CN107301478A - A kind of cable run short-term load forecasting method - Google Patents
A kind of cable run short-term load forecasting method Download PDFInfo
- Publication number
- CN107301478A CN107301478A CN201710495777.3A CN201710495777A CN107301478A CN 107301478 A CN107301478 A CN 107301478A CN 201710495777 A CN201710495777 A CN 201710495777A CN 107301478 A CN107301478 A CN 107301478A
- Authority
- CN
- China
- Prior art keywords
- load current
- network
- load
- moment
- neutral net
- 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 10
- 238000012549 training Methods 0.000 claims abstract description 27
- 239000002245 particle Substances 0.000 claims abstract description 17
- 230000007935 neutral effect Effects 0.000 claims abstract description 15
- 238000013528 artificial neural network Methods 0.000 claims abstract description 7
- 238000005457 optimization Methods 0.000 claims abstract description 7
- 238000000034 method Methods 0.000 abstract description 9
- 238000013459 approach Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 240000002853 Nelumbo nucifera Species 0.000 description 1
- 235000006508 Nelumbo nucifera Nutrition 0.000 description 1
- 235000006510 Nelumbo pentapetala Nutrition 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of cable run short-term load forecasting method.A kind of cable run short-term load forecasting method, with load prediction moment corresponding temperature on average and weather characteristics value, and its preceding four moment corresponding load current is input, and load prediction moment corresponding load current is output, builds BP neural network;Initial training is carried out to network using training sample, particle swarm optimization algorithm is reused further to train network, by load prediction moment corresponding temperature on average and weather characteristics value, and its preceding four moment corresponding load current is input to the neutral net finally trained, you can predict the load current at the moment.The Forecasting Methodology of the present invention can calculate to a nicety load current, can provide foundation for electric power system dispatching.
Description
Technical field
The present invention relates to load prediction technical field, more particularly, to a kind of cable run short-term load forecasting method.
Background technology
Load prediction by the time of prediction can be divided into for a long time, mid-term and short-term load forecasting.Wherein, the standard of short term
Really prediction is conducive to improving the on-road efficiency of electric power management body.
At present, the method for power-system short-term load forecasting mainly includes statistical technique, expert system approach and neutral net
Method.Short term model used generally can be divided into time system model and regression model in statistical technique.Time system model
The climatic information having a significant impact to load performance and other factors can not be made full use of, forecasting inaccuracy is true and unstable.Return
Model needs to know the functional relation between load and meteorological variables in advance, and computationally intensive, it is impossible to handle Climatic and negative
Non-equilibrium transient state relation between lotus.Expert system approach utilize expert Heuristics and inference rule, improve festivals or holidays or
The load prediction precision of occasion day, but expertise and experience etc. are converted into series of rules in the presence of very big tired exactly
It is difficult.
Load curve is a nonlinear function related to several factors.For nonlinear prediction and reasoning, nerve
Network is a kind of suitable method.Recent studies indicate that, relative to statistical technique and expert system approach, utilize neutral net
Method is predicted to power-system short-term load can obtain higher precision.
The content of the invention
It is an object of the invention to provide a kind of cable run short-term load forecasting method, deficiency of the prior art is overcome,
A kind of cable run short-term load forecasting method is provided, cable run short-term load forecasting is solved the problems, such as, precision of prediction is improved.
In order to solve the above technical problems, the technical solution adopted by the present invention is:A kind of cable run short-term load forecasting side
Method, wherein, comprise the following steps:
(1) training sample is chosen:Include load current and meteorological significant condition amount, wherein load current inside every group of sample
Including moment t and its preceding four moment t-1, t-2, t-3, the corresponding load current I of t-4t, It-1, It-2It-3, It-4;It is meteorological special
Levying quantity of state includes the corresponding temperature on average K of moment t1tWith weather characteristics value K2t;
(2) network initial training:With the load current I in step (1)t-1, It-2It-3, It-4, temperature on average K1tAnd day
Gas characteristic value K2tAs input, load current ItAs output, BP neural network is built, and network is carried out using training sample
Training, obtains the good neutral net of initial training;
(3) network is further trained:The good neutral net of the initial training that obtains for step (2), keeps network structure
It is constant, using training sample and particle swarm optimization algorithm is combined, the connection weight and threshold value to network are further optimized,
The neutral net finally trained;
(4) load prediction:By load prediction moment corresponding temperature on average and weather characteristics value, and its preceding four moment
Corresponding load current is input to the neutral net that step (3) is finally trained, you can predict the load current at the moment.
Compared with prior art, its advantage is the present invention:
The present invention only sets up BP neural network, and use training sample pair not against expertise using the data of monitoring
Network carries out initial training, network connection weights and threshold value is further optimized using particle swarm optimization algorithm afterwards, energy
Enough improve the precision of load prediction.
Brief description of the drawings
Fig. 1 is a kind of schematic flow sheet of cable run short-term load forecasting method of the present invention.
Fig. 2 is 4 layers of BP neural network topology diagram.
Embodiment
Accompanying drawing being given for example only property explanation, it is impossible to be interpreted as the limitation to this patent;It is attached in order to more preferably illustrate the present embodiment
Scheme some parts to have omission, zoom in or out, do not represent the size of actual product;To those skilled in the art,
Some known features and its explanation may be omitted and will be understood by accompanying drawing.Being given for example only property of position relationship described in accompanying drawing
Explanation, it is impossible to be interpreted as the limitation to this patent.
As shown in figure 1, a kind of cable run short-term load forecasting method, wherein, comprise the following steps:
(1) training sample is chosen:Training sample is chosen:Include load current and meteorological significant condition inside every group of sample
Amount, wherein load current include moment t and its preceding four moment t-1, t-2, t-3, the corresponding load current I of t-4t, It-1, It- 2It-3, It-4;Meteorological Characteristics quantity of state includes the corresponding temperature on average K of moment t1tWith weather characteristics value K2t, weather characteristics value K2t
0,0.5,1 is can be taken as, fine day, cloudy day and rainy day are represented respectively.
The corresponding temperature on average K of moment t1tCalculation formula be:
K1t=(K1max+K1min)/2
In formula, K1maxFor the corresponding highest temperatures of moment t, K1minFor the corresponding lowest temperatures of moment t.
(2) network initial training:With the load current I in step (1)t-1, It-2It-3, It-4, temperature on average K1tAnd day
Gas characteristic value K2tAs input, load current ItAs output, BP neural network is built, and network is carried out using training sample
Training, obtains the good neutral net of initial training.
As shown in Fig. 24 layers of BP neural network in the present embodiment include input layer, two hidden layers and output layer, network
It is output as:
X, Y represent input and the output variable of network, W respectively1For input layer and the weight matrix of the first hidden layer, W2For
The weight matrix of first hidden layer and the second hidden layer, W3For the second hidden layer and the weight matrix of output layer, f1And f2It is implicit
The activation primitive of layer, the present invention elects sigmoid functions as, and the activation primitive of output layer elects purelin linear functions, a as1,a2For
The input vector of hidden layer, a3For the output vector of output layer.(n1,n2) and (b1,b2) be respectively hidden layer input vector and
Threshold vector, n3, b3The respectively input vector and threshold vector of output layer.
(3) network is further trained:The good neutral net of the initial training that obtains for step (2), keeps network structure
Constant, i.e., hidden layer is kept for two layers, and the neuron number of each hidden layer keeps constant, using training sample and combines particle
Colony optimization algorithm, connection weight and threshold value to network are further optimized, the neutral net finally trained.
Using training sample and particle swarm optimization algorithm is combined, the connection weight and threshold value to network carry out further excellent
Change step as follows:
Step S1:Population is initialized, as each group weights, threshold value assign initial value.Define the fitness function of population
For the output error of network, it is shown below, and calculates the fitness of each particle.
Wherein, in formula, N is the number of training of network, and O exports for network, and d exports for target, and p is that network is exported and mesh
Mark the dimension of output.
Step S2:Determine local extremum and global extremum;Compare first the current fitness of each particle with it is previous
Optimal adaptation degree, then the current local extremum of each particle be taken as the smaller value in the two;It is determined that during global extremum, taking all
Minimum one in the current fitness of particle.
Step S3:It is iterated in calculating, each iteration, the local extremum and global extremum obtained using step (2), point
The not speed of more new particle and position according to the following formula.
The speed of particle more new formula is as follows:
vij(k+1)=wvij(k)+c1r1[Qij(k)-xij(k)]
+c2r2[Qgj(k)-xij(k)]
The location updating formula of particle is as follows:
xij(k+1)=xij(k)+vij(k+1)
In formula, vijAnd xijThe speed of respectively i-th particle jth dimension space and position;QijIt is empty for i-th of particle jth dimension
Between local extremum, QgjFor the global extremum of all particle jth dimension spaces;W is inertia weight coefficient, c1And c2It is normal to accelerate
Number, r1And r2For random number, its interval is [0 1].
Step (4):When algorithm meets error precision or reaches maximum iteration, particle swarm optimization algorithm is exited.
(4) load prediction:By load prediction moment corresponding temperature on average and weather characteristics value, and its preceding four moment
Corresponding load current is input to the neutral net that step (3) is finally trained, you can predict the load current at the moment.Its
In, load prediction moment corresponding temperature on average and weather characteristics value are obtained by weather forecast information.
Obviously, the above embodiment of the present invention is just for the sake of clearly demonstrating example of the present invention, and is not
Restriction to embodiments of the present invention.For those of ordinary skill in the field, on the basis of the above description also
It can make other changes in different forms.There is no necessity and possibility to exhaust all the enbodiments.It is all
Any modifications, equivalent substitutions and improvements made within the spirit and principles in the present invention etc., should be included in right of the present invention will
Within the protection domain asked.
Claims (1)
1. a kind of cable run short-term load forecasting method, it is characterised in that comprise the following steps:
(1) training sample is chosen:Include load current and meteorological significant condition amount inside every group of sample, wherein load current includes
Moment t and its preceding four moment t-1, t-2, t-3, the corresponding load current I of t-4t, It-1, It-2It-3, It-4;Meteorological Characteristics shape
State amount includes the corresponding temperature on average K of moment t1tWith weather characteristics value K2t;
(2) network initial training:With the load current I in step (1)t-1, It-2It-3, It-4, temperature on average K1tAnd weather is special
Value indicative K2tAs input, load current ItAs output, BP neural network is built, and network is instructed using training sample
Practice, obtain the good neutral net of initial training;
(3) network is further trained:The good neutral net of the initial training that obtains for step (2), keeps network structure constant,
Using training sample and particle swarm optimization algorithm is combined, the connection weight and threshold value to network are further optimized, and are obtained
The neutral net finally trained;
(4) load prediction:By load prediction moment corresponding temperature on average and weather characteristics value, and its preceding four moment correspondence
Load current be input to the neutral net that step (3) is finally trained, you can predict the load current at the moment.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710495777.3A CN107301478A (en) | 2017-06-26 | 2017-06-26 | A kind of cable run short-term load forecasting method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710495777.3A CN107301478A (en) | 2017-06-26 | 2017-06-26 | A kind of cable run short-term load forecasting method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107301478A true CN107301478A (en) | 2017-10-27 |
Family
ID=60135497
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710495777.3A Pending CN107301478A (en) | 2017-06-26 | 2017-06-26 | A kind of cable run short-term load forecasting method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107301478A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110880044A (en) * | 2019-10-23 | 2020-03-13 | 广东电网有限责任公司 | Markov chain-based load prediction method |
CN111931973A (en) * | 2020-06-16 | 2020-11-13 | 广东电网有限责任公司 | Method for improving short-time load prediction precision of cable line |
CN113112085A (en) * | 2021-04-22 | 2021-07-13 | 国网山东省电力公司德州市陵城区供电公司 | New energy station power generation load prediction method based on BP neural network |
CN113657032A (en) * | 2021-08-12 | 2021-11-16 | 国网安徽省电力有限公司 | Low-frequency load shedding method and system for pre-centralized coordination and real-time distributed control |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103093288A (en) * | 2013-02-21 | 2013-05-08 | 江苏省电力公司电力科学研究院 | Partition power grid bus load prediction system based on weather information |
CN103729695A (en) * | 2014-01-06 | 2014-04-16 | 国家电网公司 | Short-term power load forecasting method based on particle swarm and BP neural network |
CN104331737A (en) * | 2014-11-21 | 2015-02-04 | 国家电网公司 | Office building load prediction method based on particle swarm neural network |
CN106600050A (en) * | 2016-12-10 | 2017-04-26 | 国网辽宁省电力有限公司锦州供电公司 | BP neural network-based ultra-short load prediction method |
-
2017
- 2017-06-26 CN CN201710495777.3A patent/CN107301478A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103093288A (en) * | 2013-02-21 | 2013-05-08 | 江苏省电力公司电力科学研究院 | Partition power grid bus load prediction system based on weather information |
CN103729695A (en) * | 2014-01-06 | 2014-04-16 | 国家电网公司 | Short-term power load forecasting method based on particle swarm and BP neural network |
CN104331737A (en) * | 2014-11-21 | 2015-02-04 | 国家电网公司 | Office building load prediction method based on particle swarm neural network |
CN106600050A (en) * | 2016-12-10 | 2017-04-26 | 国网辽宁省电力有限公司锦州供电公司 | BP neural network-based ultra-short load prediction method |
Non-Patent Citations (3)
Title |
---|
姜林: "基于粒子群优化神经网络的电力短期负荷预测研究", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
杨应高: "基于粒子群BP神经网络的短期电力负荷预测", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
陈帅 等: "基于粒子群BP神经网络的短期电力负荷预测", 《上海电力学院学报》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110880044A (en) * | 2019-10-23 | 2020-03-13 | 广东电网有限责任公司 | Markov chain-based load prediction method |
CN110880044B (en) * | 2019-10-23 | 2023-01-24 | 广东电网有限责任公司 | Markov chain-based load prediction method |
CN111931973A (en) * | 2020-06-16 | 2020-11-13 | 广东电网有限责任公司 | Method for improving short-time load prediction precision of cable line |
CN113112085A (en) * | 2021-04-22 | 2021-07-13 | 国网山东省电力公司德州市陵城区供电公司 | New energy station power generation load prediction method based on BP neural network |
CN113657032A (en) * | 2021-08-12 | 2021-11-16 | 国网安徽省电力有限公司 | Low-frequency load shedding method and system for pre-centralized coordination and real-time distributed control |
CN113657032B (en) * | 2021-08-12 | 2023-11-24 | 国网安徽省电力有限公司 | Low-frequency load shedding method and system for pre-centralized coordination real-time distribution control |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Pousinho et al. | A hybrid PSO–ANFIS approach for short-term wind power prediction in Portugal | |
Xiao et al. | BP neural network with rough set for short term load forecasting | |
Catalao et al. | An artificial neural network approach for short-term wind power forecasting in Portugal | |
Jilani et al. | Multivariate high order fuzzy time series forecasting for car road accidents | |
CN107301478A (en) | A kind of cable run short-term load forecasting method | |
CN106709820A (en) | Power system load prediction method and device based on deep belief network | |
CN107967803A (en) | Traffic congestion Forecasting Methodology based on multi-source data and variable-weight combined forecasting model | |
CN103440370A (en) | Transmission and transformation project construction cost assessment method and device | |
CN110866640A (en) | Power load prediction method based on deep neural network | |
CN108280998A (en) | Short-time Traffic Flow Forecasting Methods based on historical data dynamic select | |
Chitsazan et al. | Wind speed forecasting using an echo state network with nonlinear output functions | |
Xiaojian et al. | A traffic flow forecasting model based on BP neural network | |
Shen et al. | Short-term load forecasting of power system based on similar day method and PSO-DBN | |
CN105389625B (en) | Active power distribution network ultra-short term load prediction method | |
CN116632834A (en) | Short-term power load prediction method based on SSA-BiGRU-Attention | |
Zhengxiang et al. | Transport volume forecast based on GRNN network | |
Bashir et al. | Short-term load forecasting using artificial neural network based on particle swarm optimization algorithm | |
Wang et al. | Short term load forecasting: A dynamic neural network based genetic algorithm optimization | |
Zhang et al. | A hybrid load forecasting method based on neural network in smart grid | |
CN114444763A (en) | Wind power prediction method based on AFSA-GNN | |
Niu et al. | Research on power load forecasting based on combined model of Markov and BP neural networks | |
Babu et al. | Fuzzy-C Means Clustering Based ANFIS wind speed forecast | |
Saravanan et al. | Modeling and prediction of India's electricity demand using fuzzy logic | |
Yu et al. | Short-term Load Forecasting of Power System Based on Weighted Improved Adaptive Neuro-Fuzzy Inference System and Random Forest | |
Saravanan et al. | Prediction of India's electricity consumption in industrial sector using soft computing techniques |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171027 |
|
RJ01 | Rejection of invention patent application after publication |