Raindrop distribution trueness evaluation method and device for site simulated rainfall

Abstract

The invention relates to a raindrop distribution trueness evaluation method and device oriented to site simulated rainfall. The method comprises the following steps: collecting microscopic data of real rainfall at each level; on the basis of microscopic data of real rainfall, calculating an average raindrop distribution matrix of different grades to serve as an evaluation benchmark; microcosmic data of simulated rainfall of a site are collected, and a raindrop distribution matrix of the simulated rainfall is calculated; and calculating the similarity between the raindrop distribution matrix of the simulated rainfall and the average raindrop distribution matrix of each grade, and determining a truth degree evaluation result of the simulated rainfall of the field according to the similarity. Compared with the prior art, the method provided by the invention starts from the level of the microscopic raindrop distribution matrix of rainfall, constructs the evaluation criterion based on the raindrop distribution matrix of the real rainfall sample, and can realize the quantitative evaluation of the raindrop distribution trueness of the simulated rainfall by quantifying the similarity between the simulated rainfall sample and the evaluation criterion.

Description

Technical Field

[0001] The present invention relates to the technical field of closed-field simulated rainfall testing, and in particular to a method and device for evaluating the authenticity of raindrop distribution for field simulated rainfall. Background Art

[0002] With the in-depth development of autonomous driving technology, the safety and reliability of automobiles will face new and more severe challenges. Unlike traditional cars, the focus of autonomous driving car testing is on the vehicle's perception and response capabilities to the traffic environment. It is a test for the human-vehicle-road-environment coupling system, and pays more attention to the construction of test scenarios. In particular, testing under special meteorological conditions such as rain, fog, and light is an indispensable part of the development of autonomous driving technology.

[0003] As a key module for autonomous driving vehicles to obtain external environmental information, the environmental perception system is easily affected by complex external environmental triggering conditions and is the main source of expected functional safety issues. Rainfall is a high-risk triggering condition. Therefore, it is necessary to explore the impact of rainfall on the environmental perception system. Most existing technologies use real road tests, closed field tests, and simulation tests. Among them, field simulation rainfall is a closed field test. It is the main research method at present because of its controllability, repeatability, and suitability for sensor testing. Compared with the meteorological scene test of autonomous driving in the natural environment, building a meteorological simulation scene in a closed environment can accurately control environmental parameters such as rainfall, fog visibility, and illumination, making the test background environment more accurate and reproducible, helping developers to better understand the performance and responsiveness of the autonomous driving system in different situations, thereby optimizing the design and control strategy of the system and promoting the application and popularization of autonomous driving technology.

[0004] Existing studies have explored the impact of rainfall triggering conditions on sensors through field simulated rainfall. Among them, the results of the tests based on field simulated rainfall are mostly analyzed based on the settings of field simulated rainfall. However, there are many differences between the actual presentation of field simulated rainfall and real weather, such as fogging. The current method of evaluating field simulated rainfall equipment is only based on a single indicator such as rain intensity or falling speed. Referring to relevant meteorological research, rainfall itself should be compared from the microscopic distribution level in space. Differences in rainfall at the microscopic level may lead to deviations in experimental results, which may easily make the confidence level of the evaluation of the expected functional safety of the perception system low.

[0005] Therefore, it is necessary to evaluate the realism of raindrop distribution of simulated rainfall on site from a microscopic level in order to quantify the realism of simulated rainfall on site and ensure the confidence of the test results of autonomous driving vehicles. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a method and device for evaluating the authenticity of raindrop distribution for field simulated rainfall, which can realize quantitative evaluation of the authenticity of raindrop distribution for simulated rainfall.

[0007] The purpose of the present invention can be achieved by the following technical solution: A method for evaluating the authenticity of raindrop distribution for field simulated rainfall, comprising the following steps:

[0008] S1. Collect micro data of real rainfall at all levels;

[0009] S2. Based on the microscopic data of real rainfall, the average raindrop distribution matrix of different levels is calculated as an evaluation benchmark;

[0010] S3, collecting microscopic data of simulated rainfall at the site and calculating the raindrop distribution matrix of the simulated rainfall;

[0011] S4, calculating the similarity between the raindrop distribution matrix of the simulated rainfall and the average raindrop distribution matrix of each level;

[0012] S5. Based on the similarity calculated in step S4, determine the authenticity evaluation result of the simulated rainfall on site.

[0013] Furthermore, the microscopic data of the real rainfall in step S1 and the microscopic data of the simulated rainfall in step S3 both include the number of raindrop particles of different diameters and speeds and rainfall intensity data.

[0014] Furthermore, the specific process of step S2 is: slicing the microscopic data of real rainfall at each level into different samples according to minutes, each sample contains a corresponding diameter velocity-particle number distribution matrix, and the corresponding level is defined based on the rainfall intensity data within one minute;

[0015] The diameter velocity-particle number distribution matrices of the same level are superimposed to obtain the average distribution, and the average distribution heat map of real rainfall of different levels is used as the authenticity evaluation benchmark, and the benchmark is iteratively updated as the real rainfall samples are supplemented.

[0016] Furthermore, the step S4 comprises the following steps:

[0017] S41. Based on the formula of structural similarity SSIM (Structural Similarity), the basic similarity between the raindrop distribution matrix of simulated rainfall and the average raindrop distribution matrix of each level is calculated. The similarity result is a value between 0 and 1. The calculation formula is as follows:

[0018]

[0019] Among them, μ x is the element mean of matrix x; μy is the element mean of matrix y; x is the element variance of the matrix x; σ y is the element variance of matrix y; σ xy is the covariance of matrices x and y; C1 and C2 are constants to avoid the denominator being 0;

[0020] S42: For pixels with abnormal distribution and abnormal values ​​in the raindrop distribution matrix of simulated rainfall, a penalty term y is set. dist With y val , in order to distinguish the simulated rainfall samples with significant differences in raindrop distribution, the calculation formula is as follows:

[0021]

[0022] Among them, X is the raindrop distribution matrix of simulated rainfall, Y is the reference raindrop distribution matrix, P dist The number of abnormal distribution elements whose element values ​​in the simulated rainfall samples are non-zero but whose corresponding position elements in the benchmark distribution matrix are zero is counted, and P val Then the number of abnormal numerical elements whose numerical values ​​in the sample to be tested differ from the corresponding elements in the reference matrix is ​​counted;

[0023] S43, based on two penalty terms P dist With P val , and get the total penalty coefficient P total , multiply the total penalty coefficient by the basic similarity calculation result to obtain the total similarity S between the raindrop distribution matrix of simulated rainfall and the average raindrop distribution matrix of each level.

[0024] Furthermore, the calculation formula of the total similarity S in step S43 is:

[0025] S(X,Y)=P total SSIM(X,Y)

[0026]

[0027] Among them, N X is the number of non-zero pixels in the raindrop distribution matrix of simulated rainfall.

[0028] Furthermore, the step S5 specifically includes the following steps:

[0029] S51, based on the rainfall intensity data of the simulated rainfall, setting the fuzzy membership of the rainfall level, thereby fuzzifying the division boundary of the rainfall level, the fuzzy membership is calculated by the following method:

[0030]

[0031]

[0032] Among them, μ little , μ mid , μ heavy They are the fuzzy membership of light, medium and heavy rain, and R is the rainfall intensity value of simulated rainfall;

[0033] S52, using the fuzzy membership as the weight, weighted summing up the raindrop distribution similarity between the simulated rainfall and the benchmarks at all levels, and comprehensively obtaining the raindrop distribution reality RRD (Realism of Raindrop Distribution);

[0034] S53. Based on the RRD evaluation results of simulated rainfall, by horizontally comparing the RRD index calculation results of simulated rainfall with approximate rainfall intensity, or by comparing with the authenticity threshold, it is evaluated whether the simulated rainfall on the site meets the raindrop distribution authenticity requirements based on the high or low values.

[0035] Furthermore, the calculation formula of the raindrop distribution authenticity RRD in step S52 is:

[0036] RRD=μ little *S little +μ mid *S mid +μ heavy *S heavy

[0037] Among them, S little , S mid , S heavy They represent the similarity of raindrop distribution between simulated rainfall and small, medium and large levels of realism evaluation benchmarks.

[0038] Furthermore, in step S53, the authenticity threshold is defined based on the numerical distribution of the similarity between the real rainfall sample and the distribution benchmark, and the lower limit of the 95% confidence interval based on the similarity distribution result of the real rainfall sample is used as the reference value of the authenticity threshold.

[0039] A raindrop distribution authenticity assessment device for field simulated rainfall comprises an overall frame consisting of a base, a roller and a column, wherein the roller is mounted at the bottom of the base, a battery box is arranged on the top of the base, an industrial computer, a display screen and a laser raindrop spectrometer are mounted on the column, the laser raindrop spectrometer is located at the top of the column, the laser raindrop spectrometer is used to collect microscopic data of real rainfall and simulated rainfall and transmit the data to the industrial computer, the industrial computer is used to execute a preset program to calculate the authenticity assessment result of the field simulated rainfall, and transmit the result to the display screen for display.

[0040] Furthermore, the preset program in the industrial computer includes a first program and a second program with switchable modes, wherein the first program processes microscopic data of real rainfall, takes real rainfall as the composition of the evaluation benchmark, divides the real rainfall sample levels based on rainfall intensity, and iteratively updates the average raindrop distribution matrix of different levels;

[0041] The second program processes the micro data of simulated rainfall, takes the simulated rainfall on site as the evaluation object, calculates the similarity between the raindrop distribution matrix of the simulated rainfall and the average raindrop distribution matrix of each level, and determines the authenticity evaluation result of the simulated rainfall on site.

[0042] Compared with the prior art, the present invention has the following advantages:

[0043] The present invention first collects microscopic data of real rainfall at all levels to calculate the average raindrop distribution matrix of different levels as an evaluation benchmark; then collects microscopic data of simulated rainfall on site and calculates the raindrop distribution matrix of simulated rainfall; then calculates the similarity between the raindrop distribution matrix of simulated rainfall and the average raindrop distribution matrix of each level to determine the authenticity evaluation result of simulated rainfall on site. In this way, the authenticity of raindrop distribution of simulated rainfall on site can be evaluated from a microscopic level, the authenticity of simulated rainfall on site can be quantified, and the confidence of the test results of autonomous driving vehicles can be ensured.

[0044] In the present invention, the microscopic data of rainfall include the number of raindrop particles of different diameters and speeds and rainfall intensity data, and the microscopic data are sliced ​​into different samples according to minutes, each sample contains the corresponding diameter speed-particle number distribution matrix, and the level is delineated based on the rainfall intensity value within one minute. For the microscopic data of real rainfall, the raindrop distribution matrix of the rainfall intensity data within the same level is also superimposed to obtain the average distribution, and the average raindrop distribution matrix of real rainfall of different levels is used as the authenticity evaluation benchmark, and the benchmark is iteratively updated as the real rainfall samples are supplemented. On the one hand, since the diameter speed-particle number distribution matrix contains most of the information of the microscopic distribution in the rainfall space, the evaluation result can reflect the authenticity of the simulated rainfall at the microscopic level of the site. On the other hand, using the average raindrop distribution matrix of each level of real rainfall as the evaluation benchmark can avoid the contingency of individual rainfall samples, obtain the common characteristics of the microscopic distribution of rainfall of a specific level in space, and have more reliability.

[0045] The present invention proposes a raindrop distribution authenticity assessment device for field simulated rainfall. The device uses a laser raindrop spectrometer to collect real rainfall and simulated rainfall micro data, uses an industrial computer to receive the rainfall micro data output by the laser raindrop spectrometer, and executes a preset program and adopts a raindrop distribution authenticity assessment method to assess the authenticity of field simulated rainfall. The preset program can switch between different modes, collect real rainfall samples to iteratively update the assessment benchmark; and conduct authenticity assessment on field simulated rainfall samples based on the stored assessment benchmark to ensure real-time updating, storage and output of rainfall data. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0047] Figure 2 This is a heat map of the distribution of some real rainfall samples in the embodiment;

[0048] Figure 3 The rainfall level classification standard in the embodiment;

[0049] Figure 4 To construct a three-level average benchmark distribution heat map in the embodiment;

[0050] Figure 5 This is a heat map of simulated rainfall sample distribution in some sites in the embodiment;

[0051] Figure 6 The distribution of the evaluation results of the real rainfall samples in the embodiment;

[0052] Figure 7 This is a case study on the authenticity of simulated rainfall samples in the embodiment;

[0053] Figure 8 It is a schematic diagram of the overall structure of the device of the present invention;

[0054] Explanation of the markings in the figure: 11. base, 12. roller, 13. column, 2. battery box, 3. industrial computer, 4. display screen, 5. laser raindrop spectrometer. DETAILED DESCRIPTION

[0055] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] Example

[0057] like Figure 1 As shown, a method for evaluating the authenticity of raindrop distribution for field rainfall simulation includes the following steps:

[0058] S1. Collect micro data of real rainfall at all levels;

[0059] S2. Based on the microscopic data of real rainfall, the average raindrop distribution matrix of different levels is calculated as an evaluation benchmark;

[0060] S3, collecting microscopic data of simulated rainfall at the site and calculating the raindrop distribution matrix of the simulated rainfall;

[0061] S4, calculating the similarity between the raindrop distribution matrix of the simulated rainfall and the average raindrop distribution matrix of each level;

[0062] S5. Based on the similarity calculated in step S4, determine the authenticity evaluation result of the simulated rainfall on site.

[0063] This embodiment applies the above solution, and the main contents are:

[0064] 1. Use equipment to collect micro data of real rainfall at all levels;

[0065] In this embodiment, micro data samples of real rainfall are collected (the collected micro data include the number of raindrop particles of different diameters and speeds and rainfall intensity data), where the micro data of rainfall is sliced ​​into different samples according to minutes, and the levels are delineated based on the rainfall intensity value within the corresponding minute. Finally, a total of 712 groups of real rainfall samples are collected, including 391 groups of light rain samples, 141 groups of moderate rain samples, and 180 groups of heavy rain samples. The heat map of the distribution of some samples is as follows: Figure 2 As shown, the horizontal axis represents different particle size levels, the vertical axis represents different speed levels, and the pixel points represent the number of raindrop particles of corresponding particle size speed.

[0066] Second, based on the real rainfall micro data, the average distribution heat map of different levels (i.e., the average raindrop distribution matrix of different levels) is calculated as the evaluation benchmark. Specifically, the raindrop distribution matrix of the rainfall intensity samples in the same level is superimposed to obtain the average distribution, and the average distribution heat map of real rainfall at different levels is used as the authenticity evaluation benchmark, and the benchmark is iteratively updated as the real rainfall samples are supplemented;

[0067] In this embodiment, the real rainfall samples are divided into three levels according to the rainfall intensity: small, medium and large. The classification criteria are as follows: Figure 3 As shown in the figure, the average distribution heat map is calculated from the distribution heat map of samples in the same rainfall level as the benchmark distribution for subsequent authenticity evaluation. The three-level average distribution heat map obtained based on the summary of real rainfall samples is shown in Figure 4 shown.

[0068] 3. Collect micro data of simulated rainfall at the site;

[0069] In this embodiment, micro data samples of simulated rainfall at the site are collected to obtain a total of 34 groups of simulated rainfall samples. The distribution heat map of simulated rainfall samples at some sites is as follows: Figure 5 shown.

[0070] Fourth, the authenticity evaluation result is determined by calculating the similarity between the simulated rainfall distribution heat map of the site and the benchmark distribution heat maps of different levels;

[0071] In this embodiment, the similarity between the simulated rainfall samples on site and the evaluation benchmark distribution heat map is calculated, and the result is a value from 0 to 1, which is convenient for intuitive analysis and comparison. The value can directly reflect the distinguishability. The closer to 1, the more realistic the simulated rainfall on site.

[0072] 5. Based on the authenticity assessment results, determine whether the simulated rainfall on site meets the raindrop distribution authenticity requirements.

[0073] When calculating the similarity between the simulated rainfall sample on site and the evaluation benchmark distribution heat map, the specific process is:

[0074] First, based on the formula of structural similarity SSIM, the basic similarity between the simulated rainfall samples at the site and the benchmark raindrop distribution matrices at all levels of evaluation is calculated. The similarity result is a value from 0 to 1. The calculation formula is as follows:

[0075]

[0076] Where: μ x is the element mean of matrix x; μ y is the element mean of matrix y; σ x is the element variance of the matrix x; σ y is the element variance of matrix y; σ xy is the covariance of matrices x and y; C1 and C2 are constants to avoid the denominator being zero.

[0077] Afterwards, for the pixels with abnormal distribution and abnormal values ​​in the simulated rainfall sample distribution heat map, a penalty term P is set dist With P val , in order to distinguish the simulated rainfall samples with significant differences in raindrop distribution, the calculation formula is as follows:

[0078]

[0079] Where: X is the raindrop distribution matrix of the simulated rainfall sample to be evaluated, Y is the benchmark raindrop distribution matrix, P dist The number of abnormal distribution elements whose element values ​​in the simulated rainfall samples are non-zero but whose corresponding position elements in the benchmark distribution matrix are zero is counted, and P val Then the number of abnormal numerical elements whose numerical magnitudes differ between the numerical values ​​in the sample to be tested and the corresponding elements in the reference matrix is ​​counted.

[0080] Combining the above two penalty items, determine the total penalty coefficient P total, multiply the total penalty coefficient by the basic similarity calculation result to obtain the raindrop distribution similarity S between the simulated rainfall and the benchmark distribution heat maps at all levels. The calculation formula is as follows:

[0081]

[0082] S(Y,Y)=P total SSIM(Y,Y)

[0083] Where: N X is the number of non-zero pixels in the distribution heat map of simulated rainfall samples to be evaluated in the simulated rainfall.

[0084] When determining the authenticity assessment results, the specific process is:

[0085] Based on the rainfall intensity information of the simulated rainfall samples to be evaluated, the fuzzy membership of the rainfall level is set, thereby fuzzifying the division boundary of the rainfall level.

[0086] The fuzzy membership is calculated as follows:

[0087]

[0088] Where: μ little , μ mid , μ heavy are the fuzzy membership of light, medium and heavy rain, and R is the rainfall intensity value of the rainfall sample to be tested.

[0089] Using fuzzy membership as weight, weighted sum of raindrop distribution similarities between the sample to be evaluated and the benchmarks at all levels is used to comprehensively obtain the raindrop distribution authenticity (RRD). The calculation formula for raindrop distribution authenticity is as follows:

[0090] RRD=μ little *S little +μ mid *S mid +μ heavy *S heavy

[0091] Where: S little , S mid , S heavy They represent the raindrop distribution similarity between the sample to be evaluated and the small, medium and large levels of authenticity evaluation benchmarks.

[0092] In this embodiment, the similarity between the real rainfall sample and the distribution benchmark is calculated by the above method, and the following is obtained: Figure 6 Similarity distribution of real samples shown.

[0093] The KS (Kolmogorov-Smirnov) test was used for the sample similarity distribution. The test results showed that the RRD calculation results satisfied the normal distribution, and the significance value P was greater than 0.2. Therefore, based on the distribution of the real rainfall similarity, this embodiment defined the lower limit of the 95% confidence interval as the authenticity threshold, and the numerical result was 0.76.

[0094] In this embodiment, the raindrop distribution authenticity is evaluated for all simulated rainfall samples, and the samples with the highest raindrop distribution authenticity are as follows: Figure 7 As shown in the figure, its RRD value is 0.74, which is closest to the raindrop distribution of real rainfall. It can be found that the high-fidelity samples screened can conform to the overall distribution of real rainfall under the corresponding rainfall intensity, especially in terms of the overall raindrop order, which is highly consistent with the real rainfall.

[0095] This embodiment also proposes a raindrop distribution authenticity assessment device for field rainfall simulation, such as Figure 8 As shown, the device includes a base 11, a roller 12, a column 13, a battery box 2, an industrial computer 3, a display screen 4 and a laser raindrop spectrometer 5, wherein the base 11, the roller 12 and the column 13 together constitute the frame structure of the device, the battery box 2 is fixedly mounted on the base 11, the industrial computer 3 and the display screen 4 are mounted on the central part of the column 13, and the laser raindrop spectrometer 5 is mounted on the top of the column 13.

[0096] Among them, the battery box 2 supplies power to the industrial computer 3, the display screen 4 and the laser raindrop spectrometer 5 through the wiring harness. The battery box 2 is installed at a lower position, which can correspondingly lower the center of gravity of the entire device. At the same time, the shell of the battery box 2 is waterproof to ensure the safety and reliability of power supply.

[0097] The laser raindrop spectrometer 5 is used to collect microscopic data of real rainfall and simulated rainfall, including the number of raindrop particles with different speed diameters and the fitted rainfall intensity information, and transmit the collected microscopic data to the industrial computer 3. The industrial computer 3 stores and processes the rainfall data, and transmits the processed authenticity evaluation results to the display screen 4 in real time for result display.

[0098] In actual applications, the industrial computer 3 is preset with a corresponding program. After receiving the rainfall samples, the industrial computer 3 will adopt different processing methods accordingly: the real rainfall samples are divided by minutes, and divided into different levels based on the rainfall intensity, and the average distribution heat map is calculated as the evaluation benchmark; the simulated rainfall samples are evaluated through the authenticity evaluation method to calculate the authenticity evaluation results.

[0099] In addition, the display screen 4 displays the authenticity evaluation result and the distribution heat map information of the collected samples.

[0100] In this embodiment, the industrial computer 3, the display screen 4 and the laser raindrop spectrometer 5 are connected by a wire harness to achieve real-time data transmission.

[0101] In summary, this scheme starts from the level of the microscopic raindrop distribution matrix of rainfall, constructs an evaluation benchmark based on the raindrop distribution matrix of real rainfall samples, and realizes quantitative evaluation of the authenticity of the simulated rainfall raindrop distribution by quantifying the similarity between the simulated rainfall samples and the evaluation benchmark.

Claims

1. A method for evaluating the authenticity of raindrop distribution for field rainfall simulation, characterized in that: The following steps are involved: S1. Collect micro data of real rainfall at all levels; S2. Based on the microscopic data of real rainfall, the average raindrop distribution matrix of different levels is calculated as an evaluation benchmark; S3, collecting microscopic data of simulated rainfall at the site and calculating the raindrop distribution matrix of the simulated rainfall; S4, calculating the similarity between the raindrop distribution matrix of the simulated rainfall and the average raindrop distribution matrix of each level; S5. Based on the similarity calculated in step S4, determine the authenticity evaluation result of the simulated rainfall on site.

2. The method for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 1, characterized in that: The microscopic data of the real rainfall in step S1 and the microscopic data of the simulated rainfall in step S3 both include the number of raindrop particles of different diameters and speeds and rainfall intensity data.

3. The method for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 2, characterized in that: The specific process of step S2 is: slicing the microscopic data of real rainfall at each level into different samples according to minutes, each sample contains a corresponding diameter velocity-particle number distribution matrix, and the corresponding level is defined based on the rainfall intensity data within one minute; The diameter velocity-particle number distribution matrices of the same level are superimposed to obtain the average distribution, and the average distribution heat map of real rainfall of different levels is used as the authenticity evaluation benchmark, and the benchmark is iteratively updated as the real rainfall samples are supplemented.

4. The method for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 1, characterized in that: The step S4 comprises the following steps: S41. Based on the formula of structural similarity SSIM, the basic similarity between the raindrop distribution matrix of simulated rainfall and the average raindrop distribution matrix of each level is calculated. The similarity result is a value from 0 to 1. The calculation formula is as follows: Among them, μ x is the element mean of matrix x; μ y is the element mean of matrix y; x is the element variance of the matrix x; σ y is the element variance of matrix y; σ xy is the covariance of matrices x and y; C1 and C2 are constants to avoid the denominator being 0; S42: For pixels with abnormal distribution and abnormal values ​​in the raindrop distribution matrix of simulated rainfall, a penalty term P is set. dist With P val , in order to distinguish the simulated rainfall samples with significant differences in raindrop distribution, the calculation formula is as follows: Among them, X is the raindrop distribution matrix of simulated rainfall, Y is the reference raindrop distribution matrix, P dist The number of abnormal distribution elements whose element values ​​in the simulated rainfall samples are non-zero but whose corresponding position elements in the benchmark distribution matrix are zero is counted, and P val Then the number of abnormal numerical elements whose numerical values ​​in the sample to be tested differ from the corresponding elements in the reference matrix is ​​counted; S43, based on two penalty terms P dist With P val , and get the total penalty coefficient P total , multiply the total penalty coefficient by the basic similarity calculation result to obtain the total similarity S between the raindrop distribution matrix of simulated rainfall and the average raindrop distribution matrix of each level.

5. The method for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 4, characterized in that: The calculation formula of the total similarity S in step S43 is: S(X,Y)=P total ·SSIM(X,Y) Among them, N X is the number of non-zero pixels in the raindrop distribution matrix of simulated rainfall.

6. The method for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 5, characterized in that: The step S5 specifically comprises the following steps: S51, based on the rainfall intensity data of the simulated rainfall, setting the fuzzy membership of the rainfall level, thereby fuzzifying the division boundary of the rainfall level, the fuzzy membership is calculated by the following method: Among them, μ little , μ mid , μ heavy are the fuzzy membership of light, medium and heavy rain, respectively, and R is the rainfall intensity value of simulated rainfall; S52, using the fuzzy membership as the weight, weighted summing up the raindrop distribution similarity between the simulated rainfall and the benchmarks at all levels, and comprehensively obtaining the raindrop distribution authenticity RRD; S53. Based on the RRD evaluation results of simulated rainfall, by horizontally comparing the RRD index calculation results of simulated rainfall with approximate rainfall intensity, or by comparing with the authenticity threshold, it is evaluated whether the simulated rainfall on the site meets the raindrop distribution authenticity requirements based on the high or low values.

7. The method for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 6, characterized in that: The calculation formula of the raindrop distribution authenticity RRD in step S52 is: RRD=μ little *S little +m mid *S mid +m heavy *S heavy Among them, S little , S mid , S heavy They represent the similarity of raindrop distribution between simulated rainfall and small, medium and large levels of realism evaluation benchmarks.

8. The method for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 6, characterized in that: In step S53, the authenticity threshold is defined based on the numerical distribution of the similarity between the real rainfall sample and the distribution benchmark, and the lower limit of the 95% confidence interval based on the similarity distribution result of the real rainfall sample is used as the reference value of the authenticity threshold.

9. A device for evaluating the authenticity of raindrop distribution for field-based rainfall simulation, using the method for evaluating the authenticity of raindrop distribution for field-based rainfall simulation as claimed in claim 1, characterized in that: The invention comprises an integral frame consisting of a base (11), a roller (12) and a column (13), wherein the roller (12) is mounted at the bottom of the base (11), a battery box (2) is arranged at the top of the base (11), an industrial computer (3), a display screen (4) and a laser raindrop spectrometer (5) are mounted on the column (13), and the laser raindrop spectrometer (5) is located at the top of the column (13). The laser raindrop spectrometer (5) is used to collect microscopic data of real rainfall and simulated rainfall and transmit the data to the industrial computer (3), and the industrial computer (3) is used to execute a preset program to calculate the authenticity evaluation result of the simulated rainfall on the site, and transmit the result to the display screen (4) for display.

10. The device for evaluating the authenticity of raindrop distribution for field rainfall simulation according to claim 9, characterized in that: The preset programs in the industrial control computer (3) include a first program and a second program with switchable modes, wherein the first program processes microscopic data of real rainfall, takes real rainfall as the composition of an evaluation benchmark, divides real rainfall samples into grades based on rainfall intensity, and iteratively updates average raindrop distribution matrices of different grades; The second program processes the micro data of simulated rainfall, takes the simulated rainfall on site as the evaluation object, calculates the similarity between the raindrop distribution matrix of the simulated rainfall and the average raindrop distribution matrix of each level, and determines the authenticity evaluation result of the simulated rainfall on site.

Images