Design rainstorm process calculation method fusing rainfall random variables

Abstract

The invention discloses a design rainstorm process calculation method fused with rainfall random variables, which comprises the following steps: acquiring long-series rainfall information of a hydrometric station in a research area, performing frequency arrangement analysis on information data, and deducing design surface rainstorm of a specific frequency of a drainage basin above a section; selecting a rainfall process in a specific time period according to an engineering design condition, and performing distribution fitting on each interval time period sequence of the selected rainfall of multiple sessions; extracting characteristic parameters of each rainfall by adopting a two-dimensional copula function, and establishing a relationship with a fitting optimal distribution function of each time period; randomly generating a sequence meeting copula extraction parameter distribution and each time period distribution, generating a random rain pattern, and obtaining a random rain pattern set; and assimilating the rainfall process according to the design surface rainstorm to obtain a design rainstorm process line. According to the method, the multi-session rainfall characteristics of the research area are comprehensively considered, and a new thought is provided for randomly generating design rainstorm calculation conforming to the local rainfall characteristics.

Description

Technical Field

[0001] The invention relates to a method for calculating a designed rainstorm process in a watershed, in particular to a method for calculating a designed rainstorm process integrating rainfall random variables, and belongs to the technical field of flood control safety design of water conservancy projects. Background Art

[0002] The "Specification for Calculation of Design Floods for Water Conservancy and Hydropower Engineering: SL44-2006" proposes two basic methods for basin design floods: calculation of design floods based on flow data and calculation of design floods based on rainstorm data. Among them, when calculating the design flood based on flow data, the annual peak flow and flood volume series in different periods should be composed of the maximum values ​​of each year when performing frequency calculations. When the flood and rainstorm data of the design basin are scarce, the statistical parameters of the peak and flood volume calculated by the analysis of the neighboring areas, or the peak modulus of the same frequency, etc., can be used for regional synthesis and used for the design basin. When there is a lack of flood and rainstorm data in the design flow city, but reliable historical floods have been investigated near the project site and their recurrence period is close to the design flood standard of the project, the historical flood level can be directly used with appropriate adjustments as the design flood of the project; in calculating the design flood based on rainstorm data, when the various duration surface rainstorm series of the flow city are long, the frequency analysis method can be used for calculation; when the basin area is small and various duration surface rainstorm series are scarce, the corresponding duration design point rainstorm and the rainstorm point-surface relationship can be used for indirect calculation; the rainstorm point-surface relationship should adopt the comprehensive fixed-point fixed-surface relationship of the region; when the data conditions are not available, the moving-point-moving-surface relationship can also be used, but it should be appropriately corrected; when the basin area is very small, the design point rainstorm can be used as the basin design surface rainstorm; when the rainstorm data in the waterlogged area is scarce, the typical year method can be used for calculation; when the elevation gradient of the design basin changes greatly, the design surface rainstorm should be checked for rationality based on the law of rainfall changes with elevation, and appropriate corrections should be made if necessary.

[0003] In the prior art, a method for identifying a typical rainstorm process in a time period, such as the method disclosed in the publication number CN110457374A, includes determining the time period length; calculating the accumulated rainfall within the time period length; extracting typical rainfall process samples; determining typical classes; establishing a characteristic index system of a typical rainfall process in a time period; calculating the weighted average of the characteristic index values ​​of a typical rainfall process in a time period; and determining a typical rainfall process sequence in a time period; in order to solve the problems of subjective arbitrariness and large manual workload in the traditional method, the selection and identification of the typical rainstorm process is objective, quantitative and automated, and the work efficiency is improved. Rainstorms have obvious randomness, but the prior art uses a specific analysis method and a specific weighting function for calculation, so it cannot be applied to various different rain types for calculation. When the design rainstorm is used to deduce the design flood, the rainstorm type has an impact on the process line calculation of the design flood. Currently, fixed rain types are used, which are inconsistent with natural rainfall conditions. Therefore, an objective and reasonable design rainstorm calculation method that can take into account the change of rain types is urgently needed. Summary of the invention

[0004] The purpose of the present invention is to provide a method for calculating a designed rainstorm process by integrating rainfall random variables in order to solve at least one of the above technical problems.

[0005] The present invention achieves the above-mentioned purpose through the following technical scheme: a design rainstorm process calculation method integrating rainfall random variables, the design rainstorm process calculation method comprising the following steps:

[0006] Step 1: Collect long series rainfall data from hydrological stations in the study area and deduce the design surface rainstorm P with a specific frequency in the basin above the section;

[0007] Step 2: According to the engineering design situation, a specific rainfall period T is selected for period distribution fitting, and the total duration of all rainfall in the rainfall data is selected as the rainfall in the specific period, and each rainfall duration data is arranged in order from large to small;

[0008] Step 3: Perform distribution fitting on the rainfall characteristic values ​​of each rainfall event, select a specific rainfall period T for rain type fitting according to the engineering design, select different distribution functions to fit the sequence values ​​of each period, and determine the best fitting of each period sequence through evaluation indicators as the basis for generating random functions;

[0009] Step 4: The copula function establishes the relationship between the rainfall characteristic parameters and the optimal distribution function for each At period;

[0010] Step 5: Randomly generate a sequence that satisfies the distribution of copula extraction parameters and the distribution of each time period to generate a random rain model;

[0011] Step 6: Assimilate the rainfall process according to the design surface rainstorm to obtain the design rainstorm process line:

[0012] P i,j =P×m i,j

[0013] Where P i,j is the rainfall distribution result of random rain type; P is the total rainfall; m i,j Assign a proportion of rainfall patterns that satisfy the distribution function to each time period.

[0014] As a further solution of the present invention: in step 1, the design rainfall value of the specific return period of the research site is derived using P-III distribution, generalized Pareto distribution or lognormal distribution, and whether the rainfall value is a KC rainstorm mode is determined based on the intensity-duration-frequency curve under the specific return period;

[0015] By introducing the rainstorm peak position coefficient r to characterize the peak time, the duration of the KC rainstorm mode is divided into two parts: before and after the peak. The specific process is as follows:

[0016] (1) Divide the sample of the annual maximum 24-hour rainfall process and the duration of each rainfall into several segments at 1-hour intervals and calculate r;

[0017] (2) The instantaneous intensity before and after the peak is i(t b )、i(t a ), the corresponding durations are t b ,t a ; The rainstorm intensity formula under a certain return period T is The instantaneous rainfall intensity before and after the peak can be calculated as follows:

[0018]

[0019] Where A, b and n are the parameters in the formula of rainstorm intensity with a specific return period.

[0020] As a further solution of the present invention: in step 2, the specific time period is set to 1h, 6h, 12h, 24h, 3d, 7d, 15d, 30d; the matrix calculation formula of the proportion of rainfall in each rainfall period Δt to the total rainfall is as follows:

[0021]

[0022] Where: m i,j is the proportion of the j-th rainfall in the i-th period.

[0023] As a further solution of the present invention: in step 2, the object of the distribution function fitting is the time course distribution ratio, and the time course distribution ratio calculation formula is:

[0024]

[0025] As a further solution of the present invention: in step 3, the selected distribution function includes but is not limited to Gauss distribution function, Gamma distribution function, Poisson distribution function, t distribution function or F distribution function, and the distribution function can also select Weibull distribution, LogNormal distribution and generalized extreme value distribution, wherein:

[0026] The formula for the Gauss distribution function is:

[0027]

[0028] The formula for the Gamma distribution function is:

[0029]

[0030] The formula for the Poisson distribution function is:

[0031]

[0032] The formula for the t distribution function is:

[0033]

[0034] The formula for the F distribution function is:

[0035]

[0036] When determining the optimal distribution function for each time period series, it is necessary to select appropriate evaluation indicators, and the selected evaluation indicators include but are not limited to mean square error MSE, root mean square error RMSE, mean absolute error MAE, mean absolute percentage error MAPE or AIC criteria, where:

[0037] Mean square error MSE formula:

[0038]

[0039] Root mean square error RMSE formula:

[0040]

[0041] Mean absolute error MAE formula:

[0042]

[0043] Mean absolute percentage error MAPE formula:

[0044]

[0045] AIC criterion formula:

[0046]

[0047] As a further solution of the present invention: in step 4, for each m i and m i-1 There are different marginal distributions F i (m) and F i-1 (m), so m i The conditional distribution function of is:

[0048]

[0049] Use two-dimensional FrankCopula function to connect each m i Fitting distribution function:

[0050]

[0051] Copula functions can also choose GumbelCopula and ClaytonCopula, where:

[0052]

[0053] As a further solution of the present invention: in step 5, the random rain pattern model based on the Copula function is used to generate the random rain pattern P s (t), and obtain a random rain type set; among them, the rain type time distribution ratio m i,,j Satisfy the constraints:

[0054]

[0055] The random rainfall pattern generated by copula satisfies both the optimal distribution of the sequence in each time period and the optimal distribution of the characteristic value of the rainfall process.

[0056] A device for constructing a design rainstorm process model integrating rainfall random variables, comprising:

[0057] The acquisition module is used to obtain long-series rainfall data information of different time lengths at the hydrological stations in the study area;

[0058] An analysis module is used to perform frequency calculation on the data information to obtain surface rainfall information;

[0059] A fitting module, used to perform distribution fitting on the data information in the acquisition module, and comprehensively select a fitting function through multiple indicators;

[0060] The calculation module is used to calculate the random rainfall process in the study area according to the results of the analysis module and the fitting module.

[0061] A computer device comprises: a memory and a processor, the memory and the processor are connected to each other for communication, the memory stores computer instructions, and the processor executes the computer instructions to extract random distribution characteristics of rainfall and design a rainstorm calculation method.

[0062] A computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable a computer to execute a method for extracting random distribution characteristics of rainfall and designing a rainstorm calculation method.

[0063] The beneficial effects of the present invention are as follows: the present invention obtains long-series rainfall information from the hydrological station in the study area, performs frequency analysis on the information data, and deduces the design surface rainstorm of the specific frequency of the basin above the section; selects the rainfall process in a specific period according to the engineering design situation, and performs distribution fitting on each interval period sequence of the selected multiple rainfall sessions; uses a two-dimensional copula function to extract the characteristic parameters of each rainfall, and establishes a relationship with the optimal distribution function for each period; randomly generates a sequence that satisfies the distribution of the copula extraction parameters and the distribution of each period, generates a random rain type, and obtains a random rain type set; assimilates the rainfall process according to the design surface rainstorm to obtain the design rainstorm process line. Comprehensively considering the characteristics of multiple rainfall sessions in the study area, it provides a new idea for the random generation of design rainstorm calculations that meet the local rainfall characteristics;

[0064] The present invention fully considers the impact of differences in rainfall patterns on the design flood, selects a specific rainfall period for rainfall pattern fitting according to the engineering design situation, selects different distribution functions to fit the sequence values ​​of each period, and determines the best fitting of each period sequence through evaluation indicators as the basis for generating random functions, so as to be able to select the most suitable distribution function for different situations, so as to improve the accuracy of rainfall pattern fitting. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 is a flow chart of a method according to an embodiment of the present invention;

[0066] Figure 2 It is a schematic diagram of the structure of a model building device for extracting random distribution characteristics of rainfall in a watershed and designing calculations of a rainstorm process in an embodiment of the present invention;

[0067] Figure 3 A schematic diagram of the structure of an electronic device in an embodiment of the present invention;

[0068] Figure 4 Fitting diagrams of different distributions output by the electronic device of the present invention;

[0069] Figure 5 It is the random rainfall process output by the electronic device of the present invention. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] Embodiment 1, as Figures 1 to 5 As shown, a design rainstorm process calculation method integrating rainfall random variables, the design rainstorm process calculation method comprises the following steps:

[0072] Step 1: Collect long-series rainfall data from hydrological stations in the study area and deduce the design surface rainstorm P with a specific frequency in the basin above the section (e.g., once in a hundred years);

[0073] Step 2: According to the engineering design situation, a specific rainfall period T is selected for period distribution fitting. The total duration of all rainfall in the rainfall data is selected as the rainfall in a specific period, and each rainfall duration data is arranged in order from large to small.

[0074] Step 3: Perform distribution fitting on the rainfall characteristic values ​​of each rainfall event, select a specific rainfall period T for rain type fitting according to the engineering design, select different distribution functions to fit the sequence values ​​of each period, and determine the best fitting of each period sequence through evaluation indicators as the basis for generating random functions;

[0075] Step 4: The copula function establishes the relationship between the rainfall characteristic parameters and the optimal distribution function for each At period;

[0076] Step 5: Randomly generate a sequence that satisfies the distribution of copula extraction parameters and the distribution of each time period to generate a random rain model;

[0077] Step 6: Assimilate the rainfall process according to the design surface rainstorm to obtain the design rainstorm process line:

[0078] P i,j =P×m i,j

[0079] Where: P i,j is the rainfall distribution result of random rain type; P is the total rainfall; m i,j Assign a proportion of rainfall patterns that satisfy the distribution function to each time period.

[0080] Embodiment 2: In addition to all the technical features of Embodiment 1, this embodiment also includes:

[0081] In step 1, the design rainfall value of the specific return period of the research site is derived using the P-III distribution, generalized Pareto distribution or log-normal distribution, and the rainfall value is judged to be the KC rainstorm mode based on the intensity-duration-frequency curve under the specific return period;

[0082] By introducing the rainstorm peak position coefficient r to characterize the peak time, the duration of the KC rainstorm mode is divided into two parts: before and after the peak. The specific process is as follows:

[0083] (1) Divide the sample of the annual maximum 24-hour rainfall process and the duration of each rainfall into several segments at 1-hour intervals and calculate r;

[0084] (2) The instantaneous intensity before and after the peak is i(t b )、i(t a ), and the corresponding durations are t b ,t a ; The rainstorm intensity formula under a certain return period T is The instantaneous rainfall intensity before and after the peak can be calculated as follows:

[0085]

[0086] Among them, A, b and n are parameters in the formula of rainstorm intensity with a specific return period.

[0087] Furthermore, in step 2, the specific time period is set to 1h, 6h, 12h, 24h, 3d, 7d, 15d, and 30d; the matrix calculation formula for the proportion of rainfall in each rainfall period Δt to the total rainfall is as follows:

[0088]

[0089] Where: m i,j is the proportion of the j-th rainfall in the i-th period.

[0090] Furthermore, in step 2, the object of the distribution function fitting is the time course distribution ratio, and the time course distribution ratio calculation formula is:

[0091]

[0092] Further, in step 3, the selected distribution function includes but is not limited to Gauss distribution function, Gamma distribution function, Poisson distribution function, t distribution function or F distribution function, and the distribution function can also select Weibull distribution, LogNormal distribution and generalized extreme value distribution, where:

[0093] The formula for the Gauss distribution function is:

[0094]

[0095] The formula for the Gamma distribution function is:

[0096]

[0097] The formula for the Poisson distribution function is:

[0098]

[0099] The formula for the t distribution function is:

[0100]

[0101] The formula for the F distribution function is:

[0102]

[0103] Among them, when determining the optimal distribution function for each time period series fitting, it is necessary to select appropriate evaluation indicators, and the selected evaluation indicators include but are not limited to mean square error MSE, root mean square error RMSE, mean absolute error MAE, mean absolute percentage error MAPE or AIC criterion, where:

[0104] Mean square error MSE formula:

[0105]

[0106] Root mean square error RMSE formula:

[0107]

[0108] Mean absolute error MAE formula:

[0109]

[0110] Mean absolute percentage error MAPE formula:

[0111]

[0112] AIC criterion formula:

[0113]

[0114] In step 4, for each m i and m i-1 There are different marginal distributions F i (m) and F i-1 (m), so m i The conditional distribution function of is:

[0115]

[0116] Use two-dimensional FrankCopula function to connect each m i Fitting distribution function:

[0117]

[0118] Copula functions can also choose GumbelCopula and ClaytonCopula, where:

[0119]

[0120] In step 5, the random rain pattern model based on the Copula function is used to generate the random rain pattern P s (t), and obtain a random rain type set; among them, the rain type time distribution ratio m i,,j Satisfy the constraints:

[0121]

[0122] The random rainfall pattern generated by copula satisfies both the optimal distribution of the sequence in each time period and the optimal distribution of the characteristic value of the rainfall process.

[0123] Embodiment three, as Figures 2 to 5 As shown, a device for constructing a design rainstorm process model integrating rainfall random variables comprises:

[0124] The acquisition module is used to obtain long-series rainfall data information of different time lengths at the hydrological stations in the study area;

[0125] An analysis module is used to perform frequency calculation on the data information to obtain surface rainfall information;

[0126] A fitting module, used to perform distribution fitting on the data information in the acquisition module, and comprehensively select a fitting function through multiple indicators;

[0127] The calculation module is used to calculate the random rainfall process in the study area according to the results of the analysis module and the fitting module.

[0128] A computer device comprises: a memory and a processor, the memory and the processor are connected to each other for communication, the memory stores computer instructions, and the processor executes the computer instructions to extract random distribution characteristics of rainfall and design a rainstorm calculation method.

[0129] A computer-readable storage medium stores computer instructions, wherein the computer instructions are used to enable a computer to execute a method for extracting random distribution characteristics of rainfall and designing a rainstorm calculation method.

[0130] Working principle: obtain long series rainfall information from the hydrological station in the study area, perform frequency analysis on the information data, and deduce the design surface rainstorm with a specific frequency in the basin above the section; select the rainfall process in a specific period according to the engineering design situation, and perform distribution fitting on each interval period sequence of the selected multiple rainfall events; use the two-dimensional copula function to extract the characteristic parameters of each rainfall event, and establish the relationship with the optimal distribution function for each t period; randomly generate a sequence that satisfies the distribution of the copula extraction parameters and the distribution of each period, generate a random rain pattern, and obtain a random rain pattern set; assimilate the rainfall process according to the design surface rainstorm to obtain the design rainstorm process line.

[0131] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0132] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A design rainstorm process calculation method integrating rainfall random variables, characterized in that: The design rainstorm process calculation method comprises the following steps: Step 1: Collect long series rainfall data from hydrological stations in the study area and deduce the design surface rainstorm P with a specific frequency in the basin above the section; Step 2: According to the engineering design situation, a specific rainfall period T is selected for period distribution fitting, and the total duration of all rainfall in the rainfall data is selected as the rainfall in the specific period, and each rainfall duration data is arranged in order from large to small; Step 3: Perform distribution fitting on the rainfall characteristic values ​​of each rainfall event, select a specific rainfall period T for rain pattern fitting according to the engineering design, select different distribution functions to fit the sequence values ​​of each period, and determine the best fitting of each period sequence through evaluation indicators as the basis for generating random functions; Step 4: The copula function establishes the relationship between the rainfall characteristic parameters and the optimal distribution function for each At period; Step 5: Randomly generate a sequence that satisfies the distribution of copula extraction parameters and the distribution of each time period to generate a random rain model; Step 6: Assimilate the rainfall process according to the design surface rainstorm to obtain the design rainstorm process line: P i,j =P×m i,j Where: P i,j is the rainfall distribution result of random rain type; P is the total rainfall; m i,j Assign a proportion of rainfall patterns that satisfy the distribution function to each time period.

2. The design rainstorm process calculation method according to claim 1 is characterized in that: In the step 1, the design rainfall value of the specific return period of the research site is derived using the P-III distribution, the generalized Pareto distribution or the log-normal distribution, and whether the rainfall value is a KC rainstorm mode is determined based on the intensity-duration-frequency curve under the specific return period; By introducing the rainstorm peak position coefficient r to characterize the peak time, the duration of the KC rainstorm mode is divided into two parts: before and after the peak. The specific process is as follows: (1) Divide the sample of the annual maximum 24-hour rainfall process and the duration of each rainfall into several segments at 1-hour intervals and calculate r; (2) The instantaneous intensity before and after the peak is i(t b )、i(t a ), the corresponding durations are t b ,t a ; The rainstorm intensity formula under a certain return period T is The instantaneous rainfall intensity before and after the peak is calculated as follows: Where A, b and n are the parameters in the formula of rainstorm intensity with a specific return period.

3. The design rainstorm process calculation method according to claim 2 is characterized in that: In step 2, the specific time period is set to 1h, 6h, 12h, 24h, 3d, 7d, 15d, and 30d; the matrix calculation formula for the proportion of rainfall in each rainfall period Δt to the total rainfall is as follows: Where: m i,j is the proportion of the j-th rainfall in the i-th period.

4. The design rainstorm process calculation method according to claim 3 is characterized by: In step 2, the object of the distribution function fitting is the time course distribution ratio, and the time course distribution ratio calculation formula is:

5. The design rainstorm process calculation method according to claim 4 is characterized in that: In step 3, the selected distribution function includes but is not limited to Gauss distribution function, Gamma distribution function, Poisson distribution function, t distribution function or F distribution function. The distribution function can also select Weibull distribution, LogNormal distribution and generalized extreme value distribution, where: The formula for the Gauss distribution function is: The formula for the Gamma distribution function is: The formula for the Poisson distribution function is: The formula for the t distribution function is: The formula for the F distribution function is: Among them, when determining the optimal distribution function for each time period series fitting, it is necessary to select appropriate evaluation indicators, and the selected evaluation indicators include but are not limited to mean square error MSE, root mean square error RMSE, mean absolute error MAE, mean absolute percentage error MAPE or AIC criterion, where: Mean square error MSE formula: Root mean square error RMSE formula: Mean absolute error MAE formula: Mean absolute percentage error MAPE formula: AIC criterion formula:

6. The design rainstorm process calculation method according to claim 5 is characterized by: In step 4, for each m i and m i-1 There are different marginal distributions F i (m) and F i-1 (m), so m i The conditional distribution function of is: Use two-dimensional FrankCopula function to connect each m i Fitting distribution function: Copula functions select GumbelCopula and ClaytonCopula, where:

7. The design rainstorm process calculation method according to claim 6 is characterized by: In step 5, a random rain pattern model based on the Copula function is used to generate a random rain pattern P s (t), and obtain a random rain type set; among them, the rain type time distribution ratio m i,j Satisfy the constraints: The random rainfall pattern generated by copula satisfies both the optimal distribution of the sequence in each time period and the optimal distribution of the characteristic value of the rainfall process.

8. A device for constructing a design rainstorm process model integrating rainfall random variables, characterized in that: include: The acquisition module is used to obtain long-series rainfall data information of different time lengths at the hydrological stations in the study area; An analysis module is used to perform frequency calculation on the data information to obtain surface rainfall information; A fitting module, used to perform distribution fitting on the data information in the acquisition module, and comprehensively select a fitting function through multiple indicators; The calculation module is used to calculate the random rainfall process in the study area according to the results of the analysis module and the fitting module.

9. A computer device, characterized in that: include: The memory and the processor are communicatively connected with each other, the memory stores computer instructions, and the processor executes any one of claims 1-7 to extract the random distribution characteristics of rainfall and design a rainstorm calculation method by executing the computer instructions.

10. A computer-readable storage medium, characterized in that: The readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute any one of claims 1-7 to extract the random distribution characteristics of rainfall and design a rainstorm calculation method.

Images