A medium-term runoff prediction method and system considering time-schedule distribution law

Abstract

The present invention provides a medium-term runoff prediction method and system taking into account the time-course distribution law. For the research basin, the long-sequence hydrological runoff data of the main control sections are collected and sorted, including: hydrological station flow data and reservoir station inflow data at daily, ten-day, monthly, seasonal and annual scales. Based on the water balance principle, the runoff restoration calculation is carried out to obtain the natural runoff sequence of any node in the basin; the water inflow frequency analysis is carried out to determine the quarterly and monthly water inflow frequency of any node under different water inflow years; the distribution ratio of the upper, middle and lower ten-day scales of the runoff of any node is calculated, and the runoff time-course distribution coefficient of the month is defined; based on the natural flow at the monthly scale of a certain section, similar characteristic historical years are found, and the time-course distribution coefficient of the historical runoff sequence is matched to obtain the medium-term natural flow prediction. The time-course distribution of the historical runoff sequence is matched to obtain the monthly runoff ten-day scale distribution plan, which provides important technical support for the medium- and long-term forecast of hydrological meteorology.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrology and water resources forecasting and dispatching, and in particular to a medium-term runoff prediction method and system taking into account time schedule distribution rules. Background Art

[0002] As a core link in river basin water resources management, runoff forecasting is particularly important in flood prevention and disaster reduction, hydropower generation benefits, and agricultural irrigation. In the current context of increasingly tight water resources and intensified climate change, high-precision medium- and long-term hydrological forecasts have become a key tool for guiding the optimal allocation of water resources, optimizing reservoir scheduling strategies, and effectively responding to flood control and drought relief challenges. Faced with the urgent needs and challenges of water resources management, the country's requirements for efficient water resources management and intelligent reservoir scheduling are constantly increasing, and higher requirements are placed on the accuracy and reliability of medium- and long-term runoff forecasts.

[0003] Extreme weather events caused by climate change are frequent, human activities have exacerbated the uncertainty of runoff changes, and river runoff changes frequently, which limits the applicability and accuracy of traditional prediction methods. Although research in the field of medium- and long-term runoff forecasts has made some progress, it is still in the stage of continuous exploration and improvement. Compared with short-term runoff forecasts, medium- and long-term forecasts still have a certain gap in accuracy, timeliness and meeting actual production needs. Therefore, improving the accuracy and reliability of runoff forecasts and making more accurate runoff forecasts are still important directions and challenges for current research.

[0004] The vigorous development of diversified models and theories in the field of runoff forecasting has provided new ideas for the research of medium- and long-term runoff forecasting, but each model and technology has its advantages and limitations. In order to meet the demand for runoff forecast accuracy, it is still necessary to explore more effective prediction methods and provide technical support for medium- and long-term runoff forecasting models and methods. Summary of the invention

[0005] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art and provide a medium-term runoff prediction method and system taking into account the time distribution law, define the monthly runoff time distribution coefficient, match similar runoff processes in historical years, and realize the prediction of natural flow at the ten-day scale of monthly runoff.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] The present invention provides a medium-term runoff prediction method taking into account the time schedule distribution law, comprising:

[0008] S1. Collect and organize long-sequence hydrological runoff data of the control section of the research basin, including flow data of hydrological stations at daily, ten-day, monthly, seasonal and annual scales. and the inflow data of the reservoir station ;

[0009] S2. For any node, based on the principle of water balance, runoff restoration calculation is carried out to restore the impact of reservoir group dispatch above the section where the basin node is located to the section where it is located, and obtain the natural runoff sequence of the basin node, specifically:

[0010] ;

[0011] or ;

[0012] in, For cross section No. Natural flow of time, ; Hydrological station section No. Live traffic at all times; Reservoir station section No. Live traffic at all times; For the Time section The above reservoir groups have cross sections The impact of traffic flow; The influence of propagation time is always considered, that is, the discharge of the upstream reservoir group is at time Propagate to section ;

[0013] S3. Carry out water inflow frequency analysis for the natural runoff sequence of any node year, quarter, and month to determine the water inflow frequency in different quarters and months in different water inflow years of abundant, normal, and dry seasons. The calculation formula is:

[0014] ;

[0015] Among them, different years, quarters or months are sorted from largest to smallest according to their average runoff; It is the rank of a year, quarter or month; Totals for the year, quarter or month to be sorted; is the water inflow frequency;

[0016] S4. For the natural runoff sequence of each node, the distribution of the monthly runoff in different scales of the first ten days, the middle ten days and the last ten days of each month is calculated respectively, and the distribution ratio of the three ten days of each month is defined as the runoff time distribution coefficient of the corresponding month;

[0017] S5. Based on the monthly natural flow of a section in the research basin obtained from hydrological and meteorological forecasts, the time distribution coefficient of the historical runoff sequence is matched according to the characteristics of the year, quarter, and month and the similarity with historical years to obtain the monthly runoff ten-day scale distribution plan, that is, the predicted value of the natural flow in each ten-day period.

[0018] Furthermore, the long-sequence hydrological runoff data in S1 refers to flood report data or compiled data. If compiled data exists, the compiled data shall be used; if not, the report data shall be used. Among the data of different scales, the compiled or report data of the daily, ten-day, or monthly scales shall prevail. If there is no compiled or report data of the ten-day or monthly scale, the daily scale data shall prevail. Calculate and obtain supplementary data on a ten-day and monthly basis. The specific formula is:

[0019] ;

[0020] ;

[0021] in, is the natural flow rate of a certain section at a ten-day scale, m 3 / s; is the monthly natural flow rate of a certain section, m 3 / s; is the daily natural flow of a certain section, m 3 / s; They are the number of days included in the ten-day and month scales respectively.

[0022] Furthermore, the runoff time distribution coefficient in S4 is a multidimensional vector, specifically:

[0023] ;

[0024] in, is the runoff time distribution coefficient of a certain section in a certain month; The natural flow in the first, middle and last ten days of a month, m 3 / s.

[0025] Furthermore, the matching in S5 refers to the matching process between the predicted object and the historical runoff process.

[0026] Furthermore, a medium-term runoff prediction system taking into account the time-course distribution law comprises: at least one processor, and a memory connected to at least one of the processors in communication; wherein,

[0027] The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the medium-term runoff prediction method taking into account the time schedule distribution law.

[0028] The beneficial effects of the present invention are as follows: for the study basin, the long-sequence hydrological runoff data of the main control sections are collected and sorted, including: hydrological station flow data and reservoir station inflow data at daily, ten-day, monthly, seasonal and annual scales; based on the water balance principle, the runoff restoration calculation is carried out to obtain the natural runoff sequence of any node in the basin; the water inflow frequency analysis is carried out to determine the quarterly and monthly water inflow frequency of any node in different water inflow years; the distribution ratio of the upper, middle and lower ten-day scales of the runoff of any node is calculated, and the runoff time-course distribution coefficient of the month is defined; based on the natural flow at the monthly scale of a certain section, similar characteristic historical years are found, the time-course distribution coefficient of the historical runoff sequence is matched, and the natural flow prediction at the ten-day scale is obtained. The time-course distribution of the historical runoff sequence is matched to obtain the ten-day scale distribution plan of the monthly runoff, providing important technical support for the medium- and long-term forecast of hydrological meteorology. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flow chart of a medium-term runoff prediction method taking into account the time distribution law. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0031] See also Figure 1 , a medium-term runoff prediction method taking into account the time distribution law, including:

[0032] S1. Collect and organize long-sequence hydrological runoff data of the control section of the research basin, including flow data of hydrological stations at daily, ten-day, monthly, seasonal and annual scales. and the inflow data of the reservoir station ;

[0033] S2. For any node, based on the principle of water balance, runoff restoration calculation is carried out to restore the impact of reservoir group dispatch above the section where the basin node is located to the section where it is located, and obtain the natural runoff sequence of the basin node, specifically:

[0034] ;

[0035] or ;

[0036] in, For cross section No. Natural flow of time, ; Hydrological station section No. Live traffic at all times; Reservoir station section No. Live traffic at all times; For the Time section The above reservoir groups have cross sections The impact of traffic flow; The influence of propagation time is always considered, that is, the discharge of the upstream reservoir group is at time Propagate to section ;

[0037] S3. Carry out water inflow frequency analysis for the natural runoff sequence of any node year, quarter, and month to determine the water inflow frequency in different quarters and months in different water inflow years of abundant, normal, and dry seasons. The calculation formula is:

[0038] ;

[0039] Among them, different years, quarters or months are sorted from largest to smallest according to their average runoff; It is the rank of a year, quarter or month; Totals for the year, quarter or month to be sorted; is the water inflow frequency;

[0040] Among them, the frequency analysis of water inflow is used to judge the similarity between the water inflow in the predicted year and the typical historical year, which can be divided into extremely flood year, flood year, normal water year, dry year and extremely dry year. The frequency analysis of monthly and quarterly runoff is used to judge the similarity between the month and quarter of the predicted decade and the typical historical process.

[0041] S4. For the natural runoff sequence of each node, the distribution of the monthly runoff in different scales of the first ten days, the middle ten days and the last ten days of each month is calculated respectively, and the distribution ratio of the three ten days of each month is defined as the runoff time distribution coefficient of the corresponding month;

[0042] S5. Based on the monthly natural flow of a section in the research basin obtained from hydrological and meteorological forecasts, the time distribution coefficient of the historical runoff sequence is matched according to the characteristics of the year, quarter, and month and the similarity with historical years to obtain the monthly runoff ten-day scale distribution plan, that is, the predicted value of the natural flow in each ten-day period.

[0043] The long-sequence hydrological runoff data in S1 refers to flood report data or compiled data. If compiled data exists, it shall be used; if not, the report data shall be used. Among the data of different scales, the compiled or report data of the daily, ten-day or monthly scales shall prevail. If there is no compiled or report data of the ten-day or monthly scale, the daily scale data shall prevail. Calculate and obtain supplementary data on a ten-day and monthly basis. The specific formula is:

[0044] ;

[0045] ;

[0046] in, is the natural flow rate of a certain section at a ten-day scale, m 3 / s; is the monthly natural flow rate of a certain section, m 3 / s; is the daily natural flow of a certain section, m 3 / s; They are the number of days included in the ten-day and month scales respectively.

[0047] The runoff time distribution coefficient in S4 is a multidimensional vector, specifically:

[0048] ;

[0049] in, is the runoff time distribution coefficient of a certain section in a certain month; The natural flow in the first, middle and last ten days of a month, m 3 / s.

[0050] The matching in S5 refers to the matching process between the predicted object and the historical runoff process.

[0051] Specific matching methods include multi-year mean matching method, typical year matching method, similar year matching method, and critical period matching method;

[0052] The multi-year mean matching method refers to selecting the average water inflow under multi-year conditions and using the time distribution of the multi-year average water inflow process as the basis for prediction;

[0053] The typical year matching method refers to selecting water inflows from similar historical typical years and using the runoff time distribution coefficient of similar typical years as the basis for prediction;

[0054] The similar year matching method refers to selecting historical years with similar water inflow frequencies and using the runoff time distribution of similar years as the basis for prediction;

[0055] The critical period matching method refers to matching the period of the prediction object with the historical period, and selecting the runoff time distribution in a similar historical period as the basis for prediction.

[0056] A medium-term runoff prediction system taking into account the time-course distribution law comprises: at least one processor and a memory connected to the at least one processor in communication; wherein:

[0057] The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the medium-term runoff prediction method taking into account the time schedule distribution law.

[0058] Example:

[0059] The runoff data of Danjiangkou Reservoir section in the Hanjiang River Basin from 1956 to 2024 were collected, and the data of February each year were selected as calculation examples; runoff restoration calculations were carried out to restore the impact of the reservoir group dispatching above the section where the node was located in the basin to the section, as shown in Table 1; the data series from 1956 to 2023 were used as the calculation period, and the data from 2024 were used as the verification period; the inflow frequency of the natural runoff series from 1956 to 2023 was analyzed, as shown in Table 2; the February runoff time-series distribution coefficient of the historical data series was calculated; the similar year matching method was used to match the historical runoff time-series distribution, and the inflow frequency in February 2024 was equivalent to that in February 2012. The runoff time-series distribution coefficient in February 2012 was selected as the basis for the prediction of inflow in each decade of February 2024. The specific prediction results are shown in Table 3; the absolute error was used as the accuracy assessment indicator, and the accuracy assessment results are shown in Table 3. The absolute error of each decade was within 10%.

[0060] Table 1 Natural runoff series of Danjiangkou Reservoir in February

[0061]

[0062] Table 2 Natural runoff series of Danjiangkou Reservoir in February

[0063]

[0064] Table 3 Analysis of water inflow frequency in Danjiangkou Reservoir in February

[0065]

[0066] Table 4 Analysis of water inflow frequency in Danjiangkou Reservoir in February

[0067]

[0068] Table 5 Danjiangkou Reservoir water inflow forecast for each decade in February

[0069]

[0070] As shown in Tables 1 to 5, the method of the present invention adopts the calculation ideas of water frequency analysis, historical runoff time distribution coefficient definition, similar year matching, and ten-day natural flow prediction, which can quickly realize medium-term (ten-day scale) runoff prediction. With full reference to the runoff distribution of historical data series, the runoff prediction results are obtained, which also proves the feasibility and effectiveness of the method. It can be seen that the method has excellent application effect in medium-term (ten-day scale) runoff prediction.

[0071] According to the above analysis, the method of the present invention is highly practical and can effectively solve the problem of medium-term (ten-day scale) runoff prediction method.

[0072] In summary, the present invention has the advantages of practicality and strong operability. It can quickly match the time distribution of historical runoff sequences, obtain the runoff prediction results of important sections, and provide a more scientific and efficient new method for basin hydrological forecasting.

[0073] The above-mentioned embodiments only express the implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be based on the attached claims.

Claims

1. A medium-term runoff prediction method taking into account the time distribution law, characterized in that: include: S1. Collect and organize long-sequence hydrological runoff data of the control section of the research basin, including flow data of hydrological stations at daily, ten-day, monthly, seasonal and annual scales. and the inflow data of the reservoir station ; S2. For any node, based on the principle of water balance, runoff restoration calculation is carried out to restore the impact of reservoir group dispatch above the section where the basin node is located to the section where it is located, and obtain the natural runoff sequence of the basin node, specifically: ； or ; in, For cross section No. Natural flow of time, ; Hydrological station section No. Live traffic at all times; Reservoir station section No. Live traffic at all times; For the Time section The above reservoir groups have cross sections The impact of traffic flow; The influence of propagation time is always considered, that is, the discharge of the upstream reservoir group is at time Propagate to section ; S3. Carry out water inflow frequency analysis for the natural runoff sequence of any node year, quarter, and month to determine the water inflow frequency in different quarters and months in different water inflow years of abundant, normal, and dry seasons. The calculation formula is: ； Among them, different years, quarters or months are sorted from largest to smallest according to their average runoff; It is the rank of a year, quarter or month; Totals for the year, quarter, or month to be sorted; is the water inflow frequency; S4. For each node's natural runoff sequence, calculate the distribution of the monthly runoff in the first ten days, the middle ten days, and the last ten days of each month respectively, and define the distribution ratio of the three ten days of each month as the runoff time distribution coefficient of the corresponding month; S5. Based on the monthly natural flow of a section in the research basin obtained from hydrological and meteorological forecasts, the time distribution coefficient of the historical runoff sequence is matched according to the characteristics of the year, quarter, and month and the similarity with historical years to obtain the monthly runoff ten-day scale distribution plan, that is, the predicted value of the natural flow in each ten-day period.

2. A medium-term runoff prediction method taking into account the time distribution law according to claim 1, characterized in that: The long-sequence hydrological runoff data in S1 refers to flood report data or compiled data. If compiled data exists, it shall be used; if not, the report data shall be used. Among the data of different scales, the compiled or report data of the daily, ten-day or monthly scales shall prevail. If there is no compiled or report data of the ten-day or monthly scale, the daily scale data shall prevail. Calculate and obtain supplementary data on a ten-day and monthly basis. The specific formula is: ； ； in, is the natural flow rate of a certain section at a ten-day scale, m 3 / s; is the monthly natural flow rate of a certain section, m 3 / s; is the daily natural flow of a certain section, m 3 / s; They are the number of days included in the ten-day and month scales respectively.

3. A medium-term runoff prediction method taking into account the time distribution law according to claim 2, characterized in that: The runoff time distribution coefficient in S4 is a multidimensional vector, specifically: ； in, is the runoff time distribution coefficient of a certain section in a certain month; The natural flow in the first, middle and last ten days of a month, m 3 / s.

4. A medium-term runoff prediction method taking into account the time distribution law according to claim 3, characterized in that: The matching in S5 refers to the matching process between the predicted object and the historical runoff process.

5. A medium-term runoff prediction system taking into account the time distribution law, characterized in that: include: at least one processor, and a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement a medium-term runoff prediction method taking into account a time schedule distribution law as described in any one of claims 1 to 4.

Images