CN103984988B - Light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method - Google Patents
Light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method Download PDFInfo
- Publication number
- CN103984988B CN103984988B CN201410188522.9A CN201410188522A CN103984988B CN 103984988 B CN103984988 B CN 103984988B CN 201410188522 A CN201410188522 A CN 201410188522A CN 103984988 B CN103984988 B CN 103984988B
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- model
- data
- photovoltaic
- 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.)
- Active
Links
- 238000012937 correction Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims abstract description 28
- 241001123248 Arma Species 0.000 title claims abstract description 26
- 238000012544 monitoring process Methods 0.000 claims abstract description 19
- 238000005259 measurement Methods 0.000 claims abstract description 12
- 238000012549 training Methods 0.000 claims description 8
- 230000005855 radiation Effects 0.000 claims description 6
- 238000005311 autocorrelation function Methods 0.000 claims description 3
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 3
- 230000007547 defect Effects 0.000 abstract description 2
- 241000196324 Embryophyta Species 0.000 description 24
- 230000005611 electricity Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 4
- 238000009434 installation Methods 0.000 description 4
- 230000006872 improvement Effects 0.000 description 2
- 230000001373 regressive effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 206010068052 Mosaicism Diseases 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000005622 photoelectricity Effects 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 210000003765 sex chromosome Anatomy 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Photovoltaic Devices (AREA)
Abstract
The invention discloses a kind of light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method, including input data to obtain autoregressive moving-average model parameter;Input light source monitor system data and operation monitoring system data, and according to operational monitoring real-time correction start capacity;Autoregressive moving-average model is established so as to obtain photovoltaic power ultra-short term prediction result;Introduce light simultaneous measurement station data and real time correction is carried out to photovoltaic power ultra-short term prediction result.Real time correction is carried out to photovoltaic generation power ultra-short term prediction result by introducing light simultaneous measurement station data, overcomes the defects of photovoltaic generation power ultra-short term precision of prediction is low in existing ARMA technologies, reaches the purpose of high-precision photovoltaic generation power ultra-short term prediction.
Description
Technical field
The present invention relates to photovoltaic power electric powder prediction during generation of electricity by new energy, more particularly to a kind of light-metering network
The arma modeling photovoltaic power ultra-short term prediction method of real time correction.
Background technology
Caused large-scale new energy base majority is located at " three Norths after China's photovoltaic generation enters the large-scale development stage
Area " (northwest, northeast, North China), large-scale new energy base are generally off-site from load center, and its electric power is needed through long-distance, height
Voltage is transported to load center and dissolved.Due to wind, the intermittence of light resource, randomness and fluctuation, cause extensive new
Large range of fluctuation can occur therewith for the wind-powered electricity generation of Energy Base, photovoltaic generation output, further result in power transmission network charging work(
The fluctuation of rate, a series of problems is brought to safe operation of electric network.
By in April, 2014, Photovoltaic generation installed capacity has reached 4,350,000 kilowatts, accounts for Gansu Power Grid total installation of generating capacity
13%, while Gansu turns into China's photovoltaic generation and installed largest province.At present, Gansu Power Grid wind-powered electricity generation, photovoltaic generation
1/3 of installation more than Gansu Power Grid total installation of generating capacity.As the continuous improvement of new-energy grid-connected scale, photovoltaic generation are uncertain
Problems are brought with safety and stability economical operation of the uncontrollability to power network.It is pair accurately to estimate available generating light resource
The basis of large-scale photovoltaic generation optimization scheduling.Photovoltaic generation power during photovoltaic generation is predicted, can be new energy
Source generating Real-Time Scheduling, generation of electricity by new energy are planned a few days ago, generation of electricity by new energy monthly plan, generation of electricity by new energy capability evaluation and abandon light
Electricity estimation provides key message.
ARMA (autoregressive moving-average model) is widely used in photovoltaic generation as a kind of ripe machine learning method
Power ultra-short term is predicted.Arma modeling is made up of autoregression model (AR) and moving average model (MA), using to historical power
Carry out autoregressive operation and moving average is carried out to white noise sequence to predict that the photovoltaic generation in following 0-4 hours is contributed.
ARMA methods have many good qualities, therefore are widely used in the prediction of photovoltaic generation power ultra-short term, but are exactly it the shortcomings that ARMA maximums
The hysteresis quality of prediction --- i.e. when photovoltaic generation, which is contributed, to change, the pace of change of the result of ARMA predictions is generally slower than reality
Border photovoltaic generation output pace of change.Therefore, ARMA precision of prediction is had a strong impact on.
The content of the invention
It is an object of the present invention in view of the above-mentioned problems, propose a kind of light-metering network real time correction arma modeling photovoltaic work(
Rate ultra-short term prediction method, to realize the advantages of high-precision photovoltaic generation power ultra-short term is predicted.
To achieve the above object, the technical solution adopted by the present invention is:
A kind of light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method, including input data obtain certainly
Regressive average model parameters;
Input light source monitor system data and operation monitoring system data, and opened according to operational monitoring real-time correction
Machine capacity;
Autoregressive moving-average model is established so as to obtain photovoltaic power ultra-short term prediction result;
Introduce light simultaneous measurement station data and real time correction is carried out to photovoltaic power ultra-short term prediction result.
According to a preferred embodiment of the invention, the input data, which obtains autoregressive moving-average model parameter, includes, defeated
Enter model training basic data;
Model order;
ARMA (p, the q) model parameter for determining rank is estimated using moment estimation method.
According to a preferred embodiment of the invention, the input model training basic data, input data include, history radiation
Data and historical power data.
According to a preferred embodiment of the invention, the model order is specially:
Model order, x are carried out using residual variance figure methodt+ITo need the item estimated, x1,x2,...,xtFor known history
Power sequence, for ARMA (p, q) model, model order is to determine Model Parameter p and q value;
With the gradual incremental models fitting original series of serial exponent number, residual sum of squares (RSS) is calculated every timeThen draw
Exponent number andFigure, when exponent number is by small increase,It can be remarkably decreased, after reaching true exponent numberValue can gradually tend to
Gently, or even on the contrary increase,
Actual observed value number refers to the observed value item number actually used during model of fit, for the sequence with N number of observed value
Row, AR (p) model is fitted, then the observed value actually used is up to N-p, and model parameter number refers to actual in established model
Comprising number of parameters, for the model containing average, model parameter number be model order add 1, for the sequence of N number of observation
Row, the residual error estimator of arma modeling are:
Wherein, Q is the sum of squares function of error of fitting,And θj(1≤j≤q) is model coefficient, and N is
Observation sequence length,It is the constant term in model parameter.
According to a preferred embodiment of the invention, it is described that ARMA (p, the q) model parameter for determining rank is entered using moment estimation method
Row estimation concretely comprises the following steps:
Photovoltaic plant historical power data are utilized into data sequence x1,x2,...,xtRepresent, the definition of its sample auto-covariance
For
Wherein, k=0,1,2 ..., n-1, xtAnd xt-kIt is data sequence x1,x2,...,xtIn numerical value;
Then
Then historical power data sample auto-correlation function is:
Wherein, k=0,1,2 ..., n-1;
The moments estimation of AR parts is,
Order
Then covariance function is
Use γkEstimationTo replace γk,
Parameter can be obtained
To MA (q) model coefficients θ1,θ2,...,θqHad using moments estimation
Until
Wherein l=1,2 ..., m,
M+1 equation nonlinear equation of the above, solved using iterative method and obtain autoregressive moving-average model ginseng
Number.
According to a preferred embodiment of the invention,
The smooth source monitor system data include the real-time survey that the light-metering station related to photovoltaic plant to be predicted is monitored
The photovoltaic plant average radiation of light data and numerical weather forecast data prediction, the operation monitoring system data are light to be predicted
The real-time monitoring information of overhead utility photovoltaic module, including photovoltaic DC-to-AC converter stop machine status information in real time.
According to a preferred embodiment of the invention, in addition to,
Prediction result is exported into database, and prediction result is shown by chart and curve and shows prediction and actual measurement
As a result contrast.
According to a preferred embodiment of the invention, the autoregressive moving-average model is:
Wherein,And θj(1≤j≤q) is coefficient, αtIt is white noise sequence.
According to a preferred embodiment of the invention, the introducing light simultaneous measurement station data are to photovoltaic power ultra-short term prediction result
Carrying out real time correction is specially:
If t1Moment, the photovoltaic plant average irradiance that light-metering station monitors to obtain are I1, the photovoltaic of data of weather forecast prediction
Power station average irradiance is J1, the actual output of photovoltaic plant is p1;Next time point t2Moment, data of weather forecast prediction
Photovoltaic plant average irradiance be J2, then the actual irradiation level I of photovoltaic plant2For,
I2=I1+(J2-J1)
Then the parameters revision amount of predicting power of photovoltaic plant is
According to a preferred embodiment of the invention, the final prediction result of output is:
Wherein, it is the output prediction of t-i moment photovoltaic plant;It is that t-i moment photovoltaic plants go out
Power is predictedIt is the white noise sequence of t;αt-jIt is Wt-iThe white noise sequence at moment;θjIt is coefficient, 1≤i
≤ p and 1≤j≤q
λ is weight coefficient, ItIt is the average irradiance of t photovoltaic plant
Technical scheme has the advantages that:
Technical scheme passes through introducing by being predicted to the photovoltaic generation power during photovoltaic generation
Light simultaneous measurement station data carry out real time correction to photovoltaic generation power ultra-short term prediction result, overcome photovoltaic in existing ARMA technologies
The defects of generated output ultra-short term precision of prediction is low, reach the purpose of high-precision photovoltaic generation power ultra-short term prediction.
Below by drawings and examples, technical scheme is described in further detail.
Brief description of the drawings
Fig. 1 is the light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method described in the embodiment of the present invention
Theory diagram.
Embodiment
The preferred embodiments of the present invention are illustrated below in conjunction with accompanying drawing, it will be appreciated that described herein preferred real
Apply example to be merely to illustrate and explain the present invention, be not intended to limit the present invention.
A kind of light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method, including input data obtain certainly
Regressive average model parameters;
Input light source monitor system data and operation monitoring system data, and opened according to operational monitoring real-time correction
Machine capacity;
Autoregressive moving-average model is established so as to obtain photovoltaic power ultra-short term prediction result;
Introduce light simultaneous measurement station data and real time correction is carried out to photovoltaic power ultra-short term prediction result.
As shown in figure 1, the photovoltaic generation power ultra-short term prediction that technical solution of the present invention proposes can be divided into two stages:Mould
Type training stage and power prediction stage.
Stage 1:Model training
Step 1.1:Model training basic data inputs
Photovoltaic power generation power prediction system model training required input data include history radiation data, historical power data
Deng.Basic data is input in forecast model and carries out model training.
Step 1.2:Model order
Estimation function is established due to that can not determine to need to use the item of how many known time sequences in advance, so needing pair
Model carries out determining rank judgement.
xt+1To need the item estimated, x1,x2,...,xtFor known historical power sequence, for ARMA (p, q) model, mould
Type determines the value that rank is just to determine Model Parameter p and q.
Model order is carried out using residual variance figure method.Hypothetical model is limited rank autoregression model, if the rank set
Number is less than true exponent number, then is a kind of deficiency fitting, thus regression criterion quadratic sum must be bigger than normal, now can by improving exponent number
To significantly reduce residual sum of squares (RSS)., whereas if exponent number has reached actual value, then exponent number is further added by, is exactly overfitting,
Now increase exponent number will not make residual sum of squares (RSS) be substantially reduced, or even can be increased slightly.
So with a series of exponent numbers, gradually incremental model is fitted original series, calculates residual sum of squares (RSS) every time
Then draw exponent number andFigure.When exponent number is by small increase,It can be remarkably decreased, after reaching true exponent numberValue
It can gradually tend towards stability, increase sometimes or even on the contrary.The estimator of residual variance is:
" actual observed value number " refers to the observed value item number actually used during model of fit, for N number of observed value
Sequence, be fitted AR (p) model, then the observed value actually used is up to N-p.
" model parameter number " refers to the number of parameters actually included in established model, for the mould containing average
Type, model parameter number are that model order adds 1.For the sequence of N number of observation, the residual error estimator of corresponding arma modeling is:
Step 1.3:Model parameter estimation
ARMA (p, q) model parameter is estimated using moment estimation method.First, by photovoltaic plant historical power number
According to utilization data sequence x1,x2,...,xtRepresent, its sample auto-covariance is defined as
Wherein, k=0,1,2 ..., n-1, xtAnd xt-kIt is data sequence x1,x2,...,xtIn numerical value.
Particularly,
Then historical power data sample auto-correlation function is:
Wherein, k=0,1,2 ..., n-1.
The moments estimation of AR parts is
Order
Then covariance function is
WithEstimation replace γk, have
Parameter can be obtained
To MA (q) model coefficients θ1,θ2,...,θqHad using moments estimation
……
……
……
Wherein l=1,2 ..., m
Include m+1 equation altogether above, for its parameter, equation is non-linear, is solved using iterative method.
Comprise the following steps that, equation is deformed into:
Given θ1,θ2,...,θqWithOne group of initial value, such as
More than substituting on the right of two formulas, the value obtained by the left side is first step iterative value, is designated as
The value is substituted into the right side of upper two formula successively again, just obtains second step iterative value,The like,
When adjacent iteration result twice is less than given threshold value, approximate solution of the result of gained as parameter is taken.
Found by above-mentioned solution procedure, it is desirable to solve the exponent numbers of time series models it is necessary to obtain the prediction of time series
Value;Obtain the predicted value of time series, it is necessary to first establish specific anticipation function;Establish specific anticipation function, it is necessary to
Know the exponent number of model.
According to experiment, time series models exponent number is usually no more than 5 ranks.So when the algorithm implements, can be first
First hypothesized model is 1 rank, obtains the parameter of first order modeling using the method for parameter estimation in step 1.3, and then establish estimation letter
Number can be estimated to obtain the predicted value of each in the hope of first order modeling time series models, so as to try to achieve the residual error of first order modeling
Variance;Afterwards, it is assumed that model is second order, tries to achieve the residual error of second-order model in aforementioned manners;By that analogy, 1 to 5 ranks can be obtained
The residual error of model, select exponent number of the exponent number of the minimum model of residual error as final mask.After determining model order, it can calculate
To parameter θ1,θ2,...,θqValue.
Stage 2:Power prediction
Step 2.1:Light source monitor system data input
Light source monitor system data mainly include the real-time survey that the light-metering station related to photovoltaic plant to be predicted is monitored
Light data and the photovoltaic plant average radiation of NWP (numerical weather forecast data) predictions.
Step 2.2:Operation monitoring system data input
Operation monitoring system data refer to the real-time monitoring information of photovoltaic components in photovoltaic plant to be predicted, and it is inverse mainly to include photovoltaic
Become device and stop the status informations such as switch on condition in real time.
Step 2.3:Operational monitoring real-time correction start capacity
In photovoltaic plant running, because a variety of causes can cause shutdown situation, such as typical 50,000 kilowatts of installations
Photoelectricity station, capacity of averagely starting shooting generally are known that photovoltaic electric less than 50,000 kilowatts, therefore by real-time photovoltaic operational monitoring data
The start capacity for reality of standing, rather than photovoltaic generation power ultra-short term prediction is carried out with the installed capacity of photovoltaic plant, so as to obtain
Higher precision of prediction.
Step 2.4:Photovoltaic generation power ultra-short term prediction based on arma modeling
After model parameter estimation is come out, with reference to the model order estimated, it can obtain being used for photovoltaic generation power
The time series equation of ultra-short term prediction.P the and q values drawn according to above-mentioned steps, andθ1,
θ2,...,θqValue establish autoregressive moving-average model;
Autoregressive moving-average model is as follows:
Wherein,And θj(1≤j≤q) is coefficient, αtIt is white noise sequence.
Step 2.5:Monitoring resource real-time correction photovoltaic generation power ultra-short term prediction result
Real-time change for photovoltaic generation power can be seen that by above-mentioned ARMA forecast models, above-mentioned model is always
With hysteresis quality, the present invention carries out real-time school by introducing light simultaneous measurement station data to photovoltaic generation power ultra-short term prediction result
Just.
If t1Moment, the photovoltaic plant average irradiance that light-metering station monitors to obtain are I1, NWP prediction photovoltaic plant be averaged
Irradiation level is J1, the actual output of photovoltaic plant is p1;Next time point t2Moment, the photovoltaic plant average lamp of NWP predictions
Spend for J2, then the actual irradiation level I of photovoltaic plant2For,
I2=I1+(J2-J1) (formula 15)
Then the parameters revision amount of predicting power of photovoltaic plant is
Step 2.6:Final prediction result output and displaying
Then the arma modeling photovoltaic generation power ultra-short term prediction result of light-metering network real time correction is
Wherein,
WtIt is the output prediction of t-i moment photovoltaic plant;Wt-iIt is the output prediction of t-i moment photovoltaic plant;αtIt is the white of t
Noise sequence;αt-jIt is the white noise sequence at t-j moment;θjIt is coefficient, 1≤i≤p and 1≤j≤q;
λ is weight coefficient, ItIt is the average irradiance of t photovoltaic plant.
By introducing the revised prediction irradiation level of light-metering station Real-time Monitoring Data, arma modeling can be predicted in next step
Weighting adjustment is made, so as to solve the hysteresis sex chromosome mosaicism of arma modeling prediction.
Prediction result is exported into database, and prediction result, displaying prediction and actual measurement are shown by chart and curve
As a result contrast.
Finally it should be noted that:The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention,
Although the present invention is described in detail with reference to the foregoing embodiments, for those skilled in the art, it still may be used
To be modified to the technical scheme described in foregoing embodiments, or equivalent substitution is carried out to which part technical characteristic.
Within the spirit and principles of the invention, any modification, equivalent substitution and improvements made etc., it should be included in the present invention's
Within protection domain.
Claims (6)
1. a kind of light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method, it is characterised in that including input
Data obtain autoregressive moving-average model parameter;
Input light source monitor system data and operation monitoring system data, and started shooting and held according to operational monitoring real-time correction
Amount;The smooth source monitor system data include the light simultaneous measurement data that the light-metering station related to photovoltaic plant to be predicted is monitored
And the photovoltaic plant average radiation of numerical weather forecast data prediction, the operation monitoring system data are photovoltaic plants to be predicted
The real-time monitoring information of photovoltaic module, including photovoltaic DC-to-AC converter stop machine status information in real time;
Autoregressive moving-average model is established so as to obtain photovoltaic power ultra-short term prediction result;
Introduce light simultaneous measurement station data and real time correction is carried out to photovoltaic power ultra-short term prediction result;
Also include,
Prediction result is exported into database, and prediction result is shown by chart and curve and shows prediction and measured result
Contrast;
The autoregressive moving-average model is:
Wherein,θjIt is coefficient, 1≤i≤p and 1≤j≤q, αtIt is white noise sequence;
The introducing light simultaneous measurement station data carry out real time correction to photovoltaic power ultra-short term prediction result:
If t1Moment, the photovoltaic plant average irradiance that light-metering station monitors to obtain are I1, the photovoltaic plant of data of weather forecast prediction
Average irradiance is J1, the actual output of photovoltaic plant is p1;Next time point t2Moment, the light of data of weather forecast prediction
Overhead utility average irradiance is J2, then the actual irradiation level I of photovoltaic plant2For,
I2=I1+(J2-J1)
Then the parameters revision amount of predicting power of photovoltaic plant is
Wherein I1≠0。
2. light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method according to claim 1, it is special
Sign is that the input data, which obtains autoregressive moving-average model parameter, to be included, input model training basic data;
Model order;
ARMA (p, the q) model parameter for determining rank is estimated using moment estimation method.
3. light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method according to claim 2, it is special
Sign is that the input model trains basic data, and input data includes, history radiation data and historical power data.
4. light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method according to claim 3, it is special
Sign is that the model order is specially:
Model order is carried out using residual variance figure method, specially sets xt+1To need the item estimated, x1,x2,...,xtTo be known
Historical power sequence, for ARMA (p, q) model, model order is to determine Model Parameter p and q value;
With the gradual incremental models fitting original series of serial exponent number, residual sum of squares (RSS) is calculated every timeThen exponent number is drawn
WithFigure, when exponent number is by small increase,It can be remarkably decreased, after reaching true exponent numberValue can gradually tend to be flat
It is slow, or even increase on the contrary,
Actual observed value number refers to the observed value item number actually used during model of fit, for the sequence with N number of observed value, intends
AR (p) model is closed, then the observed value actually used is up to N-p, and model parameter number refers to actually to be included in established model
Number of parameters, for the model containing average, model parameter number is that model order adds 1, for the sequence of N number of observation,
The residual error estimator of arma modeling is:
Wherein, Q is the sum of squares function of error of fitting,θjIt is model coefficient, 1≤i≤p and 1≤j≤q are model coefficients, N
It is observation sequence length,It is the constant term in model parameter.
5. light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method according to claim 4, it is special
Sign is, described to carry out estimation to ARMA (p, the q) model parameter for determining rank using moment estimation method and concretely comprise the following steps:
Photovoltaic plant historical power data are utilized into data sequence x1,x2,...,xtRepresent, its sample auto-covariance is defined as
<mrow>
<msub>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mi>k</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>n</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msub>
<mi>x</mi>
<mi>t</mi>
</msub>
<msub>
<mi>x</mi>
<mrow>
<mi>t</mi>
<mo>-</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>,</mo>
</mrow>
Wherein, k=0,1,2 ..., n-1, xtAnd xt-kIt is data sequence x1,x2,...,xtIn numerical value;
Then
Then historical power data sample auto-correlation function is:
<mrow>
<msub>
<mover>
<mi>&rho;</mi>
<mo>^</mo>
</mover>
<mi>k</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mi>k</mi>
</msub>
<msub>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mn>0</mn>
</msub>
</mfrac>
<mo>=</mo>
<mfrac>
<mrow>
<mfrac>
<mn>1</mn>
<mi>n</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msub>
<mi>x</mi>
<mi>t</mi>
</msub>
<msub>
<mi>x</mi>
<mrow>
<mi>t</mi>
<mo>-</mo>
<mi>k</mi>
</mrow>
</msub>
</mrow>
<mrow>
<mfrac>
<mn>1</mn>
<mi>n</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msubsup>
<mi>x</mi>
<mi>t</mi>
<mn>2</mn>
</msubsup>
</mrow>
</mfrac>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msub>
<mi>x</mi>
<mi>t</mi>
</msub>
<msub>
<mi>x</mi>
<mrow>
<mi>t</mi>
<mo>-</mo>
<mi>k</mi>
</mrow>
</msub>
</mrow>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msubsup>
<mi>x</mi>
<mi>t</mi>
<mn>2</mn>
</msubsup>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
Wherein, k=0,1,2 ..., n-1;
The moments estimation of AR parts is,
Order
Then covariance function is
Use γkEstimationTo replace γk,
Parameter can be obtained
To MA (q) model coefficients θ1,θ2,...,θqHad using moments estimation
Until
<mrow>
<msub>
<mi>&gamma;</mi>
<mi>l</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>y</mi>
<mi>t</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mo>-</mo>
<msub>
<mi>&theta;</mi>
<mi>l</mi>
</msub>
<mo>+</mo>
<msub>
<mi>&theta;</mi>
<mn>1</mn>
</msub>
<msub>
<mi>&theta;</mi>
<msup>
<mi>l</mi>
<mrow>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msup>
</msub>
<mo>+</mo>
<mo>...</mo>
<mo>+</mo>
<msub>
<mi>&theta;</mi>
<mrow>
<mi>q</mi>
<mo>-</mo>
</mrow>
</msub>
<msub>
<mmultiscripts>
<mi>&theta;</mi>
<mi>l</mi>
</mmultiscripts>
<mi>q</mi>
</msub>
<mo>)</mo>
</mrow>
<msubsup>
<mi>&sigma;</mi>
<mi>a</mi>
<mn>2</mn>
</msubsup>
</mrow>
Wherein l=1,2 ..., m,
M+1 equation of the above is nonlinear equation, is solved using iterative method and obtains autoregressive moving-average model parameter.
6. light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method according to claim 5, it is special
Sign is that the final prediction result of output is:
Wherein, WtIt is the output prediction of t-i moment photovoltaic plant;Wt-iIt is the output prediction of t-i moment photovoltaic plant;αtIt is t
White noise sequence;αt-jIt is the white noise sequence at t-j moment;θjIt is coefficient, 1≤i≤p and 1≤j≤q;
λ is weight coefficient, ItIt is the average irradiance of t photovoltaic plant.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410188522.9A CN103984988B (en) | 2014-05-06 | 2014-05-06 | Light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410188522.9A CN103984988B (en) | 2014-05-06 | 2014-05-06 | Light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103984988A CN103984988A (en) | 2014-08-13 |
CN103984988B true CN103984988B (en) | 2018-03-02 |
Family
ID=51276949
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410188522.9A Active CN103984988B (en) | 2014-05-06 | 2014-05-06 | Light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103984988B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108268963B (en) * | 2016-12-30 | 2022-08-19 | 中国电力科学研究院 | Self-correction-inspection photovoltaic power short-term prediction method |
CN108038569A (en) * | 2017-12-08 | 2018-05-15 | 北京汇通金财信息科技有限公司 | A kind of grid-connected power predicating method of distributed photovoltaic and system |
CN109038646B (en) * | 2018-06-29 | 2022-11-25 | 中国电力科学研究院有限公司 | Photovoltaic power generation scene construction method and system |
CN112070319B (en) * | 2020-09-21 | 2023-09-19 | 西安交通大学 | Ultra-short-term photovoltaic power generation power prediction method based on time sequence dynamic regression |
CN114253302A (en) * | 2021-12-15 | 2022-03-29 | 阳光电源(上海)有限公司 | Tracking control method and device |
CN117833346B (en) * | 2024-03-04 | 2024-05-14 | 湖南璟泰信息***有限公司 | Photovoltaic and power grid complementary direct-current energy-saving air conditioner power supply method and system |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102102626A (en) * | 2011-01-30 | 2011-06-22 | 华北电力大学 | Method for forecasting short-term power in wind power station |
CN102567809A (en) * | 2011-11-18 | 2012-07-11 | 中国电力科学研究院 | Power generation output power prediction system of photovoltaic power station |
JP2013009492A (en) * | 2011-06-23 | 2013-01-10 | Hitachi Ltd | Supply power control system, supply power control method, and supply power control program |
CN103208037A (en) * | 2013-04-26 | 2013-07-17 | 国电南瑞南京控制***有限公司 | Online correction based power prediction method applicable to new energy power station |
CN103279804A (en) * | 2013-04-29 | 2013-09-04 | 清华大学 | Super short-period wind power prediction method |
CN103390199A (en) * | 2013-07-18 | 2013-11-13 | 国家电网公司 | Photovoltaic power generation capacity/power prediction device |
CN103390197A (en) * | 2013-07-04 | 2013-11-13 | 云南电网公司北京能源新技术研究发展中心 | Photovoltaic power station power prediction system based on time period clustering |
CN103473322A (en) * | 2013-09-13 | 2013-12-25 | 国家电网公司 | Photovoltaic generation power ultra-short term prediction method based on time series model |
CN103500365A (en) * | 2013-09-18 | 2014-01-08 | 广州供电局有限公司 | Photovoltaic power generation power prediction method and system |
-
2014
- 2014-05-06 CN CN201410188522.9A patent/CN103984988B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102102626A (en) * | 2011-01-30 | 2011-06-22 | 华北电力大学 | Method for forecasting short-term power in wind power station |
JP2013009492A (en) * | 2011-06-23 | 2013-01-10 | Hitachi Ltd | Supply power control system, supply power control method, and supply power control program |
CN102567809A (en) * | 2011-11-18 | 2012-07-11 | 中国电力科学研究院 | Power generation output power prediction system of photovoltaic power station |
CN103208037A (en) * | 2013-04-26 | 2013-07-17 | 国电南瑞南京控制***有限公司 | Online correction based power prediction method applicable to new energy power station |
CN103279804A (en) * | 2013-04-29 | 2013-09-04 | 清华大学 | Super short-period wind power prediction method |
CN103390197A (en) * | 2013-07-04 | 2013-11-13 | 云南电网公司北京能源新技术研究发展中心 | Photovoltaic power station power prediction system based on time period clustering |
CN103390199A (en) * | 2013-07-18 | 2013-11-13 | 国家电网公司 | Photovoltaic power generation capacity/power prediction device |
CN103473322A (en) * | 2013-09-13 | 2013-12-25 | 国家电网公司 | Photovoltaic generation power ultra-short term prediction method based on time series model |
CN103500365A (en) * | 2013-09-18 | 2014-01-08 | 广州供电局有限公司 | Photovoltaic power generation power prediction method and system |
Non-Patent Citations (2)
Title |
---|
"甘肃电网风电功率超短期预测预报***建设";赵龙 等;《甘肃省电机工程学会2012年学术年会》;20120930;第430页第2段,第432页第1-5、10段,第433页第2段至第436页第1段 * |
"短期风电功率预测误差分布研究";刘立阳 等;《电力***保护与控制》;20130616;第41卷(第12期);第65页第1段至第67页第1段 * |
Also Published As
Publication number | Publication date |
---|---|
CN103984988A (en) | 2014-08-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103996073B (en) | Light-metering network real time correction self study arma modeling photovoltaic power Forecasting Methodology | |
CN103927695B (en) | Ultrashort-term wind power prediction method based on self study complex data source | |
CN103984988B (en) | Light-metering network real time correction arma modeling photovoltaic power ultra-short term prediction method | |
CN105429129B (en) | Intermittent energy power generation capacity confidence evaluation method considering network constraints | |
CN103679282B (en) | The Forecasting Methodology of wind power climbing | |
CN103984987B (en) | A kind of arma modeling ultrashort-term wind power prediction method of wind measurement network real time correction | |
CN103559561A (en) | Super-short-term prediction method of photovoltaic power station irradiance | |
CN104766175A (en) | Power system abnormal data identifying and correcting method based on time series analysis | |
CN103473322A (en) | Photovoltaic generation power ultra-short term prediction method based on time series model | |
Raptis et al. | Total power quality index for electrical networks using neural networks | |
CN103150635B (en) | Power equipment O&M method | |
CN103984986B (en) | The self study arma modeling ultrashort-term wind power prediction method of real time correction | |
CN103996079B (en) | Wind power weighting predication method based on conditional probability | |
CN104573876A (en) | Wind power plant short-period wind speed prediction method based on time sequence long memory model | |
CN103399491B (en) | Parameter identification method for photovoltaic module mechanism model of photovoltaic power generation system | |
CN104037790B (en) | A kind of new forms of energy based on sequential Monte Carlo simulation receive capability assessment method | |
CN103455716B (en) | A kind of power system voltage stabilization margin calculation method based on super short-period wind power prediction | |
CN109978242A (en) | The photovoltaic power generation cluster power forecasting method and device of scale are risen based on statistics | |
CN106203693A (en) | A kind of system and method for Power Output for Wind Power Field climbing event prediction | |
CN103927597A (en) | Ultra-short-term wind power prediction method based on autoregression moving average model | |
CN105373967A (en) | Game theory combined weighting-based photovoltaic power plant performance evaluation method | |
CN105225006A (en) | A kind of short-term wind-electricity power nonparametric probability forecasting method | |
CN105631550A (en) | Short-term wind power combined prediction method | |
Chen et al. | Distributional analysis for model predictive deferrable load control | |
CN103106314A (en) | Time sequence probability modeling method for output power of solar photovoltaic power supply |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |