CN111611741B - Time sequence wind speed simulation method and system based on finite state Markov sequence - Google Patents

Time sequence wind speed simulation method and system based on finite state Markov sequence Download PDF

Info

Publication number
CN111611741B
CN111611741B CN202010488905.3A CN202010488905A CN111611741B CN 111611741 B CN111611741 B CN 111611741B CN 202010488905 A CN202010488905 A CN 202010488905A CN 111611741 B CN111611741 B CN 111611741B
Authority
CN
China
Prior art keywords
time series
wind speed
random
normal distribution
markov
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
CN202010488905.3A
Other languages
Chinese (zh)
Other versions
CN111611741A (en
Inventor
李玉敦
杨超
史方芳
张国辉
刘萌
李宽
赵斌超
王昕�
李广磊
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.)
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Shandong Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
Electric Power Research Institute of State Grid Shandong Electric Power Co Ltd
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 State Grid Corp of China SGCC, Electric Power Research Institute of State Grid Shandong Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN202010488905.3A priority Critical patent/CN111611741B/en
Publication of CN111611741A publication Critical patent/CN111611741A/en
Application granted granted Critical
Publication of CN111611741B publication Critical patent/CN111611741B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • 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
    • G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Aerodynamic Tests, Hydrodynamic Tests, Wind Tunnels, And Water Tanks (AREA)
  • Control Of Eletrric Generators (AREA)

Abstract

本发明提出了基于有限状态马尔科夫序列的时序风速模拟方法和系统，该方法接收原始风速时间序列，利用多项式正态变换技术将原始风速时间序列变换为服从正态分布的时间序列；利用相邻K个时刻服从正态分布的时间序列构造k维随机向量；随机向量服从协方差矩阵C的多元正态分布；根据多元正态分布构造状态转移函数；利用状态转移函数进行随机抽样得到设定时间长度的马尔科夫时间序列；利用多项式正态变换将马尔科夫时间序列变换得到模拟风速时间序列。基于该方法，还提出了模拟系统。本发明在连续状态空间建立风速的时间序列模型，具有更高的仿真精度；在正态空间构建风速时间序列的马尔科夫随机模型，降低了直接建模的复杂度，建模更加灵活方便。

The present invention proposes a time series wind speed simulation method and system based on finite state Markov sequences. The method receives the original wind speed time series and uses polynomial normal transformation technology to transform the original wind speed time series into a time series that obeys a normal distribution; K-dimensional random vectors are constructed from the time series of adjacent K moments that obey the normal distribution; the random vector obeys the multivariate normal distribution of the covariance matrix C; the state transition function is constructed according to the multivariate normal distribution; the state transition function is used for random sampling to obtain the setting Markov time series of time length; use polynomial normal transformation to transform the Markov time series to obtain simulated wind speed time series. Based on this approach, a simulation system was also proposed. The present invention establishes a time series model of wind speed in the continuous state space, which has higher simulation accuracy; and builds a Markov random model of the wind speed time series in the normal space, which reduces the complexity of direct modeling and makes the modeling more flexible and convenient.

Description

Time sequence wind speed simulation method and system based on finite state Markov sequence
Technical Field
The invention belongs to the technical field of power systems and automation thereof, and particularly relates to a time sequence wind speed simulation method and system based on a finite state Markov sequence.
Background
Wind speed data is the basis for wind energy development and utilization. An accurate wind speed model is established and a key of wind speed simulation is developed. At present, a great deal of research is carried out on wind speed models at home and abroad, and a plurality of wind speed simulation methods are provided. At present, the wind speed time sequence simulation mainly comprises a time sequence model and a Markov model, the time sequence model needs more model parameters, so that the modeling is complex, the probability distribution is error, and the Markov process has higher precision because the probability distribution characteristic and the autocorrelation characteristic of a sample can be kept, so that the Markov model has more practical application value.
The wind speed is a physical process which is continuous in time and space, and can be statistically described as a random process with continuous time and continuous state, the current Markov model mainly adopts a discrete Markov chain to simulate the wind speed, the model discretizes the wind speed state, a discrete state probability matrix is established, the model precision depends on a discrete scale, but the high discretization can improve the model precision, but too many model parameters can be caused, and in most cases, the model precision is sacrificed in order to reduce the model complexity, so that the applicability is reduced. It can be seen that the existing wind speed time series simulation has the difficulty that the precision and the simplicity of the model are difficult to ensure simultaneously, and the bottleneck of the model constructs any type of state transfer function in a continuous state space.
Disclosure of Invention
The invention provides a time sequence wind speed simulation method and a time sequence wind speed simulation system based on a finite state Markov sequence, which are used for establishing a time sequence model of wind speed in a continuous state space and have higher simulation precision; the Markov random model of the wind speed time sequence is built in a normal space, the complexity of direct modeling is reduced, and modeling is more flexible and convenient.
In order to achieve the above object, the present invention proposes a time-series wind speed simulation method based on a finite state markov sequence, the method comprising the steps of:
s1: transforming the original wind speed time sequence into a time sequence obeying normal distribution by using a polynomial normal transformation technology;
s2: constructing a K-dimensional random vector by using a time sequence of which adjacent K moments obey normal distribution; the k-dimensional random vector obeys the multi-element normal distribution of the covariance matrix C; constructing a state transfer function according to the multi-element normal distribution;
s3: randomly sampling by using a state transfer function to obtain a Markov time sequence with a set time length;
s4: and transforming the Markov time sequence by using polynomial normal transformation to obtain a simulated wind speed time sequence.
Further, before step S1, the method further includes receiving a time sequence of the original wind speed.
Further, in step S1, the step of transforming the original wind speed time sequence into a time sequence compliant with a normal distribution by using a polynomial normal transformation technique is as follows:
s11: establishing an original wind speed time sequence W t Random variable Z t A polynomial functional relationship between; w (W) t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 ；
wherein ,c0 ,c 1 ,c 2 ,c 3 Are all function model parameters; c 0 ＝μ v -σ v γ 3v /6,c 1 ＝σ v (33-3γ 4v )/24,c 2 ＝σ v γ 3v /6,c 3 ＝σ v (3γ 4v -3)/24；μ v Mathematical expectations for a random time sequence; sigma (sigma) v Standard deviation as random time series; gamma ray 3v A bias value which is a random time sequence; gamma ray 4v Kurtosis values that are random time series;
s12, utilizing the formula W t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 Time-series W of original wind speed t Transforming into a time series Y subject to normal distribution t The method comprises the steps of carrying out a first treatment on the surface of the Wherein t=1, 2.
Further, the process of step S2 is as follows:
s21, constructing a k-dimensional random vector Y= (Y) by using a time sequence of adjacent k moments obeying normal distribution t-(k-1) ,...,Y t ,Y t+1 ) Then the k-dimensional random vector obeys a multivariate normal distribution with covariance matrix C:
s22: constructing random variable Y at t+1 time according to multi-element normal distribution t+1 Conditional probability distribution functions of (2); the conditional probability distribution function is a state transfer function;
the random variable Y t+1 The method comprises the following steps: wherein ,
further, the process of step S3 is as follows:
s31: randomly generating k-1 random numbers obeying normal distribution as initial values;
s32: bringing the k-1 initial values into the formulaCalculating the average value mu of the kth random variable;
s33: substituting μ into formulaObtaining a probability distribution function of a kth random variable, and randomly sampling to generate a kth moment wind speed random variable;
s34: repeating steps S32 to S33n times to obtain a Markov sequence X with length of n t (t＝1,2,....,n)。
Further, the process of step S4 is as follows: using formula W t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 The normal distribution is distributed along with the time sequence W t (t=1, 2,) n) conversion to an analog wind speed time series V t (t＝1,2,....,n)。
The invention also provides a time sequence wind speed simulation system based on the finite state Markov sequence, which comprises a receiving module, a first transformation module, a solving module, a sampling module and a second transformation module;
the receiving module is used for receiving the original wind speed time sequence;
the first transformation module is used for transforming the original wind speed time sequence into a time sequence obeying normal distribution by using a polynomial normal transformation technology;
the acquisition module is used for constructing a K-dimensional random vector by utilizing a time sequence of which adjacent K moments obey normal distribution; the k-dimensional random vector obeys the multi-element normal distribution of the covariance matrix C; constructing a state transfer function according to the multi-element normal distribution;
the sampling module is used for randomly sampling by using a state transfer function to obtain a Markov time sequence with set time length;
the second transformation module is used for transforming the Markov time sequence into an analog wind speed time sequence by using polynomial normal transformation.
Further, the implementation process of the first transformation module is as follows:
establishing an original wind speed time sequence W t Random variable Z t A polynomial functional relationship between; w (W) t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3； wherein ,c0 ,c 1 ,c 2 ,c 3 Are all function model parameters; c 0 ＝μ v -σ v γ3v/6,c 1 ＝σ v (33-3γ 4v )/24,c 2 ＝σ v γ3v/6,C 3 ＝σ v (3γ 4v -3)/24；μ v Mathematical expectations for a random time sequence; sigma (sigma) v Standard deviation as random time series; gamma ray 3v A bias value which is a random time sequence; gamma ray 4v Kurtosis values that are random time series;
using formula W t ＝c 0 +c 1 Z t +c 2 Z t 2+c 3 Z t 3 Time-series W of original wind speed t Transforming into a time series Y subject to normal distribution t The method comprises the steps of carrying out a first treatment on the surface of the Wherein t=1, 2.
Further, the implementation process of the solving module is as follows: constructing a k-dimensional random vector y= (Y) using a time series of adjacent k moments subject to normal distribution t-(k-1) ,...,Y t ,Y t+1 ) Then the k-dimensional random vector obeys a multivariate normal distribution with covariance matrix C:
s22: constructing random variable Y at t+1 time according to multi-element normal distribution t+1 Conditional probability distribution functions of (2); the conditional probability distribution function is a state transfer function; the method comprises the steps of carrying out a first treatment on the surface of the
The random variable Y t+1 Is that wherein ,
further, the implementation process of the sampling module is as follows:
s31: randomly generating k-1 random numbers obeying normal distribution as initial values;
s32: bringing the k-1 initial values into the formulaCalculating the average value mu of the kth random variable;
s33: substituting μ into formulaObtaining a probability distribution function of a kth random variable, and randomly sampling to generate a kth moment wind speed random variable;
s34: repeating steps S32 to S33n times to obtain a Markov sequence X with length of n t (t＝1,2,....,n)。
The effects provided in the summary of the invention are merely effects of embodiments, not all effects of the invention, and one of the above technical solutions has the following advantages or beneficial effects:
the embodiment of the invention provides a time sequence wind speed simulation method and a time sequence wind speed simulation system based on a finite state Markov sequence, wherein after receiving an original wind speed time sequence, the method transforms the original wind speed time sequence into a time sequence obeying normal distribution by utilizing a polynomial normal transformation technology; constructing a K-dimensional random vector by using a time sequence of which adjacent K moments obey normal distribution; the k-dimensional random vector obeys the multi-element normal distribution of the covariance matrix C; constructing a state transfer function according to the multi-element normal distribution; randomly sampling by using a state transfer function to obtain a Markov time sequence with a set time length; and transforming the Markov time sequence by using polynomial normal transformation to obtain a simulated wind speed time sequence. The invention provides a time sequence wind speed simulation method based on a finite state Markov sequence, and also provides a time sequence wind speed simulation system based on the finite state Markov sequence. The method establishes a time sequence model of the wind speed in a continuous state space, accords with the physical change process of the actual wind speed, and has higher simulation precision; the Markov random model of the wind speed time sequence is built in a normal space, the complexity of direct modeling is reduced, and the modeling is more flexible and convenient, so that the method has good engineering application prospect.
Drawings
FIG. 1 is a flow chart of a time-series wind speed simulation method based on a finite state Markov sequence according to embodiment 1 of the present invention;
FIG. 2 is a schematic diagram showing 8760 hour wind speed time series of a certain region according to embodiment 1 of the present invention;
FIG. 3 shows a simulated wind speed frequency histogram based on the time-series wind speed simulation method based on finite state Markov sequence proposed in embodiment 1 of the present invention;
FIG. 4 shows a simulated wind speed autocorrelation function based on the finite state Markov sequence-based time-series wind speed simulation method proposed in embodiment 1 of the present invention;
fig. 5 shows a schematic diagram of a time-series wind speed simulation system based on a finite state markov sequence according to embodiment 1 of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
The invention provides a time sequence wind speed simulation method based on a finite state Markov sequence, as shown in FIG. 1, a flow chart of the time sequence wind speed simulation method based on the finite state Markov sequence is provided;
in step S101, the flow of processing is started.
In step S102, an original wind speed time series W is received t 。
In step S103, transforming the original wind speed time series into a time series compliant with a normal distribution using a polynomial normal transformation technique;
first, an original wind speed time sequence W is established t Random variable Z t A polynomial functional relationship between; w (W) t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 ；
wherein ,c0 ,c 1 ,c 2 ,c 3 Are all function model parameters; c 0 ＝μ v -σvγ 3v /6,c 1 ＝σ v (33-3γ 4v )/24,c 2 ＝σ v γ 3v /6,c 3 ＝σ v (3γ 4v -3)/24；μ v Mathematical expectations for a random time sequence; sigma (sigma) v Standard deviation as random time series; gamma ray 3v A bias value which is a random time sequence; gamma ray 4v Kurtosis values that are random time series;
then utilize equation W t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 Time-series W of original wind speed t Transforming into a time series Y subject to normal distribution t The method comprises the steps of carrying out a first treatment on the surface of the Wherein t=1, 2.
In step S104, a K-dimensional random vector is constructed using a time sequence in which K adjacent moments are subjected to normal distribution; the k-dimensional random vector obeys the multi-element normal distribution of the covariance matrix C; and constructing a state transfer function according to the multivariate normal distribution.
First, a k-dimensional random vector y= (Y) is constructed using a time series of adjacent k times subject to normal distribution t-(k-1) ,...,Y t ,Y t+1 ) Then the k-dimensional random vector obeys a multivariate normal distribution with covariance matrix C:
s22: constructing random variable Y at t+1 time according to multi-element normal distribution t+1 Conditional probability distribution functions of (2);the conditional probability distribution function is a state transfer function;
the random variable Y t+1 The method comprises the following steps: wherein ,
in step S105, random sampling is performed by using the state transfer function to obtain a markov time series with a set time length;
s31: randomly generating k-1 random numbers obeying normal distribution as initial values;
s32: bringing the k-1 initial values into the formulaCalculating the average value mu of the kth random variable;
s33: substituting μ into formulaObtaining a probability distribution function of a kth random variable, and randomly sampling to generate a kth moment wind speed random variable;
s34: repeating steps S32 to S33n times to obtain a Markov sequence X with length of n t (t＝1,2,....,n)。
In step S106: using formula W t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 The normal distribution is distributed along with the time sequence W t (t=1, 2,) n) conversion to an analog wind speed time series V t (t＝1,2,....,n)。
In step S107, the flow ends.
A time series diagram of 8760 hours wind speed in a certain region is shown in fig. 2. Using time series sample data V of wind speed t (t=1, 2,) the polynomial transformation function relationship is found, 8760, and the settlement result is as follows:
using formula V t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 Time series of wind speeds V t Conversion to a time series Z subject to normal distribution t The method comprises the steps of carrying out a first treatment on the surface of the Wherein t=1, 2..8760.
The state transfer function is obtained by using a normal distribution time sequence, a 2-order Markov sequence is set and adopted, and the settlement result is shown as follows:
random sampling by using state transfer function to obtain a normal distribution time sequence, assuming sampling time is 87600 time points (10 years) Z t (t＝1,2,....,87600)。
For normally distributed time series Z t (t=1, 2.,. 87600) according to the polynomial transformation function relationship V described above t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 Can obtain wind speed time sequence V t (t=1, 2.,. 87600). FIG. 3 shows a simulated wind speed frequency histogram based on the time-series wind speed simulation method based on finite state Markov sequence proposed in embodiment 1 of the present invention; fig. 4 shows a simulated wind speed autocorrelation function based on the time-series wind speed simulation method based on the finite state markov sequence proposed in embodiment 1 of the present invention.
The invention also provides a time sequence wind speed simulation system based on the finite state Markov sequence, as shown in FIG. 5, the time sequence wind speed simulation system based on the finite state Markov sequence provided by the embodiment 1 of the invention comprises a receiving module, a first transformation module, a solving module, a sampling module and a second transformation module.
For receiving modulesReceiving the original wind speed time sequence W t 。
The first transformation module is used for transforming the original wind speed time sequence into a time sequence obeying normal distribution by using polynomial normal transformation technology.
The implementation process of the first transformation module is as follows: first, an original wind speed time sequence W is established t Random variable Z t A polynomial functional relationship between; w (W) t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3； wherein ,c0 ,c 1 ,c 2 ,c 3 Are all function model parameters; c 0 ＝γ v -σ v γ 3v /6,c 1 ＝σ v (33-3γ 4v )/24,c 2 ＝σ v γ 3v /6,c 3 ＝σ v (3γ 4v -3)/24；μ v Mathematical expectations for a random time sequence; sigma (sigma) v Standard deviation as random time series; gamma ray 3v A bias value which is a random time sequence; gamma ray 4v Kurtosis values that are random time series; then, using formula W t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 Time-series W of original wind speed t Transforming into a time series Y subject to normal distribution t The method comprises the steps of carrying out a first treatment on the surface of the Wherein t=1, 2.
The solving module is used for constructing a K-dimensional random vector by utilizing a time sequence of adjacent K moments obeying normal distribution; the k-dimensional random vector obeys the multi-element normal distribution of the covariance matrix C; and constructing a state transfer function according to the multivariate normal distribution.
The implementation process of the solving module is as follows: first, a k-dimensional random vector y= (Y) is constructed using a time series of adjacent k times subject to normal distribution t-(k-1) ,...,Y t ,Y t+1 ) Then the k-dimensional random vector obeys a multivariate normal distribution with covariance matrix C:then, the time t+1 is constructed according to the multivariate normal distributionRandom variable Y of (2) t+1 Conditional probability distribution functions of (2); the conditional probability distribution function is a state transfer function;
random variable Y t+1 The method comprises the following steps: wherein ,
the sampling module is used for randomly sampling by using the state transfer function to obtain a Markov time sequence with set time length.
The concrete implementation process of the sampling module is as follows: randomly generating k-1 random numbers obeying normal distribution as initial values; bringing the k-1 initial values into the formulaCalculating the average value mu of the kth random variable; substituting μ into the formula +.>Obtaining a probability distribution function of a kth random variable, and randomly sampling to generate a kth moment wind speed random variable; repeating the process of obtaining the average value mu of the kth random variable and substituting mu into the probability distribution function of the kth random variable n times to obtain a Markov sequence X with the length of n t (t＝1,2,....,n)。
The second transformation module is used for transforming the Markov time sequence into an analog wind speed time sequence by using polynomial normal transformation, and using a formula W t ＝c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 The normal distribution is distributed along with the time sequence W t (t=1, 2,) n) conversion to an analog wind speed time series V t (t＝1,2,....,n)。
The foregoing is merely illustrative and explanatory of the invention, as it is well within the scope of the invention as claimed, as it relates to various modifications, additions and substitutions for those skilled in the art, without departing from the inventive concept and without departing from the scope of the invention as defined in the accompanying claims.

Claims (8)

1.基于有限状态马尔科夫序列的时序风速模拟方法，其特征在于，包括以下步骤：1. The time series wind speed simulation method based on finite state Markov sequence is characterized by including the following steps: S1：利用多项式正态变换技术将原始风速时间序列变换为服从正态分布的时间序列；S1: Use polynomial normal transformation technology to transform the original wind speed time series into a time series that obeys a normal distribution; S2：利用相邻k个时刻服从正态分布的时间序列构造k维随机向量；k维随机向量服从协方差矩阵C的多元正态分布；根据多元正态分布构造状态转移函数；步骤S2的过程为：S2: Construct a k-dimensional random vector using the time series of k adjacent k moments that obey the normal distribution; the k-dimensional random vector obeys the multivariate normal distribution of the covariance matrix C; construct the state transition function according to the multivariate normal distribution; the process of step S2 for: S21:利用相邻k个时刻服从正态分布的时间序列构造一个k维随机向量Y＝(Yt-(k-1),...,Yt,Yt+1)，则k维随机向量服从协方差矩阵为C的多元正态分布：S21: Construct a k-dimensional random vector Y=(Y t-(k-1) ,...,Y t ,Y t+1 ) using k adjacent time series that obey normal distribution, then the k-dimensional random vector The vector follows a multivariate normal distribution with covariance matrix C: S22：根据多元正态分布构造t+1时刻的随机变量Yt+1的条件概率分布函数；所述条件概率分布函数为状态转移函数；S22: Construct a conditional probability distribution function of the random variable Y t+1 at time t+1 according to the multivariate normal distribution; the conditional probability distribution function is a state transition function; 所述随机变量Yt+1为其中，The random variable Y t+1 is in, S3：利用状态转移函数进行随机抽样得到设定时间长度的马尔科夫时间序列；S3: Use the state transition function to conduct random sampling to obtain a Markov time series of a set time length; S4：利用多项式正态变换将马尔科夫时间序列变换得到模拟风速时间序列。S4: Use polynomial normal transformation to transform the Markov time series to obtain simulated wind speed time series. 2.根据权利要求1所述的基于有限状态马尔科夫序列的时序风速模拟方法，其特征在于，在步骤S1之前，还包括接收原始风速时间序列。2. The time series wind speed simulation method based on finite state Markov sequence according to claim 1, characterized in that before step S1, it also includes receiving the original wind speed time series. 3.根据权利要求1所述的基于有限状态马尔科夫序列的时序风速模拟方法，其特征在于，步骤S1中，所述利用多项式正态变换技术将原始风速时间序列变换为服从正态分布的时间序列的步骤为：3. The time series wind speed simulation method based on finite state Markov sequence according to claim 1, characterized in that, in step S1, the original wind speed time series is transformed into a normal distribution by using polynomial normal transformation technology. The steps of the time series are: S11：建立原始风速时间序列Wt和随机变量Zt之间的多项式函数关系；Wt＝c0+c1Zt+c2Zt 2+c3Zt 3；其中，c0,c1,c2,c3均为函数模型参数；c0＝μv-σvγ3v/6,c1＝σv(33-3γ4v)/24,c2＝σvγ3v/6,c3＝σv(3γ4v-3)/24；μv为随机时间序列的数学期望；σv为随机时间序列的标准差；γ3v为随机时间序列的偏度值；γ4v为随机时间序列的峰度值；S11: Establish the polynomial function relationship between the original wind speed time series W t and the random variable Z t ; W t =c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 ; where, c 0 ,c 1 , c 2 , c 3 are all function model parameters; c 0 ＝μ v -σ v γ 3v /6, c 1 ＝σ v (33-3γ 4v )/24, c 2 ＝σ v γ 3v /6, c 3 =σ v (3γ 4v -3)/24; μ v is the mathematical expectation of the random time series; σ v is the standard deviation of the random time series; γ 3v is the skewness value of the random time series; γ 4v is the random time The kurtosis value of the sequence; S12:利用公式Wt＝c0+c1Zt+c2Zt 2+c3Zt 3，将原始风速时间序列Wt变换为服从正态分布的时间序列Yt；其中t＝1,2..n。S12: Use the formula W t =c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 to transform the original wind speed time series W t into a time series Y t that obeys the normal distribution; where t=1 ,2..n. 4.根据权利要求3所述的基于有限状态马尔科夫序列的时序风速模拟方法，其特征在于，步骤S3的过程为：4. The time series wind speed simulation method based on finite state Markov sequence according to claim 3, characterized in that the process of step S3 is: S31：随机产生k-1个服从正态分布的随机数，作为初始值；S31: Randomly generate k-1 random numbers obeying the normal distribution as initial values; S32：将这k-1个初始值带入公式求取第k个随机变量的平均值μ；S32: Bring these k-1 initial values into the formula Find the average μ of the kth random variable; S33：将μ代入到公式得到第k个随机变量的概率分布函数，随机抽样产生第k时刻风速随机变量；S33: Substitute μ into the formula Obtain the probability distribution function of the kth random variable, and randomly sample the wind speed random variable at the kth moment; S34：重复步骤S32至S33n次，得到一条长度为n的马尔科夫序列Xt(t＝1,2,....,n)。S34: Repeat steps S32 to S33n times to obtain a Markov sequence X t of length n (t=1,2,....,n). 5.根据权利要求3所述的基于有限状态马尔科夫序列的时序风速模拟方法，其特征在于，步骤S4的过程为：利用公式Wt＝c0+c1Zt+c2Zt 2+c3Zt 3，将正态分布随时间序列Wt(t＝1,2,....,n)变换为模拟风速时间序列Vt(t＝1,2,....,n)。5. The time series wind speed simulation method based on finite state Markov sequence according to claim 3, characterized in that the process of step S4 is: using the formula W t =c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 , transform the normal distribution with time series W t (t＝1,2,....,n) into simulated wind speed time series V t (t＝1,2,...., n). 6.基于有限状态马尔科夫序列的时序风速模拟系统，其特征在于，包括接收模块、第一变换模块、求取模块、抽样模块、第二变换模块；6. A time-series wind speed simulation system based on finite state Markov sequence, which is characterized by including a receiving module, a first transformation module, an acquisition module, a sampling module, and a second transformation module; 所述接收模块用于接收原始风速时间序列；The receiving module is used to receive the original wind speed time series; 所述第一变换模块用于利用多项式正态变换技术将原始风速时间序列变换为服从正态分布的时间序列；The first transformation module is used to transform the original wind speed time series into a time series that obeys a normal distribution using polynomial normal transformation technology; 所述求取模块用于利用相邻k个时刻服从正态分布的时间序列构造k维随机向量；k维随机向量服从协方差矩阵C的多元正态分布；根据多元正态分布构造状态转移函数；所述求取模块的实现过程为：利用相邻k个时刻服从正态分布的时间序列构造一个k维随机向量Y＝(Yt-(k-1),...,Yt,Yt+1)，则k维随机向量服从协方差矩阵为C的多元正态分布：The obtaining module is used to construct a k-dimensional random vector using a time series that obeys the normal distribution at k adjacent moments; the k-dimensional random vector obeys the multivariate normal distribution of the covariance matrix C; and constructs the state transition function according to the multivariate normal distribution. ; The implementation process of the obtaining module is: construct a k-dimensional random vector Y=(Y t-(k-1) ,...,Y t ,Y by using the time series of k adjacent k moments obeying the normal distribution) t+1 ), then the k-dimensional random vector obeys the multivariate normal distribution with a covariance matrix C: 根据多元正态分布构造t+1时刻的随机变量Yt+1的条件概率分布函数；所述条件概率分布函数为状态转移函数；Construct a conditional probability distribution function of the random variable Y t+1 at time t+1 according to the multivariate normal distribution; the conditional probability distribution function is a state transition function; 所述随机变量Yt+1为其中，The random variable Y t+1 is in, 所述抽样模块用于利用状态转移函数进行随机抽样得到设定时间长度的马尔科夫时间序列；The sampling module is used to perform random sampling using the state transition function to obtain a Markov time series of a set time length; 所述第二变换模块用于利用多项式正态变换将马尔科夫时间序列变换得到模拟风速时间序列。The second transformation module is used to transform the Markov time series using polynomial normal transformation to obtain a simulated wind speed time series. 7.根据权利要求6所述的基于有限状态马尔科夫序列的时序风速模拟系统，其特征在于，所述第一变换模块的实现过程为：7. The time series wind speed simulation system based on finite state Markov sequence according to claim 6, characterized in that the implementation process of the first transformation module is: 建立原始风速时间序列Wt和随机变量Zt之间的多项式函数关系；Wt＝c0+c1Zt+c2Zt 2+c3Zt 3；其中，c0,c1,c2,c3均为函数模型参数；c0＝μv-σvγ3v/6,c1＝σv(33-3γ4v)/24,c2＝σvγ3v/6,c3＝σv(3γ4v-3)/24；μv为随机时间序列的数学期望；σv为随机时间序列的标准差；γ3v为随机时间序列的偏度值；γ4v为随机时间序列的峰度值；Establish the polynomial functional relationship between the original wind speed time series W t and the random variable Z t ; W t =c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 ; where, c 0 ,c 1 , c 2 and c 3 are both function model parameters; c 0 =μ v -σ v γ 3v /6,c 1 =σ v (33-3γ 4v )/24,c 2 =σ v γ 3v /6,c 3 =σ v (3γ 4v -3)/24; μ v is the mathematical expectation of the random time series; σ v is the standard deviation of the random time series; γ 3v is the skewness value of the random time series; γ 4v is the skewness value of the random time series kurtosis value; 利用公式Wt＝c0+c1Zt+c2Zt 2+c3Zt 3，将原始风速时间序列Wt变换为服从正态分布的时间序列Yt；其中t＝1,2..n。Using the formula W t =c 0 +c 1 Z t +c 2 Z t 2 +c 3 Z t 3 , the original wind speed time series W t is transformed into a time series Y t that obeys the normal distribution; where t=1,2 ..n. 8.根据权利要求7所述的基于有限状态马尔科夫序列的时序风速模拟系统，其特征在于，所述抽样模块的实现过程为：8. The time series wind speed simulation system based on finite state Markov sequence according to claim 7, characterized in that the implementation process of the sampling module is: S31：随机产生k-1个服从正态分布的随机数，作为初始值；S31: Randomly generate k-1 random numbers obeying the normal distribution as initial values; S32：将这k-1个初始值带入公式求取第k个随机变量的平均值μ；S32: Bring these k-1 initial values into the formula Find the average μ of the kth random variable; S33：将μ代入到公式得到第k个随机变量的概率分布函数，随机抽样产生第k时刻风速随机变量；S33: Substitute μ into the formula Obtain the probability distribution function of the kth random variable, and randomly sample the wind speed random variable at the kth moment; S34：重复步骤S32至S33n次，得到一条长度为n的马尔科夫序列Xt(t＝1,2,....,n)。S34: Repeat steps S32 to S33n times to obtain a Markov sequence X t of length n (t=1,2,....,n).
CN202010488905.3A 2020-06-02 2020-06-02 Time sequence wind speed simulation method and system based on finite state Markov sequence Active CN111611741B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010488905.3A CN111611741B (en) 2020-06-02 2020-06-02 Time sequence wind speed simulation method and system based on finite state Markov sequence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010488905.3A CN111611741B (en) 2020-06-02 2020-06-02 Time sequence wind speed simulation method and system based on finite state Markov sequence

Publications (2)

Publication Number Publication Date
CN111611741A CN111611741A (en) 2020-09-01
CN111611741B true CN111611741B (en) 2023-10-10

Family

ID=72197782

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010488905.3A Active CN111611741B (en) 2020-06-02 2020-06-02 Time sequence wind speed simulation method and system based on finite state Markov sequence

Country Status (1)

Country Link
CN (1) CN111611741B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104850710A (en) * 2015-05-26 2015-08-19 河海大学 Stochastic partial differential equation based wind speed fluctuation characteristic modeling method
CN105939014A (en) * 2016-06-24 2016-09-14 中国电力科学研究院 Wind power station correlation index acquisition method
CN106532688A (en) * 2016-11-22 2017-03-22 国电南瑞科技股份有限公司 Method and system for evaluating operation reliability of micro-grid
CN107103411A (en) * 2017-04-08 2017-08-29 东北电力大学 Based on the markovian simulation wind power time series generation method of improvement
CN107765347A (en) * 2017-06-29 2018-03-06 河海大学 A kind of Gaussian process returns and the short-term wind speed forecasting method of particle filter
CN109492315A (en) * 2018-11-19 2019-03-19 西安交通大学 A kind of temporal and spatial correlations scene series model method based on Copula function
CN109872248A (en) * 2018-12-18 2019-06-11 国网青海省电力公司经济技术研究院 A method and system for calculating the output of a wind farm cluster
CN110717277A (en) * 2019-10-14 2020-01-21 河北工业大学 A time series wind speed simulation method

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104850710A (en) * 2015-05-26 2015-08-19 河海大学 Stochastic partial differential equation based wind speed fluctuation characteristic modeling method
CN105939014A (en) * 2016-06-24 2016-09-14 中国电力科学研究院 Wind power station correlation index acquisition method
CN106532688A (en) * 2016-11-22 2017-03-22 国电南瑞科技股份有限公司 Method and system for evaluating operation reliability of micro-grid
CN107103411A (en) * 2017-04-08 2017-08-29 东北电力大学 Based on the markovian simulation wind power time series generation method of improvement
CN107765347A (en) * 2017-06-29 2018-03-06 河海大学 A kind of Gaussian process returns and the short-term wind speed forecasting method of particle filter
CN109492315A (en) * 2018-11-19 2019-03-19 西安交通大学 A kind of temporal and spatial correlations scene series model method based on Copula function
CN109872248A (en) * 2018-12-18 2019-06-11 国网青海省电力公司经济技术研究院 A method and system for calculating the output of a wind farm cluster
CN110717277A (en) * 2019-10-14 2020-01-21 河北工业大学 A time series wind speed simulation method

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
史可琴 ; 王方雨 ; 梁琛 ; 刘文颖 ; .基于随机过程自相关性的风速预测模型分析.电网技术.2017,(第02期),全文. *
李玉敦 ; 谢开贵 ; 胡博 ; .基于Copula函数的多维时序风速相依模型及其在可靠性评估中的应用.电网技术.2013,(第03期),全文. *
王梅 ; 李玉敦 ; .连续状态马尔科夫链风速模型及其在风电系统可靠性评估中的应用.现代电力.2013,(第06期),全文. *

Also Published As

Publication number Publication date
CN111611741A (en) 2020-09-01

Similar Documents

Publication Publication Date Title
CN110705762B (en) Ubiquitous power Internet of things perception data missing repairing method based on matrix filling
WO2017054537A1 (en) Long-time scale photovoltaic output time sequence modelling method and apparatus
CN111027732B (en) A method and system for generating output scenarios of multiple wind farms
CN113936192B (en) Method, terminal and storage medium for repairing missing data of synchronous measurement of distribution network
Payne An improved technique for transfer function synthesis from frequency response data
CN101247137A (en) A Parallel Sampling System for UWB Analog Signals Based on Random Projection
CN111273353B (en) Intelligent seismic data antialiasing method and system based on U-Net network
CN116359605B (en) Harmonic signal analysis method based on secondary weighting
CN115600051B (en) Orbit maneuver intelligent detection method and device based on short-arc space-based optical observation
CN111611741B (en) Time sequence wind speed simulation method and system based on finite state Markov sequence
CN111191186B (en) Multi-cell filtering method for positioning position of mobile robot in production workshop
CN115908051A (en) Method for determining energy storage capacity of power system
CN106992782A (en) A Method for Testing Static Parameters of Timing Synchronous DAC
CN114970240A (en) Method and equipment for rapidly evaluating load state of multi-phase composite structure image
CN110398942A (en) A Parameter Identification Method for Industrial Production Process Control
CN112733348B (en) Hybrid power grid probability calculation method and device based on polynomial and maximum entropy model
CN116522785A (en) Electromagnetic transient simulation equation set solving method and device based on quantum computation
Ren et al. Local time-varying topology identification of network with unknown parameters based on adaptive synchronization
CN115204449A (en) Track prediction method based on self-adaptive Legendre Picard iteration method
CN105162112A (en) Faure sequence based photovoltaic system power flow calculation and probabilistic power flow statistical method
CN108491606B (en) A kind of strength of materials distribution acquiring method
CN114970837B (en) DOA estimation method, device and storage medium for power Internet of things
CN119149888A (en) DSP (digital Signal processor) acceleration realization method and device for two-norm constraint elongated matrix indirect singular value decomposition
CN103957068B (en) A kind of RB particle filter method for synchronizing time based on DPM model
CN118690639A (en) Junction temperature calculation method and junction temperature calculation device of power semiconductor 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