CN112686416A - Wind power probability prediction method and system - Google Patents
Wind power probability prediction method and system Download PDFInfo
- Publication number
- CN112686416A CN112686416A CN201910986356.XA CN201910986356A CN112686416A CN 112686416 A CN112686416 A CN 112686416A CN 201910986356 A CN201910986356 A CN 201910986356A CN 112686416 A CN112686416 A CN 112686416A
- Authority
- CN
- China
- Prior art keywords
- probability
- wind power
- distribution function
- preset
- cumulative distribution
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Wind Motors (AREA)
Abstract
The invention relates to a wind power probability prediction method and a system, comprising the following steps: determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function according to the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in the historical time period; and correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function. According to the technical scheme provided by the invention, the quantile result of the wind power probability prediction can be more accurately obtained, so that the wind power can be more accurately predicted, meanwhile, the probability prediction algorithm has small calculation amount and high calculation efficiency, and the method has wide popularization and application scenes.
Description
Technical Field
The invention relates to the field of wind power probability prediction, in particular to a wind power probability prediction method and a wind power probability prediction system.
Background
The traditional wind power prediction is mainly based on deterministic prediction, and the predicted power given at each moment is a point (namely a predicted mean value), however, the prediction cannot be guaranteed to be completely accurate, so that a prediction error exists, namely the prediction has uncertainty. To describe this uncertainty, one gives the conditional probability distribution of the predicted power at each instant by probabilistic prediction.
The result form of Probability Prediction mainly includes Probability Density Function (PDF), Cumulative Distribution Function (CDF), Quantile (Quantile), and Prediction Interval (PI). The quantile is a discrete probability prediction result form common in wind power probability prediction, can be directly extracted from CDF, and the prediction interval forms an upper boundary and a lower boundary of the interval through a pair of quantiles.
After the wind power probability prediction model is deployed, with the continuous increase of new data, the model trained according to the old data is gradually not suitable for the previous prediction, the prediction quality is reduced, and modeling training needs to be performed again by combining the new data so as to adapt to the new data characteristics.
But since probabilistic predictive models are often computationally complex and time consuming, re-modeling is costly.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a wind power probability prediction method and a wind power probability prediction system.
The purpose of the invention is realized by adopting the following technical scheme:
the invention provides a wind power probability prediction method, which is improved in that the method comprises the following steps:
determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function according to the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in the historical time period;
and correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function.
Preferably, the determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted cumulative distribution function according to the quantile corresponding to the preset probability of the wind power probability predicted cumulative distribution function at each moment in the historical period includes:
step 1: determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function by utilizing the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the previous n moments in the historical time period;
step 2: correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function from the N +1 th moment to the Nth moment in the historical period by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function;
and step 3: updating quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the (N + 1) th moment to the Nth moment in the historical time period;
and 4, step 4: if the quantiles corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions from the N +1 th moment to the Nth moment in the updated historical time period simultaneously meet a first preset condition and a second preset condition, outputting regression coefficient matrixes corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions, otherwise, outputting the regression coefficient matrixes corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions to be [0 … 0 … 1, 0 ];
wherein N belongs to (1-N), N is the total time of the historical time period, and the matrix [0 … 0 … 1, 0] is a vector of 1 Xm order.
Further, the step 1 includes:
the preset probability of solving the cumulative distribution function of the wind power probability prediction shown in the following formula by using a simple search method is tauiTime-corresponding regression coefficient matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiThe coefficients of the k-th order terms of the corresponding quantiles,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiChecking function of time correspondence, ptIs the actual wind power at the historical time t, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction to be tau according to the following formulaiChecking function of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiA time-corresponding quantile polynomial matrix;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment t as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Further, the step 2 includes:
correcting the preset probability of the wind power probability prediction accumulative distribution function of the historical moment j according to the following formula to be tauiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for the wind power probability prediction of the historical moment j is tauiTime-corresponding quantile polynomial matrices, tauiThe ith preset probability of the cumulative distribution function for wind power probability prediction is represented by i ∈ (1-p), and p is the cumulative probability of the wind power probability predictionThe preset probability total number of the distribution function;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment j as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Further, the step 3 includes:
and arranging the quantiles corresponding to the preset probabilities of the wind power probability prediction accumulative distribution function at the historical moment j after correction in an ascending order, and updating the ith number in the ascending order to the quantile corresponding to the ith preset probability of the wind power probability prediction accumulative distribution function at the historical moment j.
Further, the first preset condition is determined according to the following formula:
in the formula, R2(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the N +1 th moment to the Nth moment in the updated historical period is tauiMean reliability deviation value of time-corresponding quantiles, R1(τi) Forecasting a cumulative distribution function for wind power probability forecasting from the (N + 1) th moment to the Nth moment in a historical period before correctionLet the probability be τiAverage reliability deviation values of corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the N +1 th moment to the Nth moment in the updated historical period as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles2(τi)：
In the formula, xi'j(τi) The preset probability of the cumulative distribution function for predicting the probability of the wind power at the jth moment in the updated historical time period is tauiAn indicative function of the corresponding quantiles in time, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the (N + 1) th moment to the Nth moment in the historical period before correction as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles1(τi)：
In the formula, xij(τi) The preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiAn indicative function of the corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period as tau according to the following formulaiIndication function xi 'of corresponding quantiles'j(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period is tauiA quantile corresponding to the hour;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the historical period before correction as tau according to the following formulaiIndicative function xi of quantiles corresponding to timej(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiA quantile corresponding to the hour;
determining the second preset condition according to the following formula:
in the formula, S1jQuantile loss value of cumulative distribution function for wind power probability prediction at jth moment in historical period before correction, S2jThe quantile loss value of the cumulative distribution function for wind power probability prediction at the jth moment in the updated historical time period is obtained;
determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula1j：
Determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula2j：
In the formula, j belongs to (N +1, N), N belongs to (1-N), and N is the total time in the history period.
Preferably, the correcting the quantile corresponding to the preset probability of the wind power probability predicted cumulative distribution function at the predicted time by using the regression coefficient matrix corresponding to the preset probability of the wind power probability predicted cumulative distribution function includes:
the preset probability of the wind power probability prediction accumulative distribution function of the prediction time r is corrected to be tau according to the following formulaiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for predicting the wind power probability at the predicted time r is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
wherein the prediction of the cumulative distribution function of the wind power probability prediction of the prediction time r is determined according to the following formulaLet the probability be τiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
The invention provides a wind power probability prediction system, and the improvement is that the system comprises:
the determining module is used for determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function according to the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in the historical time period;
and the correction module is used for correcting the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using the regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function.
Preferably, the determining module includes:
the first determining unit is used for determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function by utilizing the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the previous n moments in the historical time period;
the correction unit is used for correcting quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the N +1 th moment to the Nth moment in the historical period by using a regression coefficient matrix corresponding to the preset probability of the wind power probability prediction accumulative distribution function;
the updating unit is used for updating quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the (N + 1) th moment to the Nth moment in the historical time period;
the output unit is used for outputting a regression coefficient matrix corresponding to the preset probability of the wind power probability prediction cumulative distribution function if the quantile corresponding to the preset probability of the wind power probability prediction cumulative distribution function from the N +1 th moment to the Nth moment in the updated historical period meets a first preset condition and a second preset condition at the same time, otherwise, outputting the regression coefficient matrix corresponding to the preset probability of the wind power probability prediction cumulative distribution function [0 … 0 … 1, 0 ];
wherein N belongs to (1-N), N is the total time of the historical time period, and the matrix [0 … 0 … 1, 0] is a vector of 1 Xm order.
Further, the first determining unit is configured to:
the preset probability of solving the cumulative distribution function of the wind power probability prediction shown in the following formula by using a simple search method is tauiTime-corresponding regression coefficient matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiThe coefficients of the k-th order terms of the corresponding quantiles,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiChecking function of time correspondence, ptIs the actual wind power at the historical time t, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction to be tau according to the following formulaiChecking function of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiA time-corresponding quantile polynomial matrix;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment t as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Further, the correction unit is configured to:
wind power probability prediction of historical moment j is corrected according to the following formulaIs τ toiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for the wind power probability prediction of the historical moment j is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment j as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time corresponds to the k power of quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is in the historical time periodTotal number of time instants.
Further, the update unit includes:
and arranging the quantiles corresponding to the preset probabilities of the wind power probability prediction accumulative distribution function at the historical moment j after correction in an ascending order, and updating the ith number in the ascending order to the quantile corresponding to the ith preset probability of the wind power probability prediction accumulative distribution function at the historical moment j.
Further, the first preset condition is determined according to the following formula:
in the formula, R2(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the N +1 th moment to the Nth moment in the updated historical period is tauiMean reliability deviation value of time-corresponding quantiles, R1(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the (N + 1) th moment to the Nth moment in the historical period before correction is tauiAverage reliability deviation values of corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the N +1 th moment to the Nth moment in the updated historical period as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles2(τi)：
In the formula, xi'j(τi) The preset probability of the cumulative distribution function for predicting the probability of the wind power at the jth moment in the updated historical time period is tauiAn indicative function of the corresponding quantiles in time, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the (N + 1) th moment to the Nth moment in the historical period before correction as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles1(τi)：
In the formula, xij(τi) The preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiAn indicative function of the corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period as tau according to the following formulaiIndication function xi 'of corresponding quantiles'j(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period is tauiA quantile corresponding to the hour;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the historical period before correction as tau according to the following formulaiIndicative function xi of quantiles corresponding to timej(τi)：
In the formula (I), the compound is shown in the specification,
to correct forThe preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the previous historical period is tauiA quantile corresponding to the hour;
determining the second preset condition according to the following formula:
in the formula, S1jQuantile loss value of cumulative distribution function for wind power probability prediction at jth moment in historical period before correction, S2jThe quantile loss value of the cumulative distribution function for wind power probability prediction at the jth moment in the updated historical time period is obtained;
determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula1j：
Determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula2j：
In the formula, j belongs to (N +1, N), N belongs to (1-N), and N is the total time in the history period.
Preferably, the modification module is configured to:
the preset probability of the wind power probability prediction accumulative distribution function of the prediction time r is corrected to be tau according to the following formulaiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for predicting the wind power probability at the predicted time r is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the wind power probability prediction accumulative distribution function at the prediction time r as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Compared with the closest prior art, the invention has the following beneficial effects:
according to the technical scheme provided by the invention, a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function is determined according to the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in the historical time period; correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function; the quantile result of the wind power probability prediction can be obtained more accurately, the wind power can be predicted more accurately, meanwhile, the probability prediction algorithm is small in calculation amount and high in calculation efficiency, and the method has wide popularization and application scenes.
Drawings
FIG. 1 is a flow chart of a wind power probability prediction method;
FIG. 2 is a graph comparing an original probability prediction curve and an optimized probability prediction curve;
FIG. 3 is a diagram of a wind power probability prediction system.
Detailed Description
The following describes embodiments of the present invention in further detail with reference to the accompanying drawings.
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides a wind power probability prediction method, as shown in fig. 1, the method comprises the following steps:
step 101, determining a regression coefficient matrix corresponding to a preset probability of the wind power probability predicted accumulative distribution function according to a quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in a historical time period;
and 102, correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function.
In the best embodiment of the invention, the quantile form wind power probability prediction value of the prediction time output by the existing wind power probability prediction model is optimized, and a more accurate wind power probability prediction result is obtained.
Specifically, the step 101 includes:
step 1: determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function by utilizing the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the previous n moments in the historical time period;
step 2: correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function from the N +1 th moment to the Nth moment in the historical period by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function;
and step 3: updating quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the (N + 1) th moment to the Nth moment in the historical time period;
and 4, step 4: if the quantiles corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions from the N +1 th moment to the Nth moment in the updated historical time period simultaneously meet a first preset condition and a second preset condition, outputting regression coefficient matrixes corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions, otherwise, outputting the regression coefficient matrixes corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions to be [0 … 0 … 1, 0 ];
wherein N belongs to (1-N), N is the total time of the historical time period, and the matrix [0 … 0 … 1, 0] is a vector of 1 Xm order.
Further, the step 1 includes:
the preset probability of solving the cumulative distribution function of the wind power probability prediction shown in the following formula by using a simple search method is tauiTime-corresponding regression coefficient matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiThe coefficients of the k-th order terms of the corresponding quantiles,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiChecking function of time correspondence, ptIs the actual wind power at the historical time t, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
in the preferred embodiment of the present invention, there is a wide variety of software that can implement simplex search algorithms, such as the fminsearch function of MATLAB.
Determining the preset probability of the cumulative distribution function of the wind power probability prediction to be tau according to the following formulaiChecking function of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiA time-corresponding quantile polynomial matrix;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment t according to the following formulaτiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
In the preferred embodiment of the present invention, m can be different integer values, and the optimal value is selected according to the actual effect, and when m is 1, the optimal value is a unary linear regression matrix.
Further, the step 2 includes:
correcting the preset probability of the wind power probability prediction accumulative distribution function of the historical moment j according to the following formula to be tauiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for the wind power probability prediction of the historical moment j is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment j as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Further, the step 3 includes:
and arranging the quantiles corresponding to the preset probabilities of the wind power probability prediction accumulative distribution function at the historical moment j after correction in an ascending order, and updating the ith number in the ascending order to the quantile corresponding to the ith preset probability of the wind power probability prediction accumulative distribution function at the historical moment j.
Further, the first preset condition is determined according to the following formula:
in the formula, R2(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the N +1 th moment to the Nth moment in the updated historical period is tauiTime pairMean reliability deviation of corresponding quantiles, R1(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the (N + 1) th moment to the Nth moment in the historical period before correction is tauiAverage reliability deviation values of corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the N +1 th moment to the Nth moment in the updated historical period as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles2(τi)：
In the formula, xi'j(τi) The preset probability of the cumulative distribution function for predicting the probability of the wind power at the jth moment in the updated historical time period is tauiAn indicative function of the corresponding quantiles in time, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the (N + 1) th moment to the Nth moment in the historical period before correction as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles1(τi)：
In the formula, xij(τi) The preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiAn indicative function of the corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period as tau according to the following formulaiIndication function xi 'of corresponding quantiles'j(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period is tauiA quantile corresponding to the hour;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the historical period before correction as tau according to the following formulaiIndicative function xi of quantiles corresponding to timej(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiA quantile corresponding to the hour;
determining the second preset condition according to the following formula:
in the formula, S1jQuantile loss value of cumulative distribution function for wind power probability prediction at jth moment in historical period before correction, S2jThe quantile loss value of the cumulative distribution function for wind power probability prediction at the jth moment in the updated historical time period is obtained;
determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula1j：
Determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula2j：
In the formula, j belongs to (N +1, N), N belongs to (1-N), and N is the total time in the history period.
Specifically, the step 102 includes:
the preset probability of the wind power probability prediction accumulative distribution function of the prediction time r is corrected to be tau according to the following formulaiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for predicting the wind power probability at the predicted time r is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
wherein, the preset summary of the wind power probability prediction accumulative distribution function of the prediction time r is determined according to the following formulaRate of tauiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
In the best embodiment of the invention, the wind power probability prediction is mainly used for measuring the probability condition of the wind power at the prediction moment, namely providing data basis for the prediction of the wind power at the prediction moment.
When the technical scheme provided by the invention is used, online optimization is recommended to be carried out when newly-added data are accumulated for more than 20 days, on one hand, the probability prediction index calculated by too little data is unstable, and on the other hand, the problem of local overfitting easily occurs to the model.
An example of an algorithm is given below:
1) a computing platform: intel (R) core (TM) i7-6500U CPU @2.50GHz and RAM 8.00GB
2) Training data: data dimension 2253 × 21
3) The preset probabilities of the 21 wind power probability prediction accumulative distribution functions are respectively as follows:
2.5％，5％，7.5％，10％，12.5％，15％，17.5％，20％，22.5％，25％，50％，75％，77.5％，80％，82.5％，85％，87.5％，90％，92.5％，95％，97.5％
4) the head quantiles and the tail quantiles sequentially form a prediction interval of the following confidence degrees:
95％，90％，85％，80％，75％，70％，65％，60％，55％，50％
for example: the quantiles 97.5% and 2.5% make up the prediction interval of 97.5% -2.5% ═ 95%.
5) Calculation of parameter estimation 17.34s
6) The number of samples in the detection set is 2252
7)RelImp＝45.1087％，S1t＝0.027，S1t0.026, therefore, the optimized results were used.
Fig. 2 shows a prediction interval graph, prediction intervals with different confidence degrees are represented by colors with different shades, a blue star-shaped dotted line is actual power, and a red solid line is a 50% quantile.
The invention provides a wind power probability prediction system, as shown in fig. 3, the system comprises:
the determining module is used for determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function according to the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in the historical time period;
and the correction module is used for correcting the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using the regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function.
Specifically, the determining module includes:
the first determining unit is used for determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function by utilizing the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the previous n moments in the historical time period;
the correction unit is used for correcting quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the N +1 th moment to the Nth moment in the historical period by using a regression coefficient matrix corresponding to the preset probability of the wind power probability prediction accumulative distribution function;
the updating unit is used for updating quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the (N + 1) th moment to the Nth moment in the historical time period;
the output unit is used for outputting a regression coefficient matrix corresponding to the preset probability of the wind power probability prediction cumulative distribution function if the quantile corresponding to the preset probability of the wind power probability prediction cumulative distribution function from the N +1 th moment to the Nth moment in the updated historical period meets a first preset condition and a second preset condition at the same time, otherwise, outputting the regression coefficient matrix corresponding to the preset probability of the wind power probability prediction cumulative distribution function [0 … 0 … 1, 0 ];
wherein N belongs to (1-N), N is the total time of the historical time period, and the matrix [0 … 0 … 1, 0] is a vector of 1 Xm order.
Specifically, the first determining unit is configured to:
the preset probability of solving the cumulative distribution function of the wind power probability prediction shown in the following formula by using a simple search method is tauiTime-corresponding regression coefficient matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiThe coefficients of the k-th order terms of the corresponding quantiles,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiChecking function of time correspondence, ptIs the actual wind power at the historical time t, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction to be tau according to the following formulaiChecking function of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiA time-corresponding quantile polynomial matrix;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment t as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Specifically, the correction unit is configured to:
correcting the preset probability of the wind power probability prediction accumulative distribution function of the historical moment j according to the following formula to be tauiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for the wind power probability prediction of the historical moment j is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment j as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Specifically, the update unit includes:
and arranging the quantiles corresponding to the preset probabilities of the wind power probability prediction accumulative distribution function at the historical moment j after correction in an ascending order, and updating the ith number in the ascending order to the quantile corresponding to the ith preset probability of the wind power probability prediction accumulative distribution function at the historical moment j.
Specifically, the first preset condition is determined according to the following formula:
in the formula, R2(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the N +1 th moment to the Nth moment in the updated historical period is tauiMean reliability deviation value of time-corresponding quantiles, R1(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the (N + 1) th moment to the Nth moment in the historical period before correction is tauiAverage reliability deviation values of corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the N +1 th moment to the Nth moment in the updated historical period as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles2(τi)：
In the formula, xi'j(τi) The preset probability of the cumulative distribution function for predicting the probability of the wind power at the jth moment in the updated historical time period is tauiAn indicative function of the corresponding quantiles in time, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the (N + 1) th moment to the Nth moment in the historical period before correction as tau according to the following formulaiTime of day correspondingMean reliability deviation value R of quantile1(τi)：
In the formula, xij(τi) The preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiAn indicative function of the corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period as tau according to the following formulaiIndication function xi 'of corresponding quantiles'j(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period is tauiA quantile corresponding to the hour;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the historical period before correction as tau according to the following formulaiIndicative function xi of quantiles corresponding to timej(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiA quantile corresponding to the hour;
determining the second preset condition according to the following formula:
in the formula, S1jQuantile loss value of cumulative distribution function for wind power probability prediction at jth moment in historical period before correction, S2jThe quantile loss value of the cumulative distribution function for wind power probability prediction at the jth moment in the updated historical time period is obtained;
determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula1j：
Determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula2j：
In the formula, j belongs to (N +1, N), N belongs to (1-N), and N is the total time in the history period.
Specifically, the correction module is configured to:
the preset probability of the wind power probability prediction accumulative distribution function of the prediction time r is corrected to be tau according to the following formulaiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for predicting the wind power probability at the predicted time r is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the wind power probability prediction accumulative distribution function at the prediction time r as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.
Claims (14)
1. A wind power probability prediction method is characterized by comprising the following steps:
determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function according to the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in the historical time period;
and correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function.
2. The method of claim 1, wherein the determining a regression coefficient matrix corresponding to the preset probability of the cumulative distribution function of the wind power probability prediction according to the quantile corresponding to the preset probability of the cumulative distribution function of the wind power probability prediction at each time in the historical period comprises:
step 1: determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function by utilizing the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the previous n moments in the historical time period;
step 2: correcting quantiles corresponding to the preset probability of the wind power probability predicted accumulative distribution function from the N +1 th moment to the Nth moment in the historical period by using a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function;
and step 3: updating quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the (N + 1) th moment to the Nth moment in the historical time period;
and 4, step 4: if the quantiles corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions from the N +1 th moment to the Nth moment in the updated historical time period simultaneously meet a first preset condition and a second preset condition, outputting regression coefficient matrixes corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions, otherwise, outputting the regression coefficient matrixes corresponding to the preset probabilities of the wind power probability prediction cumulative distribution functions to be [0 … 0 … 1, 0 ];
wherein N belongs to (1-N), N is the total time of the historical time period, and the matrix [0 … 0 … 1, 0] is a vector of 1 Xm order.
3. The method of claim 2, wherein step 1 comprises:
the preset probability of solving the cumulative distribution function of the wind power probability prediction shown in the following formula by using a simple search method is tauiTime-corresponding regression coefficient matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiThe coefficients of the k-th order terms of the corresponding quantiles,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiChecking function of time correspondence, ptIs the actual wind power at the historical time t, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction to be tau according to the following formulaiChecking function of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiA time-corresponding quantile polynomial matrix;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment t as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
4. The method of claim 2, wherein step 2, comprises:
correcting the preset probability of the wind power probability prediction accumulative distribution function of the historical moment j according to the following formula to be tauiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for the wind power probability prediction of the historical moment j is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment j as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
5. The method of claim 2, wherein step 3, comprises:
and arranging the quantiles corresponding to the preset probabilities of the wind power probability prediction accumulative distribution function at the historical moment j after correction in an ascending order, and updating the ith number in the ascending order to the quantile corresponding to the ith preset probability of the wind power probability prediction accumulative distribution function at the historical moment j.
6. The method of claim 2, wherein the first predetermined condition is determined as follows:
in the formula, R2(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the N +1 th moment to the Nth moment in the updated historical period is tauiMean reliability deviation value of time-corresponding quantiles, R1(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the (N + 1) th moment to the Nth moment in the historical period before correction is tauiAverage reliability deviation values of corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the N +1 th moment to the Nth moment in the updated historical period as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles2(τi)：
In the formula, xij'(τi) The preset probability of the cumulative distribution function for predicting the probability of the wind power at the jth moment in the updated historical time period is tauiAn indicative function of the corresponding quantiles in time, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the (N + 1) th moment to the Nth moment in the historical period before correction as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles1(τi)：
In the formula, xij(τi) The preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiAn indicative function of the corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period as tau according to the following formulaiIndication function xi 'of corresponding quantiles'j(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period is tauiA quantile corresponding to the hour;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the historical period before correction as tau according to the following formulaiIndicative function xi of quantiles corresponding to timej(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiA quantile corresponding to the hour;
determining the second preset condition according to the following formula:
in the formula, S1jQuantile loss value of cumulative distribution function for wind power probability prediction at jth moment in historical period before correction, S2jThe quantile loss value of the cumulative distribution function for wind power probability prediction at the jth moment in the updated historical time period is obtained;
determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula1j：
Determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula2j：
In the formula, j belongs to (N +1, N), N belongs to (1-N), and N is the total time in the history period.
7. The method according to claim 1, wherein the correcting the quantile corresponding to the preset probability of the wind power probability predicted cumulative distribution function at the predicted time by using the regression coefficient matrix corresponding to the preset probability of the wind power probability predicted cumulative distribution function comprises:
wind power probability of predicted time r is corrected according to the following formulaThe predetermined probability of the cumulative distribution function of the rate prediction is tauiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for predicting the wind power probability at the predicted time r is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the wind power probability prediction accumulative distribution function at the prediction time r as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability at the predicted time r is tauiThe time is the k power of the corresponding quantile,k belongs to (1-m), and m is a preset polynomial order.
8. A wind power probability prediction system, the system comprising:
the determining module is used for determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function according to the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at each moment in the historical time period;
and the correction module is used for correcting the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the predicted time by using the regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function.
9. The system of claim 8, wherein the determination module comprises:
the first determining unit is used for determining a regression coefficient matrix corresponding to the preset probability of the wind power probability predicted accumulative distribution function by utilizing the quantile corresponding to the preset probability of the wind power probability predicted accumulative distribution function at the previous n moments in the historical time period;
the correction unit is used for correcting quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the N +1 th moment to the Nth moment in the historical period by using a regression coefficient matrix corresponding to the preset probability of the wind power probability prediction accumulative distribution function;
the updating unit is used for updating quantiles corresponding to the preset probability of the wind power probability prediction accumulative distribution function from the (N + 1) th moment to the Nth moment in the historical time period;
the output unit is used for outputting a regression coefficient matrix corresponding to the preset probability of the wind power probability prediction cumulative distribution function if the quantile corresponding to the preset probability of the wind power probability prediction cumulative distribution function from the N +1 th moment to the Nth moment in the updated historical period meets a first preset condition and a second preset condition at the same time, otherwise, outputting the regression coefficient matrix corresponding to the preset probability of the wind power probability prediction cumulative distribution function [0 … 0 … 1, 0 ];
wherein N belongs to (1-N), N is the total time of the historical time period, and the matrix [0 … 0 … 1, 0] is a vector of 1 Xm order.
10. The system of claim 9, wherein the first determination unit is to:
the preset probability of solving the cumulative distribution function of the wind power probability prediction shown in the following formula by using a simple search method is tauiTime-corresponding regression coefficient matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiThe coefficients of the k-th order terms of the corresponding quantiles,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiChecking function of time correspondence, ptIs the actual wind power at the historical time t, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction to be tau according to the following formulaiChecking function of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiA time-corresponding quantile polynomial matrix;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment t as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
11. The system of claim 9, wherein the modification unit is to:
correcting the preset probability of the wind power probability prediction accumulative distribution function of the historical moment j according to the following formula to be tauiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for the wind power probability prediction of the historical moment j is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the historical moment j as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
12. The system of claim 9, wherein the update unit comprises:
and arranging the quantiles corresponding to the preset probabilities of the wind power probability prediction accumulative distribution function at the historical moment j after correction in an ascending order, and updating the ith number in the ascending order to the quantile corresponding to the ith preset probability of the wind power probability prediction accumulative distribution function at the historical moment j.
13. The system of claim 9, wherein the first predetermined condition is determined as follows:
in the formula, R2(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the N +1 th moment to the Nth moment in the updated historical period is tauiMean reliability deviation value of time-corresponding quantiles, R1(τi) The preset probability of the cumulative distribution function for predicting the wind power probability from the (N + 1) th moment to the Nth moment in the historical period before correction is tauiAverage reliability deviation values of corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the N +1 th moment to the Nth moment in the updated historical period as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles2(τi)：
In the formula, xi'j(τi) The preset probability of the cumulative distribution function for predicting the probability of the wind power at the jth moment in the updated historical time period is tauiAn indicative function of the corresponding quantiles in time, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the cumulative distribution function of the wind power probability prediction from the (N + 1) th moment to the Nth moment in the historical period before correction as tau according to the following formulaiMean reliability deviation value R of time-corresponding quantiles1(τi)：
In the formula, xij(τi) The preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiAn indicative function of the corresponding quantiles;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period as tau according to the following formulaiIndication function xi 'of corresponding quantiles'j(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the updated historical time period is tauiA quantile corresponding to the hour;
determining the preset probability of the cumulative distribution function of the wind power probability prediction at the jth moment in the historical period before correction as tau according to the following formulaiIndicative function xi of quantiles corresponding to timej(τi)：
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability at the jth moment in the historical time period before correction is tauiA quantile corresponding to the hour;
determining the second preset condition according to the following formula:
in the formula, S1jQuantile loss value of cumulative distribution function for wind power probability prediction at jth moment in historical period before correction, S2jThe quantile loss value of the cumulative distribution function for wind power probability prediction at the jth moment in the updated historical time period is obtained;
determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula1j：
Determining a quantile loss value S of an accumulative distribution function of wind power probability prediction at the jth moment in a historical period before optimization according to the following formula2j：
In the formula, j belongs to (N +1, N), N belongs to (1-N), and N is the total time in the history period.
14. The system of claim 8, wherein the correction module is to:
the preset probability of the wind power probability prediction accumulative distribution function of the prediction time r is corrected to be tau according to the following formulaiQuantile of time correspondence
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for predicting the wind power probability is tauiA matrix of regression coefficients corresponding to the time,
the preset probability of the cumulative distribution function for predicting the wind power probability at the predicted time r is tauiTime-corresponding quantile polynomial matrices, tauiThe method comprises the steps that the ith preset probability of a cumulative distribution function predicted by the wind power probability is set, i belongs to (1-p), and p is the preset probability total number of the cumulative distribution function predicted by the wind power probability;
determining the preset probability of the wind power probability prediction accumulative distribution function at the prediction time r as tau according to the following formulaiTime-corresponding quantile polynomial matrix
In the formula (I), the compound is shown in the specification,
the preset probability of the cumulative distribution function for the wind power probability prediction at the historical moment t is tauiThe time is the k power of the corresponding quantile, k belongs to (1-m), m is a preset polynomial order, t belongs to (1-N), N belongs to (1-N), and N is the total time of the historical time period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910986356.XA CN112686416A (en) | 2019-10-17 | 2019-10-17 | Wind power probability prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910986356.XA CN112686416A (en) | 2019-10-17 | 2019-10-17 | Wind power probability prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112686416A true CN112686416A (en) | 2021-04-20 |
Family
ID=75444377
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910986356.XA Pending CN112686416A (en) | 2019-10-17 | 2019-10-17 | Wind power probability prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112686416A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150446A (en) * | 2023-10-30 | 2023-12-01 | 武汉华信数据系统有限公司 | Blower operation state identification method and device, electronic equipment and storage medium |
-
2019
- 2019-10-17 CN CN201910986356.XA patent/CN112686416A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117150446A (en) * | 2023-10-30 | 2023-12-01 | 武汉华信数据系统有限公司 | Blower operation state identification method and device, electronic equipment and storage medium |
| CN117150446B (en) * | 2023-10-30 | 2024-02-09 | 武汉华信数据系统有限公司 | Blower operation state identification method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108875916B (en) | Advertisement click rate prediction method based on GRU neural network | |
| CN108959728B (en) | Radio frequency device parameter optimization method based on deep learning | |
| CN113037531A (en) | Flow prediction method, device and storage medium | |
| CN110942248B (en) | Training method and device for transaction wind control network and transaction risk detection method | |
| CN109886343B (en) | Image classification method and device, equipment and storage medium | |
| CN111950810B (en) | Multi-variable time sequence prediction method and equipment based on self-evolution pre-training | |
| CN112734128A (en) | 7-day power load peak value prediction method based on optimized RBF | |
| CN116596044A (en) | Power generation load prediction model training method and device based on multi-source data | |
| CN108681751B (en) | Method for determining event influence factors and terminal equipment | |
| CN116307215A (en) | Load prediction method, device, equipment and storage medium of power system | |
| CN113361195B (en) | Power distribution network line ultra-short term load prediction method based on LSTM and similar days | |
| CN111833594A (en) | Traffic flow prediction method, traffic flow prediction device, electronic device, and storage medium | |
| JP2005004658A (en) | Change point detection device, change point detection method and change point-detecting program | |
| JPH04372046A (en) | Method and device for predicting demand amount | |
| CN115409153A (en) | Attention LSTM-based animal husbandry index prediction method and prediction system | |
| CN112686416A (en) | Wind power probability prediction method and system | |
| CN118193749A (en) | Deep learning-based relationship automatic identification method and system | |
| JPH06337852A (en) | Time series prediction method by neural network | |
| CN109783769B (en) | Matrix decomposition method and device based on user project scoring | |
| CN111768246B (en) | Data processing method, model building method, device and electronic equipment | |
| CN106067075B (en) | Building energy load prediction model building and load prediction method and device | |
| CN109829115B (en) | Search engine keyword optimization method | |
| Zhou et al. | Structural reliability analysis based on probability density evolution method and stepwise truncated variance reduction | |
| CN111832840B (en) | Ultra-short-term wind power interval prediction method and device based on micro cost function | |
| JP7214672B2 (en) | Information processing device, information processing method, and computer program |
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 |