A method for evaluating the quality of laser wind measurement observation data

Abstract

The present invention application relates to the technical field of quality assessment of meteorological observation data. The present invention application discloses a method for quality assessment of laser wind measurement observation data. By constructing an error analysis model, it analyzes the reliability of standard observation data and predicts the reasonable threshold range of the standard observation data. The frequency of the standard observation data exceeding the reasonable threshold range is used as the result of the reliability analysis. Then, using the result of the reliability analysis and combining with the Arrhenius equation, it derives the distribution range of the laser wind measurement observation data under normal environment, and based on this distribution range, it conducts reliability assessment on the real-time observation data collected by the laser wind measurement method; thereby improving the accuracy and reliability of the observation data and providing a solid foundation for research in fields such as meteorology, atmospheric physics, and wind energy.

Description

Technical Field

[0001] The present invention relates to the technical field of meteorological observation data quality assessment, and particularly to a method for assessing the quality of laser wind measurement observation data. Background Technique

[0002] As an advanced meteorological observation method, laser wind measurement can measure the wind speed and wind direction of the wind field in real time and with high precision, providing important data support for research in fields such as meteorology, atmospheric physics, and wind energy. In order to ensure the quality of the observed data, technicians need to evaluate the quality of the data observed by the laser wind measurement instrument, so as to provide effective support for research in fields such as meteorology, atmospheric physics, and wind energy.

[0003] To promote the continuous improvement of the quality of meteorological observation data and accurately, comprehensively, and objectively evaluate the quality of meteorological observation data, technicians in this field have disclosed a patent for a method for assessing the quality of meteorological observation data, and the publication number of this patent is: CN115018284A; this patent determines the evaluation items for assessing the quality of meteorological observation data by determining the type of meteorological observation, obtains the evaluation parameters corresponding to the evaluation items, and calculates the evaluation parameters to determine the evaluation data corresponding to the observation type. Thus, an accurate and effective judgment of the meteorological evaluation quality is achieved, ensuring the reliability of the meteorological evaluation results.

[0004] However, the above evaluation method has an unsatisfactory evaluation effect on the observed data. The above evaluation method mainly determines the reliability of the observed data by calculating the set evaluation parameters. This method overly relies on the setting of the evaluation parameters, and the setting of the evaluation parameters is too idealized and cannot be compatible with various influencing factors that may exist in the actual observation process, resulting in obvious errors in the data. Therefore, technicians in this field need a real-time observation data quality evaluation method that can fully consider various influencing factors in the actual observation process. Summary of the Invention

[0005] The purpose of the present invention is to solve the above problems and design a method for assessing the quality of laser wind measurement observation data.

[0006] To achieve the above purpose, the technical solution of the present invention is a method for assessing the quality of laser wind measurement observation data, and the method includes:

[0007] Collect standard observation data, and collect the standard observation data of the wind field by laser wind measurement under the condition of no external interference; it should be noted that the condition of no external interference includes but is not limited to: air density, temperature, and humidity;

[0008] Construct an error analysis model: Based on the standard observation data and the Wiener process, establish an error analysis model, and use the maximum likelihood estimation method to obtain the unknown parameters of the error analysis model;

[0009] Analyze the reliability of the standard observation data: Use the error analysis model to predict the reasonable threshold range of the standard observation data, and take the frequency of the standard observation data exceeding the reasonable threshold range as the result of the reliability analysis;

[0010] Evaluate the reliability of the observation data: Use the result of the reliability analysis and combine it with the Arrhenius equation to derive the distribution range of the laser anemometry observation data under normal conditions, and evaluate the reliability of the real-time observation data collected by the laser anemometry method based on this distribution range. The longer the time length of the real-time observation data exceeding the distribution range, the lower the reliability of the real-time observation data.

[0011] Among them, the normal condition refers to the environment where the laser anemometry device is arranged in actual observations, including but not limited to: cities, mountains, jungles, hills, and deserts.

[0012] The construction process of the error analysis model is as follows:

[0013] Construct a univariate linear Wiener model with the standard observation data as the characteristic quantity as:

[0014] (1);

[0015] Among them: t is the acquisition time of the standard observation data; λ is the drift parameter; σ is the diffusion parameter; W(t) is the standard Brownian motion; X(t) is the change in the standard observation data at time t, which follows a normal distribution with an expectation of λt and a variance of σ 2 t, that is, X(t) - N(λt, σ 2 t);

[0016] Let x ef be the change in the standard observation data of the e-th standard observation data at the f-th moment, where: e = 1, 2,..., u; j = 1, 2,..., v; x ef follows a normal distribution with an expectation of λΔt and a variance of σ 2 Δt, that is, x ef - N(λΔt, σ 2 Δt); Use the maximum likelihood estimation method to perform point estimation on the drift parameter λ and the diffusion parameter σ, and the corresponding maximum likelihood estimation function is:

[0017] (2);

[0018] Take the logarithm of equation (2) to get:

[0019] (3);

[0020] Taking the partial derivatives of the parameters λ and α in Equation (3) respectively, the maximum likelihood estimation values are:

[0021] (4);

[0022] Define the reliability analysis result as:

[0023] (5);

[0024] where V is the reasonable threshold of the standard observation data;

[0025] From Equations (4) and (5), the probability density function of the standard observation data can be obtained as:

[0026] (6);

[0027] The reliability is defined as the probability that the standard observation data exceeds the reasonable threshold range within the specified time and under the specified conditions, that is

[0028] (7);

[0029] Substituting Equation (6) into Equation (7), the reliability function is obtained as:

[0030] (8);

[0031] From the above equations, the expectation and variance of the reliability analysis result T can be obtained as:

[0032] (9).

[0033] The analysis process of the reliability of the standard observation data includes:

[0034] Parameters λ and σ can be obtained from Equation (4). Substituting parameters such as λ and σ into Equations (6) and (8), the reasonable threshold range of the standard observation data can be obtained. The reliability is determined by the frequency of the standard observation data exceeding the reasonable threshold range.

[0035] The reliability evaluation process includes:

[0036] Let the expression of the Arrhenius model be:

[0037] (10);

[0038] where: M is the distribution range of the laser wind measurement observation data under normal environment; k is the Boltzmann constant; T is the reliability analysis result; A 0is a constant; t is the observation time; ΔE is the factor coefficient affecting the distribution of standard observation data, which is determined by the types and correlations of influencing factors, and the influencing factors include but are not limited to: parameters of the laser anemometer, response speed, and errors of the instrument itself; ∂M / ∂t represents the change rate of the distribution range of laser anemometry observation data under normal conditions;

[0039] Integrate both sides of Equation (10) to obtain:

[0040] (11);

[0041] Let ΔM = M 2 - M 1 , and define L as the time length by which the real-time observation data exceeds the distribution range of the laser anemometry observation data during the time period from t 1 to t 2 , and let L = t 2 - t 1 , then:

[0042] Take the logarithm of both sides of Equation (11) to obtain:

[0043] (12);

[0044] According to Equation (12), obtain the time length by which the real-time observation data exceeds the distribution range of the laser anemometry observation data under normal conditions, and use the magnitude of this time length to evaluate the reliability of the real-time observation data.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. The present invention predicts the reasonable threshold range of standard observation data by constructing an error analysis model, then derives the distribution range of laser anemometry observation data under normal conditions using the Arrhenius equation, and evaluates the reliability of real-time observation data collected by laser anemometry based on this distribution range, thereby improving the accuracy and reliability of the observation data, effectively eliminating the errors in the laser anemometry radar observation data, and thus better meeting the requirements of meteorological observation work;

[0047] 2. The method adopted by the present invention evaluates the reliability of real-time observation data by establishing the correspondence between standard observation data and actual measurement, minimizing the influence of external factors on real-time observation data, and improving the accuracy of the reliability analysis of real-time observation data. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flowchart of a method for evaluating the quality of laser anemometry observation data according to the present invention;

[0049] Figure 2 It is the distribution diagram of the laser wind measurement observation data under the normal environment described in the present invention;

[0050] Figure 3 It is the reliability curve of the real-time observation data described in the present invention. Specific implementation mode

[0051] The present invention will be specifically described below with reference to the accompanying drawings, as Figures 1-3 shown;

[0052] The creative point of this application is that by constructing an error analysis model to analyze the reliability of standard observation data and predict the reasonable threshold range of standard observation data, the frequency of the standard observation data exceeding the reasonable threshold range is used as the reliability analysis result, and then the distribution range of the laser wind measurement observation data under the normal environment is derived by using the reliability analysis result in combination with the Arrhenius equation, and the reliability of the real-time observation data collected by the laser wind measurement method is evaluated according to this distribution range.

[0053] In the technical solution of this application, the construction process of the error analysis model is as follows:

[0054] A unary linear Wiener model is constructed with standard observation data as the characteristic quantity as:

[0055] (1);

[0056] Where: t is the acquisition time of the standard observation data; λ is the drift parameter; σ is the diffusion parameter; W(t) is the standard Brownian motion; X(t) is the change amount of the standard observation data at time t, which follows a normal distribution with an expectation of λt and a variance of σ 2 t, that is, X(t) - N(λt, σ 2 t);

[0057] Let x ef be the change amount of the standard observation data of the e-th standard observation data at the f-th moment, where: e = 1, 2,..., u; j = 1, 2,..., v; x ef follows a normal distribution with an expectation of λΔt and a variance of σ 2 Δt, that is, x ef - N(λΔt, σ 2 Δt); The point estimation of the drift parameter λ and the diffusion parameter σ is carried out by using the maximum likelihood estimation method, and the corresponding maximum likelihood estimation function is:

[0058] (2);

[0059] Taking the logarithm of formula (2) gives:

[0060] (3);

[0061] Taking the partial derivatives of the parameters λ and α in Equation (3) respectively, the maximum likelihood estimation values are obtained as follows:

[0062] (4);

[0063] Define the reliability analysis result as:

[0064] (5);

[0065] where V is the reasonable threshold of the standard observation data;

[0066] From Equation (4) and Equation (5), the probability density function of the standard observation data can be obtained as:

[0067] (6);

[0068] The reliability is defined as the probability that the standard observation data exceeds the reasonable threshold range within the specified time and under the specified conditions, that is

[0069] (7);

[0070] Substituting Equation (6) into Equation (7), the reliability function is obtained as:

[0071] (8);

[0072] From the above equations, the expectation and variance of the reliability analysis result T can be obtained as:

[0073] (9).

[0074] It should be noted that the analysis process of the reliability of the standard observation data includes:

[0075] The parameters λ and σ can be obtained from Equation (4). Substituting the parameters such as λ and σ into Equation (6) and Equation (8), the reasonable threshold range of the standard observation data can be obtained. The reliability is determined by the frequency of the standard observation data exceeding the reasonable threshold range.

[0076] The reliability evaluation process includes:

[0077] Let the expression of the Arrhenius model be:

[0078] (10);

[0079] where: M is the distribution range of the laser wind measurement observation data under normal environment; k is the Boltzmann constant; T is the reliability analysis result; A 0is a constant; t is the observation time; ΔE is the factor coefficient affecting the distribution of standard observation data, which is determined by the type and correlation of influencing factors. The influencing factors include, but are not limited to: the parameters of the laser anemometer, the response speed, and the error of the instrument itself; ∂M / ∂t represents the change rate of the distribution range of laser anemometer observation data under normal conditions;

[0080] Integrate both sides of Equation (10) to obtain:

[0081] (11);

[0082] Let ΔM = M 2 -M 1 , and define L as the time length by which the real-time observation data exceeds the distribution range of the laser anemometer observation data within the time period from t 1 to t 2 , and let L = t 2 -t 1 , then:

[0083] Take the logarithm of both sides of Equation (11) to get:

[0084] (12);

[0085] According to Equation (12), obtain the time length by which the real-time observation data exceeds the distribution range of the laser anemometer observation data under normal conditions, and use the magnitude of this time length to evaluate the reliability of the real-time observation data.

[0086] In the specific implementation process, both ln(ΔM / A0) and ΔE / k are determined constants, and the value of ln(ΔM / A0) can be negative. Then, the reliability analysis result of the time length L is determined by T. Substitute the determined constants into Equation (12) to obtain:

[0087] ;

[0088] It can be seen from the above equation that the time length represented by L is determined by the change of T.

[0089] The above technical solution only reflects the preferred technical solution of the technical solution of the present invention. Some changes that those skilled in the art of this technology may make to some parts thereof all reflect the principle of the present invention and fall within the protection scope of the present invention.

Claims

1. A method for evaluating the quality of laser wind observation data, characterized in that: The method includes: Collect standard observation data, collect standard observation data of wind field by laser wind measurement without external interference; Construct an error analysis model, establish the error analysis model based on standard observation data and Wiener process, and obtain the unknown parameters of the error analysis model; Analyze the reliability of standard observation data, use the error analysis model to predict the reasonable threshold range of standard observation data, and use the frequency of standard observation data exceeding the reasonable threshold range as the reliability analysis result; Reliability assessment of observation data: The reliability analysis results are combined with the Arrhenius equation to derive the distribution range of laser wind measurement observation data under normal conditions, and the reliability of real-time observation data collected by laser wind measurement is assessed based on the distribution range; The construction process of the error analysis model is as follows: The univariate linear Wiener model is constructed using standard observation data as the feature quantity: (1) Where: t is the acquisition time of the standard observation data; λ is the drift parameter; σ is the diffusion parameter; W(t) is the standard Brownian motion; X(t) is the change of the standard observation data at time t, which has an expectation of λt and a variance of σ 2 Normal distribution of t, that is, X(t)-N(λt,σ 2 t); Let x ef is the change in the standard observation data of the e-th standard observation data at the f-th time, where: e=1,2,…,u; j=1,2,…,v; x ef Subject to the expectation of λΔt and the variance of σ 2 The normal distribution of Δt, that is, x ef -N(λΔt,σ 2 Δt); the drift parameter λ and the diffusion parameter σ are estimated by the maximum likelihood estimation method, and the corresponding maximum likelihood estimation function is: （2） Taking the logarithm of formula (2) yields: （3） The maximum likelihood estimates of the parameters λ and α in equation (3) are obtained by taking partial derivatives: （4） The reliability analysis result is defined as: （5） Where V is a reasonable threshold of standard observation data; From equations (4) and (5), we can get the probability density function of the standard observation data: （6） Reliability is defined as the probability that the standard observation data exceeds the reasonable threshold range within the specified time and under the specified conditions, that is, （7） Substituting formula (6) into formula (7), the reliability function is obtained as follows: （8） From the above formulas, the expectation and variance of the reliability analysis result T can be obtained as follows: (9) The analysis process of the reliability of the standard observation data includes: From formula (4), we can get the parameters λ and σ. Substituting parameters such as λ and σ into formula (6) and formula (8), we can get the reasonable threshold range of the standard observation data. The reliability is determined by the frequency of the standard observation data exceeding the reasonable threshold range.

2. A laser wind measurement observation data quality assessment method according to claim 1, characterized in that: The reliability assessment process includes: Assume that the expression of the Arrhenius equation is: (10) Where: M is the distribution range of laser wind observation data under normal environment; k is the Boltzmann constant; T is the reliability analysis result; A0 is a constant; t is the observation time; ΔE is the coefficient of factors affecting the distribution of standard observation data; ∂M / ∂t represents the rate of change of the distribution range of laser wind observation data under normal environment; Integrating both sides of equation (10), we get: (11) Let ΔM = M2-M1, and define L as the length of time that the real-time observation data exceeds the distribution range of the laser wind observation data from t1 to t2, and let L = t2-t1, then: Taking the logarithm of both sides of equation (11) we get: (12) According to formula (12), the duration that the real-time observation data exceeds the distribution range of the laser wind measurement observation data under normal conditions is obtained, and the size of this duration is used to evaluate the reliability of the real-time observation data.

Images