Short-term daily flow prediction method and system based on historical statistical data

Abstract

The invention discloses a short-term daily flow prediction method and system based on historical statistical data, and relates to the technical field of basin water environment simulation and prediction. Calculating the maximum value and the minimum value of the historical day-by-day traffic according to the historical traffic data; calculating a maximum value, a minimum value and an average value of historical traffic in different average periods according to the historical traffic data; obtaining a day to be predicted, and obtaining a predicted flow initial value by adopting a random sampling method; and calculating the maximum value, the minimum value and the average value of the predicted flow initial values in different average periods according to the predicted flow initial values and the historical flow data, and calibrating the predicted flow initial values to obtain the predicted flow value of the day to be predicted. According to the method, the short-term traffic is predicted based on historical observation data for many years, the randomness of day-by-day traffic generation is reflected, the method can be constrained by statistical rules of different average periods, and the method has operability in actual work.

Description

Technical Field

[0001] The present invention relates to the technical field of basin water environment simulation prediction, and more specifically to a short-term daily flow prediction method and system based on historical statistical data. Background Art

[0002] River flow prediction, especially short-term flow prediction, is an important reference for water resources management, water environment management and even water ecological management. Water volume prediction is generally based on hydrology, using data analysis and other means to estimate and speculate on water resources in a certain area or basin in a specific period in the future. Traditional water volume prediction is based on existing hydrological forecasting means or methods, making full use of real-time and historical data of different natures, data sources and data types such as hydrological and meteorological station observation data, satellite remote sensing and radar data, numerical model output data, basic geographic information data, etc., with the help of modern information and forecasting technologies such as numerical elevation models, information fusion, data assimilation, and large-scale hydrological models, to analyze and predict the rainfall in a certain area or basin in the future and the surface water and groundwater that may be formed. The advantage of this mechanistic prediction is that it is more rigorous and consistent with the occurrence mechanism, and the disadvantage is that it has high requirements for data and real-time information, and there are certain difficulties in the actual operation process.

[0003] With the continuous accumulation of traffic data, data scalability and reliability are further increased. Using statistical data to predict traffic, especially after fully considering short-term randomness and long-term statistical laws, using big data statistical methods for data prediction and analysis has gradually become an important means. However, the big data-driven approach may have some deviations in certain extreme values ​​and extreme ranges, or when the external environment undergoes major changes.

[0004] Therefore, how to provide a short-term daily flow prediction method and system based on historical statistical data, predict short-term flow based on many years of historical observation data, which not only reflects the randomness of daily flow, but also can be constrained by the statistical laws of different averaging periods and be operational in actual work is a problem that technical personnel in this field urgently need to solve. Summary of the invention

[0005] In view of this, the present invention provides a short-term daily flow prediction method and system based on historical statistical data, taking into account the short-term management needs of water resources and water environment, predicting the daily flow of rivers to support water resources and water environment management, and predicting short-term flow based on many years of historical observation data, which not only reflects the randomness of daily flow, but also is subject to the statistical laws of different averaging periods, and is operational in actual work.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a short-term daily flow prediction method based on historical statistical data, comprising:

[0007] Collect historical flow data from river hydrological stations or sections in previous years;

[0008] Calculate the maximum and minimum values ​​of historical daily traffic according to the historical traffic data;

[0009] Calculate the maximum value, minimum value and average value of historical flow in different averaging periods according to the historical flow data;

[0010] Obtain the day to be predicted, and use random sampling method within the range of historical daily maximum and minimum flow rates to obtain the initial value of the predicted flow rate;

[0011] Calculate the maximum, minimum and average values ​​of the predicted initial flow values ​​for different averaging periods based on the predicted initial flow value and the historical flow data;

[0012] Based on the maximum, minimum and average values ​​of historical flows in different averaging periods and the maximum, minimum and average values ​​of initial values ​​of predicted flows in different averaging periods, the predicted initial values ​​of flows are calibrated to obtain the predicted flow value of the day to be predicted.

[0013] Preferably, calibrating the predicted flow initial value to obtain the predicted flow value for the day to be predicted includes:

[0014] If the average value of the initial value of the predicted flow in the average period is greater than the average value of the historical flow in the average period, adjust the initial value of the predicted flow so that the average value of the initial value of the predicted flow in the average period is equal to the average value of the historical flow in the average period;

[0015] If the maximum value of the initial value of the average period predicted flow is greater than the maximum value of the average period historical flow, adjust the initial value of the predicted flow so that the maximum value of the initial value of the average period predicted flow is equal to the maximum value of the average period historical flow;

[0016] If the minimum value of the initial value of the average period predicted flow is less than the minimum value of the average period historical flow, adjust the initial value of the predicted flow so that the minimum value of the initial value of the average period flow is equal to the minimum value of the average period historical flow.

[0017] Preferably, the maximum value of the historical daily flow is expressed as:

[0018] QMX i =max(q i,j );

[0019] The minimum value of historical daily traffic is expressed as:

[0020] QMN i =min(q i,j );

[0021] Among them, QMX i is the maximum flow on the i-th day; QMNi is the minimum flow rate on the i-th day; q i,j is the flow rate on the i-th day of the j-th year.

[0022] Preferably, the maximum value of historical flow in different averaging periods is expressed as:

[0023] Q1MX k =max(q1 k,j );

[0024] The minimum value of historical flow in different averaging periods is expressed as:

[0025] Q1N k =min(q1 k,j );

[0026] The average value of historical flow in different averaging periods is expressed as:

[0027] Q1 k =ave(q1 k,j );

[0028] Among them, Q1MX k is the maximum value of the average flow rate on the 1st to kth day; Q1MN k is the minimum value of the average flow rate on the 1st to kth day; Q1 k is the average of the average flow rates on the 1st to kth day; q1 k,j is the average flow rate on day 1-k in year j.

[0029] Preferably, the average flow rate of day 1-k in year j is expressed as:

[0030]

[0031] Among them, q i,j is the flow rate on the i-th day of the j-th year.

[0032] Preferably, the initial value of the predicted flow is expressed as:

[0033] q0 i =rand(QMN i ,QMX i );

[0034] Among them, q0 i is the initial flow value of the i-th day, QMX i is the maximum flow on the i-th day; QMN i is the minimum flow rate on the i-th day.

[0035] Preferably, a short-term daily flow prediction system based on historical statistical data comprises:

[0036] The hydrological data collection module is used to collect historical flow data of river hydrological stations or sections in previous years;

[0037] A first calculation module, used to calculate the maximum and minimum values ​​of the historical daily flow according to the historical flow data;

[0038] A second calculation module is used to calculate the maximum value, minimum value and average value of historical flow in different averaging periods according to the historical flow data;

[0039] The third calculation module is used to obtain the day to be predicted, and use a random sampling method within the range of the maximum and minimum values ​​of the historical daily flow to obtain the initial value of the predicted flow;

[0040] A fourth calculation module, used to calculate the maximum value, minimum value and average value of the initial value of the predicted flow rate in different averaging periods according to the predicted initial value of the flow rate and the historical flow data;

[0041] The flow value prediction module is used to calibrate the predicted flow initial value based on the maximum, minimum and average values ​​of historical flows in different averaging periods and the maximum, minimum and average values ​​of the predicted flow initial value in different averaging periods to obtain the predicted flow value of the day to be predicted.

[0042] Through the above technical scheme, it can be known that compared with the prior art, the present invention discloses a short-term daily flow prediction method and system based on historical statistical data, including: collecting historical flow data of river hydrological stations or sections in previous years; calculating the maximum and minimum values ​​of historical daily flow according to the historical flow data; calculating the maximum, minimum and average values ​​of historical flow in different averaging periods according to the historical flow data; obtaining the day to be predicted, using a random sampling method within the interval of the maximum and minimum values ​​of historical daily flow, to obtain the initial value of predicted flow; calculating the maximum, minimum and average values ​​of the initial values ​​of predicted flow in different averaging periods according to the initial value of predicted flow and the historical flow data; based on the maximum, minimum and average values ​​of historical flow in different averaging periods and the maximum, minimum and average values ​​of the initial values ​​of predicted flow in different averaging periods, calibrating the initial value of predicted flow to obtain the predicted flow value of the day to be predicted. The present invention integrates historical hydrological data, the distribution and variation law of daily flow, takes into account both a certain randomness and the daily value of daily flow and the statistical value of the averaging period, predicts short-term daily flow, and can better serve related applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A schematic flow chart of a short-term daily flow prediction method based on historical statistical data provided by the present invention. DETAILED DESCRIPTION

[0045] 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.

[0046] The embodiment of the present invention discloses a short-term daily flow prediction method based on historical statistical data, such as Figure 1 As shown, including:

[0047] Collect historical flow data from river hydrological stations or sections in previous years;

[0048] Calculate the maximum and minimum values ​​of historical daily traffic according to the historical traffic data;

[0049] Calculate the maximum value, minimum value and average value of historical flow in different averaging periods according to the historical flow data;

[0050] Obtain the day to be predicted, and use random sampling method within the range of historical daily maximum and minimum flow rates to obtain the initial value of the predicted flow rate;

[0051] Calculate the maximum, minimum and average values ​​of the predicted initial flow values ​​for different averaging periods based on the predicted initial flow value and the historical flow data;

[0052] Based on the maximum, minimum and average values ​​of historical flows in different averaging periods and the maximum, minimum and average values ​​of initial values ​​of predicted flows in different averaging periods, the predicted initial values ​​of flows are calibrated to obtain the predicted flow value of the day to be predicted.

[0053] Specifically, during the data collection process, daily flow data of river hydrological stations or typical sections are collected, and the data length is recommended to be no less than 35 years. For stations that have been established for less than 35 years or have monitoring data for less than 35 years, relevant hydrological methods can be used for interpolation and extension to ensure that the data length is no less than 35 years.

[0054] Specifically, the traffic statistics of the embodiment of the present invention are calculated by calculating the daily maximum value, minimum value, and average value according to the calendar year (January 1st to December 31st).

[0055] Specifically, the maximum value of historical daily traffic is expressed as:

[0056] QMX i =max(qi,j );

[0057] The minimum value of historical daily traffic is expressed as:

[0058] QMN i =min(q i,j );

[0059] Among them, QMX i is the maximum flow rate on the i-th day, m 3 / s;QMN i is the minimum flow rate on the i-th day, m 3 / s;q i,j is the flow rate on the i-th day of the j-th year, m 3 / s.

[0060] Specifically, the traffic statistics for different average periods are calculated as follows: the maximum, minimum, and average values ​​for different average periods are calculated according to the calendar year (January 1st - December 31st), such as 1 day, 2 days, 3 days...366 days.

[0061] Specifically, the maximum value of historical flow in different averaging periods is expressed as:

[0062] Q1MX k =max(q1 k,j );

[0063] The minimum value of historical flow in different averaging periods is expressed as:

[0064] Q1N k =min(q1 k,j );

[0065] The average value of historical flow in different averaging periods is expressed as:

[0066] Q1 k =ave(q1 k,j );

[0067] Among them, if you want to obtain the maximum, minimum and average values ​​of the average period from 1 to k days (1-k days), Q1MX k is the maximum value of the average flow rate on the 1st to kth day, m 3 / s;Q1MN k is the minimum value of the average flow rate on the 1st to kth day, m 3 / s;Q1 k is the average of the average flow rates on the 1st to kth day, m 3 / s;q1 k,j is the average flow rate of day 1-k in year j, m 3 / s.

[0068] Specifically, the average flow rate of day 1-k in year j is expressed as:

[0069]

[0070] Among them, q i,j is the flow rate on the i-th day of the j-th year, m 3 / s.

[0071] Specifically, the initial value of the predicted flow is expressed as:

[0072] q0 i =rand(QMN i ,QMX i );

[0073] Among them, q0 i is the initial flow value of the i-th day, m 3 / s, QMX i is the maximum flow rate on the i-th day, m 3 / s;QMN i is the minimum flow rate on the i-th day, m 3 / s.

[0074] Specifically, calibrating the predicted flow initial value to obtain the predicted flow value of the day to be predicted includes:

[0075] If the average value of the initial value of the predicted flow in the average period is greater than the average value of the historical flow in the average period, adjust the initial value of the predicted flow so that the average value of the initial value of the predicted flow in the average period is equal to the average value of the historical flow in the average period;

[0076] If the maximum value of the initial value of the average period predicted flow is greater than the maximum value of the average period historical flow, adjust the initial value of the predicted flow so that the maximum value of the initial value of the average period predicted flow is equal to the maximum value of the average period historical flow;

[0077] If the minimum value of the initial value of the average period predicted flow is less than the minimum value of the average period historical flow, adjust the initial value of the predicted flow so that the minimum value of the initial value of the average period flow is equal to the minimum value of the average period historical flow.

[0078] Specifically, in the verification of the predicted daily flow, the maximum, minimum and average values ​​from the first day to the predicted day are calculated based on the initial value of the predicted daily flow, and compared with the maximum, minimum and average values ​​of different periods of many years of statistics to adjust the initial value of the predicted day.

[0079]

[0080] Q2MX k =max(q2 k );

[0081] Q2 k =min(q2 k );

[0082] Q2 k =ave(q2 k );

[0083] (1) If Q2 k ≠Q1 k ,q01 i =q0 i +(Q2 k -Q1 k )×k,

[0084] Otherwise q01 i =q0 i ;

[0085] (2) If Q2MX k >Q1MX k ,q02 i =q0 i -(Q2MX k -Q1MX k )×k,

[0086] Otherwise q02 i =q0 i ;

[0087] (3) If Q2MN k <Q1MN k ,q03 i =q0 i +(Q2MX k -Q1MX k )×k,

[0088] Otherwise q03 i =q0 i ;

[0089] (4)q00 i =ave(q01 i ,q02 i ,q03 i );

[0090] Among them, q2 k is the initial value of the average flow rate for 1-k days, m 3 / s;Q2MX k is the maximum value of the initial flow forecast value for the 1-kth day average period, m 3 / s;Q2MN k is the minimum value of the initial flow forecast value for the 1-k-day average period, m 3 / s;Q2 k is the average value of the initial flow forecast value for the 1-kth day average period, m 3 / s;q3 iis the measured or predicted daily flow value, m 3 / s;q01 i is the daily flow forecast value after considering the correction of the mean values ​​of different averaging periods, m 3 / s;q02 i is the daily flow forecast value after correction considering the maximum value of different average periods, m 3 / s;q03 i The daily flow forecast value after considering the minimum value of different average periods, m 3 / s;q00 i is the predicted flow value of the day to be predicted, m 3 / s.

[0091] In the embodiment of the present invention, k represents a certain day in a year, and the initial prediction value of the kth day of the prediction year is randomly obtained based on the minimum and maximum values ​​of the historical daily flow, and then corrected according to different periods (such as the average value, maximum value and minimum value of 1-5 days).

[0092] Among them, q3 i is the measured or predicted daily flow value. When it is predicted for k days, the average flow from the 1st day to the kth day is obtained by {using the measured or predicted values ​​from 1 to k-1 days (the main purpose is to correct) plus the predicted value of the kth day} and finally dividing it by k.

[0093] The embodiment of the present invention also discloses a short-term daily flow prediction system based on historical statistical data, including:

[0094] The hydrological data collection module is used to collect historical flow data of river hydrological stations or sections in previous years;

[0095] A first calculation module, used to calculate the maximum and minimum values ​​of the historical daily flow according to the historical flow data;

[0096] A second calculation module is used to calculate the maximum value, minimum value and average value of historical flow in different averaging periods according to the historical flow data;

[0097] The third calculation module is used to obtain the day to be predicted, and use a random sampling method within the range of the maximum and minimum values ​​of the historical daily flow to obtain the initial value of the predicted flow;

[0098] A fourth calculation module, used to calculate the maximum value, minimum value and average value of the initial value of the predicted flow rate in different averaging periods according to the predicted initial value of the flow rate and the historical flow data;

[0099] The flow value prediction module is used to calibrate the predicted flow initial value based on the maximum, minimum and average values ​​of historical flows in different averaging periods and the maximum, minimum and average values ​​of the predicted flow initial value in different averaging periods to obtain the predicted flow value of the day to be predicted.

[0100] In the process of water resources and water environment management, more risk management is adopted, that is, a management mode in which the risk probability is acceptable, and the risks that humans can bear are controlled, such as the water supply guarantee rate reaches 99%, the ecological flow meets 90%, etc., rather than comprehensive management of extreme values ​​or extreme events. Therefore, the prediction of short-term flow based on historical statistical data proposed in the embodiment of the present invention is feasible and reliable in conventional management. Predicting short-term flow based on many years of historical observation data not only reflects the randomness of daily flow, but also can be constrained by the statistical laws of different average periods, which is very operational in actual work.

[0101] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0102] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A short-term daily flow prediction method based on historical statistical data, characterized in that: include: Collect historical flow data from river hydrological stations or sections in previous years; Calculate the maximum and minimum values ​​of historical daily traffic according to the historical traffic data; Calculate the maximum value, minimum value and average value of historical flow in different averaging periods according to the historical flow data; Obtain the day to be predicted, and use random sampling method within the range of historical daily maximum and minimum flow rates to obtain the initial value of the predicted flow rate; Calculate the maximum, minimum and average values ​​of the predicted initial flow values ​​for different averaging periods based on the predicted initial flow value and the historical flow data; Based on the maximum, minimum and average values ​​of historical flows in different averaging periods and the maximum, minimum and average values ​​of initial values ​​of predicted flows in different averaging periods, the predicted initial values ​​of flows are calibrated to obtain the predicted flow value of the day to be predicted.

2. A short-term daily flow prediction method based on historical statistical data according to claim 1, characterized in that: The predicted flow initial value is calibrated to obtain the predicted flow value of the day to be predicted, including: If the average value of the initial value of the predicted flow in the average period is greater than the average value of the historical flow in the average period, adjust the initial value of the predicted flow so that the average value of the initial value of the predicted flow in the average period is equal to the average value of the historical flow in the average period; If the maximum value of the initial value of the average period predicted flow is greater than the maximum value of the average period historical flow, adjust the initial value of the predicted flow so that the maximum value of the initial value of the average period predicted flow is equal to the maximum value of the average period historical flow; If the minimum value of the initial value of the average period predicted flow is less than the minimum value of the average period historical flow, adjust the initial value of the predicted flow so that the minimum value of the initial value of the average period flow is equal to the minimum value of the average period historical flow.

3. The short-term daily flow prediction method based on historical statistical data according to claim 1 is characterized in that: The maximum value of historical daily traffic is expressed as: QMX i =max(q i,j ); The minimum value of historical daily traffic is expressed as: QMN i min(q i,j )4 Among them, QMX i is the maximum flow on the i-th day; QMN i is the minimum flow rate on the i-th day; q i,j is the flow rate on the i-th day of the j-th year.

4. The method for short-term daily flow prediction based on historical statistical data according to claim 1 is characterized in that: The maximum value of historical flow in different averaging periods is expressed as: Q1MX k =max(q1 k,j ); The minimum value of historical flow in different averaging periods is expressed as: Q1MN k =min(q1 k,j ); The average value of historical flow in different averaging periods is expressed as: Q1 k =ave(q1 k,j ); Among them, Q1MX k is the maximum value of the average flow rate on the 1st to kth day; Q1MN k is the minimum value of the average flow rate on the 1st to kth day; Q1 k is the average value of the average flow rate on the 1st to kth day; q1 k,j is the average flow rate on day 1-k in year j.

5. The method for short-term daily flow prediction based on historical statistical data according to claim 4 is characterized in that: The average flow rate on day 1-k in year j is expressed as: Among them, q i,j is the flow rate on the i-th day of the j-th year.

6. The method for short-term daily flow prediction based on historical statistical data according to claim 1 is characterized in that: The initial value of the predicted flow is expressed as: q0 i 1 rand(QMN i ,QMX i )4 Among them, q0 i is the initial flow value of the i-th day, QMX i is the maximum flow on the i-th day; QMN i is the minimum flow rate on the i-th day.

7. A short-term daily flow prediction system based on historical statistical data, characterized in that: include: The hydrological data collection module is used to collect historical flow data of river hydrological stations or sections in previous years; A first calculation module, used to calculate the maximum and minimum values ​​of the historical daily flow according to the historical flow data; A second calculation module is used to calculate the maximum value, minimum value and average value of historical flow in different averaging periods according to the historical flow data; The third calculation module is used to obtain the day to be predicted, and use a random sampling method within the range of the maximum and minimum values ​​of the historical daily flow to obtain the initial value of the predicted flow; A fourth calculation module, used to calculate the maximum value, minimum value and average value of the initial value of the predicted flow rate in different averaging periods according to the predicted initial value of the flow rate and the historical flow data; The flow value prediction module is used to calibrate the predicted flow initial value based on the maximum, minimum and average values ​​of historical flows in different averaging periods and the maximum, minimum and average values ​​of the predicted flow initial value in different averaging periods to obtain the predicted flow value of the day to be predicted.

Images