CN111914465A - Data-free regional hydrological parameter calibration method based on clustering and particle swarm optimization - Google Patents
Data-free regional hydrological parameter calibration method based on clustering and particle swarm optimization Download PDFInfo
- Publication number
- CN111914465A CN111914465A CN202010819100.2A CN202010819100A CN111914465A CN 111914465 A CN111914465 A CN 111914465A CN 202010819100 A CN202010819100 A CN 202010819100A CN 111914465 A CN111914465 A CN 111914465A
- Authority
- CN
- China
- Prior art keywords
- parameter
- parameters
- hydrological
- data
- particle swarm
- 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 66
- 239000002245 particle Substances 0.000 title claims abstract description 45
- 238000005457 optimization Methods 0.000 title claims abstract description 40
- 238000004364 calculation method Methods 0.000 claims abstract description 32
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 23
- 239000002689 soil Substances 0.000 claims abstract description 6
- 230000008020 evaporation Effects 0.000 claims abstract description 4
- 238000001704 evaporation Methods 0.000 claims abstract description 4
- 230000008569 process Effects 0.000 claims description 14
- 238000007621 cluster analysis Methods 0.000 claims description 13
- 238000012821 model calculation Methods 0.000 claims description 7
- 238000013136 deep learning model Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 abstract description 5
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 244000144992 flock Species 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/25—Design optimisation, verification or simulation using particle-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for calibrating hydrological parameters of a data-free area based on clustering and particle swarm optimization, which comprises the following steps: collecting soil texture, vegetation coverage, land utilization rate, terrain data, runoff coefficient, total annual evaporation amount, slope and gradient data; dividing the calibration area into calculation units with the square kilometers below 30; determining the underlying surface of each parameter of each computing unit and weather related factors according to the physical characteristics of the parameters of the hydrological model; clustering, analyzing and dividing similar units for each underlying surface factor; calculating hydrological model parameters of all similar units by adopting a particle swarm optimization algorithm to obtain optimal parameters of similar watersheds; the method solves the technical problems that the traditional trial and error method is adopted for determining the hydrological model parameters of the watershed with the data, namely, the parameter values of the hydrological model are continuously adjusted manually to meet the requirement of simulation precision, and the method has artificial subjectivity, low work repeatability, low efficiency and high complexity, is not beneficial to application and popularization of the hydrological model and the like.
Description
Technical Field
The invention belongs to a hydrological parameter calibration technology, and particularly relates to a data-free regional hydrological parameter calibration method based on clustering and particle swarm optimization.
Background
The hydrological model plays an important role in hydrological law research and production practical problem solving, along with the rapid development of modern scientific technology, the information technology taking computers and communication as the core is widely applied to the fields of hydrological water resources and hydraulic engineering science, so that the research of the hydrological model is rapidly developed and is widely applied to the fields of hydrological basic law research, prevention and control of flood and drought disasters, water resource evaluation and development and utilization, water environment and ecological system protection, climate change, analysis of influences of human activities on the water resources and the water environment and the like. Therefore, the research on how to improve the prediction accuracy of the hydrological model has important scientific significance and application value.
Any model is accompanied by errors and uncertainties, and in the model modeling work, the error sources are large, and the error sources mainly have the following aspects:
(1) errors due to excluded factors
In the modeling process, each link of the whole hydrological process of precipitation-runoff production-confluence needs to be considered in the hydrological model, each link has a plurality of influence factors, and each factor cannot be introduced into the model. The selection of these influencing factors results in a certain prediction error.
(2) Error of measured historical data
The accuracy of the measured data is determined by the advancement and maturity of the measuring technology, and the fitting degree of the model simulation is influenced, so that the prediction accuracy of the model is influenced. These data include not only traditional hydrological (flow) meteorological (rainfall) data, but also factors such as geology, vegetation, soil and land utilization.
(3) Error of parameter
The distributed hydrological model parameters have relatively definite physical significance, the variation range of the parameters is easy to estimate, but the optimal values of the parameters are difficult to determine.
(4) Structural error of model
Incorrect calculation methods adopted in the model design and establishment process, improper time step, improper operation sequence, incomplete or deviated model structure and the like can cause model prediction errors.
In order to eliminate model prediction errors caused by the reasons, parameter calibration is an important link for improving the prediction accuracy of the hydrological model, most of watershed hydrological models, particularly parameters of small and medium watersheds, cannot be determined directly through observation tests, and values of the parameters have a certain relation with underlying surface characteristics of the watersheds but cannot be established with the underlying surface characteristics of the watersheds, so the parameter calibration is still a difficult problem for the watershed hydrological model.
In the prior art, when the method is specifically applied to a watershed with data, the parameter calibration of the hydrological model generally adopts a traditional trial and error method, namely, the parameter value of the hydrological model is continuously adjusted manually to meet the requirement of simulation precision, but for the calibration of the hydrological model parameter without data, the method has the problems of low calibration accuracy, serious influence on hydrological prediction precision and the like.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method is used for solving the problems that the traditional trial-and-error method is adopted for determining the hydrological model parameters of the non-data basin in the prior art, namely, the parameter values of the hydrological model are adjusted manually and continuously to meet the requirement of simulation precision, the calibration accuracy is low, the hydrological prediction precision is seriously influenced and the like.
The technical scheme of the invention is as follows:
a data-free regional hydrological parameter calibration method based on clustering and particle swarm optimization comprises the following steps:
step 1, collecting soil texture, vegetation coverage, land utilization rate, terrain data, runoff coefficient, total annual evaporation amount, slope and gradient data;
step 2, dividing the rating area into calculation units with the length of 30 square kilometers or less;
step 3, determining the underlying surface of each parameter of each computing unit and weather related factors according to the physical characteristics of the parameters of the hydrological model;
step 4, performing cluster analysis on each underlying surface factor, and taking all units with the most similar underlying surface factor area ratio as the same type of targets as similar units;
and 5, calculating hydrological model parameters of all similar units as calibrated hydrological parameters by adopting a particle swarm optimization algorithm based on clustering results.
And 3, setting each parameter underlying surface and weather related factors as follows:
step 4, the method for carrying out cluster analysis on each underlying surface factor comprises the following steps:
a. initializing, classifying each sample into one class, and calculating the distance between every two classes, namely the similarity between the samples;
b. finding the two closest classes among the classes, and classifying the two closest classes into one class;
c. recalculating the similarity between the newly generated class and each old class;
d. and repeating b and c until all sample points are classified into one type, and ending.
Step 4, the method for carrying out cluster analysis on each underlying surface factor comprises the following steps: and (3) carrying out cluster analysis on each influence factor independently, initially setting a distance step length, dividing the influence factors into N classes according to the distance step length, classifying each hydrological model calculation unit into one class according to the factor area ratio of the hydrological model calculation unit, thus obtaining a first layer, expanding the distance step length, adopting the step length x 2, reclassifying the area ratio of the underlying surface influence factor, and repeating the steps until only one class exists at the last.
Step 5, when the particle swarm optimization algorithm is adopted for calculation, the method with the priority of the parameters comprises the following steps: defining position information of particles in a particle swarm algorithm to form a one-dimensional array for all parameters needing to be optimized, determining the classification number corresponding to each parameter to obtain a total optimized parameter set PN, wherein the particles represent a feature array with a dimension of 1 and a length of N, each parameter is configured with a minimum value Pmin, a maximum value Pmax and a step length PStep, the feature array is an integer array with the length of N, Pc is a feature parameter item in the feature array, converting the feature parameter Pc item into a corresponding hydrological parameter value PVal through a decoding operation (PVal ═ Pmin + (Pmax-Pmin) × Pc × PStep), obtaining a flow process of a target node through a hydrological model calculation after the feature array is completely converted into hydrological model parameters by adopting the method, calculating a deterministic coefficient through a flow process and an actual measurement process, and setting the calculated deterministic coefficient as the fitness of the particles, and performing iterative calculation through selection and updating operations of the algorithm until the calculation is completed.
Step 5, when the particle swarm optimization algorithm is adopted for calculation, the parameter-by-parameter optimization method comprises the following steps: and generating an optimized parameter sequence in all parameter sets needing to be optimized by adopting a random method, optimizing the parameters one by one according to the sequence, calculating by adopting a particle swarm algorithm to obtain an optimal value of the parameter at the current position in the current arrangement, fixing the parameter, performing optimized calculation on the next parameter in the arrangement, and finishing the direct arrangement calculation.
And 6, adopting a deep learning model and the lower cushion surface influence factor as input, using the hydrological model parameter as output, learning the optimal parameter optimization result of the whole watershed and the lower cushion surface influence factor, establishing a hydrological model parameter and lower cushion surface influence factor regression model, and inputting the lower cushion influence factor in a data-free area to obtain the data-free watershed hydrological model parameter value.
The invention has the beneficial effects that:
the optimization area is divided into a plurality of independent calculation units, the related influence factors of each unit influencing the hydrological parameters are subjected to cluster analysis to obtain similar units, then an intelligent optimization algorithm is adopted, automatic and intelligent means are adopted to realize parameter calibration of the regions with the data, and the method has certain universality and can be suitable for most mainstream hydrological models. Therefore, the problem of difficult use of the modern hydrological model due to strong specialization can be effectively solved, and a large number of complicated steps and work for adjusting and calibrating professional manual parameters can be reduced in practical application. The method solves the technical problems that the prior art adopts a traditional trial and error method aiming at the hydrological model parameter determination of the watershed with the information, namely, the parameter value of the hydrological model is continuously adjusted manually to meet the requirement of simulation precision, and the method has artificial subjectivity, low work repeatability, low efficiency and high complexity, does not utilize the application and popularization of the hydrological model and the like.
The invention has the advantages that:
the method fully utilizes resources of weather, natural resources, hydrology and the like, and solves the problem of parameter calibration without data by using a big data thought.
The deep learning method is adopted to learn the underlying surface and better parameter data, so that the problem of accurately fitting complex and abstract correlation relations is solved, and the parameter estimation in the data-free area is realized.
Detailed Description
A data-free regional hydrological parameter calibration method based on clustering and particle swarm optimization comprises the following steps:
step 1, collecting soil texture, vegetation coverage, land utilization rate, terrain data, runoff coefficient, total annual evaporation amount, slope and gradient data;
step 2, dividing the rating area into calculation units with the length of less than 30 square kilometers,
step 3, determining the underlying surface of each parameter of each computing unit and weather related factors according to the physical characteristics of the parameters of the hydrological model;
and 3, setting each parameter underlying surface and weather related factors as follows:
step 4, performing cluster analysis on each underlying surface factor, and taking all units with the most similar underlying surface factor area ratio as the same type of targets as similar units; the basic basis for cluster analysis is "the correlation parameters of computing units with similar underlying surface structures are substantially the same". Based on the idea, clustering analysis is carried out on each underlying surface factor of all the computing units of the drainage basin, and the goal that all units with similar underlying surface factor area ratios are in the same class is achieved.
Clustering, which is a process of clustering similar objects together, is an important method for analyzing and searching data rules of big data at present. Hierarchical clustering is a method for performing clustering layer by layer, and small clusters can be merged and aggregated from bottom to top, or large clusters can be segmented from top to bottom. The invention employs clustering from bottom to top.
Step 4, the method for carrying out cluster analysis on each underlying surface factor comprises the following steps:
a. initializing, classifying each sample into one class, and calculating the distance between every two classes, namely the similarity between the samples;
b. finding the two closest classes among the classes, and classifying the two closest classes into one class;
c. recalculating the similarity between the newly generated class and each old class;
d. and repeating b and c until all sample points are classified into one type, and ending.
Step 4, the method for carrying out cluster analysis on each underlying surface factor comprises the following steps: the hydrological model underlying surface factors are numerous and comprise vegetation, soil, terrain, cultivated land, runoff coefficients and the like, each influence factor is subjected to clustering analysis independently, a distance step is initially set, then the influence factors are divided into N types according to the distance step, each hydrological model calculating unit is classified into one type according to the factor area occupation ratio of the hydrological model calculating unit, the first layer is obtained in this way, the distance step is enlarged, the step size is adopted by 2, the area occupation ratios of the underlying surface influence factors are reclassified, and the rest is done until only one type exists at last.
Since each hydrological model parameter is associated with a plurality of underlying surface factors, each calculation unit forms a table below:
the number of the first action category in the table, the number of the second action category layer, and the grey part of the table body are the category numbers. And when the parameter optimization calculation is carried out, the parameters are considered to be the same when the layer number and the class number of each underlying surface factor in each unit are the same.
And 5, calculating hydrological model parameters of all similar units as calibrated hydrological parameters by adopting a particle swarm optimization algorithm based on clustering results.
The hydrological model parameter optimization calculation is based on a particle swarm optimization algorithm, and the Particle Swarm Optimization (PSO) belongs to one of swarm intelligence algorithms and is designed by simulating the predation behavior of a bird swarm. Assuming that there is only one food in the area (i.e., the optimal solution in the optimization problem in general), the task of the flock is to find this food source. In the whole searching process of the bird group, the other birds can know the position of the bird group through mutual transmission of respective information, whether the bird group finds the optimal solution or not is judged through the cooperation, meanwhile, the information of the optimal solution is transmitted to the whole bird group, and finally, the whole bird group can be gathered around a food source, namely the optimal solution is found, namely the problem convergence.
Although the traditional particle swarm algorithm has better optimizing capability when solving the optimization function; through iterative optimization calculation, an approximate solution can be quickly found; but the basic PSO tends to fall into local optima, resulting in large errors in the results. The optimization calculation partially improves the PSO, and mainly adds the following strategies:
a. perturbation strategy
The perturbation strategy is to add a certain speed value randomly to make the particles move randomly;
wherein k, l and m are three different particles, and C is a parameter
b. Jump strategy
The jump strategy is to randomly negate a certain position value to enable the particles to move randomly;
wherein P isi,minPi,maxIs the minimum and maximum values of the i characteristic component
c. Foreign policy
The external strategy is to add new random particles into the cluster according to a certain random probability;
d. dynamic policy
The conventional PSO velocity calculation formula is:
wherein w is an inertial weight parameter, C1、C2Is a constant parameter
The dynamic strategy improves the traditional formula into:
wherein
For the value of the current speed, it is,
is the speed value of the next time interval, T is the current iteration cycle, T is the maximum iteration cycle, w is the inertia weight parameter, C1、C2Is a constant parameter.
Parameter optimization concept
The position information of the particles in the particle swarm algorithm is defined as a one-dimensional array formed by all parameters to be optimized, for example, WM, FC, and F0 are set as optimized parameters for a hui-river basin, the number of classes corresponding to each parameter is [5,8,3], that is, a total optimized parameter set PN is obtained, i.e., { WM, FC, F0} (5+8+3) ═ 16 (one), and then the particles represent a feature array with a dimension of 1 and a length of 16. Each parameter is configured with a minimum value Pmin, a maximum value Pmax and a step size PStep, the characteristic array is an integer array with the length of 16, Pc is a certain characteristic parameter item in the characteristic array, the characteristic parameter Pc item can be converted into a corresponding EC parameter value PVal through a decoding operation (PVal ═ Pmin + (Pmax-Pmin) × Pc PStep), the characteristic array is completely converted into a hydrological model parameter by adopting the method, a flow process of a target node is obtained through a hydrological model calculation, and a determinacy coefficient obtained through the calculation is set as the fitness of the particle by carrying out the calculation of the determinacy coefficient on the flow process and an actual measurement process. And performing iterative calculation through selection and updating operations of the algorithm until the calculation is completed.
The above technical steps of the invention can find a global optimal solution, but the required time and the calculation capacity are larger, and in order to improve the adoption: parameter optimization thinking-parameter-by-parameter optimization
The optimal thought of parameters one by one is as follows: firstly, a random method is adopted to generate a sequence of optimized parameters in all parameter sets needing to be optimized, then the parameters are optimized one by one according to the sequence, each time one parameter is optimized, the parameter is fixed, and then the next parameter is continuously optimized until all the parameters are optimized. At this time, the position information of the particles in the particle swarm algorithm is defined as a certain parameter to be optimized, for example, regarding a hui-river basin, the parameters of the hydrological model participating in optimization include WM, FC, and F0, the number of classes corresponding to each parameter is [5,8,3], that is, the total optimized parameter set PN { WM, FC, F0} - (5+8+3) ═ 16 (one) is obtained, a parameter sequence PS is generated by a random value, PS is a random arrangement of 16 parameters, and the arrangement of the 16 parameters is sequentially calculated, wherein the particles represent an array with a dimension of 1 and a length of 1. And calculating by adopting a particle swarm algorithm to obtain the optimal value of the parameter of the current position in the current arrangement, then fixing the parameter, performing optimization calculation on the next parameter in the arrangement, and directly finishing the arrangement calculation.
The method has the advantages that the calculation workload can be controlled on the balance time and precision requirements, and a better feasible solution can be found in a shorter time.
The method comprises the steps of adopting a deep learning model and a lower cushion surface influence factor as input, using hydrological model parameters as output, learning parameter optimization results and lower cushion surface factors of the whole basin, establishing a hydrological model parameter and lower cushion surface influence regression model, and calculating corresponding basin hydrological parameters by inputting the lower cushion surface factor in a non-material area.
Deep learning is an algorithm that attempts to perform high-level abstraction on data using multiple processing layers containing complex structures or consisting of multiple nonlinear transformations, and is a method that is improved on the basis of artificial neural networks, and is therefore also referred to as "deep neural networks". It simulates the machine learning mode of human brain's nerve structure, making the computer possess similar thinking ability to human.
In the invention, a better parameter is calculated through a watershed with data, and a corresponding underlying surface influence factor is extracted; and then, taking the lower cushion surface influence factor as input and the unit better parameter as output, training a neural network, and applying the trained neural network to a data-free area to accurately obtain the data-free earth hydrological model parameters.
Claims (6)
1. A data-free regional hydrological parameter calibration method based on clustering and particle swarm optimization comprises the following steps:
step 1, collecting soil texture, vegetation coverage, land utilization rate, terrain data, runoff coefficient, total annual evaporation amount, slope and gradient data;
step 2, dividing the rating area into calculation units with the length of 30 square kilometers or less;
step 3, determining the underlying surface of each parameter of each computing unit and weather related factors according to the physical characteristics of the parameters of the hydrological model;
step 4, performing cluster analysis on each underlying surface factor, and taking all units with the most similar underlying surface factor area ratio as the same type of targets as similar units;
step 5, calculating hydrological model parameters of all similar units by adopting a particle swarm optimization algorithm based on clustering results to obtain optimal parameters of similar watersheds;
and 6, adopting a deep learning model and the lower cushion surface influence factor as input, using the hydrological model parameter as output, learning the optimal parameter optimization result of the whole watershed and the lower cushion surface influence factor, establishing a hydrological model parameter and lower cushion surface influence factor regression model, and inputting the lower cushion influence factor in a data-free area to obtain the data-free watershed hydrological model parameter value.
2. The method for calibrating hydrological parameters in data-free areas based on clustering and particle swarm optimization according to claim 1, wherein the method comprises the following steps: and 3, setting each parameter underlying surface and weather related factors as follows:
3. the method for calibrating hydrological parameters in data-free areas based on clustering and particle swarm optimization according to claim 1, wherein the method comprises the following steps: step 4, the method for carrying out cluster analysis on each underlying surface factor comprises the following steps:
a. initializing, classifying each sample into one class, and calculating the distance between every two classes, namely the similarity between the samples;
b. finding the two closest classes among the classes, and classifying the two closest classes into one class;
c. recalculating the similarity between the newly generated class and each old class;
d. and repeating b and c until all sample points are classified into one type, and ending.
4. The method for calibrating hydrological parameters in data-free areas based on clustering and particle swarm optimization according to claim 1, wherein the method comprises the following steps: step 4, the method for carrying out cluster analysis on each underlying surface factor comprises the following steps: and (3) carrying out cluster analysis on each influence factor independently, initially setting a distance step length, dividing the influence factors into N classes according to the distance step length, classifying each hydrological model calculation unit into one class according to the factor area ratio of the hydrological model calculation unit, thus obtaining a first layer, expanding the distance step length, adopting the step length x 2, reclassifying the area ratio of the underlying surface influence factor, and repeating the steps until only one class exists at the last.
5. The method for calibrating hydrological parameters in data-free areas based on clustering and particle swarm optimization according to claim 1, wherein the method comprises the following steps: step 5, when the particle swarm optimization algorithm is adopted for calculation, the method with the priority of the parameters comprises the following steps: defining position information of particles in a particle swarm algorithm to form a one-dimensional array for all parameters needing to be optimized, determining the classification number corresponding to each parameter to obtain a total optimized parameter set PN, wherein the particles represent a feature array with a dimension of 1 and a length of N, each parameter is configured with a minimum value Pmin, a maximum value Pmax and a step length PStep, the feature array is an integer array with the length of N, Pc is a feature parameter item in the feature array, converting the feature parameter Pc item into a corresponding hydrological parameter value PVal through a decoding operation (PVal ═ Pmin + (Pmax-Pmin) × Pc × PStep), obtaining a flow process of a target node through a hydrological model calculation after the feature array is completely converted into hydrological model parameters by adopting the method, calculating a deterministic coefficient through a flow process and an actual measurement process, and setting the calculated deterministic coefficient as the fitness of the particles, and performing iterative calculation through selection and updating operations of the algorithm until the calculation is completed.
6. The method for calibrating hydrological parameters in data-free areas based on clustering and particle swarm optimization according to claim 1, wherein the method comprises the following steps: step 5, when the particle swarm optimization algorithm is adopted for calculation, the parameter-by-parameter optimization method comprises the following steps: and generating an optimized parameter sequence in all parameter sets needing to be optimized by adopting a random method, optimizing the parameters one by one according to the sequence, calculating by adopting a particle swarm algorithm to obtain an optimal value of the parameter at the current position in the current arrangement, fixing the parameter, performing optimized calculation on the next parameter in the arrangement, and finishing the direct arrangement calculation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010819100.2A CN111914465B (en) | 2020-08-14 | 2020-08-14 | Clustering and particle swarm optimization-based method for calibrating hydrologic parameters of non-data region |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010819100.2A CN111914465B (en) | 2020-08-14 | 2020-08-14 | Clustering and particle swarm optimization-based method for calibrating hydrologic parameters of non-data region |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111914465A true CN111914465A (en) | 2020-11-10 |
| CN111914465B CN111914465B (en) | 2024-04-12 |
Family
ID=73283144
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010819100.2A Active CN111914465B (en) | 2020-08-14 | 2020-08-14 | Clustering and particle swarm optimization-based method for calibrating hydrologic parameters of non-data region |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111914465B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114264356A (en) * | 2021-12-23 | 2022-04-01 | 东方世纪科技股份有限公司 | Dynamic vehicle weighing system and method based on edge calculation |
| CN117763970A (en) * | 2023-12-26 | 2024-03-26 | 水利部交通运输部国家能源局南京水利科学研究院 | Image deep learning-based hydrologic model parameter reconstruction method for scarce data area |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130039191A (en) * | 2011-10-11 | 2013-04-19 | 한국수자원공사 | Grid based long term rainfall runoff model for large scale watersheds |
| CN106682355A (en) * | 2017-01-12 | 2017-05-17 | 中国水利水电科学研究院 | Hydrological model parameter calibration method based on PSO (particle swarm optimization)-GA (genetic algorithm) mixed algorithm |
| KR20170136094A (en) * | 2016-05-31 | 2017-12-11 | 홍익대학교 산학협력단 | System for generating stochastic rainfall of poisson cluster based on optimized parameter map, and method for the same |
| CN108133292A (en) * | 2017-12-25 | 2018-06-08 | 贵州东方世纪科技股份有限公司 | A kind of water and soil balance computational methods based on artificial intelligence |
| CN108647425A (en) * | 2018-05-07 | 2018-10-12 | 中国水利水电科学研究院 | K-means high flow or low flow time forecasting procedures based on particle group optimizing |
| CN109753680A (en) * | 2018-11-20 | 2019-05-14 | 南京南瑞集团公司 | A kind of swarm of particles intelligent method based on chaos masking mechanism |
| CN109815305A (en) * | 2019-01-18 | 2019-05-28 | 三峡大学 | A kind of method of Cross Some Region Without Data play flood runoff process inverting |
| CN109840873A (en) * | 2019-02-02 | 2019-06-04 | 中国水利水电科学研究院 | A kind of Cross Some Region Without Data Hydro-Model Parameter Calibration Technology fields method based on machine learning |
-
2020
- 2020-08-14 CN CN202010819100.2A patent/CN111914465B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130039191A (en) * | 2011-10-11 | 2013-04-19 | 한국수자원공사 | Grid based long term rainfall runoff model for large scale watersheds |
| KR20170136094A (en) * | 2016-05-31 | 2017-12-11 | 홍익대학교 산학협력단 | System for generating stochastic rainfall of poisson cluster based on optimized parameter map, and method for the same |
| CN106682355A (en) * | 2017-01-12 | 2017-05-17 | 中国水利水电科学研究院 | Hydrological model parameter calibration method based on PSO (particle swarm optimization)-GA (genetic algorithm) mixed algorithm |
| CN108133292A (en) * | 2017-12-25 | 2018-06-08 | 贵州东方世纪科技股份有限公司 | A kind of water and soil balance computational methods based on artificial intelligence |
| CN108647425A (en) * | 2018-05-07 | 2018-10-12 | 中国水利水电科学研究院 | K-means high flow or low flow time forecasting procedures based on particle group optimizing |
| CN109753680A (en) * | 2018-11-20 | 2019-05-14 | 南京南瑞集团公司 | A kind of swarm of particles intelligent method based on chaos masking mechanism |
| CN109815305A (en) * | 2019-01-18 | 2019-05-28 | 三峡大学 | A kind of method of Cross Some Region Without Data play flood runoff process inverting |
| CN109840873A (en) * | 2019-02-02 | 2019-06-04 | 中国水利水电科学研究院 | A kind of Cross Some Region Without Data Hydro-Model Parameter Calibration Technology fields method based on machine learning |
Non-Patent Citations (3)
| Title |
|---|
| 尹鑫卫;李晓玲;王琦;张永梅;: "垄沟集雨系统Laio土壤水分动态随机模型参数敏感性分析及优化", 中国生态农业学报, vol. 26, no. 05, pages 746 - 758 * |
| 强玲娟: "基于机器学习的混合智能算法求解参数优化问题", 中国优秀硕士学位论文全文数据库信息科技专辑, pages 1 - 50 * |
| 郭旖琪;马立山;王利民;杨国丽;张海平;: "SWMM模型参数敏感度分析应用研究", 河北建筑工程学院学报, vol. 38, no. 01, pages 134 - 137 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114264356A (en) * | 2021-12-23 | 2022-04-01 | 东方世纪科技股份有限公司 | Dynamic vehicle weighing system and method based on edge calculation |
| CN114264356B (en) * | 2021-12-23 | 2024-04-26 | 东方世纪科技股份有限公司 | Dynamic vehicle weighing system and method based on edge calculation |
| CN117763970A (en) * | 2023-12-26 | 2024-03-26 | 水利部交通运输部国家能源局南京水利科学研究院 | Image deep learning-based hydrologic model parameter reconstruction method for scarce data area |
| CN117763970B (en) * | 2023-12-26 | 2024-08-02 | 水利部交通运输部国家能源局南京水利科学研究院 | Image deep learning-based hydrologic model parameter reconstruction method for scarce data area |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111914465B (en) | 2024-04-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109858647B (en) | Regional flood disaster risk evaluation and estimation method coupled with GIS and GBDT algorithm | |
| CN110619432B (en) | Feature extraction hydrological forecasting method based on deep learning | |
| CN104408518B (en) | Based on the neural network learning optimization method of particle swarm optimization algorithm | |
| CN109214579B (en) | BP neural network-based saline-alkali soil stability prediction method and system | |
| CN111914487B (en) | Data-free regional hydrological parameter calibration method based on antagonistic neural network | |
| CN116227692B (en) | Crop heavy metal enrichment risk quantification method, system and storable medium | |
| CN111914465B (en) | Clustering and particle swarm optimization-based method for calibrating hydrologic parameters of non-data region | |
| CN109858180B (en) | Regional crop water consumption space pattern optimization design method | |
| CN109754122A (en) | A kind of Numerical Predicting Method of the BP neural network based on random forest feature extraction | |
| CN112712169A (en) | Model building method and application of full residual depth network based on graph convolution | |
| CN108733952B (en) | Three-dimensional characterization method for spatial variability of soil water content based on sequential simulation | |
| CN113837352A (en) | Rainfall-runoff space-time relation simulation method based on long-short term memory neural network | |
| CN110874454A (en) | Method for accurately measuring and calculating regional scale moso bamboo carbon reserves based on mixed probability density | |
| CN111914488B (en) | Data area hydrologic parameter calibration method based on antagonistic neural network | |
| CN116796799A (en) | Method for creating small-river basin flood rainfall threshold model in area without hydrologic data | |
| CN117370766A (en) | Satellite mission planning scheme evaluation method based on deep learning | |
| Antonopoulos et al. | Evaluation of different methods to estimate monthly reference evapotranspiration in a Mediterranean area | |
| CN111914430B (en) | Clustering and particle swarm optimization-based hydrologic parameter calibration method for data-bearing region | |
| CN107688702B (en) | Lane colony algorithm-based river channel flood flow evolution law simulation method | |
| CN117494042A (en) | Soil-water fusion method for data-missing area based on climate similarity migration | |
| CN117010274A (en) | Intelligent early warning method for harmful elements in underground water based on integrated incremental learning | |
| CN111126827A (en) | Input-output accounting model construction method based on BP artificial neural network | |
| Zhao et al. | Using artificial neural networks to produce high-resolution soil property maps | |
| Yang et al. | An improved probabilistic neural network with ga optimization | |
| CN117807510B (en) | River network dynamic partitioning method and system combining multi-source data processing |
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 |