific steps of S1 are as follows:
s11, acquiring power data and related information data sets of the concerned region, wherein the data sets comprise time information (corresponding to data time), calendar information (whether special days and specific days in a week) and weather information (temperature, air humidity, wind speed, somatosensory temperature and the like). Wherein the somatosensory temperature is obtained by the following formula:
wherein AT (DEG C) is the somatosensory temperature, T (DEG C) is the temperature, and ws (m/s) is the wind speed.
S12, supplementing the missing data by using the similarity between the missing data sample in the data set and other samples for the data obtained in the S11.
The day type similarity calculation formula of the ith sample and other samples is as follows:
m id ＝1-|f(X i )-f(X d )|；
wherein m is id The day type correlation degree of the ith sample and the ith sample is represented, and the function f represents the specific date X in one week i The mapping relation with the numerical values can be obtained from table 1.
TABLE 1
Special days refer to specific dates of significantly different loads. Mainly comprises the occurrence time of emergency such as a certain holiday and other special event, such as a primordial day, a spring festival, a national celebration or an earthquake, fire disaster and the like. The similarity of the ith sample and other samples on a special day can be given by four cases: the same special day; all special days but different types of special days; the same is not a special day; one special day, one non-special day.
The calculation formula of the similarity between the ith sample and the date of other samples is as follows:
wherein m is id The date distance similarity between the ith sample and the d sample; k is the date distance between the ith sample and other samples; beta is an attenuation coefficient, the value is between 0.90 and 0.98, and t is a date distance threshold; a is the lowest similarity of the factors, and a is less than or equal to beta t 。
Each weather information in the sample is arranged to form a weather feature vector X and normalized, and the expression is:
X′ i (k)＝[X i (k)-X i min (k)]/[X i max (k)-X i min (k)]；
wherein X 'is' i To normalize the post-vector, X i min And X i max Maximum and minimum values for the kth feature component of all data. The similarity of the weather features of the ith sample and other samples can be obtained by a gray correlation analysis method, and the expression is as follows:
wherein X 'is' i (k) Representing the kth component of the weather feature vector of the ith sample after normalization; ρ is [0,1 ]]The resolution coefficient is usually 0.5; epsilon d (k) A gray correlation coefficient between the kth component of the weather feature vector of the ith sample and the kth component of the weather feature vector of the d sample is represented, r is the number of the weather feature vectors selected, m id And (5) the date meteorological characteristic similarity of the ith sample and the d th sample.
The final obtained similarity is the product of the similarities, wherein N is the number of missing features in the predicted sample:
further, according to the calculation method, the similarity between other samples and the missing samples in the data set is obtained, n samples with the maximum similarity with the missing samples are selected, and one sample is randomly selected to fill the missing data.
In this embodiment, the S2 spatial grid transformation includes the following steps:
s21, sequencing N power supply grids concerned according to longitude and latitude of the center of the area, and sequentially marking the position of each power supply grid as (i, j), so that a position matrix with the size of m x N can be obtained. Where n=m×n.
S22, calculating the renewable energy permeability of each power supply grid, wherein the calculation formula is as follows:
p＝W n /W total ；
wherein p represents the permeability of renewable energy sources in a certain grid, W n Represents the energy generation capacity of renewable energy sources in the grid, W total Representing the total power generation in the grid.
S23, calculating a temperature prediction value of the temperature of each power supply grid in each period according to temperature and humidity information provided by weather forecast, wherein the calculation formula is as follows:
wherein AT ' (. Degree.C.) is the predicted somatosensory temperature, T ' (. Degree.C.) is the predicted temperature, ws ' (m/s) is the predicted wind speed;
and S24, according to the results of S22 and S23, combining the data in the data set to obtain characteristic information C= [ historical load, renewable energy permeability at the prediction time, prediction date and time information, prediction target time, prediction date and holiday information and prediction time temperature sensitivity prediction value ].
Further, the model input can be determined as [ m×n×t ij ]. Where i, j represents the calculated grid position (i, j), T ij Characteristic information representing a grid of positions (i, j).
In this embodiment, the result obtained in S2 is encoded in S3 using conv-LSTM as a temporal encoder. The convolution long-short-term memory network (conv-LSTM) realizes the time series prediction related to the space dimension by replacing the feedforward in the original LSTM network by the convolution.
The conv-LSTM core formula is as follows:
i t ＝σ(W xi *X t +W hi *H t-1 +W ci ·C t-1 +b i )；
f t ＝σ(W xf *X t +W hf *H t-1 +W cf ·C t-1 +b f )；
C t ＝f t ·C t-1 +i t ·tanh(W XC *X t +W hc *H t-1 +b c )；
o t ＝σ(W XO *X t +W ho *H t-1 +W co ·C t +b o )；
H t ＝o t ·tanh(C t )；
wherein, represents convolution operation; sigma represents a tanh or sigmoid function; x is X t Input of neurons at time t; i.e t ，f t Respectively controlling an input door and a forget door; o (o) t Is an output door; c (C) t The state of the neuron at the time t; w (W) xi ，W xf ，W xc ，W XO Representing the weights input to the respective gates; w (W) hi ，W hf ，W hc ，W ho Weights representing hidden states to the various gates; w (W) ci ，W cf ，W co Weights representing cell states to respective gates; b represents bias; h t Indicating the neuron output at time t. In the three gating units, the forgetting gate controls how much information is discarded/inherited from the previous memory; the input gate determines how much input information at the current moment is added into the memory information stream; the output gate determines the amount of information transferred to the hidden state in the memory cell at the current time. Each gate is covered by the hidden state H at the previous moment t-1 Cell state C t-1 And input X at the current time t Is composed of a linear combination of activation functions.
In this embodiment, in S4, the output in step S3 is aggregated using a location-sensitive self-focusing module. For a given training sample D N The aggregation at position o (i, j) can be expressed as:wherein the m×n set is all spatial positions, and the query term q o ＝W Q x o Key item k q ＝W K x o Value term v p ＝W V x o Is input x at o (i, j) o Linear transformation of W Q 、W K 、W V Is a matrix of parameters that can be learned. softmax p Representing a softmax function operation at all spatial locations. The module aggregates the global features of the locations and the convolution in S3 only extracts the context of the time dimension.
In the above embodiment, the inputs are established according to steps S2-S4Entering the matrix X, in step S5, a probabilistic predictive model is built, at a given training sample D N In (1) a set of parameters θ is found such that the objective functionThe minimum, i.e. the difference between the expected values of the measured and predicted loads is the minimum. The final output of the model is the load probability distribution at the predicted time. The probability distribution is discrete, at Load max ，Load min ]In the section, each load segment l with Deltal as the load segment step length i Is a probability size of (c).
Wherein D is N The space-time input matrix X, p (load|x) =dnn established for S2-S4 θ (X), p (load|X) is the probability distribution of the Load at the predicted time, DNN θ (X) is a deep learning neural network with H layers of error forward propagation, θ is a set of parameters that can be trained to learn, including the weights and bias of each layer of the network. By DNN θ (X) the output of the previous layer calculates the output of the next layer. Wherein the input of the 1 st layer is X, and the output of the h layer is a h ，a h ＝σ(W h a h-1 +b h )；W h For each layer of weight matrix, for each load segment l i The sum of the probability of (2) is 1; sigma is an activation function, and a Softmax function is adopted; b h For the bias matrix of each layer, multi-class cross entropy is used. The parameter set with the minimum training error can be solved by a gradient descent method. That is, the model can predict the target power supply area at the target time T according to the input matrix X x The implementation effect is shown in fig. 2, with probability of load distribution.
The invention also discloses a multi-element load space-time probability distribution prediction system for different permeabilities, which comprises the following steps:
and a data acquisition module: the power supply station is used for collecting power data and related information data sets of each power supply station area;
and a pretreatment module: the system is used for supplementing missing data to the power data and related information data sets;
the space gridding module is used for constructing a space grid according to the power supply grids, and converting the preprocessed power data and related information data sets according to the structure of the space grid to obtain data characteristic groups of each power supply grid;
the coding module is used for coding the data feature group sequences of the power supply grids by adopting a conv-LSTM network to obtain coded data;
location-sensitive self-attention module: the method comprises the steps of performing space aggregation on coded data of each grid in space grids to obtain aggregated data;
load distribution prediction module: the method is used for obtaining the discrete load probability distribution of each power supply grid through aggregation data according to the probability prediction model.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. A multi-element load space-time probability distribution prediction method for different permeability is characterized by comprising the following specific steps:
collecting power data and related information data sets of each power supply station area, and preprocessing the power data and the related information data sets;
constructing a space grid according to the power supply grids, transforming the preprocessed power data and related information data sets according to the structure of the space grid, and extracting the data characteristic groups of each power supply grid;
adopting a conv-LSTM network to encode the data feature group sequences of all power supply grids to obtain encoded data;
spatially aggregating the encoded data of each grid in the spatial grid by using a position-sensitive self-attention module to obtain aggregated data;
and inputting the aggregated data into a probability prediction model to obtain the discrete load probability distribution of each power supply grid.
2. The method of claim 1, wherein the power data and related information data sets of each power supply station area comprise historical load, time information, calendar information and weather information.
3. The method for predicting the multi-element load space-time probability distribution for different permeabilities according to claim 1, wherein the preprocessing is as follows: defining any piece of missing data in the power data and related information data set as i; and supplementing the missing data i according to the scene similarity of the missing data i and other data by adopting a hot card filling method.
4. The method for predicting multi-element load space-time probability distribution for different permeabilities according to claim 2, wherein the step of obtaining the data feature sets of each power supply grid is as follows:
sequencing each power supply grid according to the longitude and latitude of the center of the region to obtain the horizontal space grid position of the power supply grid;
calculating the renewable energy permeability of each power supply grid according to the power data of each power supply grid;
calculating a temperature prediction sequence of the temperature of each power supply grid according to weather information;
and establishing a data characteristic set of each power supply grid based on the renewable energy permeability, the temperature sensing prediction sequence, the power data and the related information data set of each power supply grid, wherein the data characteristic set comprises historical load, power supply grid position, renewable energy permeability at the predicted time, month and day information, predicted target time, predicted time, holiday information and predicted time temperature sensing prediction value.
5. The method for predicting multi-element load space-time probability distribution for different permeabilities according to claim 1, wherein the step of obtaining encoded data is: and (3) encoding the data feature group sequences of the power supply grids by adopting a conv-LSTM network, and extracting load association information at different times.
6. The method for predicting multi-element load space-time probability distribution for different permeabilities according to claim 1, wherein the step of acquiring the aggregate data is: based on the coded data, load related information of different spatial positions is extracted by adopting a position sensitive self-focusing module, and the expression is as follows:
wherein m×n represents all spatial positions, softmax p Representing a softmax function operation in all spatial locations; q o Representing query terms, k p Representing a key item; v p And representing a value item, wherein T represents characteristic information of each power supply grid.
7. The method for predicting multi-element load space-time probability distribution for different permeabilities according to claim 1, wherein the training step of the probability prediction model is as follows:
constructing a probability prediction model based on a deep learning neural network of error forward propagation;
and carrying out parameter optimization on the probability prediction model according to the expected value difference value of the actual measured load and the predicted load.
8. A multi-element load space-time probability distribution prediction system for different permeabilities, comprising:
and a data acquisition module: the power supply station is used for collecting power data and related information data sets of each power supply station area;
and a pretreatment module: the system is used for supplementing missing data to the power data and related information data sets;
the space gridding module is used for constructing a space grid according to the power supply grids, and converting the preprocessed power data and related information data sets according to the structure of the space grid to obtain data characteristic groups of each power supply grid;
the coding module is used for coding the data feature group sequences of the power supply grids by adopting a conv-LSTM network to obtain coded data;
location-sensitive self-attention module: the method comprises the steps of performing space aggregation on coded data of each grid in space grids to obtain aggregated data;
load distribution prediction module: the method is used for obtaining the discrete load probability distribution of each power supply grid through aggregation data according to the probability prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311621657.5A CN117674098B (en) | 2023-11-29 | Multi-element load space-time probability distribution prediction method and system for different permeability |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311621657.5A CN117674098B (en) | 2023-11-29 | Multi-element load space-time probability distribution prediction method and system for different permeability |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117674098A true CN117674098A (en) | 2024-03-08 |
| CN117674098B CN117674098B (en) | 2024-06-07 |
Family
ID=
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0720952A (en) * | 1993-06-23 | 1995-01-24 | Canon Inc | Power unit |
| JP2012125113A (en) * | 2010-12-10 | 2012-06-28 | Fujitsu Ltd | Electric power source switching method and electric power source switching device |
| EP2778824A2 (en) * | 2013-03-15 | 2014-09-17 | Omron Corporation | Power conditioner and program |
| CN112016734A (en) * | 2020-04-07 | 2020-12-01 | 沈阳工业大学 | Stack type self-coding multi-model load prediction method and system based on LSTM |
| AU2020104000A4 (en) * | 2020-12-10 | 2021-02-18 | Guangxi University | Short-term Load Forecasting Method Based on TCN and IPSO-LSSVM Combined Model |
| CN113673775A (en) * | 2021-08-25 | 2021-11-19 | 武汉科技大学 | Time-space combination prediction method based on CNN-LSTM and deep learning |
| CN114091766A (en) * | 2021-11-24 | 2022-02-25 | 东北电力大学 | Space load prediction method based on CEEMDAN-LSTM |
| US20220197233A1 (en) * | 2020-12-18 | 2022-06-23 | Wuhan University | Wind power prediction method and system for optimizing deep transformer network |
| WO2022241932A1 (en) * | 2021-05-21 | 2022-11-24 | 同济大学 | Prediction method based on non-intrusive attention preprocessing process and bilstm model |
| CN116014724A (en) * | 2023-01-16 | 2023-04-25 | 华北电力大学（保定） | Active power distribution network gridding load prediction method |
| CN116090602A (en) * | 2022-11-30 | 2023-05-09 | 国网内蒙古东部电力有限公司供电服务监管与支持中心 | Power load prediction method and system |
| KR20230089652A (en) * | 2021-12-14 | 2023-06-21 | 한국광기술원 | System for non-intrusive load monitoring of electronic device and method thereof |
| CN116388150A (en) * | 2023-02-15 | 2023-07-04 | 国网辽宁省电力有限公司电力科学研究院 | Matching short-term load prediction method based on digital twin and neural network technology |
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0720952A (en) * | 1993-06-23 | 1995-01-24 | Canon Inc | Power unit |
| JP2012125113A (en) * | 2010-12-10 | 2012-06-28 | Fujitsu Ltd | Electric power source switching method and electric power source switching device |
| EP2778824A2 (en) * | 2013-03-15 | 2014-09-17 | Omron Corporation | Power conditioner and program |
| CN112016734A (en) * | 2020-04-07 | 2020-12-01 | 沈阳工业大学 | Stack type self-coding multi-model load prediction method and system based on LSTM |
| AU2020104000A4 (en) * | 2020-12-10 | 2021-02-18 | Guangxi University | Short-term Load Forecasting Method Based on TCN and IPSO-LSSVM Combined Model |
| US20220197233A1 (en) * | 2020-12-18 | 2022-06-23 | Wuhan University | Wind power prediction method and system for optimizing deep transformer network |
| WO2022241932A1 (en) * | 2021-05-21 | 2022-11-24 | 同济大学 | Prediction method based on non-intrusive attention preprocessing process and bilstm model |
| CN113673775A (en) * | 2021-08-25 | 2021-11-19 | 武汉科技大学 | Time-space combination prediction method based on CNN-LSTM and deep learning |
| CN114091766A (en) * | 2021-11-24 | 2022-02-25 | 东北电力大学 | Space load prediction method based on CEEMDAN-LSTM |
| KR20230089652A (en) * | 2021-12-14 | 2023-06-21 | 한국광기술원 | System for non-intrusive load monitoring of electronic device and method thereof |
| CN116090602A (en) * | 2022-11-30 | 2023-05-09 | 国网内蒙古东部电力有限公司供电服务监管与支持中心 | Power load prediction method and system |
| CN116014724A (en) * | 2023-01-16 | 2023-04-25 | 华北电力大学（保定） | Active power distribution network gridding load prediction method |
| CN116388150A (en) * | 2023-02-15 | 2023-07-04 | 国网辽宁省电力有限公司电力科学研究院 | Matching short-term load prediction method based on digital twin and neural network technology |
Non-Patent Citations (4)
| Title |
|---|
| YUQIANG WANG: ""Load Forecasting Method of Integrated Energy System Based on CNN-BiLSTM with Attention Mechanism"", 《 2021 3RD INTERNATIONAL CONFERENCE ON SMART POWER & INTERNET ENERGY SYSTEMS (SPIES)》, 9 December 2021 (2021-12-09) * |
| 穆晨宇: ""基于VMD⁃LSTM⁃Attention模型的短期负荷预测研究"", 《现代电子技术》, 1 September 2023 (2023-09-01), pages 174 - 178 * |
| 童占北 等: ""基于相似日和CNN-LSTM的短期负荷预测"", 《电工电气》, 30 August 2022 (2022-08-30), pages 17 - 22 * |
| 谢建文;张元标;: "农村电力系统负荷预测方法研究", 安徽农业科学, no. 21, 20 July 2008 (2008-07-20) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Deep-learning-based probabilistic forecasting of electric vehicle charging load with a novel queuing model | |
| Ke et al. | Short-term electrical load forecasting method based on stacked auto-encoding and GRU neural network | |
| Xuan et al. | A multi-energy load prediction model based on deep multi-task learning and ensemble approach for regional integrated energy systems | |
| Rahman et al. | Predicting electricity consumption for commercial and residential buildings using deep recurrent neural networks | |
| CN107590567A (en) | A kind of Recognition with Recurrent Neural Network short-term load forecasting method based on comentropy cluster and notice mechanism | |
| CN108711847B (en) | A kind of short-term wind power forecast method based on coding and decoding shot and long term memory network | |
| CN108280551A (en) | A kind of photovoltaic power generation power prediction method using shot and long term memory network | |
| Du et al. | Power load forecasting using BiLSTM-attention | |
| CN110222901A (en) | A kind of electric load prediction technique of the Bi-LSTM based on deep learning | |
| CN113128113B (en) | Lean information building load prediction method based on deep learning and transfer learning | |
| CN113554466B (en) | Short-term electricity consumption prediction model construction method, prediction method and device | |
| CN106600041A (en) | Restricted Boltzmann machine-based photovoltaic power generation short-term power probability prediction method | |
| Liu et al. | Heating load forecasting for combined heat and power plants via strand-based LSTM | |
| CN112418526A (en) | Comprehensive energy load control method and device based on improved deep belief network | |
| CN116090602A (en) | Power load prediction method and system | |
| Li et al. | An effective deep learning neural network model for short‐term load forecasting | |
| CN117674098B (en) | Multi-element load space-time probability distribution prediction method and system for different permeability | |
| CN115481788B (en) | Phase change energy storage system load prediction method and system | |
| CN116404637A (en) | Short-term load prediction method and device for electric power system | |
| CN116402194A (en) | Multi-time scale load prediction method based on hybrid neural network | |
| CN117674098A (en) | Multi-element load space-time probability distribution prediction method and system for different permeability | |
| CN115456235A (en) | Photovoltaic power prediction system based on multi-mode fusion | |
| CN113537539B (en) | Multi-time-step heat and gas consumption prediction model based on attention mechanism | |
| Wang et al. | Optimization of Convolutional Long Short-Term Memory Hybrid Neural Network Model Based on Genetic Algorithm for Weather Prediction | |
| Manoj et al. | Deep Feature Selection for Wind Forecasting‐II |
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 | ||
| GR01 | Patent grant |