CN116205138B - Wind speed forecast correction method and device - Google Patents

Wind speed forecast correction method and device Download PDF

Info

Publication number
CN116205138B
CN116205138B CN202310088210.XA CN202310088210A CN116205138B CN 116205138 B CN116205138 B CN 116205138B CN 202310088210 A CN202310088210 A CN 202310088210A CN 116205138 B CN116205138 B CN 116205138B
Authority
CN
China
Prior art keywords
wind speed
correction
equation
correction result
forecast
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202310088210.XA
Other languages
Chinese (zh)
Other versions
CN116205138A (en
Inventor
田群
朱宇
龙雨青
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Institute Of Tropical Marine Meteorology China Meteorological Administration (guangdong Meteorology Science Institute)
Original Assignee
Guangzhou Institute Of Tropical Marine Meteorology China Meteorological Administration (guangdong Meteorology Science Institute)
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangzhou Institute Of Tropical Marine Meteorology China Meteorological Administration (guangdong Meteorology Science Institute) filed Critical Guangzhou Institute Of Tropical Marine Meteorology China Meteorological Administration (guangdong Meteorology Science Institute)
Priority to CN202310088210.XA priority Critical patent/CN116205138B/en
Publication of CN116205138A publication Critical patent/CN116205138A/en
Application granted granted Critical
Publication of CN116205138B publication Critical patent/CN116205138B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
    • G01P5/00—Measuring speed of fluids, e.g. of air stream; Measuring speed of bodies relative to fluids, e.g. of ship, of aircraft
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
    • H03H17/00—Networks using digital techniques
    • H03H17/02—Frequency selective networks
    • H03H17/0248—Filters characterised by a particular frequency response or filtering method
    • H03H17/0255—Filters based on statistics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/08—Probabilistic or stochastic CAD
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/08—Fluids
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A30/00—Adapting or protecting infrastructure or their operation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Artificial Intelligence (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Probability & Statistics with Applications (AREA)
  • Mathematical Physics (AREA)
  • Feedback Control In General (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The embodiment of the specification provides a wind speed forecast correction method and device, wherein the method comprises the following steps: carrying out linear regression correction on the numerical mode wind speed forecast to obtain a linear regression correction result; correcting the linear regression correction result by adopting a pre-trained Kalman filtering model to obtain a Kalman filtering correction result; and correcting the Kalman filtering correction result by adopting a pre-trained Bayesian inference model to obtain a final wind speed forecast correction result. The error of wind speed forecast can be further reduced, and the effect of wind speed forecast is improved.

Description

Wind speed forecast correction method and device
Technical Field
The present document relates to the field of computer technologies, and in particular, to a method and an apparatus for correcting a wind speed forecast.
Background
The numerical mode utilizes a hydrodynamic and thermodynamic basic equation set and combines a numerical calculation method to forecast the movement and change of the earth atmosphere, which is called a numerical weather forecast technology. Through the development of the last century, the numerical weather forecast technology tends to be mature, and is the most main tool for weather forecast worldwide. However, due to technical limitations, the prediction result of the numerical mode is different from the actual result by a certain amount, so that correction is needed to achieve better application effect, and the wind speed prediction of the numerical mode is also the same.
The prior art scheme is a wind speed forecast correction technology based on a Kalman-Bayesian method, and the scheme simultaneously considers long-term systematic errors and long-term probability distribution characteristics of the errors to correct the wind speed forecast of a numerical mode. The method comprises the following steps: first, kalman filtering is carried out, and then Bayesian inference correction is carried out.
Most of the existing wind speed forecasting correction techniques only consider long-term systematic errors, such as a wind speed forecasting correction technique based on Kalman filtering, or long-term probability distribution characteristics of errors, such as a wind speed forecasting correction technique based on Bayesian inference, or a combination of the two, such as a wind speed forecasting correction technique based on a Kalman-Bayesian method. However, in addition to the long-term nature of the error, there is also a feature of the error of the single wind speed forecast, which is not considered in the correction methods of the prior art.
Disclosure of Invention
The invention aims to provide a wind speed forecast correction method and device, and aims to solve the problems in the prior art.
The invention provides a wind speed forecast correction method, which comprises the following steps:
carrying out linear regression correction on the numerical mode wind speed forecast to obtain a linear regression correction result;
correcting the linear regression correction result by adopting a pre-trained Kalman filtering model to obtain a Kalman filtering correction result;
and correcting the Kalman filtering correction result by adopting a pre-trained Bayesian inference model to obtain a final wind speed forecast correction result.
The invention provides a wind speed forecast correction device, comprising:
the linear regression correction module is used for carrying out linear regression correction on the numerical mode wind speed forecast to obtain a linear regression correction result;
the Kalman filtering correction module is used for correcting the linear regression correction result by adopting a Kalman filtering model trained in advance to obtain a Kalman filtering correction result;
and the Bayesian inference correction module is used for correcting the Kalman filtering correction result by adopting a pre-trained Bayesian inference model to obtain a final wind speed forecast correction result.
The embodiment of the invention also provides electronic equipment, which comprises: the wind speed forecast correction method comprises the steps of a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the computer program is executed by the processor to realize the wind speed forecast correction method.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium stores an information transmission implementation program, and the program realizes the steps of the wind speed forecast correction method when being executed by a processor.
By adopting the embodiment of the invention, the error of wind speed forecast can be further reduced, and the effect of wind speed forecast can be improved.
Drawings
For a clearer description of one or more embodiments of the present description or of the solutions of the prior art, the drawings that are necessary for the description of the embodiments or of the prior art will be briefly described, it being apparent that the drawings in the description that follow are only some of the embodiments described in the description, from which, for a person skilled in the art, other drawings can be obtained without inventive faculty.
FIG. 1 is a flow chart of a wind speed forecast correction method in accordance with an embodiment of the present invention;
FIG. 2 is a flowchart of a preferred process for wind speed forecast correction in accordance with an embodiment of the present invention;
FIG. 3 is a schematic diagram of a wind speed forecast correction apparatus according to an embodiment of the present invention;
fig. 4 is a schematic diagram of an electronic device according to an embodiment of the invention.
Detailed Description
In order to enable a person skilled in the art to better understand the technical solutions in one or more embodiments of the present specification, the technical solutions in one or more embodiments of the present specification will be clearly and completely described below with reference to the drawings in one or more embodiments of the present specification, and it is obvious that the described embodiments are only some embodiments of the present specification, not all embodiments. All other embodiments, which can be made by one or more embodiments of the present disclosure without inventive faculty, are intended to be within the scope of the present disclosure.
Method embodiment
According to an embodiment of the present invention, there is provided a wind speed forecast correction method, and fig. 1 is a flowchart of a wind speed forecast correction method according to an embodiment of the present invention, as shown in fig. 1, and the wind speed forecast correction method according to an embodiment of the present invention specifically includes:
step 101, performing linear regression correction on the log mode wind speed forecast to obtain a linear regression correction result; the step 101 specifically includes:
assuming that the numerical mode reporting time is t0, the forecasting aging is n hours, and the time when the forecasting operation is completed is t1, establishing a regression equation for the wind speed forecasting value m and the observed value o between t0 and t1 by using a least square method according to a formula 1:
f(x i )＝wx i +b equation 1;
wherein b represents the intercept of the regression equation, w represents the slope of the regression equation, x i Representing an ith wind speed forecast value;
solving for w and b such thatMinimum, E is calculated according to equation 2 (w，b) Deriving w and b, and applying E according to equations 3 and 4 (w，b) The partial derivatives for w and b are both 0:
wherein y is i Represents the ith wind speed observation, x i Indicating the i-th wind speed forecast value,represents partial differentiation +_>Representing an average value of the wind speed forecast;
o=w×m+b is calculated, and the wind speed forecast from t1 to t0+n is corrected by using the regression equations shown in the formulas 1 to 4, and the obtained result is calculated as a linear regression correction result r.
102, correcting the linear regression correction result by adopting a pre-trained Kalman filter model to obtain a Kalman filter correction result; step 102 specifically includes:
according to the formulas 5 and 6, the error y at the time of t is established t And state vector x t ＝[x 0，t x 1，t x 2，t …x n，t ] T Sum pattern forecast result m t Is the relation of:
wherein i= 6,H t ＝[1 o t ot 2 …ot n ]O is an observation value at the time t;
iteration is performed using a kalman filter according to the following steps:
step 1, calculating a state vector X at the moment t t X is the first guess of (1) t/t-1 ＝F t ·X t-1 Wherein F is t Is a unit matrix;
step 2, calculating a corresponding covariance matrix P t Is thatWherein, the liquid crystal display device comprises a liquid crystal display device,t represents a transpose;
step 3, when obtaining a new observation vector Y t ,X t Becomes X t ＝X t/t-1 +K t ·(Y t -H t ·X t/t-1 ) WhereinAnd the result is called Kalman gain, and a Kalman filtering correction result k is finally obtained.
And step 103, correcting the Kalman filtering correction result by adopting a pre-trained Bayesian inference model to obtain a final wind speed forecast correction result. Step 103 specifically includes:
the kalman filter is the output value k at time t t ＝o t +v t Wherein o t Is the observed value of corresponding time, v t Is an error, o for wind speed data t Satisfy Weibull distribution, v t The normal distribution is satisfied: p (o) t )～W(a，b)，P(v t )～N(0,σ 2 )；
Determining o according to equation 7 t At k t The following conditional probabilities are:
wherein e represents a natural constant, a represents a scale parameter of Weibull distribution, b represents a shape parameter of Weibull distribution, and sigma represents a standard deviation of normal distribution;
determining O according to equation 8 t The expectations of (2) are:
the following describes the above technical solution of the embodiment of the present invention in detail.
Since the prediction result is obtained with a certain hysteresis, for example, the prediction result of 24 hours from 0 point to back today can be calculated and completed by 2 points, and the observation value of 0-2 points is already available at this time, a correction function can be established by using the relation between the prediction value and the observation value in the period, and the prediction of 2-24 points can be corrected. And correcting the corrected result based on the long-term systematic error and the long-term probability distribution characteristic of the error.
The embodiment of the invention is used for correcting the numerical mode wind speed forecast and mainly comprises the following 3 steps: (1) linear regression correction, (2) Kalman filter correction, (3) Bayesian inference correction, described in detail below:
1. linear regression correction
Assuming that the numerical mode reporting time is t0, the forecasting ageing is n hours, the time for forecasting operation to be completed is t1, and a regression equation is established for the wind speed forecasting value m and the observed value o between t0 and t1 by using a least square method:
regression equation f (x i )＝wx i +b, solving for w and b such thatMinimum.
Will E (w，b) Derivative of w and b:
will E (w，b) The partial derivatives for w and b are both 0, yielding:
wherein:
o=w×m+b is obtained, and the wind speed forecast from t1 to t0+n is corrected using this regression equation, and the result obtained is denoted as r.
2. Kalman filter correction
The correction of the kalman filter is carried out on r,
establishing a t moment error y t And state vector x t ＝[x 0，t x 1，t x 2，t …x n，t ] T Sum pattern forecast result m t Is the relation of:
wherein i= 6,H t ＝[1 o t ot 2 …ot n ]O is an observation value at the time t;
iteration is performed using a kalman filter according to the following steps:
step 1, calculating a state vector X at the moment t t X is the first guess of (1) t/t-1 ＝F t ·X t-1 Wherein F is t Is a unit matrix;
step 2, calculating a corresponding covariance matrix P t Is thatWherein, the liquid crystal display device comprises a liquid crystal display device,i=6, t represents a transpose;
step 3, when obtaining a new observation vector Y t ,X t Becomes X t ＝X t/t-1 +K t ·(Y t -H t ·X t/t-1 ) WhereinAnd the result is called Kalman gain, and a Kalman filtering correction result k is finally obtained.
3. Bayesian inference correction
Bayesian inference correction is carried out on k, and the Kalman filter is an output value k at the moment t t ＝o t +v t Wherein o t Is the observed value of corresponding time, v t Is an error, o for wind speed data t Satisfy Weibull distribution, v t The normal distribution is satisfied: p (o) t )～W(a，b)，P(v t )～N(0，σ 2 )；
Determining o t At k t The following conditions are outlinedThe ratio is:
determination of O t The expectations of (2) are:
the result obtained was calculated as ba.
In practical applications, the models of the second and third steps need to be trained in advance (the model of the first step does not need to be trained in advance). Step two, the r of the first 20 days can be used for pre-training, step three, the k of the first 20 days can be used for pre-training, and the length of training data used specifically can be adjusted according to actual conditions.
The embodiment of the invention can correct the numerical mode wind speed forecasting result and improve the accuracy of the forecasting result; meanwhile, the single forecast error characteristic, the long-term systematic error and the probability distribution characteristic of the long-term error are considered; linear regression, kalman filtering and Bayesian inference correction techniques are fused. According to the technical scheme provided by the embodiment of the invention, on the basis of the best prior art, the error characteristics of single prediction are considered, and the error is corrected, so that the accuracy of a wind speed prediction result can be further improved.
Device embodiment 1
According to an embodiment of the present invention, there is provided a wind speed forecast correction apparatus, and fig. 3 is a schematic diagram of the wind speed forecast correction apparatus according to the embodiment of the present invention, as shown in fig. 3, the wind speed forecast correction apparatus according to the embodiment of the present invention specifically includes:
the linear regression correction module 30 is used for carrying out linear regression correction on the numerical mode wind speed forecast to obtain a linear regression correction result; the linear regression correction module 30 specifically includes:
assuming that the numerical mode reporting time is t0, the forecasting aging is n hours, and the time when the forecasting operation is completed is t1, establishing a regression equation for the wind speed forecasting value m and the observed value o between t0 and t1 by using a least square method according to a formula 1:
f(x i )＝wx i +b equation 1;
wherein b represents the intercept of the regression equation, w represents the slope of the regression equation, x i Representing an ith wind speed forecast value;
solving for w and b such thatMinimum, E is calculated according to equation 2 (w，b) Deriving w and b, and applying E according to equations 3 and 4 (w，b) The partial derivatives for w and b are both 0:
wherein y is i Represents the ith wind speed observation, x i Indicating the i-th wind speed forecast value,represents partial differentiation +_>Representing an average value of the wind speed forecast;
o=w×m+b is calculated, and the wind speed forecast from t1 to t0+n is corrected by using the regression equations shown in the formulas 1 to 4, and the obtained result is calculated as a linear regression correction result r.
A kalman filter correction module 32, configured to correct the linear regression correction result with a pre-trained kalman filter model to obtain a kalman filter correction result; the kalman filter correction module 32 specifically includes:
according to the formulas 5 and 6, the error y at the time of t is established t And state vector x t ＝[x 0，t x 1，t x 2，t …x n，t ] T Sum pattern forecast result m t Is the relation of:
wherein i= 6,H t ＝[1 o t ot 2 …ot n ]O is an observation value at the time t;
iteration is performed using a kalman filter according to the following steps:
step 1, calculating a state vector X at the moment t t X is the first guess of (1) t/t-1 ＝F t ·X t-1 Wherein F is t Is a unit matrix;
step 2, calculating a corresponding covariance matrix P t Is thatWherein, the liquid crystal display device comprises a liquid crystal display device,i=6, t represents a transpose;
step 3, when obtaining a new observation vector Y t ,X t Becomes X t ＝X t/t-1 +K t ·(Y t -H t ·X t/t-1 ) WhereinReferred to as Kalman gain, ultimatelyAnd obtaining a Kalman filtering correction result.
The bayesian inference correcting module 34 is configured to correct the kalman filter correcting result by using a pre-trained bayesian inference model, so as to obtain a final wind speed forecast correcting result. The bayesian inference correction module 34 specifically includes:
the kalman filter is the output value k at time t t ＝o t +v t Wherein o t Is the observed value of corresponding time, v t Is an error, o for wind speed data t Satisfy Weibull distribution, v t The normal distribution is satisfied: p (o) t )～W(a，b)，P(v t )～N(0，σ 2 )；
Determining o according to equation 7 t At k t The following conditional probabilities are:
wherein e represents a natural constant, a represents a scale parameter of Weibull distribution, b represents a shape parameter of Weibull distribution, and sigma represents a standard deviation of normal distribution;
determining O according to equation 8 t The expectations of (2) are:
the embodiment of the present invention is an embodiment of a device corresponding to the embodiment of the method, and specific operations of each module may be understood by referring to descriptions of the embodiment of the method, which are not repeated herein.
Device example two
An embodiment of the present invention provides an electronic device, as shown in fig. 4, including: memory 40, processor 42, and a computer program stored on the memory 40 and executable on the processor 42, which when executed by the processor 42, performs the steps as described in the method embodiments.
Device example III
Embodiments of the present invention provide a computer-readable storage medium having stored thereon a program for carrying out information transmission, which when executed by the processor 42, carries out the steps as described in the method embodiments.
The computer readable storage medium of the present embodiment includes, but is not limited to: ROM, RAM, magnetic or optical disks, etc.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention.

Claims (8)

1. A method of correcting a wind speed forecast, comprising:
carrying out linear regression correction on the numerical mode wind speed forecast to obtain a linear regression correction result, wherein the method specifically comprises the following steps of:
assuming that the numerical mode reporting time is t0, the forecasting aging is n hours, and the time when the forecasting operation is completed is t1, establishing a regression equation for the wind speed forecasting value m and the observed value o between t0 and t1 by using a least square method according to a formula 1:
f(x i )＝wx i +b equation 1;
wherein b represents the intercept of the regression equation, w represents the slope of the regression equation, x i Representing an ith wind speed forecast value;
solving for w and b such thatMinimum, E is calculated according to equation 2 (w，b) Deriving w and b, and applying E according to equations 3 and 4 (w，b) The partial derivatives for w and b are both 0:
wherein y is i Represents the ith wind speed observation, x i Indicating the i-th wind speed forecast value,represents partial differentiation +_>Representing an average value of the wind speed forecast;
calculating o=w×m+b, correcting the wind speed forecast of t1 to t0+n by using a regression equation shown in the formulas 1 to 4, and calculating the obtained result as a linear regression correction result r;
correcting the linear regression correction result by adopting a pre-trained Kalman filtering model to obtain a Kalman filtering correction result;
and correcting the Kalman filtering correction result by adopting a pre-trained Bayesian inference model to obtain a final wind speed forecast correction result.
2. The method of claim 1, wherein correcting the linear regression correction result using a pre-trained kalman filter model to obtain a kalman filter correction result specifically comprises:
according to the formula 5 and the formula 6, the total error y at the moment t is established t And state vector x t ＝[x 0，t x 1，t x 2，t … x n，t ] T Sum pattern forecast result m t Is the relation of:
y t ＝x 0,t +x 1，t ·m t +x 2，t ·m t 2 +…+x n，t +v t equation 5;
wherein v is t As errors, estimation was performed using the total error, state vector and observations 7 times before time t:
wherein H is t ＝[1 o t o t 2 … o t n ]，o t The observation value at the time t;
iteration is performed using a kalman filter according to the following steps:
step 1, calculating a state vector X at the moment t t X is the first guess of (1) t/t-1 ＝F t ·X t-1 Wherein F is t Is a unit matrix;
step 2, calculating a corresponding covariance matrix P t Is P t/t-1 ＝F t ·P t-1 ·F t T +W t Wherein T represents transpose, W t Estimation was performed using the state vector at time t and 7 times before:
step 3, when obtaining a new observation vector Y t ,X t Becomes X t ＝X t/t-1 +K t ·(Y t -H t ·X t/t-1 ) Wherein, the method comprises the steps of, wherein,and (5) obtaining a Kalman filtering correction result finally.
3. The method according to claim 2, wherein correcting the kalman filter correction result by using a pre-trained bayesian inference model, the obtaining a final wind speed forecast correction result specifically comprises:
the kalman filter is the output value k at time t t ＝o t +v t Wherein o t Is the observed value of corresponding time, v t Is an error, o for wind speed data t Satisfy Weibull distribution, v t The normal distribution is satisfied: p (o) t )～W(a,b),P(v t )～N(0,σ 2 )；
Determining o according to equation 7 t At k t The following conditional probabilities are:
wherein e represents a natural constant, a represents a scale parameter of Weibull distribution, b represents a shape parameter of Weibull distribution, and sigma represents a standard deviation of normal distribution;
determining o according to equation 8 t The expectations of (2) are:
4. a wind speed forecast correction apparatus, comprising:
the linear regression correction module is used for carrying out linear regression correction on the numerical mode wind speed forecast to obtain a linear regression correction result, and is particularly used for:
assuming that the numerical mode reporting time is t0, the forecasting aging is n hours, and the time when the forecasting operation is completed is t1, establishing a regression equation for the wind speed forecasting value m and the observed value o between t0 and t1 by using a least square method according to a formula 1:
f(x i )＝wx i +b equation 1;
wherein b represents the intercept of the regression equation, w represents the slope of the regression equation, x i Representing an ith wind speed forecast value;
solving for w and b such thatMinimum, E is calculated according to equation 2 (w，b) Deriving w and b, and applying E according to equations 3 and 4 (w，b) The partial derivatives for w and b are both 0:
wherein y is i Represents the ith wind speed observation, x i Indicating the i-th wind speed forecast value,represents partial differentiation +_>Representing an average value of the wind speed forecast;
calculating o=w×m+b, correcting the wind speed forecast of t1 to t0+n by using a regression equation shown in the formulas 1 to 4, and calculating the obtained result as a linear regression correction result r;
the Kalman filtering correction module is used for correcting the linear regression correction result by adopting a Kalman filtering model trained in advance to obtain a Kalman filtering correction result;
and the Bayesian inference correction module is used for correcting the Kalman filtering correction result by adopting a pre-trained Bayesian inference model to obtain a final wind speed forecast correction result.
5. The apparatus of claim 4, wherein the kalman filter correction module is specifically configured to:
according to the formulas 5 and 6, the total error yt and the state vector x at the time t are established t ＝[x 0，t x 1，t x 2，t ... x n，t ] T Relationship with the pattern forecast result mt:
y t ＝x 0，t +x 1，t ·m t +x 2，t ·m t 2 +…+x n，t ·m t n +v t equation 5;
wherein v is t As errors, estimation was performed using the total error, state vector and observations 7 times before time t:
wherein H is t ＝[1 o t o t 2 … o t n ]，o t The observation value at the time t;
iteration is performed using a kalman filter according to the following steps:
step 1, calculating a state vector X at the moment t t X is the first guess of (1) t/t-1 ＝F t ·X t-1 Wherein F is t Is a unit matrix;
step 2, calculating a corresponding covariance matrix P t Is P t/t-1 ＝F t ·P t-1 ·F t T +W t Wherein T represents transpose, W t Estimation was performed using the t-moment and the previous 7 moment state vectors:
step 3, when obtaining a new observation vector Y t ,X t Becomes X t ＝X t/t-1 +K t ·(Y t -H t ·X t/t-1 ) Wherein, the method comprises the steps of, wherein,and (5) obtaining a Kalman filtering correction result finally.
6. The apparatus of claim 5, wherein the bayesian inference correction module is specifically configured to:
the kalman filter is the output value k at time t t ＝o t +v t Wherein o t Is the observed value of corresponding time, v t Is an error, o for wind speed data t Satisfy Weibull distribution, v t The normal distribution is satisfied: p (o) t )～W(a,b),P(v t )～N(0,σ 2 )；
Determining o according to equation 7 t At k t The following conditional probabilities are:
wherein e represents a natural constant, a represents a scale parameter of Weibull distribution, b represents a shape parameter of Weibull distribution, and sigma represents a standard deviation of normal distribution;
determining o according to equation 8 t The expectations of (2) are:
7. an electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, which when executed by the processor carries out the steps of the wind speed forecast correction method as claimed in any of claims 1 to 3.
8. A computer-readable storage medium, wherein a program for realizing information transfer is stored on the computer-readable storage medium, and when executed by a processor, the program realizes the steps of the wind speed forecast correction method according to any one of claims 1 to 3.
CN202310088210.XA 2023-01-16 2023-01-16 Wind speed forecast correction method and device Active CN116205138B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310088210.XA CN116205138B (en) 2023-01-16 2023-01-16 Wind speed forecast correction method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310088210.XA CN116205138B (en) 2023-01-16 2023-01-16 Wind speed forecast correction method and device

Publications (2)

Publication Number Publication Date
CN116205138A CN116205138A (en) 2023-06-02
CN116205138B true CN116205138B (en) 2023-11-03

Family

ID=86516706

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310088210.XA Active CN116205138B (en) 2023-01-16 2023-01-16 Wind speed forecast correction method and device

Country Status (1)

Country Link
CN (1) CN116205138B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014215822A (en) * 2013-04-25 2014-11-17 日本電信電話株式会社 State estimating apparatus, method, and program
CN111105068A (en) * 2019-11-01 2020-05-05 复旦大学 Numerical value mode correction method based on sequence regression learning
WO2020103677A1 (en) * 2018-11-21 2020-05-28 国网青海省电力公司 Method and device for processing meteorological element data of numerical weather prediction
CN112132365A (en) * 2020-11-24 2020-12-25 中国电力科学研究院有限公司 Multi-scale feature-based forecast wind speed correction method and system
CN112508437A (en) * 2020-12-16 2021-03-16 国网安徽省电力有限公司 Method for correcting time sequence similarity error of wind speed forecast in micro-climate environment
CN112561125A (en) * 2020-11-20 2021-03-26 浙江省气象科学研究所 Wind speed forecast multi-data fusion correction method
CN112633544A (en) * 2019-11-28 2021-04-09 北京金风慧能技术有限公司 Predicted wind speed correction method and device
CN114545528A (en) * 2022-03-09 2022-05-27 北京墨迹风云科技股份有限公司 Meteorological numerical model element forecasting and post-correcting method and device based on machine learning
CN115358151A (en) * 2022-08-25 2022-11-18 兰州大学 Correction method for near-stratum wind speed product of numerical weather forecast
CN115600142A (en) * 2022-10-20 2023-01-13 山东省气象科学研究所（山东省海洋气象科学研究所、山东省气象局培训中心）（Cn） Gale fusion integrated forecasting method based on different methods

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008063598A2 (en) * 2006-11-17 2008-05-29 Honda Motor Co., Ltd. Fully bayesian linear regression
KR102311412B1 (en) * 2020-03-18 2021-10-12 전남대학교산학협력단 Parameter estimation method based on error term independence correction, and its program implementation and performance system

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014215822A (en) * 2013-04-25 2014-11-17 日本電信電話株式会社 State estimating apparatus, method, and program
WO2020103677A1 (en) * 2018-11-21 2020-05-28 国网青海省电力公司 Method and device for processing meteorological element data of numerical weather prediction
CN111105068A (en) * 2019-11-01 2020-05-05 复旦大学 Numerical value mode correction method based on sequence regression learning
CN112633544A (en) * 2019-11-28 2021-04-09 北京金风慧能技术有限公司 Predicted wind speed correction method and device
CN112561125A (en) * 2020-11-20 2021-03-26 浙江省气象科学研究所 Wind speed forecast multi-data fusion correction method
CN112132365A (en) * 2020-11-24 2020-12-25 中国电力科学研究院有限公司 Multi-scale feature-based forecast wind speed correction method and system
CN112508437A (en) * 2020-12-16 2021-03-16 国网安徽省电力有限公司 Method for correcting time sequence similarity error of wind speed forecast in micro-climate environment
CN114545528A (en) * 2022-03-09 2022-05-27 北京墨迹风云科技股份有限公司 Meteorological numerical model element forecasting and post-correcting method and device based on machine learning
CN115358151A (en) * 2022-08-25 2022-11-18 兰州大学 Correction method for near-stratum wind speed product of numerical weather forecast
CN115600142A (en) * 2022-10-20 2023-01-13 山东省气象科学研究所（山东省海洋气象科学研究所、山东省气象局培训中心）（Cn） Gale fusion integrated forecasting method based on different methods

Non-Patent Citations (7)

* Cited by examiner, † Cited by third party
Title
Jun Chen ; Hui Liu ; Zhu Duan.Wind speed forecasting using multi-scale feature adaptive extraction ensemble model with error regression correction.《Expert Systems with Applications》.2022,1-26. *
全球谱模式T639产品在南汇站风速预报中的统计释用;朱智慧;黄宁立;;大气科学研究与应用(01);30-37 *
基于多元线性回归与时间序列的风力发电中风速捕捉综合预测研究;董浩文;张倩颖;陈兆艺;赵益;钱禹清;;机电信息(27);8-9 *
多元逐步回归与卡尔曼滤波法在霾预报中应用;咸云浩;张恒德;谢永华;杨乐;;系统仿真学报(04);1482-1489 *
多种订正技术在风电场风速预报订正中的应用;张铁军;颜鹏程;张正英;段海霞;王有生;李耀辉;;干旱气象(05);835-844 *
线性滚动极值处理方法对数值模拟风速的订正研究;祝赢;柳艳香;程兴宏;宋丽莉;黄凤新;;热带气象学报(04);681-686 *
采用贝叶斯克里金卡尔曼模型的多风电场风速短期预测;卿湘运;杨富文;王行愚;;中国电机工程学报(35);107-114 *

Also Published As

Publication number Publication date
CN116205138A (en) 2023-06-02

Similar Documents

Publication Publication Date Title
KR20190028531A (en) Training machine learning models for multiple machine learning tasks
US11863397B2 (en) Traffic prediction method, device, and storage medium
US8589329B2 (en) Method and apparatus for incremental tracking of multiple quantiles
US10962936B2 (en) Method and device for post-correction of predicted parameters by using a H-infinity filter
CN111985523A (en) Knowledge distillation training-based 2-exponential power deep neural network quantification method
JP2020506488A (en) Batch renormalization layer
KR20210017342A (en) Time series prediction method and apparatus based on past prediction data
US20180314978A1 (en) Learning apparatus and method for learning a model corresponding to a function changing in time series
WO2022095432A1 (en) Neural network model training method and apparatus, computer device, and storage medium
US11599746B2 (en) Label shift detection and adjustment in predictive modeling
CN114694379B (en) Traffic flow prediction method and system based on self-adaptive dynamic graph convolution
CN109253727B (en) Positioning method based on improved iteration volume particle filter algorithm
CN116205138B (en) Wind speed forecast correction method and device
CN110852415B (en) Vegetation index prediction method, system and equipment based on neural network algorithm
CN112270343A (en) Image classification method and device and related components
CN107247996A (en) A kind of Active Learning Method applied to different distributed data environment
CN116796649A (en) SPEI coarse resolution data space downscaling method and device based on machine learning
CN116306873A (en) Deep learning quantized perception training method, device, equipment and medium
CN115952916A (en) Artificial intelligence-based wind power prediction error correction method, device and equipment
US20220076060A1 (en) Information processing apparatus, information processing method, and non-transitory computer readable medium
CN114819398A (en) Beidou satellite clock error sequence combination prediction method based on gray Markov chain
CN112613909A (en) Agricultural product short-term price prediction method and device based on improved LSTM model
CN107180227A (en) Cloud prediction and forecasting procedure based on DTMGM+ gray models
Li A damped diffusion framework for financial modeling and closed-form maximum likelihood estimation
CN113344245A (en) Hybrid deep learning short-term prediction model, method, storage medium, and computing device

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
GR01 Patent grant
GR01 Patent grant