CN106447133A - Short-term electric load prediction method based on deep self-encoding network - Google Patents
Short-term electric load prediction method based on deep self-encoding network Download PDFInfo
- Publication number
- CN106447133A CN106447133A CN201610955493.3A CN201610955493A CN106447133A CN 106447133 A CN106447133 A CN 106447133A CN 201610955493 A CN201610955493 A CN 201610955493A CN 106447133 A CN106447133 A CN 106447133A
- Authority
- CN
- China
- Prior art keywords
- load
- short
- network
- model
- training
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000012549 training Methods 0.000 claims abstract description 21
- 239000011159 matrix material Substances 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 8
- 238000010606 normalization Methods 0.000 claims abstract description 4
- 238000005457 optimization Methods 0.000 claims abstract description 4
- 238000013277 forecasting method Methods 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 6
- 240000002853 Nelumbo nucifera Species 0.000 claims description 4
- 235000006508 Nelumbo nucifera Nutrition 0.000 claims description 4
- 235000006510 Nelumbo pentapetala Nutrition 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000003672 processing method Methods 0.000 claims description 3
- 239000013598 vector Substances 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 claims description 2
- 238000013135 deep learning Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000005611 electricity Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 235000001674 Agaricus brunnescens Nutrition 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000012706 support-vector machine Methods 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 provides a short-term electric load prediction method based on a deep self-encoding network. The method includes the steps that firstly, based on an automatic encoder and in combination with load relevant factors including the air temperature, date and historical loads, a deep self-encoding network load prediction model is established; then, data including the air temperature, date and historical loads is processed in a structured and standardized mode to form a sample matrix, this is, the sample matrix formed through the method including the steps of fuzzification, encoding and normalization processing is a high-capacity structured matrix; then, the model is trained through a training method that a non-supervision mode and a supervision mode are combined, and global optimization is achieved by adjusting relevant parameters and controlling the number of times of iterations; finally, the short-term electric load prediction result is analyzed by simultaneously selecting two indexes of the relative error and the mean absolute error, and the feasibility and an optimal parameter feasible region of the mode are evaluated.
Description
Technical field
The invention belongs to power grid control technical field, bear particularly to a kind of short term power based on depth autoencoder network
Lotus Forecasting Methodology.
Background technology
Load forecast is the premise realizing Automatic Generation Control and economic dispatch control, is Power System Planning and tune
The foundation of degree, predict the outcome is whether accurate, and the safety of China's Operation of Electric Systems, stability, economy are had very
Important impact.For a long time, Chinese scholars have carried out numerous studies to Methods of electric load forecasting and theory, generally may be used
To be divided into traditional method and artificial intelligence approach.Traditional method mainly includes time series method, Regression Forecast, gray system reason
By etc., their models are simple, fast operation, but cannot simulate complicated and changeable in the case of electric load, prediction accuracy
Generally low；Artificial intelligence approach is primarily referred to as some machine learning methods with artificial neural network as representative, including BP god
Through network, support vector machine, extreme learning machine etc.,, compared to traditional method, prediction accuracy all has and carries to some extent for they
Height, but these methods belong to shallow-layer machine learning method, do not possess the conversion to big data quantity sample and process strongly
Ability, so that learning capacity is limited, is difficult to when predictablity rate reach a certain height be improved it is impossible to meet actual production pair again
The higher and higher requirement of prediction accuracy.
Emerging in large numbers it is necessary to find a kind of new intelligent method with electricity consumption big data, is capable of power load under big data
The high accuracy of lotus and accuracy prediction, preferably to meet power system Production requirement.2006, University of Toronto was taught
Award the thought that Geoffrey Hinton proposes deep learning, open the wave in academia and industrial quarters for the deep learning from this
Tide.Deep learning from propose since, be successfully applied to speech recognition, target recognition (recognition of face, handwriting recongnition) and
The aspects such as natural language processing, all show very excellent performance, additionally, in medical treatment ＆ health, financial risks assessment, biology
It is like a raging fire that the research of the aspects such as pharmacy, search engine, smart machine, Internet of Things application is also carried out, many similar projects
Emerge in large numbers like the mushrooms after rain.
Depth autoencoder network (Deep Auto-Encoder Network, DAEN) is one kind of deep learning method, it
By building the machine learning model with many hidden layers, successively eigentransformation is carried out to training sample, by sample in former space
Character representation transforms to a new feature space, so that prediction is more prone to, the accuracy of final lifting prediction.Advise with artificial
Then the method for structural features is compared, and the method more can portray the abundant internal information of data, is more hopeful to improve the standard of prediction
Really rate, is the hot fields of machine learning research in the world at present, has been successfully applied to the side such as image recognition, text-processing
Face, but there is not yet it is applied to load forecast aspect.
Application in terms of Load Prediction In Power Systems for the depth of investigation learning art has important academic significance and engineering
Meaning, has certain feasibility.With the increase of power consumption and electricity consumption species, electricity consumption big data is emerged in large numbers, load prediction problem
Complexity increase therewith.And the inherent ability just with process large-scale data of depth autoencoder network, with data it is
Drive " raw material ", data is bigger, and performance is more excellent, shows very excellent performance in terms of Engineering prediction.
Content of the invention
It is an object of the invention to provide a kind of Short-Term Load Forecasting Method based on depth autoencoder network.
A kind of Short-Term Load Forecasting Method based on depth autoencoder network, comprises the following steps：
First, based on autocoder, in conjunction with the load correlative factor including temperature, date and historical load, build deep
Degree autoencoder network load forecasting model；
Secondly, by sample is constituted to the data structured including temperature, date and historical load value and standardization
Matrix, that is, the sample matrix passing through the method formation of Fuzzy processing, coding, normalized is Large Copacity structured matrix；
Again, using training method that is unsupervised and having two kinds of training methods combinations of supervision, model is trained, passes through
Adjust relevant parameter and control iterationses to realize global optimum；
Finally, by choosing relative error and mean absolute error two indices to short-term electric load prediction result simultaneously
It is analyzed, the feasibility of evaluation profile and optimized parameter feasible zone.
Depth autoencoder network load forecasting model modeling method is：
If the front portion of model is formed by dried layer autocoder stacking, top increases by one layer and represents the last of desired output
Layer, i.e. prediction interval,
The input of model includes 3 characteristic vectors：Historical load value L, temperature T, date D,
L=[L0,L1,···,L47], represent history day 0:00、0:30、···23:Power load charge values when 30；
T represents the value after daily mean temperature t obfuscation；
D=[d0,d1,d2,d3,d4,d5,d6], represent week encoded radio；
Output Output=[the F of model0,F1,···,F23], represent prediction day 0:00、1:00、···23:When 00
Power load charge values,
By large-scale training sample, realize the Accurate Prediction of short term by the ability that DAENs simulates complicated function.
Data structured and standardization processing method are to carry out pretreatment to historical load value, temperature and day categorical data:
For historical load value, make normalized so as to scope is between 0 to 1 using normalization formula logarithm value；
For temperature, by defining suitable membership function, it is made with Fuzzy processing, the numerical range after process is 0
To between 1；
For day type, by the different day type of coded representation, from Monday value Sunday, realize the structuring of data
Represent.
Load forecasting model training is divided into pre-training and two stages of tuning：
Pre-training is the process of initialization network parameter, using a large amount of unlabeled exemplars using successively unsupervised characteristic optimization
Algorithm is carried out；
Tuning is using a small amount of exemplar, the whole network including prediction interval to be finely adjusted so that network performance
Tend to global optimum.
Load prediction results analysis method is to choose load prediction relative error and mean absolute error as prediction simultaneously
Evaluation of result index.
The present invention proposes the short-term load forecasting pattern based on depth autoencoder network.First, based on autocoder,
In conjunction with the load correlative factor such as temperature, date, historical load, construct depth autoencoder network load forecasting model.In this base
On plinth, by sample matrix is constituted to the data structureds such as temperature, date, historical load value and standardization, here mainly
It is achieved by the method for Fuzzy processing, coding, normalized, the sample matrix being formed is Large Copacity structuring square
Battle array.Further, using training method that is unsupervised and having two kinds of training methods combinations of supervision, model is trained, by adjusting
Relevant parameter and control iterationses realize global optimum, and this training method can be prevented effectively from local optimum and suboptimum situation
Generation.Then, by choosing relative error and mean absolute error two indices to short-term electric load prediction result simultaneously
It is analyzed, the feasibility of evaluation profile and optimized parameter feasible zone.Having of this pattern is demonstrated by reality/analogue system
Effect property and reliability.
Brief description
Accompanying drawing 1 is the Short-term Load Forecasting Model based on depth autoencoder network for the present invention.
DAEN, BPNN predicted load and actual negative charge values comparison diagram in accompanying drawing 2 embodiment of the present invention.
Specific embodiment
The Short-Term Load Forecasting Method based on depth autoencoder network of the present invention, including step：
1, modeling method.
If the front portion of the model building is formed by dried layer autocoder (Auto-Encoder, AE) stacking, top increases
One layer of final layer representing desired output, i.e. prediction interval.The input of model includes 3 characteristic vectors：Historical load value L, temperature
T, date D, L=[L0,L1,···,L47], represent history day 0:00、0:30、···23:Power load charge values when 30；T
Represent the value after daily mean temperature t obfuscation；D=[d0,d1,d2,d3,d4,d5,d6], represent week encoded radio；The output of model
Output=[F0,F1,···,F23], represent prediction day 0:00、1:00、···23:Power load charge values when 00.By
Large-scale training sample, realizes the Accurate Prediction of short term by the ability that DAENs simulates complicated function.Forecast model is shown in attached
Fig. 1.
2, data structured and standardization processing method.
Mainly pretreatment is carried out to historical load value, temperature and day categorical data:1. historical load value, using normalization
Formula logarithm value makees normalized so as to scope is between 0 to 1；2. temperature, by defining suitable membership function to it
Make Fuzzy processing, the numerical range after process is between 0 to 1；3. day type, by the different day type of coded representation, from
Monday value Sunday, realizes the structured representation of data.
3, load forecasting model training method.
It is divided into pre-training and two stages of tuning：It is exactly the process of initialization network parameter on the process nature of pre-training,
Carried out using successively unsupervised feature optimization algorithm using a large amount of unlabeled exemplars；Tuning is then to bag using a small amount of exemplar
Include prediction interval to be finely adjusted so that network performance tends to global optimum in interior whole network.
4, load prediction results analysis method.
Choose load prediction relative error error and mean absolute error mean absolute error as pre- simultaneously
Survey evaluation of result index, predict the outcome and see attached list 1；Load actual value, the predicted load based on DAEN and based on tradition before to
The predicted load Comparative result of Feedback Neural Network (Back Propagation Neural Network, BPNN) is shown in accompanying drawing
2.
The predicted load based on DAEN for the subordinate list 1 and load actual value deck watch
Claims (5)
1. a kind of Short-Term Load Forecasting Method based on depth autoencoder network is it is characterised in that comprise the following steps：
First, based on autocoder, in conjunction with the load correlative factor including temperature, date and historical load, build depth certainly
Coding network load forecasting model；
Secondly, by sample matrix is constituted to the data structured including temperature, date and historical load value and standardization,
The sample matrix passing through the method formation of Fuzzy processing, coding, normalized is Large Copacity structured matrix；
Again, using training method that is unsupervised and having two kinds of training methods combinations of supervision, model is trained, by adjusting
Relevant parameter and control iterationses realize global optimum；
Finally, by choosing relative error and mean absolute error two indices simultaneously, short-term electric load prediction result is carried out
Analysis, the feasibility of evaluation profile and optimized parameter feasible zone.
2. the Short-Term Load Forecasting Method based on depth autoencoder network as claimed in claim 1 is it is characterised in that depth
Spending autoencoder network load forecasting model modeling method is：
If the front portion of model is formed by dried layer autocoder stacking, top increases by one layer of final layer representing desired output, that is,
Prediction interval,
The input of model includes 3 characteristic vectors：Historical load value L, temperature T, date D,
L=[L0,L1,…,L47], represent history day 0:00、0:30、…23:Power load charge values when 30；
T represents the value after daily mean temperature t obfuscation；
D=[d0,d1,d2,d3,d4,d5,d6], represent week encoded radio；
Output Output=[the F of model0,F1,…,F23], represent prediction day 0:00、1:00、…23:Power load charge values when 00,
By large-scale training sample, realize the Accurate Prediction of short term by the ability that DAENs simulates complicated function.
3. the Short-Term Load Forecasting Method based on depth autoencoder network as claimed in claim 2 is it is characterised in that count
It is that pretreatment is carried out to historical load value, temperature and day categorical data according to structuring and standardization processing method:
For historical load value, make normalized so as to scope is between 0 to 1 using normalization formula logarithm value；
For temperature, by defining suitable membership function, it is made with Fuzzy processing, the numerical range after process is 0 to 1
Between；
For day type, by the different day type of coded representation, from Monday value Sunday, realize the structuring table of data
Show.
4. the Short-Term Load Forecasting Method based on depth autoencoder network as claimed in claim 3 is it is characterised in that bear
The training of lotus forecast model is divided into pre-training and two stages of tuning：
Pre-training is the process of initialization network parameter, using a large amount of unlabeled exemplars using successively unsupervised feature optimization algorithm
Carry out；
Tuning is using a small amount of exemplar, the whole network including prediction interval to be finely adjusted so that network performance tends to
Global optimum.
5. the Short-Term Load Forecasting Method based on depth autoencoder network as claimed in claim 4 is it is characterised in that bear
The lotus analysis method that predicts the outcome is to choose load prediction relative error simultaneously and mean absolute error refers to as the evaluation that predicts the outcome
Mark.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610955493.3A CN106447133A (en) | 2016-11-03 | 2016-11-03 | Short-term electric load prediction method based on deep self-encoding network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610955493.3A CN106447133A (en) | 2016-11-03 | 2016-11-03 | Short-term electric load prediction method based on deep self-encoding network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106447133A true CN106447133A (en) | 2017-02-22 |
Family
ID=58180846
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610955493.3A Pending CN106447133A (en) | 2016-11-03 | 2016-11-03 | Short-term electric load prediction method based on deep self-encoding network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106447133A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107248740A (en) * | 2017-06-15 | 2017-10-13 | 贵州电网有限责任公司电力科学研究院 | A kind of household electricity machine utilization Forecasting Methodology |
| CN107423839A (en) * | 2017-04-17 | 2017-12-01 | 湘潭大学 | A kind of method of the intelligent building microgrid load prediction based on deep learning |
| CN109034453A (en) * | 2018-06-21 | 2018-12-18 | 南京邮电大学 | A kind of Short-Term Load Forecasting Method based on multiple labeling neural network |
| CN110119826A (en) * | 2018-02-06 | 2019-08-13 | 天津职业技术师范大学 | A kind of power-system short-term load forecasting method based on deep learning |
| CN110321390A (en) * | 2019-06-04 | 2019-10-11 | 上海电力学院 | Based on the load curve data visualization method for thering is supervision and unsupervised algorithm to combine |
| CN110648248A (en) * | 2019-09-05 | 2020-01-03 | 广东电网有限责任公司 | Control method, device and equipment for power station |
| CN110703899A (en) * | 2019-09-09 | 2020-01-17 | 创新奇智(南京)科技有限公司 | Data center energy efficiency optimization method based on transfer learning |
| CN112308342A (en) * | 2020-11-25 | 2021-02-02 | 广西电网有限责任公司北海供电局 | Daily load prediction method based on deep time decoupling and application |
| CN116865246A (en) * | 2023-06-27 | 2023-10-10 | 广东电网有限责任公司广州供电局 | Industrial user load feasible domain prediction method and system based on quick response |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105608512A (en) * | 2016-03-24 | 2016-05-25 | 东南大学 | Short-term load forecasting method |
| CN105631483A (en) * | 2016-03-08 | 2016-06-01 | 国家电网公司 | Method and device for predicting short-term power load |
| CN105930955A (en) * | 2016-04-07 | 2016-09-07 | 浙江万马新能源有限公司 | Deep learning-based charging network operation situation analysis method and apparatus |
-
2016
- 2016-11-03 CN CN201610955493.3A patent/CN106447133A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631483A (en) * | 2016-03-08 | 2016-06-01 | 国家电网公司 | Method and device for predicting short-term power load |
| CN105608512A (en) * | 2016-03-24 | 2016-05-25 | 东南大学 | Short-term load forecasting method |
| CN105930955A (en) * | 2016-04-07 | 2016-09-07 | 浙江万马新能源有限公司 | Deep learning-based charging network operation situation analysis method and apparatus |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107423839A (en) * | 2017-04-17 | 2017-12-01 | 湘潭大学 | A kind of method of the intelligent building microgrid load prediction based on deep learning |
| CN107248740A (en) * | 2017-06-15 | 2017-10-13 | 贵州电网有限责任公司电力科学研究院 | A kind of household electricity machine utilization Forecasting Methodology |
| CN107248740B (en) * | 2017-06-15 | 2020-03-24 | 贵州电网有限责任公司电力科学研究院 | Load prediction method for household electric equipment |
| CN110119826A (en) * | 2018-02-06 | 2019-08-13 | 天津职业技术师范大学 | A kind of power-system short-term load forecasting method based on deep learning |
| CN109034453A (en) * | 2018-06-21 | 2018-12-18 | 南京邮电大学 | A kind of Short-Term Load Forecasting Method based on multiple labeling neural network |
| CN110321390A (en) * | 2019-06-04 | 2019-10-11 | 上海电力学院 | Based on the load curve data visualization method for thering is supervision and unsupervised algorithm to combine |
| CN110648248A (en) * | 2019-09-05 | 2020-01-03 | 广东电网有限责任公司 | Control method, device and equipment for power station |
| CN110648248B (en) * | 2019-09-05 | 2023-04-07 | 广东电网有限责任公司 | Control method, device and equipment for power station |
| CN110703899A (en) * | 2019-09-09 | 2020-01-17 | 创新奇智(南京)科技有限公司 | Data center energy efficiency optimization method based on transfer learning |
| CN112308342A (en) * | 2020-11-25 | 2021-02-02 | 广西电网有限责任公司北海供电局 | Daily load prediction method based on deep time decoupling and application |
| CN116865246A (en) * | 2023-06-27 | 2023-10-10 | 广东电网有限责任公司广州供电局 | Industrial user load feasible domain prediction method and system based on quick response |
| CN116865246B (en) * | 2023-06-27 | 2023-12-26 | 广东电网有限责任公司广州供电局 | Industrial user load feasible domain prediction method and system based on quick response |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106447133A (en) | Short-term electric load prediction method based on deep self-encoding network | |
| Lin et al. | An improved moth-flame optimization algorithm for support vector machine prediction of photovoltaic power generation | |
| Wu et al. | Evolving RBF neural networks for rainfall prediction using hybrid particle swarm optimization and genetic algorithm | |
| Pai | System reliability forecasting by support vector machines with genetic algorithms | |
| CN108280551A (en) | A kind of photovoltaic power generation power prediction method using shot and long term memory network | |
| Shang et al. | Short-term load forecasting based on PSO-KFCM daily load curve clustering and CNN-LSTM model | |
| CN104268627A (en) | Short-term wind speed forecasting method based on deep neural network transfer model | |
| Tian | Backtracking search optimization algorithm-based least square support vector machine and its applications | |
| CN105631483A (en) | Method and device for predicting short-term power load | |
| CN107730006A (en) | A kind of nearly zero energy consumption controller of building based on regenerative resource big data deep learning | |
| CN109146121A (en) | The power predicating method stopped in the case of limited production based on PSO-BP model | |
| CN110276472A (en) | A kind of offshore wind farm power ultra-short term prediction method based on LSTM deep learning network | |
| CN109934422A (en) | Neural network wind speed prediction method based on time series data analysis | |
| CN109034500A (en) | A kind of mid-term electric load forecasting method of multiple timings collaboration | |
| CN110717581A (en) | Short-term load prediction method based on temperature fuzzy processing and DBN | |
| CN104050505A (en) | Multilayer-perceptron training method based on bee colony algorithm with learning factor | |
| CN104732067A (en) | Industrial process modeling forecasting method oriented at flow object | |
| Uguz et al. | A new algorithm based on artificial bee colony algorithm for energy demand forecasting in Turkey | |
| CN113762591B (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM countermeasure learning | |
| CN102663493A (en) | Delaying nerve network used for time sequence prediction | |
| Wang et al. | Sustainable development early warning and financing risk management of resource-based industrial clusters using optimization algorithms | |
| Bi et al. | Self-adaptive Teaching-learning-based Optimizer with Improved RBF and Sparse Autoencoder for Complex Optimization Problems | |
| CN115481788B (en) | Phase change energy storage system load prediction method and system | |
| Cai et al. | Short-term forecasting of user power load in China based on XGBoost | |
| CN108805358A (en) | A kind of Gray Dynamic reinforcing prediction technique based on multiple factors |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Robert Caiming Qiu Shixin Chulei He Xingling Zenan Liu Haichun Inventor after: Shi Xin Inventor after: Chu Lei Inventor after: He Xing Inventor after: Ling Zenan Inventor after: Liu Haichun Inventor before: Robert Caiming Qiu Shixin Chulei He Xinglin Zenan Liu Haichun Inventor before: Shi Xin Inventor before: Chu Lei Inventor before: He Xing Inventor before: Lin Zenan Inventor before: Liu Haichun |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170222 |