CN115408485A - Runoff reconstruction method and system based on small sample observation of site water collection area - Google Patents
Runoff reconstruction method and system based on small sample observation of site water collection area Download PDFInfo
- Publication number
- CN115408485A CN115408485A CN202211067784.0A CN202211067784A CN115408485A CN 115408485 A CN115408485 A CN 115408485A CN 202211067784 A CN202211067784 A CN 202211067784A CN 115408485 A CN115408485 A CN 115408485A
- Authority
- CN
- China
- Prior art keywords
- runoff
- training
- monthly
- estimation
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 54
- 238000012549 training Methods 0.000 claims abstract description 108
- 238000004088 simulation Methods 0.000 claims abstract description 56
- 238000010801 machine learning Methods 0.000 claims abstract description 23
- 230000001419 dependent effect Effects 0.000 claims abstract description 13
- 230000035772 mutation Effects 0.000 claims abstract description 13
- 238000007689 inspection Methods 0.000 claims abstract description 4
- 238000012545 processing Methods 0.000 claims description 31
- 238000004590 computer program Methods 0.000 claims description 11
- 238000001556 precipitation Methods 0.000 claims description 9
- 238000010998 test method Methods 0.000 claims description 9
- 238000003860 storage Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 5
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000013277 forecasting method Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 239000005442 atmospheric precipitation Substances 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000012950 reanalysis Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2474—Sequence data queries, e.g. querying versioned data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Tourism & Hospitality (AREA)
- Mathematical Physics (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Probability & Statistics with Applications (AREA)
- Fuzzy Systems (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Computational Linguistics (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Remote Sensing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a runoff reconstruction method and system based on small sample observation of a site water collection area. The method comprises the following steps: selecting meteorological hydrological time series data of preset n years and observation data of monthly runoff as simulation samples, selecting maximum and minimum year data as a group after sorting according to the annual runoff, dividing the simulation samples into a training set and an estimation set, traversing the simulation samples to obtain an estimation set formed by a T group of training sets and other samples, and constructing a machine learning simulation model by using the training set to obtain the root mean square error of the monthly runoff of the estimation set; determining a reconstructed training sample by using a mutation point inspection method in the T group training set by taking the group number of the T group training set as an independent variable and taking the root mean square error as a dependent variable; constructing a machine learning estimation model by using the reconstructed training sample; and reconstructing a sequence of the monthly runoff by taking the meteorological hydrological time series data as an argument. The method realizes runoff forecast of a long-time sequence, and reduces runoff forecast deviation caused by sample training of the water-rich years and the dry water years.
Description
Technical Field
The invention belongs to the field of watershed water resource assessment, and particularly relates to a runoff reconstruction method and system based on small sample observation of a site water-collecting area.
Background
River runoff changes have great influence on regional economic development and production and life of people. In the national level, a method for reconstructing the runoff of a typical main tributary site is urgently needed to provide a decision basis for the management of a drainage basin.
The process of converting atmospheric precipitation into runoff is complex, and the space-time heterogeneity of the runoff is more complex than that of the precipitation. The hydrological model simulation method can approximately quantitatively characterize complex hydrological phenomena and processes by generalization. At present, a plurality of development bottlenecks are faced, a hydrological conceptual model cannot be completely suitable for hydrological process simulation of all regions, and a physical-based distributed and semi-distributed hydrological model cannot be directly measured due to a plurality of parameters, and model calibration is carried out by a common trial-and-error method and a mathematical optimization method. The manual trial and error method requires that raters have abundant hydrological forecasting experiences and strong subjectivity, and optimization calculation is carried out on hydrological model parameters by using an optimization algorithm. With the rapid global development of big data technology, the machine learning can deeply mine the deep value and the internal relation of big data, and the method is well applied to interdisciplinary in various fields. The big data-based machine learning is simple to operate in runoff prediction, and simple and empirical processing is performed on complex hydrological processes, so that a new theory and method are provided for reconstructing the space-time pattern of runoff and knowing and analyzing the change rule of the runoff.
The forecasting capability of the runoff based on the machine learning is greatly influenced by forecasting elements and training samples. In actual work, complete site runoff actual observation data is often lacked, and an effective solution is not provided at present for forecasting long-time sequence runoff by digging effective information from data of limited small samples.
Disclosure of Invention
In order to solve the technical problems, the invention provides a technical scheme of a runoff reconstruction method based on small sample observation of a site water collection area, so as to solve the technical problems.
The invention discloses a runoff reconstruction method based on site water-collecting area small sample observation in a first aspect, which comprises the following steps:
s1, acquiring monthly runoff data of a hydrological site, and determining a catchment area range;
s2, acquiring a geographical water collection area range of the site according to the geographical water collection area range, and acquiring meteorological hydrological time sequence data of grid month scale one by one in the geographical water collection area range;
s3, selecting meteorological hydrological time series data of preset n years and observation data of the monthly runoff as simulation samples, selecting the maximum and minimum year data as a group in the simulation samples after sorting according to the annual runoff, dividing the simulation samples into a training set and a prediction set, traversing the simulation samples to obtain T groups of training sets and corresponding prediction sets, constructing a machine learning simulation model by using the training sets to obtain the monthly runoff of the prediction set, and then applying the monthly runoff of the prediction set to obtain the root-mean-square error of the monthly runoff of the prediction set;
s4, determining a reconstructed training sample in the T group training set by using a mutation point test method by using the group number of the T group training set as an independent variable and the root-mean-square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable;
s5, applying the reconstructed training sample to construct a machine learning estimation model;
and S6, reconstructing a sequence of the monthly runoff by taking the meteorological hydrological time series data as an argument.
According to the method of the first aspect of the present invention, in the step S1, the data of the monthly runoff includes data of monthly runoff of discontinuous years, and the data of the monthly runoff cannot have a default value.
According to the method of the first aspect of the present invention, in step S2, the method for obtaining the geographical coverage area range of the site according to the geographical coverage area range includes:
and determining the water system flow direction in the catchment area range by using the DEM topographic map, and combining river flow areas collected at the sites according to the water system flow direction to obtain the catchment area geographical range.
According to the method of the first aspect of the present invention, in the step S2, the weather hydrological time-series data includes:
the difference between the grid monthly rainfall and the evapotranspiration and the grid monthly yield.
According to the method of the first aspect of the present invention, in step S3, the method of selecting the largest and smallest year data as one group after sorting according to the annual runoff, dividing the simulation samples into a training set and an estimation set, and traversing the simulation samples to obtain T groups of training sets and corresponding estimation sets includes:
s31, sorting the observation data of the monthly runoff of n years according to the annual runoff;
s32, grouping the years ranked as i and the years ranked as n-i +1, wherein the simulated monthly runoff sample is divided into a training set and an estimation set;
and S33, traversing the simulation samples by applying the method in the step S32 to obtain T groups of training sets, wherein the rest samples are estimation sets.
According to the method of the first aspect of the present invention, in the step S3, the n.gtoreq.3.
According to the method of the first aspect of the present invention, in the step S4, the method for determining the reconstructed training sample in the T-group training set by using the mutation point inspection method with the group number of the T-group training set as an independent variable and the root mean square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable includes:
s41, dividing the root mean square error time sequence of the monthly runoff of the estimation set formed by the rest samples into two groups of time sequences of 1-T and T-T by taking T as a node, and calculating a first statistic U t,T ；
Wherein,
sgn (·) is a sign function, and RMSE is the root mean square error of the monthly runoff of the estimation set;
step S42, applying the first statistic U t,T Calculating a significance level ρ;
wherein,
K T is the second statistic, K T ＝max|U t,T |；
Step S43, obtaining the current K according to the significance level rho T And when the rho is a preset value and is less than 0.01, taking the training set with the group number being t as a reconstruction training sample.
The second aspect of the invention discloses a runoff reconstruction system based on site water-collecting area small sample observation, which comprises:
the first processing module is configured to acquire monthly runoff data of the hydrological site and determine a catchment area range;
the second processing module is configured to obtain a geographical water collection area range of the site according to the geographical water collection area range, and obtain meteorological hydrological time sequence data of grid month scale by grid month scale in the geographical water collection area range;
the third processing module is configured to select meteorological hydrological time series data of preset n years and observation data of the monthly runoff as simulation samples, select the largest and smallest year data as one group in the simulation samples after sorting according to the monthly runoff, divide the simulation samples into a training set and a pre-estimation set, traverse the simulation samples to obtain T groups of training sets and corresponding pre-estimation sets, use the training sets to construct a machine learning simulation model to obtain the monthly runoff of the pre-estimation sets, and use the monthly runoff of the pre-estimation sets to obtain the root mean square error of the monthly runoff of the pre-estimation sets;
the fourth processing module is configured to determine a reconstructed training sample in the T groups of training sets by using the group number of the T groups of training sets as an independent variable and the root mean square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable;
a fifth processing module configured to apply the reconstructed training samples to construct a machine learning prediction model;
and the sixth processing module is configured to reconstruct the sequence of the monthly runoff by taking the meteorological hydrological time series data as an argument.
According to the system of the second aspect of the present invention, the first processing module is configured to determine that the data of the monthly runoff includes data of monthly runoff of discontinuous years, and the data of the monthly runoff cannot have a default value.
According to the system of the second aspect of the present invention, the second processing module is configured to obtain the geographical range of the catchment area of the site according to the geographical range of the catchment area, and the step of obtaining the geographical range of the catchment area of the site according to the geographical range of the catchment area comprises:
and determining the water system flow direction in the catchment area range by using the DEM topographic map, and combining river flow areas collected at the sites according to the water system flow direction to obtain the catchment area geographical range.
According to the system of the second aspect of the present invention, the second processing module is configured to, the weather hydrological time-series data includes:
the net monthly precipitation, the difference between the net monthly precipitation and the evapotranspiration and the net monthly yield.
According to the system of the second aspect of the present invention, the third processing module is configured to, the traversing the simulation samples to obtain the T training sets and the corresponding estimation sets includes:
sequencing the observation data of the monthly runoff of n years according to the annual runoff;
then, the years ranked as i and the years ranked as n-i +1 are combined into a group, and the simulated monthly runoff sample is divided into a training set and an estimation set;
and traversing the simulation samples in a mode that the year ranked as i and the year ranked as n-i +1 are grouped into a training set and an estimation set by using the application, so as to obtain a T group of training sets, wherein the rest samples are the estimation set.
According to the system of the second aspect of the present invention, the third processing module is configured to, n ≧ 3.
According to the system of the second aspect of the present invention, the fourth processing module is configured to determine the reconstructed training sample in the T training set by using a mutation point test method by using the group number of the T training set as an independent variable and the root mean square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable, where the step of determining the reconstructed training sample in the T training set by using the mutation point test method includes:
dividing the root mean square error time sequence of the monthly runoff of the estimation set formed by other samples into 1-T and T-T two groups of sub time sequences by taking T as a node, and calculating a first statistic U t,T ；
Wherein,
sgn (·) is a sign function, and RMSE is the root mean square error of the monthly runoff of the estimation set;
applying the first statistic U t,T Calculating a significance level ρ;
wherein,
K T is the second statistic, K T ＝max|U t,T |；
From the significance level ρ, the value of K is obtained T And when the value is a preset value and rho is less than 0.01, taking the training set with the group number being t as a reconstruction training sample.
A third aspect of the invention discloses an electronic device. The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor executes the computer program to realize the steps of the runoff reconstruction method based on the small sample observation of the site water collection area in any one of the first aspect of the disclosure.
A fourth aspect of the invention discloses a computer-readable storage medium. The computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of a method for runoff reconstruction based on site catchment area small sample observation according to any one of the first aspect of the present disclosure.
The proposal of the invention is that the device comprises a power supply,
(1) The runoff forecasting method realizes long-time sequence runoff forecasting, and reduces runoff forecasting deviation caused by sample training of the water-rich years and the dry water years.
(2) The hydrological meteorological time series data of all grids in the site water collection area are used for carrying out combined forecasting on the site runoff, and the accuracy of the combined forecasting is more reliable than the hydrological meteorological average value forecasting accuracy of the site water collection area.
(3) The method is suitable for forecasting the monthly scale runoff of any watershed site by various machine learning methods, and has strong universality.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a runoff reconstruction method based on site water-collection area small sample observation according to an embodiment of the present invention;
fig. 2 is a structural diagram of a runoff reconstruction system based on site water collection area small sample observation according to an embodiment of the present invention;
fig. 3 is a block diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention discloses a runoff reconstruction method based on small sample observation of a site water collection area in a first aspect. Fig. 1 is a flowchart of a runoff reconstruction method based on small sample observation of a site water collection area according to an embodiment of the present invention, as shown in fig. 1, the method includes:
s1, acquiring monthly runoff data of a hydrological site, and determining a catchment area range;
s2, acquiring a geographical water collection area range of the site according to the geographical water collection area range, and acquiring meteorological hydrological time sequence data of grid month scale one by one in the geographical water collection area range;
s3, selecting meteorological hydrological time series data of preset n years and observation data of monthly runoff as simulation samples, selecting the maximum and minimum years after sorting from the simulation samples, wherein the annual runoff of the two years is a group, dividing the simulation samples into a training set and an estimation set, traversing the simulation samples to obtain T groups of training sets, constructing a machine learning simulation model by using the training sets to obtain the monthly runoff of the estimation set, and then applying the monthly runoff of the estimation set to obtain the root-mean-square error of the monthly runoff of the estimation set;
s4, determining a reconstructed training sample in the T group training set by using a mutation point test method by using the group number of the T group training set as an independent variable and the root-mean-square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable;
s5, applying the reconstructed training sample to construct a machine learning estimation model;
and S6, reconstructing a sequence of the monthly runoff by taking the meteorological hydrological time series data as an argument.
In step S1, monthly runoff data of the hydrological site is obtained, and a water collecting area range is determined.
In the step S1, the data of the monthly runoff is data of monthly runoff of discontinuous years, but the data of the monthly runoff cannot have a default value.
And S2, acquiring a geographical water collection area range of the site according to the geographical water collection area range, and acquiring meteorological hydrological time sequence data of grid month scale one by one in the geographical water collection area range.
In step S2, the method for obtaining the geographical range of the water-collecting area of the site according to the geographical range of the water-collecting area includes:
and determining the water system flow direction in the catchment area range by using the DEM topographic map, and combining river flow areas collected at the sites according to the water system flow direction to obtain the catchment area geographical range.
The weather hydrology time-series data includes:
the net monthly precipitation, the difference between the net monthly precipitation and the evapotranspiration and the net monthly yield.
Specifically, the meteorological hydrological time series data of grid month scale one by one can be meteorological observation site rainfall, evapotranspiration grid interpolation data, ERA5-Land, GLDAS, FLDAS, CLDAS rainfall, evapotranspiration and runoff reanalysis data, and can be GPM, TRMM rainfall data and MODIS evapotranspiration products based on satellite remote sensing images.
In step S3, selecting meteorological hydrological time series data of preset n years and observation data of the monthly runoff as simulation samples, selecting the monthly runoff of the maximum and minimum two years as a group in the simulation samples according to the ranking of the annual runoff, dividing the simulation samples into a training set and an estimation set, traversing the simulation samples to obtain T groups of training sets and corresponding estimation sets, constructing a machine learning simulation model by using the training sets to obtain the monthly runoff of the estimation sets, and then applying the monthly runoff of the estimation sets to obtain the root mean square error of the monthly runoff of the estimation sets.
In some embodiments, in step S3, the method for selecting the largest and smallest year data as a group after sorting according to the annual radial flow, dividing the simulation samples into a training set and an estimation set, and traversing the simulation samples to obtain T groups of training sets and corresponding estimation sets includes:
s31, sorting the observation data of the monthly runoff for n years according to the annual runoff, wherein n is more than or equal to 3;
s32, grouping the years which are ranked as i and the years which are ranked as n-i +1, wherein the simulation samples are divided into a training set and an estimation set;
and S33, traversing the simulation samples by applying the method in the step S32 to obtain a T group training set and a corresponding estimation set.
And S4, determining a reconstructed training sample in the T groups of training sets by using the group number of the T groups of training sets as an independent variable and the root-mean-square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable and using a mutation point test method.
In some embodiments, in step S4, the method for determining the reconstructed training samples in the T training sets by using the mutation point test method with the group number of the T training sets as an independent variable and the root mean square error of the path flow of the estimation set corresponding to the group number as a dependent variable includes:
s41, dividing the root mean square error time sequence of the monthly runoff of the estimation set formed by other samples into 1-T and T-T two-group sub-time sequences by taking T as a node, and calculating a first statistic U t,T ；
Wherein,
sgn (·) is a sign function, and RMSE is the root mean square error of the monthly runoff of the estimation set;
step S42, applying the first statistic U t,T Calculating a significance level ρ;
wherein,
K T is the second statistic, K T ＝max|U t,T |；
Step S43, obtaining K through the significance level rho T And when the value is a preset value and rho is less than 0.01, taking the training set with the group number being t as a reconstruction training sample.
And in step S5, applying the reconstructed training sample to construct a machine learning estimation model.
In some embodiments, the machine learning estimation model constructed by applying the reconstructed training samples is objective and has low influence of sample selection.
In some embodiments, the machine learning prediction model is an artificial neural network regression model, a random forest regression model, a decision tree regression model, a support vector machine regression model, a K-nearest neighbor regression model, an Adaboost regression, a Bagging regression model, or an ExtraTree regression model.
In summary, the solution proposed by the present invention can
(1) The runoff forecasting method realizes long-time sequence runoff forecasting, and reduces runoff forecasting deviation caused by sample training of the water-rich years and the dry water years.
(2) The hydrological meteorological time series data of all grids in the site water collection area are used for carrying out combined forecasting on the site runoff, and the accuracy of the combined forecasting is more reliable than the hydrological meteorological average value forecasting accuracy of the site water collection area.
(3) The method is suitable for forecasting the monthly scale runoff of any watershed site by various machine learning methods, and has strong universality.
The invention discloses a runoff reconstruction system based on site water-collecting area small sample observation in a second aspect. Fig. 2 is a structural diagram of a runoff reconstruction system based on site water collection area small sample observation according to an embodiment of the present invention; as shown in fig. 2, the system 100 includes:
the first processing module 101 is configured to obtain the monthly runoff data of the hydrological site and determine the range of the catchment area;
the second processing module 102 is configured to obtain a water-collecting area geographical range of the site according to the water-collecting area range, and obtain meteorological hydrological time series data of grid month scale by grid month scale in the water-collecting area geographical range;
the third processing module 103 is configured to select meteorological hydrological time series data of preset n years and observation data of the monthly runoff as simulation samples, select the largest and smallest year data as a group in the simulation samples after sorting according to the annual runoff, divide the simulation samples into a training set and an estimation set, traverse the simulation samples to obtain T groups of training sets and corresponding estimation sets, construct a machine learning simulation model by using the training sets to obtain the monthly runoff of the estimation sets, and then apply the monthly runoff of the estimation sets to obtain the root mean square error of the monthly runoff of the estimation sets;
a fourth processing module 104, configured to determine a reconstructed training sample in the T group training set by using the group number of the T group training set as an independent variable and the root mean square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable, and using a mutation point inspection method;
a fifth processing module 105, configured to apply the reconstructed training samples to construct a machine learning prediction model;
a sixth processing module 106 configured to reconstruct the sequence of the monthly runoff using the weather hydrologic time-series data as an argument.
According to the system of the second aspect of the present invention, the first processing module 101 is configured to determine that the profile of the monthly runoff includes data of monthly runoff of discontinuous years, and the profile of the monthly runoff cannot have a default value.
According to the system of the second aspect of the present invention, the second processing module 102 is configured to obtain the geographical coverage area of the site according to the geographical coverage area, including:
and determining the water system flow direction in the catchment area range by using the DEM topographic map, and combining river flow areas collected at the sites according to the water system flow direction to obtain the catchment area geographical range.
According to the system of the second aspect of the present invention, the second processing module 102 is configured to, the weather hydrological time-series data includes:
the net monthly precipitation, the difference between the net monthly precipitation and the evapotranspiration and the net monthly yield.
According to the system of the second aspect of the present invention, the third processing module 103 is configured to, the traversing the simulation samples to obtain T groups of training sets and corresponding estimation sets includes:
sequencing the observation data of the monthly runoff of n years according to the annual runoff;
then, the year ranked as i and the year ranked as n-i +1 are combined into a group, and the simulated monthly runoff sample is divided into a training set and an estimation set;
and traversing the simulation samples in a mode that the years ranked as i and the years ranked as n-i +1 are combined into one group, the simulation monthly runoff sample is divided into a training set and an estimation set to obtain a T group of training sets, and the rest samples are the estimation sets.
According to the system of the second aspect of the present invention, the third processing module 103 is configured to, n ≧ 3.
According to the system of the second aspect of the present invention, the fourth processing module 104 is configured to determine, by using the group number of the T-group training set as an independent variable and the root mean square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable, a reconstructed training sample in the T-group training set by using a mutation point test method includes:
dividing the root mean square error time sequence of the monthly runoff of the estimation set formed by other samples into 1-T and T-T two groups of sub time sequences by taking T as a node, and calculating a first statistic U t,T ；
Wherein,
sgn (·) is a sign function, and RMSE is the root mean square error of the monthly runoff of the estimation set;
applying the first statistic U t,T Calculating a significance level ρ;
wherein,
K T is the second statistic, K T ＝max|U t,T |；
From the significance level ρ, the value of K is obtained T And when the value is a preset value and rho is less than 0.01, taking the training set with the group number being t as a reconstruction training sample.
A third aspect of the invention discloses an electronic device. The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor executes the computer program to realize the steps of the runoff reconstruction method based on the site catchment area small sample observation in any one of the first aspect of the disclosure.
Fig. 3 is a block diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 3, the electronic device includes a processor, a memory, a communication interface, a display screen, and an input device, which are connected by a system bus. Wherein the processor of the electronic device is configured to provide computing and control capabilities. The memory of the electronic equipment comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The communication interface of the electronic device is used for communicating with an external terminal in a wired or wireless mode, and the wireless mode can be realized through WIFI, an operator network, near Field Communication (NFC) or other technologies. The display screen of the electronic equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the electronic equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the electronic equipment, an external keyboard, a touch pad or a mouse and the like.
It will be understood by those skilled in the art that the structure shown in fig. 3 is only a partial block diagram related to the technical solution of the present disclosure, and does not constitute a limitation to the electronic device to which the solution of the present disclosure is applied, and a specific electronic device may include more or less components than those shown in the figure, or combine some components, or have different arrangements of components.
A fourth aspect of the invention discloses a computer-readable storage medium. The computer readable storage medium has stored thereon a computer program which, when executed by the processor, implements the steps in a runoff reconstruction method based on site catchment area small sample observation according to any one of the first aspect of the disclosure.
It should be noted that the technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, however, as long as there is no contradiction between the combinations of the technical features, the scope of the present description should be considered. The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not to be construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A runoff reconstruction method based on small sample observation of a site water collection area is characterized by comprising the following steps:
s1, acquiring monthly runoff data of a hydrological site, and determining a catchment area range;
s2, acquiring a geographical range of a water collecting area of the site according to the geographical range of the water collecting area, and acquiring meteorological hydrological time sequence data of grid month scales one by one in the geographical range of the water collecting area;
s3, selecting meteorological hydrological time sequence data and observation data of the month and diameter flow of preset n years as simulation samples, selecting the largest and smallest year data as one group after sorting according to the year and diameter flow in the simulation samples, dividing the simulation samples into a training set and a pre-estimation set, traversing the simulation samples to obtain T groups of training sets and corresponding pre-estimation sets, constructing a machine learning simulation model by using the training sets to obtain the month and diameter flow of the pre-estimation sets, and then applying the month and diameter flow of the pre-estimation sets to obtain the root mean square error of the month and diameter flow of the pre-estimation sets;
s4, determining a reconstructed training sample in the T group training set by using a mutation point test method by using the group number of the T group training set as an independent variable and the root-mean-square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable;
s5, constructing a machine learning estimation model by applying the reconstructed training sample;
and S6, reconstructing a sequence of the monthly runoff by taking the meteorological hydrological time series data as an independent variable.
2. The runoff rebuilding method based on small sample observation of a water collection area of a site as claimed in claim 1, wherein in said step S1, said monthly runoff data includes monthly runoff data of discontinuous years, and said monthly runoff data cannot have default.
3. A runoff rebuilding method based on small sample observation of a site water-collecting area according to claim 1, wherein in said step S2, said method for obtaining a geographical range of a water-collecting area of a site according to said geographical range of a water-collecting area comprises:
and determining the water system flow direction by using the DEM topographic map in the catchment area range, and combining river flowing areas converged at the sites according to the water system flow direction to obtain the catchment area geographical range.
4. The runoff rebuilding method based on site water collection area small sample observation according to claim 1, wherein in said step S2, said meteorological hydrological time series data comprises:
the net monthly precipitation, the difference between the net monthly precipitation and the evapotranspiration and the net monthly yield.
5. The method for reconstructing runoff based on small sample observation of a site water-collecting area according to claim 1, wherein in the step S3, the maximum and minimum year data is selected as one group after being sorted according to the annual runoff volume, the simulation samples are divided into a training set and an estimation set, and the method for traversing the simulation samples to obtain T groups of training sets and corresponding estimation sets comprises:
s31, sequencing the observation data of the monthly runoff for n years according to the annual runoff;
s32, grouping the years ranked as i and the years ranked as n-i +1, wherein the simulated monthly runoff sample is divided into a training set and an estimation set;
and S33, traversing the simulation samples by applying the method in the step S32 to obtain a T group training set, wherein the rest samples are estimation sets.
6. A runoff rebuilding method based on small sample observation of a site water collection area according to claim 1, wherein in step S3, n is greater than or equal to 3.
7. The method for reconstructing runoff based on small sample observation of a site water-collecting area according to claim 1, wherein in the step S4, the method for determining the reconstructed training sample in the T group of training sets by using a mutation point inspection method with the group number of the T group of training sets as an independent variable and the root mean square error of the runoff volume of the estimation set corresponding to the group number as a dependent variable comprises:
s41, dividing the root mean square error time sequence of the monthly runoff of the estimation set formed by other samples into 1-T and T-T two-group sub-time sequences by taking T as a node, and calculating a first statistic U t,T ；
Wherein,
sgn (·) is a sign function, and RMSE is the root mean square error of the monthly runoff of the estimation set;
step S42, applying the first statistic U t,T Calculating a significance level ρ;
wherein,
K T is the second statistic, K T ＝max|U t,T |；
Step S43, obtaining the current K according to the significance level rho T And when the value is a preset value and rho is less than 0.01, taking the training set with the group number being t as a reconstruction training sample.
8. A runoff reconstruction system for site-based catchment area small sample observation, the system comprising:
the first processing module is configured to acquire the monthly runoff data of the hydrological site and determine the range of the catchment area;
the second processing module is configured to obtain a geographical water collection area range of the site according to the geographical water collection area range, and obtain meteorological hydrological time sequence data of grid month scale by grid month scale in the geographical water collection area range;
the third processing module is configured to select meteorological hydrological time series data of preset n years and observation data of the monthly runoff as simulation samples, select the largest and smallest annual data as a group in the simulation samples after sorting according to the annual runoff, divide the simulation samples into a training set and a prediction set, traverse the simulation samples to obtain T groups of training sets and corresponding prediction sets, use the training sets to construct a machine learning simulation model to obtain the monthly runoff of the prediction sets, and use the monthly runoff of the prediction sets to obtain the root mean square error of the monthly runoff of the prediction sets;
the fourth processing module is configured to determine a reconstructed training sample in the T group training set by using a mutation point test method by using the group number of the T group training set as an independent variable and the root mean square error of the monthly runoff of the estimation set corresponding to the group number as a dependent variable;
a fifth processing module, configured to apply the reconstructed training samples to construct a machine learning prediction model;
and the sixth processing module is configured to reconstruct the sequence of the monthly runoff by taking the meteorological hydrological time series data as an argument.
9. An electronic device, comprising a memory storing a computer program and a processor, wherein the processor, when executing the computer program, implements the steps of any one of claims 1 to 7 in a runoff reconstruction method based on site water collection area small sample observations.
10. A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, performs the steps of any one of claims 1 to 7 of a method for runoff reconstruction based on site water-collection area small sample observations.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211067784.0A CN115408485B (en) | 2022-09-01 | 2022-09-01 | Runoff reconstruction method and system based on small sample observation of site water collection area |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211067784.0A CN115408485B (en) | 2022-09-01 | 2022-09-01 | Runoff reconstruction method and system based on small sample observation of site water collection area |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115408485A true CN115408485A (en) | 2022-11-29 |
| CN115408485B CN115408485B (en) | 2023-04-04 |
Family
ID=84164506
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211067784.0A Active CN115408485B (en) | 2022-09-01 | 2022-09-01 | Runoff reconstruction method and system based on small sample observation of site water collection area |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115408485B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115829163A (en) * | 2023-01-16 | 2023-03-21 | 河海大学 | Multi-mode integration-based runoff prediction method and system for middle and lower reaches of Yangtze river |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108876021A (en) * | 2018-05-31 | 2018-11-23 | 华中科技大学 | A kind of Medium-and Long-Term Runoff Forecasting method and system |
| CN110555561A (en) * | 2019-09-06 | 2019-12-10 | 清华大学 | Medium-and-long-term runoff ensemble forecasting method |
| CN111159149A (en) * | 2019-12-13 | 2020-05-15 | 国网浙江省电力有限公司紧水滩水力发电厂 | River flow prediction method based on three-dimensional convolutional neural network |
| WO2021217776A1 (en) * | 2020-04-27 | 2021-11-04 | 中山大学 | Hydro-meteorological time sequence-based runoff simulation method and system |
-
2022
- 2022-09-01 CN CN202211067784.0A patent/CN115408485B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108876021A (en) * | 2018-05-31 | 2018-11-23 | 华中科技大学 | A kind of Medium-and Long-Term Runoff Forecasting method and system |
| CN110555561A (en) * | 2019-09-06 | 2019-12-10 | 清华大学 | Medium-and-long-term runoff ensemble forecasting method |
| CN111159149A (en) * | 2019-12-13 | 2020-05-15 | 国网浙江省电力有限公司紧水滩水力发电厂 | River flow prediction method based on three-dimensional convolutional neural network |
| WO2021217776A1 (en) * | 2020-04-27 | 2021-11-04 | 中山大学 | Hydro-meteorological time sequence-based runoff simulation method and system |
Non-Patent Citations (2)
| Title |
|---|
| D.K. DWIVED等: "Rainfall and runoff estimation of micro watersheds of coastal Navsari" * |
| 赵阳: "密云水库集水区变化环境下的小流域径流演变规律研究" * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115829163A (en) * | 2023-01-16 | 2023-03-21 | 河海大学 | Multi-mode integration-based runoff prediction method and system for middle and lower reaches of Yangtze river |
| CN115829163B (en) * | 2023-01-16 | 2023-04-11 | 河海大学 | Multi-mode integration-based runoff prediction method and system for middle and lower reaches of Yangtze river |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115408485B (en) | 2023-04-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Arnal et al. | Skilful seasonal forecasts of streamflow over Europe? | |
| Nourani et al. | ANN-based statistical downscaling of climatic parameters using decision tree predictor screening method | |
| CN112765808B (en) | Ecological drought monitoring and evaluating method | |
| CN105243435B (en) | A kind of soil moisture content prediction technique based on deep learning cellular Automation Model | |
| CN113902580B (en) | Historical farmland distribution reconstruction method based on random forest model | |
| Liu et al. | Support vector machine and data assimilation framework for Groundwater Level Forecasting using GRACE satellite data | |
| CN111428942A (en) | Line icing thickness prediction method for extracting micro-terrain factors based on variable grid technology | |
| Hussein et al. | Rainfall prediction using machine learning models: literature survey | |
| CN115079307A (en) | Wind power plant weather forecasting method and system based on multi-mode optimal integration | |
| Kittel et al. | Informing hydrological models of poorly gauged river catchments–a parameter regionalization and calibration approach | |
| CN115408485B (en) | Runoff reconstruction method and system based on small sample observation of site water collection area | |
| CN118350678B (en) | Water environment monitoring data processing method and system based on Internet of things and big data | |
| CN115496148A (en) | Method and device for evaluating large-scale hydrological situation change of river | |
| CN116955450A (en) | Hydrologic forecasting method and system for fusion of drainage basin production confluence space-time process | |
| Xue et al. | Climate variability impacts on runoff projection under quantile mapping bias correction in the support CMIP6: An investigation in Lushi basin of China | |
| Stein | Some statistical issues in climate science | |
| Sahu et al. | Density-based spatial clustering of application with noise approach for regionalisation and its effect on hierarchical clustering | |
| Saliha et al. | Estimation of flow in ungauged catchments by coupling a hydrological model and neural networks: case study | |
| Dayal et al. | Deep learning for Multi-horizon Water levelForecasting in KRS reservoir, India | |
| Muluye | Comparison of statistical methods for downscaling daily precipitation | |
| CN113762615A (en) | Flood prediction method and device, computer equipment and storage medium | |
| Wang et al. | Statistical prediction of typhoon-induced total accumulated rainfall in the Western North Pacific using typhoon track similarity indices | |
| Zhang et al. | High resolution annual irrigation water use maps in China based-on input variables selection and convolutional neural networks | |
| Zanella et al. | Sensor networks, data processing, and inference: the hydrology challenge | |
| Peeters et al. | Conceptual evaluation of continental land-surface model behaviour |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |