CN105590174A - Enterprise power consumption load prediction method based on K-means clustering RBF neural network - Google Patents
Enterprise power consumption load prediction method based on K-means clustering RBF neural network Download PDFInfo
- Publication number
- CN105590174A CN105590174A CN201511018470.1A CN201511018470A CN105590174A CN 105590174 A CN105590174 A CN 105590174A CN 201511018470 A CN201511018470 A CN 201511018470A CN 105590174 A CN105590174 A CN 105590174A
- Authority
- CN
- China
- Prior art keywords
- load
- prediction
- data
- rbf
- day
- 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 28
- 238000003064 k means clustering Methods 0.000 title claims abstract description 12
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 11
- 238000004519 manufacturing process Methods 0.000 claims abstract description 10
- 238000004364 calculation method Methods 0.000 claims abstract description 5
- 230000007935 neutral effect Effects 0.000 claims description 22
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 238000013277 forecasting method Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 7
- 238000007726 management method Methods 0.000 claims description 5
- 238000003062 neural network model Methods 0.000 claims description 4
- 210000002569 neuron Anatomy 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 2
- 230000005484 gravity Effects 0.000 claims description 2
- 238000013507 mapping Methods 0.000 claims description 2
- 238000013519 translation Methods 0.000 claims description 2
- 230000004069 differentiation Effects 0.000 claims 1
- 238000012937 correction Methods 0.000 abstract description 6
- 238000005516 engineering process Methods 0.000 abstract description 6
- 230000006870 function Effects 0.000 description 12
- 238000012549 training Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 235000013350 formula milk Nutrition 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 241001269238 Data Species 0.000 description 3
- 230000002354 daily effect Effects 0.000 description 3
- 238000004134 energy conservation Methods 0.000 description 3
- 230000003203 everyday effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000013213 extrapolation Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000000611 regression analysis Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 235000020213 comfort milk formula Nutrition 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000013439 planning Methods 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- NHDHVHZZCFYRSB-UHFFFAOYSA-N pyriproxyfen Chemical compound C=1C=CC=NC=1OC(C)COC(C=C1)=CC=C1OC1=CC=CC=C1 NHDHVHZZCFYRSB-UHFFFAOYSA-N 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
- G06Q10/06375—Prediction of business process outcome or impact based on a proposed change
-
- 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)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses an enterprise power consumption load prediction method based on a k-means clustering radial basis function (RBF) neural network. The method includes steps: historical load data acquisition, meteorological data acquisition, date discrimination, neural network prediction, error calculation and correction, load curve drawing, and prediction data export. A prediction result is obtained by employing the neural network via the historical load data and meteorological factor input quantity, and correction is realized via an error correction module. Based on the requirement control technology of load prediction, with the combination of an industrial enterprise production plan and the condition of power consumption load usage, the system performs requirement control via a built-in requirement curve node determination method at a control point before the load prediction value reaches the maximum requirement, whether unnecessary loads are removed is determined, the current most appropriate energy-saving scheme is automatically selected, the maximum requirement is controlled in advance, over-load operation and even tripping operation can be effectively avoided, and safety and energy-saving production is guaranteed.
Description
Technical field
The invention belongs to load forecast and control technology field, be specially a kind of power load short-term forecast and intelligent demand control technology that is applicable to typical industry enterprise, load forecasting method is the neutral net based on k mean cluster radial basis RBF function.
Background technology
Short-term electric load prediction is to utilize historical data, in conjunction with feature and the influence factor of prognoses system, the load data of predict future 1-7 day. Mostly load prediction work is in the past that for industrial enterprise, load prediction has larger positive role for aspects such as energy conservation, energy-saving and emission-reduction, cost optimizations by being that Electric Power Network Planning department completes. On the basis of short-term load forecasting, understand the development and change of loading recent future, dsm measure is proposed targetedly, Based Intelligent Control is carried out load scheduling. Load prediction is as the important component part of enterprise's energy guard system, and the links such as enterprise operation optimum management provide important decision support.
Traditional load forecasting method mainly contains trend extrapolation, regression analysis, time series method etc., the simple matching accuracy of extrapolation is lower, regression analysis is not considered dynamic, the nonlinear relation between the variablees such as load and weather, and time series method is not considered the impact of the meteorologic factors such as weather on load. How effectively to predict electric load, the advantage of performance intelligent algorithm, improves precision and the efficiency of prediction, is a problem that needs solution.
Implement demand Side Management (IDSM) evaluation method according to national industrial enterprise in 2015, enterprise customer manages the electricity consumption situation of own enterprise by building energy conservation platform, carry out the technical measures such as scrap build, global optimization, improve self electricity consumption situation, thereby reach the object of energy-saving and emission-reduction. Be directed to enterprise's Techniques for Prediction of Electric Loads, the emphasis point predicting the outcome is intelligent demand control, and directly load prediction results being applied to energy conservation and demand control becomes another problem that needs solution.
Summary of the invention
For solving above-mentioned existing problems, the present invention be divided into apply the load forecasting method of neutral net and directly application load predict the outcome and carry out two parts of demand control.
A kind of business electrical Short Term Load method based on k-mean cluster radial basis (RBF) function neutral net, comprises that historical load data acquisition step, meteorological data obtaining step, date discriminating step, neural network prediction step, error calculating correction step, load curve plot step and prediction data derive step.
Historical load data acquisition step, is mainly used energy management system (EMS) SCADA real-time data base, obtains whole industrial enterprise power utilization load data; Meteorological data obtaining step mainly obtains meteorological data and comprises historical meteorological data and to be measured day weather prognosis; The polytypes such as date discriminating step is mainly passed through date day to be measured, and identification and classification prediction, is mainly divided into common working day, partly overtime work day, and complete day off; Neural network prediction step, as Forecasting Methodology core cell, according to historical meteorology and load data, carries out load forecasting model by k-mean cluster radial basis RBF function neutral net; Error is calculated and is revised step, contrasts in real time predicted value and actual value, judges whether to exist wrong distortion, obtains error and revises in time data; Load curve plot step and prediction data derive step, the load data obtaining is intuitively displayed, and can export to database from predicting platform, then for demand control.
Prediction is carried out based on k mean cluster radial basis RBF function neutral net, RBF neutral net is structurally made up of input layer, hidden layer and output layer conventionally as shown in Figure 2, belong to Multilayer Feedforward Neural Networks, what the action function of RBF network adopted is RBF, it has the feature of output to parameter local linear, network training can be avoided nonlinear optimization, thereby does not have local minimum problem. In learning process, can determine topology of networks, the complexity that network weight coefficient is calculated is minimized, and learning process is accelerated. In RBF neutral net, hidden layer is to adopt nonlinear optimization strategy, and the parameter of mapping function is adjusted, and pace of learning is slow relatively; Output layer is application linear optimization strategy, linearity power is adjusted, thereby pace of learning is very fast.
Hidden layer RBF, adopts Gaussian function
In formula: || || be European norm,WeiRBF center, δ is neuron XiMean square deviation, be fixed as
In formula: dm---the ultimate range between selected center
M---middle calculation (being hidden layer unit number)
Be output as
In formula: ωiFor implicit unit is to the weights between output unit.
Because K-means clustering algorithm has succinct and high efficiency, so it is most widely used general in all clustering algorithms. K-means clustering algorithm can be used for adjusting cluster centre in the choosing of RBF neutral net center, and makes choosing of network center more accurate. The application is by k-means clustering algorithm, and the input vector that all input layers are obtained carries out cluster, can obtain the Basis Function Center vector in hidden layer. Concrete steps are:
1) the initial center C of given each hidden nodei(0), conventionally get front k value of input sample vector.
2) calculating Euclidean distance is
di(t)＝||x(t)-Ci(t-1)||,i＝1,2,…,k
3) obtaining minimum range node is
di(t)＝mindi(t)
4) adjusting center of gravity is
Ci(t)＝Ci(t-1),1≤i≤k,i≠r
Ci(t)＝Cr(t-1)＋β[x(t)-Cr(t-1)],i＝r
5) differentiate, if Ci(t)＝Ci(t-1), termination of iterations, otherwise turn (2)
6) work as CiAfter determining, can adopt least square method to ask the weights between hidden layer and output layer,
Complete whole load prediction work. Mainly by several main points once:
1. data acquisition and pretreatment
1) obtain historical load data and corresponding historical meteorological data, wherein load data derives from energy management system SCADA database, corresponding meteorological data derives from weather station actual measurement on the same day, and meteorological data comprises: daily maximum temperature, day lowest temperature, maximal humidity, minimum humidity and maximum wind velocity;
2) meteorological effect factor is used human comfort as standard, human comfort (ssd) computing formula is: ssd=(1.818t+18.18) (0.88+0.002f)+(t-32)/(45-t)-3.2v+18.2 wherein t be temperature on average, f is relative humidity, and v is wind speed;
3) historical load data are got a point for 15 minutes, and whole day is totally 96 points, comprise all load datas on the same day as a certain day data of input quantity, and the same day, meteorologic factor human comfort and a needs prediction day human comfort, did normalized by data;
4) historical data is divided three classes according to date type in conjunction with industrial enterprise's actual conditions, as similar day input quantity, be respectively common working day, part overtime work day (being mainly Saturday), and complete day off (comprising the festivals or holidays such as Sunday and long holidays on National Day), by the method for cluster, set up 3 kinds of similar day models for prediction;
2. set up forecast model
1) determine the prediction day date, and definite date type, thereby corresponding similar day forecast model used;
2) input quantity is imported to model, utilize k-mean cluster radial basis RBF function neutral net to be predicted the outcome;
3. error correction
Obtained preliminary load prediction results is proofreaied and correct by VEC, specifically will obtain load prediction results and compare in actual load, obtain the margin of error, after stack, revise forecast model, thereby predicted the outcome;
4. load curve is drawn and prediction data derivation
Prediction data imports database, and draws load curve by interface, prediction data is displayed intuitively, as the foundation of demand control.
The present invention discloses a kind of demand control method based on load prediction, obtain predicted value by load prediction, in conjunction with industrial enterprise's production schedule and power load service condition, find maximum demand main structure because of. Control point place before load prediction value reaches maximum demand, carries out demand control, judges whether to excise nonessential load or takes energy-saving scheme. It is enough according to adjusting next period output plan and effectively controlling next period maximum demand that this control mode allows production schedule maker have.
Also can realize automated intelligent control simultaneously, in the time that prediction requirement arrives the maximum tolerance limit of enterprise, system is by interior requirement curve model criterion of establishing, automatically for current most suitable energy-saving scheme is selected by enterprise, control in advance maximum demand, effectively avoid the generations such as overload operation even trips, ensure energy-saving safe production.
The invention has the advantages that:
1) hourly weather factors is considered in load prediction, has adopted human comfort formula to weigh the correlation of the meteorologic parameters such as the temperature of every day, humidity, wind speed and power load, has improved accuracy and the reasonability of prediction;
2) adopt according to the method for date type cluster, be divided into multiple similar day model according to industrial enterprise is actual, use the model of prediction day corresponding similar day of institute to predict, prediction accuracy is high;
3) the K-means clustering algorithm RBF neural network algorithm adopting, the neuron of the input layer of training network is whole day business electrical load totally 96 points of N days in similar day model, and the same day consider the human comfort data of meteorologic factor with prediction day, and input quantity is carried out to circuit training, adopt the neutral net training, prediction obtains the prediction every 15 minutes load values of day again;
4) demand control taking load prediction as foundation, can intelligence realize sharing of load and optimal control, reduces peak load, effectively avoids tripping operation to wait the loss producing.
Brief description of the drawings
Fig. 1 is business electrical Short Term Load and the demand control technology modules schematic diagram of the embodiment of the present invention.
Fig. 2 is the RBF neural network structure schematic diagram of the embodiment of the present invention.
Fig. 3 is load prediction and the demand control schematic diagram of the embodiment of the present invention.
Fig. 4 is the load demand control curve synoptic diagram of the embodiment of the present invention.
Fig. 5 is that the demand control curve model of the embodiment of the present invention judges schematic diagram.
Detailed description of the invention
Below in conjunction with specific embodiments and the drawings, technical scheme of the present invention is described further.
As shown in Figure 1, be the business electrical Short Term Load of the embodiment of the present invention based on k mean cluster radial basis RBF function neutral net and the module frame chart of demand control technical method. The concrete process of implementing is
1. obtain similar with predicted composition day day model of data pretreatment
Within similar day of prediction day, refer to, with the date this prediction day with same type, and within the same period, load variations presents similar rule of conversion to prediction day. Because the moment of load generation every day " sudden change " is incomplete same, so in the time that load is undergone mutation, the predicated error of load also may be very large. In nearest same period several of the same type day range prediction day, the load of similar day can be to present close Changing Pattern. Therefore carry out load prediction by similar day data and can improve the precision predicting the outcome.
According to the enterprise of embodiment, according to the date type similar day model that is divided three classes, be respectively common working day, part overtime work day (being mainly Saturday), and day off (comprising the festivals or holidays such as Sunday and long holidays on National Day) completely. By the method for cluster, the history data set of predicted composition.
Among model, history data set, except load data, also should be considered meteorological data. In the embodiment of the present invention taking 5 data such as daily maximum temperature, day lowest temperature, maximal humidity, minimum humidity and maximum wind velocities as main reference data. Because meteorologic factor is comparatively complicated on power load impact, and then use human comfort as standard. Human comfort is temperature, the human comfort of the combined influences such as humidity and wind speed, in the present embodiment, closely-related with meteorologic factor is air conditioner load, thus weigh with human comfort more accurate. Human comfort (ssd) computing formula is:
ssd＝(1.818t+18.18)(0.88+0.002f)+(t-32)/(45-t)-3.2v+18.2
Wherein t is temperature on average, and f is relative humidity, and v is wind speed.
Every day, load data when 1-24, got a point for every 15 minutes, and whole day is totally 96 points, as historical load input quantity. A certain day data comprise all load datas on the same day, and the same day, meteorologic factor human comfort and need a prediction day human comfort, did normalized by data, to eliminate unnecessary factor impact.
2. utilize k-mean cluster radial basis RBF function neural network model to complete load prediction
Adopt k-means clustering algorithm to predict RBF (RBF) network, radial basis RBF function neural network model as shown in Figure 2. By k-means clustering algorithm, all input samples are unified to cluster, try to achieve the RBF central value C of all hidden layer nodes, and carry out the weights adjustment of RBF network by least square method (LMS), by setting up RBF neural network model, use MATLAB software to carry out prediction and calculation work.
The neuron of the input layer of training network is whole day business electrical load totally 96 points of N days in similar day model, and the human comfort data of the same day and prediction day. The load data that training network process is the original data set of a day and obtains second day, carries out circuit training by input quantity, then adopts the neutral net training, prediction to obtain the load value sequence of prediction day. Taking embodiment as example, first obtain Japan-China similar day group whole day load value of selected common work and the same day human comfort, each daily load data and the same day and second day human comfort are a list entries totally 98 data, the data group of front n-1 day is as neutral net input quantity, the every 15 minutes prediction load datas of n day are neutral net output quantity, circulation, trains this neutral net according to this; Then adopt the neutral net training, obtain the prediction load value sequence of day.
Fig. 3 is load prediction and the demand control schematic diagram of the embodiment of the present invention.
3. error correction process and obtaining of finally predicting the outcome
Obtained preliminary load prediction results is proofreaied and correct by VEC, be specially after obtaining actual load data, to obtain load prediction results compares in actual load, calculate the error amount obtaining between load prediction data and actual load data, and error amount is fed back to neutral net. Repeatedly after comparison, form VEC, after tentative prediction data are proofreaied and correct by VEC, obtain less the predicting the outcome of error, and result is imported to database.
4. load prediction data and curves is drawn and intelligent demand control
Draw load curve by corresponding display interface, prediction data is displayed intuitively, the contrast of the real time data of importing into as SCADA system, the foundation of formation demand control.
Use the demand control technology based on load prediction proposed by the invention, effectively realize intelligent demand control. As shown in Figure 3, according to the load prediction data that obtain, in conjunction with requirement setting, status of equipment and the production schedule, can complete effective demand control work, mainly refer to sharing of load. And sharing of load is divided into:
1) ensure type load, i.e. the load of necessary guarantee, as base lighting-load etc.;
2) regulate type load, the load that can regulate, as decorative lighting, air conditioner load etc.;
3) translation type load, can avoid the peak hour at the load of operation, as the plan of can arranging production, at paddy phase some main equipment that reruns.
Control point place before load prediction value reaches maximum demand, as shown in Fig. 4 curve map, the load demand control curve synoptic diagram of the embodiment of the present invention, by demand control, Optimum Regulation, makes actual load can not be greater than maximum demand. The demand control of intelligence, system is by interior requirement curve model criterion of establishing, automatically for selecting current most suitable energy-saving scheme, as shown in Figure 5, select to operate a switch circuit according to drawing brake string road collection, current environment condition under circuit priority, present mode, excise some nonessential loads and carry out demand control, effectively avoid the generations such as overload operation even trips, ensure energy-saving safe production. Fig. 5 is that the demand control curve model of the embodiment of the present invention judges schematic diagram.
Although the invention is implemented preferably and is disclosed as above; but they are not for limiting the invention; anyly have the knack of this skill person; not departing from the spirit and scope of the invention; can make various changes or retouch from working as, what therefore the protection domain of the invention should be defined with the application's claim protection domain is as the criterion.
Claims (8)
1. the business electrical load forecasting method based on k-mean cluster radial basis (RBF) function neutral net, is characterized in that: the method comprises historical load data acquisition step, meteorological data obtaining step, date discriminating step, neural network prediction step, step, load curve plot step and prediction data are revised in error calculating derives step.
2. business electrical load forecasting method according to claim 1, it is characterized in that: historical load data are used energy management system (EMS) SCADA real-time data base, obtain whole industrial enterprise power utilization load data, meteorological data is weighed by human comfort, whole historical data normalized; Use the method for similar day cluster, be divided into common working day according to the date, part overtime work day, and completely day off polytype, and differentiate and distinguish by the date; Error is calculated and is revised step, contrasts in real time predicted value and actual value, judges whether to exist wrong distortion, obtains error and revises in time data; Load curve plot step and prediction data derive step, the load data obtaining is intuitively displayed, and can export to database from predicting platform, then for demand control.
3. business electrical load forecasting method according to claim 1, it is characterized in that: adopt K-means clustering algorithm to be optimized radial basis (RBF) function neutral net, by K-means clustering algorithm, all input samples are unified to cluster, RBF central value that must all hidden layer nodes, and carry out the weights adjustment of RBF network by least square method (LMS), carry out prediction and calculation work by setting up RBF neural network model; K-means clustering algorithm is used for adjusting cluster centre in the choosing of radial basis (RBF) function neutral net center, and makes choosing of network center more accurate.
4. according to the business electrical load forecasting method one of claims 1 to 3 Suo Shu, it is characterized in that:
Radial basis RBF function neutral net is made up of input layer, hidden layer and output layer, belongs to Multilayer Feedforward Neural Networks, and hidden layer is to adopt nonlinear optimization strategy, and the parameter of mapping function is adjusted; Output layer is application linear optimization strategy, and linearity power is adjusted;
Hidden layer RBF, adopts Gaussian function
In formula: || || be European norm,WeiRBF center, δ is neuron XiMean square deviation, be fixed as
In formula: dm---the ultimate range between selected center
M---middle calculation (being hidden layer unit number)
Be output as
In formula: ωiFor implicit unit is to the weights between output unit.
5. business electrical Short Term Load technical method according to claim 4, is characterized in that: by K-means clustering algorithm, the input vector that all input layers are obtained carries out cluster, obtains the Basis Function Center vector in hidden layer;
Concrete steps are:
The initial center C of step 1, given each hidden nodei(0), conventionally get front k value of input sample vector;
Step 2, calculating Euclidean distance are
di(t)＝||x(t)-Ci(t-1)||,i＝1,2,…,k
Step 3, obtain minimum range node and be
di(t)＝mindi(t)
Step 4, adjustment center of gravity are
Ci(t)＝Ci(t-1),1≤i≤k,i≠r
Ci(t)＝Cr(t-1)+β[x(t)-Cr(t-1)],i＝r
Step 5, differentiation, if Ci(t)＝Ci(t-1), termination of iterations, otherwise go to step 2;
Step 6, work as CiAfter determining, can adopt least square method to ask the weights between hidden layer and output layer,
Complete whole load prediction work.
6. business electrical load forecasting method according to claim 4, is characterized in that: meteorological effect factor is used human comfort as standard, and human comfort (ssd) computing formula is:
ssd＝(1.818t+18.18)(0.88+0.002f)+(t-32)/(45-t)-3.2v+18.2；
Wherein t is temperature on average, and f is relative humidity, and v is wind speed.
7. the demand control method based on load prediction, is characterized in that: obtain predicted value by load prediction, in conjunction with industrial enterprise's production schedule and power load service condition, find maximum demand main structure because of; Control point place before load prediction value reaches maximum demand, carries out demand control, judges whether to excise nonessential load or takes energy-saving scheme.
8. the demand control method based on load prediction according to claim 7, it is characterized in that: according to the load prediction data that obtain, set in conjunction with requirement, status of equipment and the production schedule, complete effective demand control work, comprise and ensure type load control, regulate type load control and the control of translation type load, the demand control of intelligence, system is by interior requirement curve model criterion of establishing, automatically for selecting current most suitable energy-saving scheme, according to circuit priority, drawing brake string road collection under present mode, current environment condition selects to operate a switch circuit, excise some nonessential loads and carry out demand control, effectively avoid the overload operation generation of even tripping, guarantee energy-saving safe is produced.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511018470.1A CN105590174A (en) | 2015-12-29 | 2015-12-29 | Enterprise power consumption load prediction method based on K-means clustering RBF neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511018470.1A CN105590174A (en) | 2015-12-29 | 2015-12-29 | Enterprise power consumption load prediction method based on K-means clustering RBF neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105590174A true CN105590174A (en) | 2016-05-18 |
Family
ID=55929740
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511018470.1A Pending CN105590174A (en) | 2015-12-29 | 2015-12-29 | Enterprise power consumption load prediction method based on K-means clustering RBF neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105590174A (en) |
Cited By (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106203702A (en) * | 2016-07-11 | 2016-12-07 | 深圳库博能源科技有限公司 | A kind of cold load autocontrol method of electrically-based requirement prediction |
| CN106651200A (en) * | 2016-12-29 | 2017-05-10 | 中国西电电气股份有限公司 | Electrical load management method and system for industrial enterprise aggregate user |
| CN106886591A (en) * | 2017-02-21 | 2017-06-23 | 华南理工大学 | A kind of intelligent road-lamp energy consumption analysis system based on data mining |
| CN107025497A (en) * | 2017-03-20 | 2017-08-08 | 南京邮电大学 | A kind of electric load method for early warning and device based on Elman neutral nets |
| CN107273997A (en) * | 2016-06-30 | 2017-10-20 | 国网江苏省电力公司南通供电公司 | The Forecasting Methodology of platform area daily power consumption based on all a few coefficient linear regression model (LRM)s |
| CN107330518A (en) * | 2017-06-21 | 2017-11-07 | 国家电网公司 | Energy management control method and system based on temperature adjustment load prediction |
| CN107563575A (en) * | 2017-10-11 | 2018-01-09 | 国网湖南省电力公司 | Long-term wind power power forecasting method under more meteorological variables |
| CN107818340A (en) * | 2017-10-25 | 2018-03-20 | 福州大学 | Two-stage Air-conditioning Load Prediction method based on K value wavelet neural networks |
| CN108053066A (en) * | 2017-12-12 | 2018-05-18 | 中国西电电气股份有限公司 | The Forecasting Methodology of adjustable electric load in industrial production user's energy management system |
| CN108090619A (en) * | 2017-12-28 | 2018-05-29 | 广东电网有限责任公司 | The research of source net lotus storage collaborative planning |
| CN108615091A (en) * | 2018-04-20 | 2018-10-02 | 浙江大学 | Electric power meteorology load data prediction technique based on cluster screening and neural network |
| CN108734355A (en) * | 2018-05-24 | 2018-11-02 | 国网福建省电力有限公司 | A kind of short-term electric load method of parallel prediction and system applied to power quality harnessed synthetically scene |
| CN108805193A (en) * | 2018-06-01 | 2018-11-13 | 广东电网有限责任公司 | A kind of power loss data filling method based on mixed strategy |
| CN109002937A (en) * | 2018-09-07 | 2018-12-14 | 深圳供电局有限公司 | Load Forecasting, device, computer equipment and storage medium |
| CN109376772A (en) * | 2018-09-28 | 2019-02-22 | 武汉华喻燃能工程技术有限公司 | A kind of Combination power load forecasting method based on neural network model |
| CN109492748A (en) * | 2018-09-26 | 2019-03-19 | 广东工业大学 | A kind of Mid-long term load forecasting method for establishing model of the electric system based on convolutional neural networks |
| CN109858103A (en) * | 2019-01-10 | 2019-06-07 | 杭州市电力设计院有限公司 | Electric automobile charging station load modeling method for power distribution network |
| CN109993456A (en) * | 2019-04-11 | 2019-07-09 | 陈建华 | A kind of life integration capability scoring method based on big data and artificial intelligence |
| CN110135617A (en) * | 2019-03-31 | 2019-08-16 | 中南林业科技大学 | A kind of prediction technique of improvement short term under meteorological factor influence based on part throttle characteristics |
| CN110135655A (en) * | 2019-05-27 | 2019-08-16 | 国网上海市电力公司 | It is a kind of for determine energy source station operation control strategy method and apparatus |
| CN110288127A (en) * | 2019-05-31 | 2019-09-27 | 武汉烽火富华电气有限责任公司 | A kind of energy big data processing method |
| CN110334739A (en) * | 2019-06-05 | 2019-10-15 | 深圳供电局有限公司 | A kind of two-part component load forecasting method and device |
| CN110895721A (en) * | 2018-09-12 | 2020-03-20 | 珠海格力电器股份有限公司 | Method and device for predicting electric appliance function |
| CN111523819A (en) * | 2020-04-28 | 2020-08-11 | 重庆涪陵电力实业股份有限公司 | Energy-saving potential evaluation method considering distributed power supply output uncertainty |
| CN112232886A (en) * | 2020-10-30 | 2021-01-15 | 南方电网能源发展研究院有限责任公司 | Electricity price probability prediction method, system, computer equipment and storage medium |
| CN112529334A (en) * | 2020-12-25 | 2021-03-19 | 山东鲁能软件技术有限公司 | Energy utilization load closed-loop control method and system for comprehensive energy user |
| CN112686449A (en) * | 2020-12-31 | 2021-04-20 | 国网河南省电力公司经济技术研究院 | Residential electricity consumption prediction method based on human body weather comfort index |
| CN113191574A (en) * | 2021-05-28 | 2021-07-30 | 上海申瑞继保电气有限公司 | Daily electricity prediction method for single product production line |
| CN114781685A (en) * | 2022-03-17 | 2022-07-22 | 广西电网有限责任公司 | Big user power load prediction method and system based on big data mining technology |
| CN115660225A (en) * | 2022-12-13 | 2023-01-31 | 浙江万胜智能科技股份有限公司 | Electricity load prediction management method and system based on ammeter communication module |
| WO2023088314A1 (en) * | 2021-11-16 | 2023-05-25 | 王树松 | Object classification method, apparatus and device, and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093285A (en) * | 2013-01-22 | 2013-05-08 | 清华大学 | Short-term load forecast method based on artificial neural network |
-
2015
- 2015-12-29 CN CN201511018470.1A patent/CN105590174A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093285A (en) * | 2013-01-22 | 2013-05-08 | 清华大学 | Short-term load forecast method based on artificial neural network |
Non-Patent Citations (3)
| Title |
|---|
| 240829544: "径向基函数神经网络", 《HTTPS://WENKU.BAIDU.COM/VIEW/3A5C357BF78A6529657D5312.HTML》 * |
| 刘昊: "模糊聚类与径向基网络相结合的短期负荷预测", 《西安工业大学学报》 * |
| 郭金伟等: "基于RBF改进模型的电力系统短期负荷预测", 《ELECTRIC POWER IT》 * |
Cited By (39)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107273997B (en) * | 2016-06-30 | 2024-02-02 | 国网江苏省电力公司南通供电公司 | Method and system for predicting daily electricity consumption of platform area based on frequency coefficient linear regression model |
| CN107273997A (en) * | 2016-06-30 | 2017-10-20 | 国网江苏省电力公司南通供电公司 | The Forecasting Methodology of platform area daily power consumption based on all a few coefficient linear regression model (LRM)s |
| CN106203702B (en) * | 2016-07-11 | 2020-07-28 | 深圳库博能源科技有限公司 | Cold machine load automatic control method based on power demand prediction |
| CN106203702A (en) * | 2016-07-11 | 2016-12-07 | 深圳库博能源科技有限公司 | A kind of cold load autocontrol method of electrically-based requirement prediction |
| CN106651200A (en) * | 2016-12-29 | 2017-05-10 | 中国西电电气股份有限公司 | Electrical load management method and system for industrial enterprise aggregate user |
| CN106886591A (en) * | 2017-02-21 | 2017-06-23 | 华南理工大学 | A kind of intelligent road-lamp energy consumption analysis system based on data mining |
| CN107025497A (en) * | 2017-03-20 | 2017-08-08 | 南京邮电大学 | A kind of electric load method for early warning and device based on Elman neutral nets |
| CN107330518A (en) * | 2017-06-21 | 2017-11-07 | 国家电网公司 | Energy management control method and system based on temperature adjustment load prediction |
| CN107563575A (en) * | 2017-10-11 | 2018-01-09 | 国网湖南省电力公司 | Long-term wind power power forecasting method under more meteorological variables |
| CN107818340A (en) * | 2017-10-25 | 2018-03-20 | 福州大学 | Two-stage Air-conditioning Load Prediction method based on K value wavelet neural networks |
| CN108053066A (en) * | 2017-12-12 | 2018-05-18 | 中国西电电气股份有限公司 | The Forecasting Methodology of adjustable electric load in industrial production user's energy management system |
| CN108090619A (en) * | 2017-12-28 | 2018-05-29 | 广东电网有限责任公司 | The research of source net lotus storage collaborative planning |
| CN108615091A (en) * | 2018-04-20 | 2018-10-02 | 浙江大学 | Electric power meteorology load data prediction technique based on cluster screening and neural network |
| CN108734355A (en) * | 2018-05-24 | 2018-11-02 | 国网福建省电力有限公司 | A kind of short-term electric load method of parallel prediction and system applied to power quality harnessed synthetically scene |
| CN108805193A (en) * | 2018-06-01 | 2018-11-13 | 广东电网有限责任公司 | A kind of power loss data filling method based on mixed strategy |
| CN108805193B (en) * | 2018-06-01 | 2022-04-26 | 广东电网有限责任公司 | Electric power missing data filling method based on hybrid strategy |
| CN109002937A (en) * | 2018-09-07 | 2018-12-14 | 深圳供电局有限公司 | Load Forecasting, device, computer equipment and storage medium |
| CN110895721A (en) * | 2018-09-12 | 2020-03-20 | 珠海格力电器股份有限公司 | Method and device for predicting electric appliance function |
| CN109492748A (en) * | 2018-09-26 | 2019-03-19 | 广东工业大学 | A kind of Mid-long term load forecasting method for establishing model of the electric system based on convolutional neural networks |
| CN109376772A (en) * | 2018-09-28 | 2019-02-22 | 武汉华喻燃能工程技术有限公司 | A kind of Combination power load forecasting method based on neural network model |
| CN109376772B (en) * | 2018-09-28 | 2021-02-23 | 武汉华喻燃能工程技术有限公司 | Power load combination prediction method based on neural network model |
| CN109858103A (en) * | 2019-01-10 | 2019-06-07 | 杭州市电力设计院有限公司 | Electric automobile charging station load modeling method for power distribution network |
| CN109858103B (en) * | 2019-01-10 | 2023-10-31 | 杭州市电力设计院有限公司 | Electric vehicle charging station load modeling method for power distribution network |
| CN110135617A (en) * | 2019-03-31 | 2019-08-16 | 中南林业科技大学 | A kind of prediction technique of improvement short term under meteorological factor influence based on part throttle characteristics |
| CN109993456A (en) * | 2019-04-11 | 2019-07-09 | 陈建华 | A kind of life integration capability scoring method based on big data and artificial intelligence |
| CN110135655A (en) * | 2019-05-27 | 2019-08-16 | 国网上海市电力公司 | It is a kind of for determine energy source station operation control strategy method and apparatus |
| CN110135655B (en) * | 2019-05-27 | 2023-06-23 | 国网上海市电力公司 | Method and device for determining operation control strategy of energy station |
| CN110288127A (en) * | 2019-05-31 | 2019-09-27 | 武汉烽火富华电气有限责任公司 | A kind of energy big data processing method |
| CN110334739A (en) * | 2019-06-05 | 2019-10-15 | 深圳供电局有限公司 | A kind of two-part component load forecasting method and device |
| CN111523819B (en) * | 2020-04-28 | 2023-04-21 | 重庆涪陵电力实业股份有限公司 | Energy-saving potential evaluation method considering uncertainty of output power of distributed power supply |
| CN111523819A (en) * | 2020-04-28 | 2020-08-11 | 重庆涪陵电力实业股份有限公司 | Energy-saving potential evaluation method considering distributed power supply output uncertainty |
| CN112232886A (en) * | 2020-10-30 | 2021-01-15 | 南方电网能源发展研究院有限责任公司 | Electricity price probability prediction method, system, computer equipment and storage medium |
| CN112529334A (en) * | 2020-12-25 | 2021-03-19 | 山东鲁能软件技术有限公司 | Energy utilization load closed-loop control method and system for comprehensive energy user |
| CN112686449A (en) * | 2020-12-31 | 2021-04-20 | 国网河南省电力公司经济技术研究院 | Residential electricity consumption prediction method based on human body weather comfort index |
| CN113191574A (en) * | 2021-05-28 | 2021-07-30 | 上海申瑞继保电气有限公司 | Daily electricity prediction method for single product production line |
| WO2023088314A1 (en) * | 2021-11-16 | 2023-05-25 | 王树松 | Object classification method, apparatus and device, and storage medium |
| CN114781685A (en) * | 2022-03-17 | 2022-07-22 | 广西电网有限责任公司 | Big user power load prediction method and system based on big data mining technology |
| CN114781685B (en) * | 2022-03-17 | 2024-01-09 | 广西电网有限责任公司 | Large user electricity load prediction method and system based on big data mining technology |
| CN115660225A (en) * | 2022-12-13 | 2023-01-31 | 浙江万胜智能科技股份有限公司 | Electricity load prediction management method and system based on ammeter communication module |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105590174A (en) | Enterprise power consumption load prediction method based on K-means clustering RBF neural network | |
| Yang et al. | Day-ahead wind power forecasting based on the clustering of equivalent power curves | |
| CN103390116B (en) | Use the photovoltaic power station power generation power forecasting method of stepping mode | |
| CN110417015A (en) | Micro-capacitance sensor Multiobjective Optimal Operation method and system based on Model Predictive Control | |
| CN106651200A (en) | Electrical load management method and system for industrial enterprise aggregate user | |
| CN110474339B (en) | Power grid reactive power control method based on deep power generation load prediction | |
| WO2023004838A1 (en) | Wind power output interval prediction method | |
| CN107301472A (en) | Distributed photovoltaic planing method based on scene analysis method and voltage-regulation strategy | |
| Yuan et al. | A hybrid prediction-based microgrid energy management strategy considering demand-side response and data interruption | |
| CN103218675A (en) | Short-term load prediction method based on clustering and sliding window | |
| CN105069525A (en) | All-weather 96-point daily load curve prediction and optimization correction system | |
| CN102184453A (en) | Wind power combination predicting method based on fuzzy neural network and support vector machine | |
| CN104899665A (en) | Wind power short-term prediction method | |
| CN101888087A (en) | Method for realizing distributed super-short-term area load forecasting in distribution network terminal | |
| CN104700158A (en) | Energy management method and system for power distribution park | |
| CN105069521A (en) | Photovoltaic power plant output power prediction method based on weighted FCM clustering algorithm | |
| CN102938562B (en) | Prediction method of total wind electricity power in area | |
| CN107706932B (en) | A kind of energy method for optimizing scheduling based on dynamic self-adapting fuzzy logic controller | |
| Sülo et al. | Energy efficient smart buildings: LSTM neural networks for time series prediction | |
| CN110380444A (en) | Distributing wind-powered electricity generation orderly accesses the method for planning capacity of power grid under a kind of more scenes based on structure changes Copula | |
| CN114638502A (en) | Power distribution network line reinforcement strategy considering flexibility of demand side resources | |
| CN110298494A (en) | A kind of wind power forecasting method based on Segment Clustering and Combinatorial Optimization | |
| CN108346009A (en) | A kind of power generation configuration method and device based on user model self study | |
| CN109858668B (en) | Coordination prediction method for power load region in thunder and lightning climate | |
| CN110135617A (en) | A kind of prediction technique of improvement short term under meteorological factor influence based on part throttle characteristics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160518 |
|
| RJ01 | Rejection of invention patent application after publication |