CN109034500A - A kind of mid-term electric load forecasting method of multiple timings collaboration - Google Patents
A kind of mid-term electric load forecasting method of multiple timings collaboration Download PDFInfo
- Publication number
- CN109034500A CN109034500A CN201811028611.1A CN201811028611A CN109034500A CN 109034500 A CN109034500 A CN 109034500A CN 201811028611 A CN201811028611 A CN 201811028611A CN 109034500 A CN109034500 A CN 109034500A
- Authority
- CN
- China
- Prior art keywords
- term
- short
- electric load
- memory network
- prediction model
- 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
- 238000013277 forecasting method Methods 0.000 title claims abstract description 15
- 230000015654 memory Effects 0.000 claims abstract description 49
- 230000005611 electricity Effects 0.000 claims abstract description 25
- 238000012549 training Methods 0.000 claims description 19
- 238000000034 method Methods 0.000 claims description 18
- 230000006870 function Effects 0.000 claims description 13
- 230000004913 activation Effects 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 241000208340 Araliaceae Species 0.000 claims 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims 1
- 235000003140 Panax quinquefolius Nutrition 0.000 claims 1
- 235000008434 ginseng Nutrition 0.000 claims 1
- 239000010410 layer Substances 0.000 description 22
- 230000006872 improvement Effects 0.000 description 4
- 210000004027 cell Anatomy 0.000 description 3
- 230000006403 short-term memory Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000003822 cell turnover Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000011229 interlayer Substances 0.000 description 1
- 238000013507 mapping Methods 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
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of mid-term electric load forecasting methods of multiple timings collaboration, it is characterised in that: establishes the internal multiple timings collaborative forecasting system in conjunction with surface.Including on the basis of common Power system load data, increase the surfaces such as date, temperature, festivals or holidays;Establish the double-deck long memory network prediction model in short-term;Long memory network prediction model in short-term is trained with multiple features combined data, by inputting tested region date, temperature, festivals or holidays parameter, obtains the tested region load forecast result.Long memory network in short-term is introduced into Load Forecasting by the present invention, by the long-acting memory characteristic of its timing, realizes the prediction to the daily total electricity consumption of following a period of time region electric load.And mid-term electric load forecasting method of the invention greatly enhances the accuracy and speed of prediction.The invention is for predicting mid-term electric load situation.
Description
Technical field
The present invention relates to technical field of power systems, are more precisely related to a kind of mid-term electric load forecasting method.
External temporal aspect data are specifically added on the basis of conventional electric power load data, with the double-deck long short-term memory of this data training
Network Prediction Model.
Background technique
A relatively reasonable reference point is provided in terms of distribution network planning of the medium term load forecasting for region, is directly affected
The investment trend of electric power resource, affects the upper limit of native national's economic development, can plan electricity with auxiliary power enterprise scien`
Power resource investment.Power grid day-to-day operation more payes attention to economic and safety, efficient and accurate, the accuracy of mid-term electric load forecasting
Higher mid-term power prediction more shows more and more important meaning to the development of power grid enterprises' electricity market.
The characteristics of load forecast, there is inaccuracy, conditionity, timeliness, multi-scheme, interdependency to make
The precision of load prediction is challenging work.Load forecasting method mainly has support vector machines, BP artificial neuron at present
The Traditional Methods such as network and artificial intelligence method.Long memory network category artificial intelligence method in short-term, has elimination non-linear and all kinds of non-
The ability for specifying factor, by quantifying various environmental variances, by kinds of abstract factor measure in the form of mathematical model with
Operation, having, which can be good at the memory capability exported in the past, solves complicated problem of nonlinear mapping.
Summary of the invention
For in load forecast, data accumulating depth problem, the present invention is devised one kind and is remembered in short-term based on the double-deck length
Recall the electric load mid-term multiple timings collaborative forecasting method of network.
The present invention is mainly realized by following scheme:
Step 1. establishes the double-deck long memory network prediction model in short-term;
Step 2. uses root mean square back-propagation algorithm, in conjunction with tested region electric load history daily total electricity consumption when
Ordinal number evidence, and temperature sequence data are added, and add festivals or holidays outside time series data to long short-term memory net in a manner of one-hot encoding
Network prediction model is trained to obtain the bias of every layer of prediction model, the activation letter of the long memory network prediction model in short-term of setting
Number, learning rate, regularization mode and batch size, obtaining the length by training study, memory network prediction model inputs in short-term
Inner link between output and between previous output.
Step 3. is according to determining length memory network prediction model in short-term, to the total electricity consumption of the reality of tested region electric load
Amount is predicted, is inputted tested region date, temperature, festivals or holidays parameter, is obtained the tested region load forecast result.
As a further improvement of the above technical scheme, the long memory network prediction model in short-term of bilayer described in step 1
Contain an input layer, two hidden layers containing LSTM block structure, the neural network of an output layer, the section of input layer
Point is 3, and output node layer is 1, time step 365, and each hidden layer node is 1095, and learning rate is set as
0.0006;
Two layers of hidden layer activation primitive of the double-deck long memory network prediction model in short-term all selects softsign function
To substitute tanh function.
As a further improvement of the above technical scheme, the step 2 the following steps are included:
Step 21. is according to the training process of long memory network in short-term, and memory network is pre- in short-term for the bilayer length established to step 1
It surveys model and carries out unsupervised training using tested period previous annual data, to obtain double-deck every layer of memory network model in short-term long
Parameter value;
Step 22. is finely adjusted using learning rate of the supervised learning method to the network;
With the long output of memory network prediction model target in short-term of bilayer for supervisory signals in the step 22, building logarithm is seemingly
Right loss function carries out Training to the network, final to determine the double-deck long memory network prediction model in short-term.
As a further improvement of the above technical scheme, the daily total electricity consumption time series data of history described in step 2 is quilt
1 year every per day electricity consumption data for surveying region electric load, when the bilayer grows Network Prediction Model in short-term before input vector
12 months every daily power consumptions, output vector are 13rd month every daily power consumption.
As a further improvement of the above technical scheme, in step 3, practical total use to tested region electric load
Electricity is predicted as predicting following 1 month daily practical total electricity consumption of tested region electric load.
Beneficial outcomes of the invention are: multiple timings collaborative forecasting network is introduced into electric load power prediction by the present invention
In, by the long-acting memory characteristic of its timing, learn the inner link between input and output and between previous output, realization pair
The prediction of the daily total electricity consumption of following a period of time region electric load.And mid-term electric load forecasting method of the invention is very
The accuracy and speed of prediction is improved in big degree.The invention is for predicting mid-term electric load situation.
Detailed description of the invention
Prediction technique flow chart Fig. 1 of the invention
Network structure Fig. 2 of the invention
Specific embodiment
In order to keep the purpose of the present invention, technical solution and summary of the invention clearer, below in conjunction with will in conjunction with the embodiments and
Attached drawing carries out clear, complete description to the technical effect of design of the invention, specific structure and generation, to be completely understood by this
Purpose, feature and the effect of invention.
Referring to Fig.1, the invention discloses a kind of mid-term electric load forecasting method of multiple timings collaboration, including following
Step:
Step 1. establishes the double-deck long memory network prediction model in short-term;
Step 2. uses root mean square back-propagation algorithm, in conjunction with tested region electric load history daily total electricity consumption when
Ordinal number evidence, and temperature sequence data are added, and add festivals or holidays outside time series data to long short-term memory net in a manner of one-hot encoding
Network prediction model is trained to obtain the bias of every layer of prediction model, the activation letter of the long memory network prediction model in short-term of setting
Number, learning rate, regularization mode and batch size, obtaining the length by training study, memory network prediction model inputs in short-term
Inner link between output and between previous output.
Step 3. is according to determining length memory network prediction model in short-term, to the total electricity consumption of the reality of tested region electric load
Amount is predicted, is inputted tested region date, temperature, festivals or holidays parameter, is obtained the tested region load forecast result.
Specifically, long memory network in short-term is introduced into Load Forecasting by the present invention, long-acting by its timing
Memory characteristic learns the inner link between input and output and between previous output, realizes to following a period of time region
The prediction of the daily total electricity consumption of electric load.And mid-term electric load forecasting method of the invention greatly enhance it is pre-
The accuracy and speed of survey.
The multiple timings collaboration, which refers to, is aided in step 2 temperature added by method for the timing electricity consumption data of input
Degree, whether festivals or holidays time series data, the three classes timing data double-deck long memory network prediction model in short-term of training jointly.
The long memory network prediction model in short-term of the bilayer contains an input layer, and two contain LSTM block structure
Hidden layer, the neural network of an output layer, the node of input layer is 3, and output node layer is 1, and time step is
365, each hidden layer node is 1095, and learning rate is set as 0.0006;
Two layers of hidden layer activation primitive of the double-deck long memory network prediction model in short-term all selects softsign function
To substitute tanh function.Softgsign function as shown in expression formula 1,
Specifically, the double-deck long memory network prediction model in short-term is stacked by two long memory network in short-term,
Its advantage is that can be in high-rise more abstract expression characteristic, and reduce the number of neuron, increase predictablity rate and drop
The low training time, as shown in Figure 2.And long memory network in short-term is improved by convolutional neural networks, each replicated blocks
In include input gate, forget door and out gate.
The working method for forgeing door is as shown in expression formula 2, ft=σ (Wf·[ht-1,xt]+bf), wherein ht-1In expression
The output of one cell, xtIndicate the input of current cell, σ indicates sigmod function.The working method of the input gate such as table
Up to shown in formula 3 to expression formula 5, it=σ (Wi·[ht-1,xt]+bi);
Wherein itDetermine which information needs more for sigmoid layers
Newly,For alternatively more new content, CtCurrently to input.The working method of the out gate is as shown in expression formula 6 to 7, ot=σ
(Wo·[ht-1,xt]+bo);ht=ot·tanh(Ct), otFor sigmoid layers of determining cell turnover part, htIt is determined for cell defeated
Out.
The root mean square back-propagation algorithm is a kind of autoadapted learning rate method, is calculated with window sliding weighted average
Second order momentum, more new formula are as shown in expression formula 8 and expression formula 9, η [g2]t=0.9 η [g2]t-1+0.1gt 2;
Wherein Θt+1For the learning rate of update.
It is further used as preferred embodiment, the invention is specifically had a try in method, and the step 2 includes following step
It is rapid:
Step 21. is according to the training process of long memory network in short-term, and memory network is pre- in short-term for the bilayer length established to step 1
It surveys model and carries out unsupervised training using tested period previous annual data, to obtain double-deck every layer of memory network model in short-term long
Parameter value;
Step 22. is finely adjusted using learning rate of the supervised learning method to the network;
With the long output of memory network prediction model target in short-term of bilayer for supervisory signals in the step 22, building logarithm is seemingly
Right loss function carries out Training to the network, final to determine the double-deck long memory network prediction model in short-term.
The double-deck long memory network prediction model in short-term is a raw forming model, by training the weight of an interlayer,
Entire neural network can be allowed to receive input and exported accordingly.This training process is divided into two stages: pre-training rank
Section and fine tuning parametric step.
The fine tuning parameter is to construct logarithm with the long output of memory network prediction model target in short-term of bilayer for supervisory signals
Feel relieved loss function, Training is carried out to the network.The loss function is as shown in expression formula 10, J (Y, P (Y | X))
=-logP (Y | X), wherein Y is output variable, and X is input variable.
It is further used as preferred embodiment, in the invention specific embodiment, history described in step 2 is daily
Total electricity consumption time series data is 1 year of tested region electric load per per day electricity consumption data, the double-deck long network in short-term
12 months every daily power consumptions before input vector when prediction model, output vector are 13rd month every daily power consumption.
It is further used as preferred embodiment, in the invention specific embodiment, in step 3, described pair tested
The practical total electricity consumption of region electric load is predicted as to tested region electric load following 1 month daily practical total use
Electricity is predicted.
Better embodiment of the invention is illustrated above, but the invention is not limited to the implementation
Example, those skilled in the art can also make various equivalent modifications on the premise of without prejudice to spirit of the invention or replace
It changes, these equivalent variation or replacement are all included in the scope defined by the claims of the present application.
Claims (5)
1. a kind of mid-term electric load forecasting method of multiple timings collaboration, which comprises the following steps:
Step 1: establishing the double-deck long memory network prediction model in short-term;
Step 2: use root mean square back-propagation algorithm, in conjunction with tested region electric load history daily total electricity consumption when ordinal number
According to and adding temperature sequence data, and it is pre- to long memory network in short-term to add in a manner of one-hot encoding time series data outside festivals or holidays
It surveys model and is trained to obtain the bias of every layer of prediction model, the long activation primitive of memory network prediction model in short-term of setting,
It is defeated to obtain length memory network prediction model input in short-term by training study for learning rate, regularization mode and batch size
Inner link between out and between previous output.
Step 3: according to determining length memory network prediction model in short-term, to the practical total electricity consumption of tested region electric load into
Row prediction, inputs tested region date, temperature, festivals or holidays parameter, obtains the tested region load forecast result.
2. a kind of mid-term electric load forecasting method of multiple timings collaboration according to claim 1, which is characterized in that step
The long memory network prediction model in short-term of bilayer described in 1 contains an input layer, and two implicit containing LSTMblock structure
Layer, the neural network of an output layer, the node of input layer are 3, and output node layer is 1, time step 365, each
Hidden layer node is 1095, and learning rate is set as 0.0006;
Two layers of hidden layer activation primitive of the double-deck long memory network prediction model in short-term all selects softsign function to replace
For tanh function.
3. a kind of mid-term electric load forecasting method of multiple timings collaboration according to claim 2, which is characterized in that described
Step 2 includes with following steps:
Step 21: according to the training process of long memory network in short-term, mould being predicted to the long memory network in short-term of bilayer that step 1 is established
Type carries out unsupervised training using previous annual data of tested period, to obtain double-deck long every layer of memory network model of the ginseng in short-term
Numerical value;
Step 22: being finely adjusted using learning rate of the supervised learning method to the network;
With the long output of memory network prediction model target in short-term of bilayer for supervisory signals in the step 22, building log-likelihood damage
Function is lost, Training is carried out to the network, it is final to determine the double-deck long memory network prediction model in short-term.
4. a kind of mid-term electric load forecasting method of multiple timings collaboration according to claim 3, it is characterised in that: step
The daily total electricity consumption time series data of history described in 2 is 1 year of tested region electric load per per day electricity consumption data, institute
12 months every daily power consumptions before input vector when stating the double-deck long Network Prediction Model in short-term, output vector are 13rd month per daily
Electricity.
5. a kind of mid-term electric load forecasting method of multiple timings collaboration according to claim 4, it is characterised in that: step
In 3, the practical total electricity consumption to tested region electric load is predicted as 1 month following to tested region electric load
Daily practical total electricity consumption is predicted.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811028611.1A CN109034500A (en) | 2018-09-04 | 2018-09-04 | A kind of mid-term electric load forecasting method of multiple timings collaboration |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811028611.1A CN109034500A (en) | 2018-09-04 | 2018-09-04 | A kind of mid-term electric load forecasting method of multiple timings collaboration |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109034500A true CN109034500A (en) | 2018-12-18 |
Family
ID=64624019
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811028611.1A Pending CN109034500A (en) | 2018-09-04 | 2018-09-04 | A kind of mid-term electric load forecasting method of multiple timings collaboration |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109034500A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110909941A (en) * | 2019-11-26 | 2020-03-24 | 广州供电局有限公司 | Power load prediction method, device and system based on LSTM neural network |
CN111178612A (en) * | 2019-12-19 | 2020-05-19 | 绍兴大明电力设计院有限公司 | LSTM load prediction method of grid user based on big data ODPS engine |
CN111241755A (en) * | 2020-02-24 | 2020-06-05 | 国网(苏州)城市能源研究院有限责任公司 | Power load prediction method |
CN111899123A (en) * | 2020-07-28 | 2020-11-06 | 深圳江行联加智能科技有限公司 | Electric quantity prediction method, electric quantity prediction device and computer readable storage medium |
CN115545362A (en) * | 2022-12-05 | 2022-12-30 | 南方电网数字电网研究院有限公司 | AI and TSD combined new energy medium-term power combined prediction method |
CN116361709A (en) * | 2023-03-31 | 2023-06-30 | 山东省计算中心(国家超级计算济南中心) | Self-adaptive power load identification method and device |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101383023A (en) * | 2008-10-22 | 2009-03-11 | 西安交通大学 | Neural network short-term electric load prediction based on sample dynamic organization and temperature compensation |
CN103295081A (en) * | 2013-07-02 | 2013-09-11 | 上海电机学院 | Electrical power system load prediction method based on back propagation (BP) neural network |
CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学(保定) | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
CN106952181A (en) * | 2017-03-08 | 2017-07-14 | 深圳市景程信息科技有限公司 | Electric Load Prediction System based on long Memory Neural Networks in short-term |
CN106960252A (en) * | 2017-03-08 | 2017-07-18 | 深圳市景程信息科技有限公司 | Methods of electric load forecasting based on long Memory Neural Networks in short-term |
CN107563539A (en) * | 2017-07-24 | 2018-01-09 | 佛山市顺德区中山大学研究院 | Short-term and long-medium term power load forecasting method based on machine learning model |
CN107590567A (en) * | 2017-09-13 | 2018-01-16 | 南京航空航天大学 | A kind of Recognition with Recurrent Neural Network short-term load forecasting method based on comentropy cluster and notice mechanism |
CN107622329A (en) * | 2017-09-22 | 2018-01-23 | 深圳市景程信息科技有限公司 | The Methods of electric load forecasting of Memory Neural Networks in short-term is grown based on Multiple Time Scales |
CN107730087A (en) * | 2017-09-20 | 2018-02-23 | 平安科技(深圳)有限公司 | Forecast model training method, data monitoring method, device, equipment and medium |
-
2018
- 2018-09-04 CN CN201811028611.1A patent/CN109034500A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101383023A (en) * | 2008-10-22 | 2009-03-11 | 西安交通大学 | Neural network short-term electric load prediction based on sample dynamic organization and temperature compensation |
CN103295081A (en) * | 2013-07-02 | 2013-09-11 | 上海电机学院 | Electrical power system load prediction method based on back propagation (BP) neural network |
CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学(保定) | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
CN106952181A (en) * | 2017-03-08 | 2017-07-14 | 深圳市景程信息科技有限公司 | Electric Load Prediction System based on long Memory Neural Networks in short-term |
CN106960252A (en) * | 2017-03-08 | 2017-07-18 | 深圳市景程信息科技有限公司 | Methods of electric load forecasting based on long Memory Neural Networks in short-term |
CN107563539A (en) * | 2017-07-24 | 2018-01-09 | 佛山市顺德区中山大学研究院 | Short-term and long-medium term power load forecasting method based on machine learning model |
CN107590567A (en) * | 2017-09-13 | 2018-01-16 | 南京航空航天大学 | A kind of Recognition with Recurrent Neural Network short-term load forecasting method based on comentropy cluster and notice mechanism |
CN107730087A (en) * | 2017-09-20 | 2018-02-23 | 平安科技(深圳)有限公司 | Forecast model training method, data monitoring method, device, equipment and medium |
CN107622329A (en) * | 2017-09-22 | 2018-01-23 | 深圳市景程信息科技有限公司 | The Methods of electric load forecasting of Memory Neural Networks in short-term is grown based on Multiple Time Scales |
Non-Patent Citations (1)
Title |
---|
翁金芳等: "基于多因素加法模型的中期电力负荷预测", 《计算机系统应用》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110909941A (en) * | 2019-11-26 | 2020-03-24 | 广州供电局有限公司 | Power load prediction method, device and system based on LSTM neural network |
CN110909941B (en) * | 2019-11-26 | 2022-08-02 | 广东电网有限责任公司广州供电局 | Power load prediction method, device and system based on LSTM neural network |
CN111178612A (en) * | 2019-12-19 | 2020-05-19 | 绍兴大明电力设计院有限公司 | LSTM load prediction method of grid user based on big data ODPS engine |
CN111178612B (en) * | 2019-12-19 | 2023-06-09 | 绍兴大明电力设计院有限公司 | LSTM load prediction method of grid user based on big data ODPS engine |
CN111241755A (en) * | 2020-02-24 | 2020-06-05 | 国网(苏州)城市能源研究院有限责任公司 | Power load prediction method |
CN111899123A (en) * | 2020-07-28 | 2020-11-06 | 深圳江行联加智能科技有限公司 | Electric quantity prediction method, electric quantity prediction device and computer readable storage medium |
CN115545362A (en) * | 2022-12-05 | 2022-12-30 | 南方电网数字电网研究院有限公司 | AI and TSD combined new energy medium-term power combined prediction method |
CN116361709A (en) * | 2023-03-31 | 2023-06-30 | 山东省计算中心(国家超级计算济南中心) | Self-adaptive power load identification method and device |
CN116361709B (en) * | 2023-03-31 | 2023-10-31 | 山东省计算中心(国家超级计算济南中心) | Self-adaptive power load identification method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109034500A (en) | A kind of mid-term electric load forecasting method of multiple timings collaboration | |
CN108280551B (en) | Photovoltaic power generation power prediction method utilizing long-term and short-term memory network | |
Tang et al. | Short‐term power load forecasting based on multi‐layer bidirectional recurrent neural network | |
Li et al. | Prediction for tourism flow based on LSTM neural network | |
CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
CN105631483B (en) | A kind of short-term electro-load forecast method and device | |
CN110674999A (en) | Cell load prediction method based on improved clustering and long-short term memory deep learning | |
CN110298501A (en) | Electric load prediction technique based on long Memory Neural Networks in short-term | |
CN109359786A (en) | A kind of power station area short-term load forecasting method | |
CN107563539A (en) | Short-term and long-medium term power load forecasting method based on machine learning model | |
CN103886218B (en) | Storehouse, the lake algal bloom Forecasting Methodology compensated with neutral net and support vector machine based on polynary non-stationary time series | |
CN108876001B (en) | Short-term power load prediction method based on twin support vector machine | |
CN110751318A (en) | IPSO-LSTM-based ultra-short-term power load prediction method | |
CN109523021A (en) | A kind of dynamic network Structure Prediction Methods based on long memory network in short-term | |
CN108074004A (en) | A kind of GIS-Geographic Information System short-term load forecasting method based on gridding method | |
CN102930354A (en) | Method and device for predicating electricity consumption of residential area | |
CN107730059A (en) | The method of transformer station's electricity trend prediction analysis based on machine learning | |
CN110866640A (en) | Power load prediction method based on deep neural network | |
CN106447133A (en) | Short-term electric load prediction method based on deep self-encoding network | |
CN108694473A (en) | Building energy consumption prediction technique based on RBF neural | |
CN113554466A (en) | Short-term power consumption prediction model construction method, prediction method and device | |
CN109214565A (en) | A kind of subregion system loading prediction technique suitable for the scheduling of bulk power grid subregion | |
Akpinar et al. | Forecasting natural gas consumption with hybrid neural networks—Artificial bee colony | |
CN109993354A (en) | A method of it is predicted for energy consumption | |
CN112288140A (en) | Keras-based short-term power load prediction method, storage medium and equipment |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181218 |