WO2019015226A1 - 一种快速识别风速分布规律的方法 - Google Patents

一种快速识别风速分布规律的方法 Download PDF

Info

Publication number
WO2019015226A1
WO2019015226A1 PCT/CN2017/114935 CN2017114935W WO2019015226A1 WO 2019015226 A1 WO2019015226 A1 WO 2019015226A1 CN 2017114935 W CN2017114935 W CN 2017114935W WO 2019015226 A1 WO2019015226 A1 WO 2019015226A1
Authority
WO
WIPO (PCT)
Prior art keywords
distribution
sample
wind speed
data
reference curve
Prior art date
Application number
PCT/CN2017/114935
Other languages
English (en)
French (fr)
Inventor
林立
夏丹丹
范文亮
胡海涛
王淮峰
Original Assignee
厦门理工学院
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 厦门理工学院 filed Critical 厦门理工学院
Priority to US16/336,729 priority Critical patent/US20190228122A1/en
Publication of WO2019015226A1 publication Critical patent/WO2019015226A1/zh

Links

Images

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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/10—Office automation; Time management
    • 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

Definitions

  • the invention relates to a wind speed analysis method, and more particularly to a method for quickly identifying a wind speed distribution law.
  • wind speed must be accurately evaluated.
  • wind load is one of the most important loads.
  • the building structure not only has to withstand the wind speed at a certain time in the past, but also ensures that the wind speed that can be withstood can be safely and reliably withstand for a specified period of time.
  • the wind speed in nature has randomness and different rules at different times. Therefore, it is necessary to accurately discriminate the wind speed distribution rules of different regions according to different analysis needs, and provide reference data for the selection of wind speed of building design. At the same time, accurate estimation of wind speed distribution is of great significance for structural design, wind farm economic evaluation and wind energy resource assessment.
  • the literature has selected the wind speed distribution mainly by assuming that the wind speed data satisfies a certain distribution, such as the extreme value distribution, the Weibull distribution, etc., and the distribution parameters are fitted. Due to the great difference of regional wind fields, the possible distribution of wind speed cannot be determined. Therefore, how to select the distribution pattern quickly, intuitively and accurately is the primary key issue for wind speed data processing and the basis for all subsequent data analysis.
  • a certain distribution such as the extreme value distribution, the Weibull distribution, etc.
  • the traditional probabilistic paper method is judged by the closeness of the distribution point and the distribution reference line, and is limited by the limited probability paper type, so the optimal distribution cannot be quickly identified in many distribution laws.
  • the existing probabilistic paper has a limited type and can only be selected in a limited probability paper type, thus greatly limiting the possibility of distribution selection.
  • the object of the present invention is to overcome the deficiencies of the prior art and to provide a simple, efficient, and more accurate method for quickly identifying wind speed distribution rules for wind speed identification.
  • a method for quickly identifying the distribution law of wind speed is used to identify the optimal distribution law of known wind speed data.
  • all types of distribution laws to be selected are converted into a unified type by Rosenblatt transformation.
  • the distribution law and draw the reference curve on the probability paper; select several types of distribution law, use the known wind speed data as the sample data, and sample the sample points generated by the sample data, and compare with the reference curve; according to the comparison result, Judging the optimal distribution law among several types of distribution laws selected.
  • the steps of drawing a reference curve are as follows:
  • the step of generating a sample point set is:
  • the sample data x i is arranged in ascending order, then the n order statistics of the random variable X are x(1) ⁇ x(2) ⁇ ... ⁇ x(i) ⁇ x(i+1)... ⁇ x(n);
  • the sample data is converted into a sample conversion point that conforms to the hypothesis distribution, and ⁇ -1 [ ⁇ j (x i )] and ⁇ -1 (P i ) are the abscissas of the sample point set converted by the hypothesis distribution corresponding to the sample points, respectively. And ordinate;
  • the sample point set generated by the sample data is compared with a reference curve, and the step of performing the fitness test is:
  • the sample point set of various hypothesized distributions generated by the sample data is compared with a reference curve, and the relative distance between the sample point set and the reference line is calculated by the following formula:
  • ⁇ j (x(i)) is the actual empirical cumulative distribution function of the ith x(i) rearranged in ascending order
  • N is the number of hypothetical distribution laws to be tested
  • n is the number of samples
  • the relative distance is used as a criterion for evaluating the fitting result.
  • the sample data obeys a hypothesis distribution
  • the smaller the relative distance the better the fit degree; the hypothesis distribution with the smallest relative distance is the optimal distribution law.
  • the method for quickly identifying the wind speed distribution law quickly identifies the optimal solution of the wind speed distribution law by testing the wind speed under different distribution laws, and has the characteristics of simple, high efficiency and more accurate wind speed identification.
  • the technology of the invention is reasonable and simple, and is suitable for discriminating wind speed distribution in various ranges; it has no wide applicability for a specific probability paper.
  • the invention is fast and efficient, and can simultaneously perform multi-distribution comparison on wind speed samples, and the distribution type is not limited, and the number of distributions is not limited, and the fitting result can be visually discriminated.
  • the invention does not need By performing cumbersome calculations, the degree of fitting of multi-distribution samples can be quantitatively analyzed, so that the optimal distribution law of wind speed samples can be scientifically selected.
  • Figure 1 is a schematic diagram of raw wind data processed by the present invention
  • Figure 2 is a data cumulative probability distribution diagram
  • Figure 3 is a graph of data probability density function
  • Figure 4 is a schematic diagram of a comparison of probability plots under different hypothesis distributions
  • Figure 5 is a graphical representation of the fit of the different hypothetical distributions ( Dj value comparison chart).
  • the invention provides a method for quickly identifying the wind speed distribution law for identifying the optimal distribution of the known wind speed data, in order to solve the problem that the probabilistic paper existing in the prior art is not universal, cannot directly perform equivalent comparison, and the result is inaccurate.
  • the distribution law of all types to be selected is transformed into a uniform type of distribution law by Rosenblatt transformation, and the reference curve is drawn on the probability paper; several types of distribution laws are selected.
  • the known wind speed data is used as sample data, and the sample point set generated by the sample data is compared with the reference curve; according to the comparison result, the optimal distribution law of the selected types of distribution rules is judged.
  • a reference curve is drawn based on the probability paper used; and for the sample data to be identified, a set of sample points is generated using a hypothesized distribution rule of possible obedience.
  • the sample A is generated by using the distribution one, the distribution two, and the distribution three respectively, and the three sample point sets are theoretically different trajectories, and the trajectories and reference of the three sample point sets are used. Comparing the curves, the sample point set with the highest degree of fit indicates the sample point with the highest degree of fit among the three possible distribution rules assumed. The corresponding distribution law of the set is the best among the three distributions. In the same way, a better distribution rule can be screened through a certain number of operations.
  • the method of the present invention mainly comprises the following steps:
  • the sample data x i is arranged in ascending order, then the n order statistics of the random variable X are x(1) ⁇ x(2) ⁇ ... ⁇ x(i) ⁇ x(i+1)... ⁇ x(n);
  • the sample data is converted into a sample conversion point that conforms to the hypothesis distribution, and ⁇ -1 [ ⁇ j (x i )] and ⁇ -1 (P i ) are the abscissas of the sample point set converted by the hypothesis distribution corresponding to the sample points, respectively. And ordinate;
  • the sample point set of various hypothesized distributions generated by the sample data is compared with a reference curve, and the relative distance between the sample point set and the reference line is calculated by the following formula:
  • ⁇ j (x(i)) is the actual empirical cumulative distribution function of the ith x(i) rearranged in ascending order
  • N is the number of hypothetical distribution laws to be tested
  • n is the number of samples
  • the relative distance is used as a criterion for evaluating the fitting result.
  • a set of average wind speed data is recorded, and the data is input in the table, and a comparison of two or more distribution laws is selected.
  • the reference curve is drawn according to the proposed generalized unified probability map method, different hypothesis distributions are drawn on the same probability paper, and the sample point set generated by the sample data is compared with the reference curve. According to the proximity of the sample point set and each reference line, the relative optimal distribution is found qualitatively.
  • the method for quickly identifying the wind speed distribution law as described above includes the following steps:
  • sample data according to formula (2) in terms of the distribution is assumed to conform to the ⁇ j
  • the sample conversion point of ( ⁇ ), ⁇ -1 [ ⁇ j (x i )] and ⁇ -1 (P i ) are the abscissa and ordinate of the sample point set after the hypothesis distribution corresponding to the sample point, respectively.
  • Fitting degree test compare the set of converted sample points of multiple sets of hypothesis distribution generated by sample data with the distribution reference line, and calculate the j-th hypothesis distribution ⁇ j ( ⁇ ) with the following formula, sample point set and reference The relative distance between the lines, ie D j .
  • ⁇ j (x(i)) is the actual empirical cumulative distribution function of the ith x rearranged in ascending order
  • N is the number of hypothetical distribution rules to be tested
  • n is the number of samples.
  • the probability maps of the three distribution laws are compared, and the distance between the sample point set and the reference line is calculated according to the probability comparison map.
  • the optimal distribution is selected by comparing the fitted Dj values with the fitting degree of the selected distribution law. As shown in Fig. 5, the D j values are 0.1189, 0.1812 and .0.3984, respectively. The smaller the D j value, the higher the fit, so it can be concluded that the Gama distribution is more suitable for this parameter. data.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • Business, Economics & Management (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Operations Research (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Probability & Statistics with Applications (AREA)
  • Algebra (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Quality & Reliability (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Complex Calculations (AREA)

Abstract

一种快速识别风速分布规律的方法，用于识别已知的风速数据的分布规律，根据选用的概率纸的分布类型，将待选用的所有类型的分布规律通过Rosenblatt变换，转换为统一类型的分布规律，并在概率纸上绘制参考曲线；选用若干类型的分布规律，以已知的风速数据作为样本数据，将样本数据生成的样本点集，并与参考曲线进行比较；依据比较结果，判断所选用的若干类型的分布规律中最优的分布规律。该方案适用于各种范围的风速分布判别；不用针对特定的概率纸，可同时对风速样本进行多分布比对，分布类型不受限制，假设分布数量不受限制，可直观判别拟合结果。

Description

一种快速识别风速分布规律的方法 技术领域
本发明涉及风速分析方法，更具体地说，涉及一种快速识别风速分布规律的方法。
背景技术
我国是世界上风灾最集中的地区之一，每年风灾对我国造成了巨大的人员伤亡和经济损失，风速作为一切涉风工程中最重要的基础参数，必须得到准确的评估。在建筑设计中，风荷载是最主要的荷载之一，建筑结构不但要承受过去某一时间的风速，还要保证在某一规定的时间期限安全可靠地承受可能经受的风速。然而自然界中的风速具有随机性，不同时间有不同的规律，因此有必要根据不同的分析需要对不同区域的风速分布规律进行准确的判别，为建筑设计风速的选择提供参考数据。同时对风速分布的准确估计，对于结构设计，风电场经济评价以及风能资源评估都具有重大意义。
已有文献对风速分布的选择，主要通过假设风速数据满足某一特定分布，如极值分布，Weibull分布等，进行分布参数拟合。由于地区风场的差异性极大，风速的可能分布无法确定，因此如何快速、直观、准确地选择分布型式，是进行风速数据处理的首要关键问题，也是后继一切数据分析的基础。
传统的概率纸法都是通过分布点与分布参考线的接近程度进行判断，受限于有限的概率纸类型，因而不能在众多的分布规律里快速识别出最优分布。
对于一组特定的风场数据，传统方式的分布识别方法将风速数据根据分布函数将相应的数据点绘制在不同的概率纸上，与该类型分布的参考线进行比较判断。但是该方法存在局限性：
1现有的概率纸的类型有限，只能在有限的概率纸类型进行选择，因此大大限制了分布选择可能。
2.对两张完全不同类型的概率纸上的风速分布拟合度进行对比是比较困难的，无法进行直观的拟合优劣判断。
3.有些概率纸法，还将其它类型分布绘制在指定的分布概率纸上，由于分布曲线受到概率纸类型的限制，产生失真，无疑将导致明显的比对误差。
发明内容
本发明的目的在于克服现有技术的不足，提供一种简单、高效、对风速识别更加准确的快速识别风速分布规律的方法。
本发明的技术方案如下：
一种快速识别风速分布规律的方法，用于识别已知的风速数据的最优分布规律，根据选用的概率纸的分布类型，将待选用的所有类型的分布规律通过Rosenblatt变换，转换为统一类型的分布规律，并在概率纸上绘制参考曲线；选用若干类型的分布规律，以已知的风速数据作为样本数据，将样本数据生成的样本点集，并与参考曲线进行比较；依据比较结果，判断所选用的若干类型的分布规律中最优的分布规律。
作为优选，绘制参考曲线的步骤如下：
1.1)绘制概率图坐标：在假设的累积分布函数曲线FX(·)中选取若干个点(xi,Fi)，根据Rosenblatt变换，由Ψ-1[FX(xi)]计算出的值作为第i个点在概率图中的横坐标；再根据Ψ-1[Fi]计算结果作为第i个点在概率图中的纵坐标；
1.2)绘制参考曲线：连接所有(ΨY -1(FX(xi)),ΨY -1(Fi)点，得到参考曲线。
作为优选，生成样本点集的步骤为：
按照升序排列样本数据xi，则随机变量X的n个次序统计量x(1)＜x(2)＜ …＜x(i)＜x(i+1)...＜x(n)；
根据x(i)的次序统计量的经验累积分布函数值Pi，确定样本换算数据对(x(i),Pi)根据样本数据xi可能服从的N种假设分布类型Ψj(·)，(j＝1,2,….N)，采用样本数据的最大似然估计的结果给出假设分布类型Ψj(·)分布参数；
将样本数据换算成符合假设分布的样本换算点，Ψ-1[Ψj(xi)]和Ψ-1(Pi)分别为样本点所对应的假设分布换算后的样本点集的横坐标和纵坐标；
以此类推，得到各种假设分布类型Ψj(·)的样本点集。
作为优选，将样本数据生成的样本点集，并与参考曲线进行比较，进行拟合度检验的步骤为：
将样本数据生成的各种假设分布的样本点集与参考曲线进行比较，利用如下公式计算样本点集与参考线间的相对距离：
Figure PCTCN2017114935-appb-000001
其中，Ψj(x(i))是以递增顺序重新排列的第i个x(i)的实际经验累积分布函数，N为需要检验的假设分布规律的个数，n为样本个数；
以相对距离为评定拟合结果的标准。
作为优选，对于不同的假设分布，如果样本数据服从某个假设分布，则相对距离越小，拟合度越好；相对距离最小的假设分布，为最优分布规律。
本发明的有益效果如下：
本发明所述的快速识别风速分布规律的方法，通过对风速在不同分布规律下的检验，快速识别出风速的分布规律的最优解，具有简单、高效、对风速识别更加准确的特点。
本发明技术合理简单，适用于各种范围的风速分布判别；不用针对特定的概率纸，具有广泛适用性。本发明快速高效，可同时对风速样本进行多分布比对，分布类型不受限制，假设分布数量不受限制，可直观判别拟合结果。本发明无需 进行繁琐的计算，即可定量分析多分布样本的拟合程度，从而科学地选择风速样本的较优分布规律。
附图说明
图1是本发明处理的原始风数据示意图；
图2是数据累计概率分布图；
图3是数据概率密度函数图；
图4是不同假设分布下的概率图对比示意图；
图5是不同假设分布下的拟合度对比示意图(Dj值对比图)。
具体实施方式
以下结合附图及实施例对本发明进行进一步的详细说明。
本发明为了解决现有技术存在的概率纸不通用、无法直接进行等效对比、结果不准确等不足，提供一种快速识别风速分布规律的方法，用于识别已知的风速数据的最优分布规律，根据选用的概率纸的分布类型，将待选用的所有类型的分布规律通过Rosenblatt变换，转换为统一类型的分布规律，并在概率纸上绘制参考曲线；选用若干类型的分布规律，以已知的风速数据作为样本数据，将样本数据生成的样本点集，并与参考曲线进行比较；依据比较结果，判断所选用的若干类型的分布规律中最优的分布规律。
本发明中，以使用的概率纸为基准，绘制参考曲线；对于待识别的样本数据，利用假设的可能服从的分布规律进行生成样本点集。例如，在一次识别中，对样本A分别采用分布一、分布二、分布三分别生成样本点集，则三个样本点集理论上必然是不同的轨迹，将三个样本点集的轨迹与参考曲线进行对比，拟合度最高的样本点集即表示在所假设的三个可能服从的分布规律中，拟合度最高的样本点 集对应的分布规律为三种分布中的最优。同理，可通过一定次数的操作，筛选出较优的分布规律。
本发明所述的方法主要包括如下步骤：
1)绘制参考曲线的步骤如下：
1.1)绘制概率图坐标：在假设的累积分布函数曲线FX(·)中选取若干个点(xi,Fi)，根据Rosenblatt变换，由Ψ-1[FX(xi)]计算出的值作为第i个点在概率图中的横坐标；再根据Ψ-1[Fi]计算结果作为第i个点在概率图中的纵坐标；
1.2)绘制参考曲线：连接所有(ΨY -1(FX(xi)),ΨY -1(Fi))点，得到参考曲线。
2)生成样本点集的步骤为：
按照升序排列样本数据xi，则随机变量X的n个次序统计量x(1)＜x(2)＜…＜x(i)＜x(i+1)…＜x(n)；
根据x(i)的次序统计量的经验累积分布函数值Pi，确定样本换算数据对(x(i),Pi)；根据样本数据xi可能服从的N种假设分布类型Ψj(·)，(j＝1,2,….N)，采用样本数据的最大似然估计的结果给出假设分布类型Ψj(·)的分布参数；
将样本数据换算成符合假设分布的样本换算点，Ψ-1[Ψj(xi)]和Ψ-1(Pi)分别为样本点所对应的假设分布换算后的样本点集的横坐标和纵坐标；
以此类推，得到各种假设分布类型Ψj(·)的样本点集。
3)将样本数据生成的样本点集，并与参考曲线进行比较，进行拟合度检验的步骤为：
将样本数据生成的各种假设分布的样本点集与参考曲线进行比较，利用如下公式计算样本点集与参考线间的相对距离：
Figure PCTCN2017114935-appb-000002
其中，Ψj(x(i))是以递增顺序重新排列的第i个x(i)的实际经验累积分布函数，N为需要检验的假设分布规律的个数，n为样本个数；
以相对距离为评定拟合结果的标准。
4)对于不同的假设分布，如果样本数据服从某个假设分布，则相对距离越小，拟合度越好；相对距离最小的假设分布，为最优分布规律。
如下以正态分布的概率纸为例，对本发明所述的方法进行具体描述。
如图1所示，记录了一组平均风速数据，表中数据输入，选取某两种或多种分布规律对比。本发明中，对于已知的风速数据，根据所提出的广义统一概率图法绘制参考曲线，将不同的假设分布绘制在同一概率纸上，将样本数据生成的样本点集与参考曲线进行比较，依据样本点集与各参考线的接近程度定性的找出相对较优分布。
如上所述的快速识别风速分布规律方法包括以下步骤：
1.绘制概率图坐标：假设随机变量X和Y分别服从分布F(xi)F(xi)和Ψ(yi)Ψ(yi)，根据Rosenplatt变换原理：
当FX(xi)＝ΨY(yi)，则yi＝ΨY -1(FX(xi))；
其中，xi(i＝1,2,3,…,n)为服从分布函数F(xi)的随机变量X的n个样本，则可以得随机变量Y的n个样本的yi(i＝1,…,n)。则X是否服从FX分等价转化为Y是否服从ΨY分布。在假设累积分布函数曲线FX(·)中选取若干个点(xi,Fi)，根据yi＝ΨY -1(FX(xi))，由ΨY -1(FX(xi))计算出的值作为第i个点在概率图中的横坐标，ΨY -1(Fi)为与之相对应的概率图中的纵坐标。
2.绘制参考曲线：连接所有(ΨY -1(FX(xi)),ΨY -1(Fi))点，由于FX(xi)等于Fi，因此Ψ-1[FX(xi)]＝Ψ-1(Fi)，即任意点的纵坐标和横坐标相等，参考曲线为过原点的对角线。
3.绘制样本可能服从的假设分布的样本点集：按照升序排列样本xi，X的n个次序统计量x(1)＜x(2)＜…＜x(i)＜x(i+1)…＜x(n)，根据x(i)的次序统计量的经验累积分布函数值Pi，确定样本换算数据对(x(i),Pi)；根据不同的样本可能服从的N种分布类型Ψj(·)(j＝1,2,….N)，采用样本数据的最大似然估计的结果给出其分布参数，将样本数据根据式(2)换算成符合假设分布Ψj(·)的样本换算点，Ψ-1[Ψj(xi)]和Ψ-1(Pi)分别为样本点所对应的假设分布换算后的样本点集的横坐标和纵坐标。以此类推，可得到各种假设分布类型Ψj(·)(j＝1,2,….N)的样本点集。
4.拟合度检验：将样本数据生成的多组假设分布的换算样本点集与分布参考直线进行比较，利用如下公式计算第j个假设分布Ψj(·)情况下，样本点集与参考线间的相对距离，即Dj。
Figure PCTCN2017114935-appb-000003
其中，Ψj(x(i))是以递增顺序重新排列的第i个x的实际经验累积分布函数，N为需要检验的假设分布规律的个数，n为样本个数。
以此为评定拟合结果的标准。
如图2、图3所示，分别为基于原始数据所得的累计概率分布图及概率密度函数图，通过上述1-4步骤在换算后的概率纸上绘制出新的样本点集。
5.数据对比：对于不同的假设分布，计算结果呈现不同的Dj值，如果样本服从假设分布，则换算后的样本点集也更接近所依据的参考分布直线，Dj值越小，拟合度越好，以此为依据定性地找出相对较优的分布。
如图4所示，为三个分布规律下(Gama分布，Weibull分布和Uniform分布)的概率图对比，根据概率对比图计算样本点集与参考线间的距离。
通过输出的Dj值对比所选取的分布规律的拟合度，从而选取最优分布。如图5所示，Dj值分别为0.1189，0.1812和.0.3984，Dj值越小，则拟合度越高，因此可以得出就以上三个分布而言，Gama分布更适于此参数数据。
同理，对于一组特定数据，通过比较各种分布规律，可以定量找出最优参数分布。
上述实施例仅是用来说明本发明，而并非用作对本发明的限定。只要是依据本发明的技术实质，对上述实施例进行变化、变型等都将落在本发明的权利要求的范围内。

Claims (5)

  1. 一种快速识别风速分布规律的方法，用于识别已知的风速数据的最优分布规律，其特征在于，根据选用的概率纸的分布类型，将待选用的所有类型的分布规律通过Rosenblatt变换，转换为统一类型的分布规律，并在概率纸上绘制参考曲线；选用若干类型的分布规律，以已知的风速数据作为样本数据，将样本数据生成的样本点集，并与参考曲线进行比较；依据比较结果，判断所选用的若干类型的分布规律中最优的分布规律。
  2. 根据权利要求1所述的快速识别风速分布规律的方法，其特征在于，绘制参考曲线的步骤如下：
    1.1)绘制概率图坐标：在假设的累积分布函数曲线FX(·)中选取若干个点(xi，Fi)，根据Rosenblatt变换，由Ψ-1[FX(xi)]计算出的值作为第i个点在概率图中的横坐标；再根据Ψ-1[Fi]计算结果作为第i个点在概率图中的纵坐标；
    1.2)绘制参考曲线：连接所有(ΨY -1(FX(xi))，ΨY -1(Fi))点，得到参考曲线。
  3. 根据权利要求2所述的快速识别风速分布规律的方法，其特征在于，生成样本点集的步骤为：
    按照升序排列样本数据xi，则随机变量X的n个次序统计量x(1)＜x(2)＜…＜x(i)＜x(i+1)...＜x(n)；
    根据x(i)的次序统计量的经验累积分布函数值Pi，确定样本换算数据对(x(i)，Pi)；根据样本数据xi可能服从的N种假设分布类型Ψj(·)，(j＝1，2，....N)，采用样本数据的最大似然估计的结果给出假设分布类型Ψj(·)的分布参数；
    将样本数据换算成符合假设分布的样本换算点，Ψ-1[Ψj(xi)]和Ψ-1(Pi)分别为样本点所对应的假设分布换算后的样本点集的横坐标和纵坐标；
    以此类推，得到各种假设分布类型Ψj(·)的样本点集。
  4. 根据权利要求3所述的快速识别风速分布规律的方法，其特征在于，将样本数据生成的样本点集，并与参考曲线进行比较，进行拟合度检验的步骤为：
    将样本数据生成的各种假设分布的样本点集与参考曲线进行比较，利用如下公式计算样本点集与参考线间的相对距离：
    Figure PCTCN2017114935-appb-100001
    其中，Ψj(x(i))是以递增顺序重新排列的第i个x(i)的实际经验累积分布函数，N为需要检验的假设分布规律的个数，n为样本个数；
    以相对距离为评定拟合结果的标准。
  5. 根据权利要求4所述的快速识别风速分布规律的方法，其特征在于，对于不同的假设分布，如果样本数据服从某个假设分布，相对距离小为接近的分布规律。
PCT/CN2017/114935 2017-07-19 2017-12-07 一种快速识别风速分布规律的方法 WO2019015226A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/336,729 US20190228122A1 (en) 2017-07-19 2017-12-07 Method of fast identifying the distribution rule of wind speed

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710590948.0A CN107895058B (zh) 2017-07-19 2017-07-19 一种快速识别风速最优分布规律的方法
CN201710590948.0 2017-07-19

Publications (1)

Publication Number Publication Date
WO2019015226A1 true WO2019015226A1 (zh) 2019-01-24

Family

ID=61803365

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/114935 WO2019015226A1 (zh) 2017-07-19 2017-12-07 一种快速识别风速分布规律的方法

Country Status (3)

Country Link
US (1) US20190228122A1 (zh)
CN (1) CN107895058B (zh)
WO (1) WO2019015226A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111523223A (zh) * 2020-04-19 2020-08-11 中国电建集团华东勘测设计研究院有限公司 超长重现期极值风速的推算方法
CN111784193A (zh) * 2020-07-17 2020-10-16 中国人民解放军国防科技大学 基于正态分布的产品性能一致性检验方法
CN113030107A (zh) * 2021-03-08 2021-06-25 深圳中科飞测科技股份有限公司 检测方法、检测系统及非易失性计算机可读存储介质

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113987745A (zh) * 2021-09-22 2022-01-28 东南大学 优化载荷样本点空间布局的抽样方法
CN116736781B (zh) * 2023-08-15 2023-11-03 国网浙江省电力有限公司杭州供电公司 一种工业自动化控制设备安全状态监测方法及装置

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102738809A (zh) * 2012-06-25 2012-10-17 山东大学 计及风功率分布规律的风电场无功补偿容量优化控制方法
CN104008305A (zh) * 2014-06-11 2014-08-27 国家电网公司 用于千万千瓦风电基地的可发电风资源分布估计方法
CN105956708A (zh) * 2016-05-12 2016-09-21 扬州大学 基于灰关联时间序列的短期风速预测方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102945318B (zh) * 2012-10-29 2015-10-28 上海电力学院 一种基于级联风机的超短期风速动态预测方法
CN103473386B (zh) * 2013-06-20 2016-04-20 国家电网公司 一种确定水平运动下击暴流风剖面的方法
CN104504464A (zh) * 2014-12-11 2015-04-08 国家电网公司 基于风区风带风速规律的风功率预测方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102738809A (zh) * 2012-06-25 2012-10-17 山东大学 计及风功率分布规律的风电场无功补偿容量优化控制方法
CN104008305A (zh) * 2014-06-11 2014-08-27 国家电网公司 用于千万千瓦风电基地的可发电风资源分布估计方法
CN105956708A (zh) * 2016-05-12 2016-09-21 扬州大学 基于灰关联时间序列的短期风速预测方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
HUANG, XIANGLIU ET AL.: "The Impact of Winter Wind to the Northward Building on Xiamen Area", FUJIAN CONSTRUCTION SCIENCE & TECHNOLOGY, 31 December 2015 (2015-12-31) *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111523223A (zh) * 2020-04-19 2020-08-11 中国电建集团华东勘测设计研究院有限公司 超长重现期极值风速的推算方法
CN111784193A (zh) * 2020-07-17 2020-10-16 中国人民解放军国防科技大学 基于正态分布的产品性能一致性检验方法
CN111784193B (zh) * 2020-07-17 2024-03-26 中国人民解放军国防科技大学 基于正态分布的产品性能一致性检验方法
CN113030107A (zh) * 2021-03-08 2021-06-25 深圳中科飞测科技股份有限公司 检测方法、检测系统及非易失性计算机可读存储介质

Also Published As

Publication number Publication date
CN107895058B (zh) 2018-08-31
US20190228122A1 (en) 2019-07-25
CN107895058A (zh) 2018-04-10

Similar Documents

Publication Publication Date Title
WO2019015226A1 (zh) 一种快速识别风速分布规律的方法
TWI639824B (zh) 用於自動及手動缺陷分類之整合的方法、設備及非暫態電腦可讀取儲存媒介
TWI585397B (zh) 用於自動缺陷分類之不明缺陷拒絕率之最佳化
WO2021128576A1 (zh) 基于生成式对抗网络的刀具状态监测数据集增强方法
CN107578149B (zh) 一种电网企业关键数据分析方法
CN116559598B (zh) 一种智慧配电网故障定位方法及系统
CN111784093B (zh) 一种基于电力大数据分析的企业复工辅助判断方法
CN111797887A (zh) 一种基于密度筛选与k-均值聚类的反窃电预警方法及系统
CN111478904B (zh) 一种基于概念漂移的物联网设备通信异常检测方法及装置
CN109781658B (zh) 一种基于近红外光谱主成分分析的绝缘纸种类判别方法
WO2018006631A1 (zh) 一种用户等级自动划分方法及系统
WO2023051275A1 (zh) 基于svm的柴油机装配冷试检测方法及系统
CN112417763B (zh) 输电线路的缺陷诊断方法、装置、设备及存储介质
CN109285791A (zh) 设计布局为主的快速在线缺陷诊断、分类及取样方法及系统
CN110930057A (zh) 基于lof算法的配电变压器检验结果可信度的量化评判方法
CN116821832A (zh) 针对高压工商业用户用电负荷的异常数据辨识与修正方法
CN109150838A (zh) 一种针对慢速拒绝服务攻击的综合检测方法
CN116796271A (zh) 一种居民用能异常识别方法
Li et al. Wind pressure coefficients zoning method based on an unsupervised learning algorithm
CN111881974B (zh) 一种基于Piper-PCA-FCL判别模型的突水水源识别方法
CN112053219A (zh) 一种基于ocsvm的消费金融欺诈行为检测方法
CN116365519A (zh) 一种电力负荷预测方法、系统、存储介质及设备
CN113298148B (zh) 一种面向生态环境评价的不平衡数据重采样方法
CN108108864B (zh) 一种实验室质量控制数据管理方法
CN117236971B (zh) 多工况排放数据造假检测方法、装置及系统和存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17918349

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17918349

Country of ref document: EP

Kind code of ref document: A1