CN112734091A - Reservoir water level model prediction method - Google Patents
Reservoir water level model prediction method Download PDFInfo
- Publication number
- CN112734091A CN112734091A CN202011611776.9A CN202011611776A CN112734091A CN 112734091 A CN112734091 A CN 112734091A CN 202011611776 A CN202011611776 A CN 202011611776A CN 112734091 A CN112734091 A CN 112734091A
- Authority
- CN
- China
- Prior art keywords
- reservoir
- water level
- prediction method
- model prediction
- amount
- 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.)
- Pending
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 104
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000013528 artificial neural network Methods 0.000 claims abstract description 5
- 238000013135 deep learning Methods 0.000 claims abstract description 4
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 12
- 238000012417 linear regression Methods 0.000 claims description 11
- 230000008020 evaporation Effects 0.000 claims description 10
- 238000001704 evaporation Methods 0.000 claims description 10
- 239000008235 industrial water Substances 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 9
- 230000000694 effects Effects 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 7
- 238000003066 decision tree Methods 0.000 claims description 6
- 238000007637 random forest analysis Methods 0.000 claims description 5
- 238000013138 pruning Methods 0.000 claims description 3
- 230000001373 regressive effect Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000012544 monitoring process Methods 0.000 abstract description 8
- 238000010801 machine learning Methods 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 abstract description 2
- 238000012360 testing method Methods 0.000 description 6
- 238000002474 experimental method Methods 0.000 description 4
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 239000011295 pitch Substances 0.000 description 2
- 239000011435 rock Substances 0.000 description 2
- 235000010918 Cudrania tricuspidata Nutrition 0.000 description 1
- 241000218211 Maclura Species 0.000 description 1
- 241001523380 Maclura tricuspidata Species 0.000 description 1
- 206010033799 Paralysis Diseases 0.000 description 1
- 235000006040 Prunus persica var persica Nutrition 0.000 description 1
- 240000006413 Prunus persica var. persica Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- 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
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Water Supply & Treatment (AREA)
- Life Sciences & Earth Sciences (AREA)
- Public Health (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Primary Health Care (AREA)
- Algebra (AREA)
- Medical Informatics (AREA)
- Probability & Statistics with Applications (AREA)
- Mathematical Analysis (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
Abstract
The invention relates to the technical field of reservoir monitoring, in particular to a reservoir water level model prediction method, which comprises the following steps: step 1: collecting information data of a selected reservoir; step 2: selecting a moment t, and regressing according to the acquired information quantity to obtain a plurality of water levels of different positions at the moment t; and step 3: taking an average value of a plurality of water levels at the time t; and 4, step 4: predicting the water level amount after a certain time through the obtained water level average value and the obtained variation; the invention uses the practical technology of the front edge such as machine learning, compressed sensing, neural network, deep learning and the like, utilizes the meteorological and hydrological big data to predict the reservoir water level in real time, applies the theoretical method of big data science to the hydrological field, ensures the real-time prediction of the reservoir water level, and is convenient for precaution in advance.
Description
Technical Field
The invention relates to the technical field of reservoir monitoring, in particular to a reservoir water level model prediction method.
Background
At present, the reservoir water level is monitored mainly in China by adopting a mode of jointly monitoring sensing equipment and manually, so that the real-time control on the reservoir water level information is ensured. The real-time monitoring method can effectively monitor the reservoir, but has the defects that the time lag is mainly reflected, the two methods of equipment monitoring and artificial monitoring are both passive monitoring and can be observed only after the water level is changed, and the future water level change is difficult to be scientifically judged. In fact, when an emergency occurs, the water level changes very rapidly, and the failure to predict in advance will result in only a rush to deal with the emergency. For example, although the weather department has forecasted in advance that continuous heavy rainfall will occur in the areas along the middle and lower reaches of the Yangtze river, the areas of Jianghuai, the southwest and eastern China and the like after 7 months and 3 days in this year, by 7 months and 6 days, 190 reservoirs in the Wuhan city still exceed the flood limit water level to cause flood overflow, the whole city is seriously waterlogged, and basic paralysis such as traffic, power supply and the like causes very important negative effects and very huge economic losses. In the flood fighting and disaster relief process, reservoir scheduling can play a crucial role. For example, day 4 in 7 months, the flood-blocking and peak-clipping effects of the Kunzhuan river reservoir in the south lake, the Yuansheng five-strong river reservoir in the downstream of the Yuanjiang river, and the Jiangxi Xishui Zhenling reservoir are fully exerted, and the flood peak is respectively reduced by 15400 cubic meters per second, 11600 cubic meters per second and 5240 cubic meters per second. If no cudrania lineata reservoir is used for retaining flood, the Yiyang urban area and the Taojiang county city suffer from top-dead disasters; the Wuqiang creek reservoir reduces the Yuansheng downstream peach source, and the Changde station flood peak water level is about 4-5 m; the cudrania tricuspidata reservoir lowers the downstream water level by more than 4 meters. It can be seen that the reservoir plays a crucial role in the flood fighting process. Real-time monitoring, even prediction and early warning of the reservoir water level are necessary guarantees that the reservoir plays a role, and therefore a reservoir water level model prediction method is provided.
Disclosure of Invention
An object of the present invention is to solve the above-mentioned drawbacks of the background art by providing a reservoir water level model prediction method.
The technical scheme adopted by the invention is as follows: the method comprises the following steps:
step 1: collecting information data of a selected reservoir;
step 2: selecting a moment t, and regressing according to the acquired information quantity to obtain a plurality of water levels of different positions at the moment t;
and step 3: taking an average value of a plurality of water levels at the time t;
and 4, step 4: and predicting the water level amount after a certain time through the obtained water level average value and the change amount.
As a preferred technical scheme of the invention: and the information data of the reservoir in the step 1 are the increase amount and the decrease amount of the reservoir.
As a preferred technical scheme of the invention: the increase of the reservoir includes rainfall and inflow.
As a preferred technical scheme of the invention: the reduced amount of the reservoir comprises evaporation amount, industrial water supply amount, domestic water supply amount, agricultural water supply amount and reservoir outflow amount.
As a preferred technical scheme of the invention: the calculation formula of the reservoir water level at the time t is as follows:
RWLt＝RWLt-1+ΔtRF+ΔtRI-ΔtEVA
-ΔtIWS-ΔtDWS-ΔtAWS-ΔtRO，
wherein RW Lt, RW Lt-1Respectively represents the reservoir water level at t and t-1 times, deltatRF、ΔtRI、ΔtEV A，ΔtIW S、ΔtDW S、ΔtAW S、ΔtRO represents rainfall, reservoir inflow, evaporation, industrial water, domestic water, agricultural water and reservoir outflow in the time from t-1 to t and delta t, respectively.
As a preferred technical scheme of the invention: the variation in step 4 includes rainfall, temperature, humidity, season, and open gate data.
As a preferred technical scheme of the invention: the formula of the predicted water level is as follows:
RW Lt～RW Lt-1+ΔtRF+ΔtTEM+ΔtHUM+ΔtSEA，
wherein RW LtIndicating the reservoir level at time t.
As a preferred technical scheme of the invention: the information quantity regression can adopt any one of multiple linear regression, random forest, Gaussian process, compressed sensing, neural network and deep learning algorithm.
As a preferred technical scheme of the invention: the algorithm of the multiple linear regression algorithm comprises the following steps:
giving a random sample (Y)i，Xi1，…，Xip) I 1, …, n, a linear regression model assuming the regression sub-YiAnd the regressive quantity Xi1The relationship between is that in addition to the effect of X, there are other variables present; adding an error term epsilon in the form of a random variablei(ii) a To capture except Xi1，…，XipAny effect on Y; the expression form is as follows:
Yi＝β0+β1Xi1+β2Xi2+...+βpXip+εi，i＝1，...，n。
as a preferred technical scheme of the invention: the random forest algorithm comprises the following steps:
representing the number of training cases by N, and representing the number of features by M; inputting a characteristic number M for determining a decision result of a node on a decision tree, wherein M is far less than M; sampling N times from N training cases in a mode of sampling with a back sampling to form a training set, and using the non-sampled cases as predictions to evaluate errors of the training sets; for each node, randomly selecting m characteristics, and determining the decision of each node on the decision tree based on the characteristics; calculating the optimal splitting mode according to the m characteristics; each tree grows completely and can not be pruned, and pruning is carried out after a normal tree classifier is built.
The invention uses the practical technology of the front edge such as machine learning, compressed sensing, neural network, deep learning and the like, utilizes the meteorological and hydrological big data to predict the reservoir water level in real time, applies the theoretical method of big data science to the hydrological field, ensures the real-time prediction of the reservoir water level, and is convenient for precaution in advance.
Detailed Description
It should be noted that, in the present application, features in embodiments and embodiments may be combined with each other without conflict, and technical solutions in the embodiments of the present invention will be clearly and completely described below, 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.
Example 1:
the preferred embodiment of the invention provides a reservoir water level model prediction method, which comprises the following steps:
step 1: collecting information data of a selected reservoir;
step 2: selecting a moment t, and regressing according to the acquired information quantity to obtain a plurality of water levels of different positions at the moment t;
and step 3: taking an average value of a plurality of water levels at the time t;
and 4, step 4: and predicting the water level amount after a certain time through the obtained water level average value and the change amount.
The rise and fall of the reservoir water level is the result of the increase and decrease of water in the reservoir. Wherein the reduction of the amount of water in the reservoir comprises the amount of evaporation, the amount of industrial water supply, the amount of domestic water supply, the amount of agricultural water supply, the amount of outflow from the reservoir, etc. The amount of water in the reservoir is increased by rainfall or inflow. Ideally, if we can measure all these quantities, then predicting the reservoir level (RWL) at time t can be based on:
RWLt＝RWLt-1+ΔtRF+ΔtRI-ΔtEVA
-ΔtIWS-ΔtDWS-ΔtAWS-ΔtRO
to calculate, wherein RW Lt, RW Lt-1Respectively represents the reservoir water level at t and t-1 times, deltatRF、ΔtRI、ΔtEV A，ΔtIW S、ΔtDW S、ΔtAW S、ΔtRO represents rainfall, reservoir inflow, evaporation, industrial water, domestic water, agricultural water and reservoir outflow in the time from t-1 to t and delta t, respectively.
However, the data of the inflow volume, the evaporation volume, the industrial water, the domestic water, the agricultural water and the outflow volume of the reservoir are difficult to measure, and the data can only be actually measured at the time t within the delta t time, namely the data cannot be obtained in advance. Since the task of predicting the reservoir water level often requires a certain time as an advance, such as performing early warning, evacuating people, and the like, the t-time data of the quantities measured by various sensors cannot be used in the actual task of predicting the water level.
Due to evaporation capacity, industrial water, domestic water and agricultural water, the water is greatly influenced by meteorological factors (such as temperature, humidity, season and the like). For example, in summer or at high temperature, the evaporation capacity is large, and the domestic water and the agricultural water are both large; and in winter or cold, the domestic water is less. These meteorological factors are relatively easy to obtain, and the two factors of the inflow and outflow of the reservoir are relatively fixed without considering the water level opening gate in a unit time. Therefore, we can consider using meteorological data to predict water levels. Specifically, the rainfall, temperature, humidity, season can be used to predict the change of the water level, and the regression formula is as follows:
RW Lt～RW Lt-1+ΔtRF+ΔtTEM+ΔtHUM+ΔtSEA，
wherein RW LtReservoir indicating time tThe water level can be obtained by regression of factors such as the water level of the reservoir, rainfall, temperature, humidity, season and the like at the time t-1. In view of the above analysis, factors affecting the reservoir water level are rainfall, inflow, evaporation, industrial water supply, domestic water supply, agricultural water supply, reservoir outflow, etc., but most of them are practically unavailable or cannot be used for predictive tasks. Considering that these factors are closely related to weather data, scheduling operations such as opening a gate and the like, and the data such as weather and opening gate operations and the like are easily obtained, we can consider using the data such as weather (rainfall, temperature, humidity, season) and opening the gate to predict the change of the reservoir water level.
In this embodiment: taking a multiple linear regression algorithm as an example; the method comprises the following steps:
giving a random sample (Y)i，Xi1，…，Xip) I 1, …, n, a linear regression model assuming the regression sub-YiAnd the regressive quantity Xi1The relationship between is that in addition to the effect of X, there are other variables present; adding an error term epsilon in the form of a random variablei(ii) a To capture except Xi1，…，XipAny effect on Y; the expression form is as follows:
Yi＝β0+β1Xi1+β2Xi2+...+βpXip+εi，i＝1....，n。
example 2:
the difference from example 1 is that: adopting a random forest algorithm;
the method comprises the following steps:
representing the number of training cases by N, and representing the number of features by M; inputting a characteristic number M for determining a decision result of a node on a decision tree, wherein M is far less than M; sampling N times from N training cases in a mode of sampling with a back sampling to form a training set, and using the non-sampled cases as predictions to evaluate errors of the training sets; for each node, randomly selecting m characteristics, and determining the decision of each node on the decision tree based on the characteristics; calculating the optimal splitting mode according to the m characteristics; each tree grows completely and can not be pruned, and pruning is carried out after a normal tree classifier is built.
The following example is demonstrated by an experiment:
the water level data and rainfall data of the water reservoirs from 2014 to 2016 are preliminarily predicted according to seven pitches, rock rocks, iron hills and long-flow wavers in Shenzhen city. Since other meteorological data and mutual geographic information of reservoirs are not available, the early result is a preliminary prediction made by taking seven-pitch reservoirs which are relatively independent geographically as an example, and the result is believed to be greatly improved after a meteorological data improved model is used. Specifically, in the previous experiment, the rainfall data is used to predict the change of the reservoir water level, and the regression model is as follows:
RWLt～RWLt-1+Δt-1RF
RWLtis the mean water level of the t period, RW Lt-1Is the mean water level, Δ, of the t-1 time periodt-1Is the amount of rainfall for the t-1 period. The average water level after one hour, one day, and three days in the future was predicted in this experiment. Because a plurality of water levels can be obtained in the obtained data in the same time period, the average value of the data is obtained in the experiment, and the adjustment can be carried out according to the requirements at the later stage.
The rainfall amount of the previous hour and the water level of the previous hour may be used first to predict the water level of the next hour. Since the problem tends to be a linear model, earlier for simplicity, it may be considered to use multiple linear regression. The Adjusted R Square (Adjusted R-Square, also called goodness-of-fit, maximum of 1) value is 0.999991669, which means that our model fits data very well. Regression coefficients and associated significance tests as shown in table 1, it can be seen that both coefficient and constant terms are significant. In addition, the model is well explanatory, RW LtRW L approximately equal to one timet-1Plus 2 times the rainfall deltat-1Wherein the unit of rainfall is millimeter and the unit of water level is meter, the difference is 1000 times.
TABLE 1 one hour prediction model coefficients and test
The rainfall amount of the previous day, the rainfall amount of the next day and the water level of the previous day are used to predict the water level of the next day. Likewise, we use multiple linear regression. The Adjusted R Square (Adjusted R-Square, also called goodness-of-fit, maximum of 1) value is 0.999235, which means that this model fits data perfectly. Regression coefficients and associated significance tests as shown in table 2, it can be seen that all three coefficient and constant terms are significant.
TABLE 2 one-day prediction model coefficients and test
The same multivariate linear regression was used to predict the water level of the last three days using the rainfall and the water level of the first three days. The Adjusted R Square (Adjusted R-Square, also called goodness-of-fit, with a maximum of 1) is 0.996154475, which means that our model is ideal for comparison of the fit to the data. The regression coefficients and associated significance tests are shown in table 3, and it can be seen that the coefficient terms are significant.
| CoefficientS | Standard error of | t Stat | P-value | |
| Intercept | 0.09645 | 0.095302 | 1.012045 | 0.311931 |
| Water level at time t-3 | 0.996627 | 0.002689 | 370.6631 | 0 |
| Rainfall at time t-3 | 0.0046 | 0.000299 | 15.38885 | 3.47E-45 |
| Rainfall at time t-2 | 0.005617 | 0.000296 | 18.9623 | 5.44E-63 |
| Rainfall at time t-1 | 0.004741 | 0.000296 | 16.01507 | 3.28E-48 |
| Rainfall at time t | 0.001373 | 0.000293 | 4.68913 | 3.41E-06 |
TABLE 3 three day prediction model coefficients and test
In other embodiments, the regression algorithm may also be other algorithms, such as a gaussian algorithm:
the Gaussian process is a machine learning method developed based on a statistical learning theory and a Bayes theory, is suitable for processing complicated regression problems such as high dimensionality, small samples and nonlinearity, and has strong generalization capability. Compared with a neural network and a support vector machine, the Gaussian process has the advantages of easiness in implementation, super-parameter self-adaptive acquisition, flexibility in non-parameter inference, probability significance in output and the like.
The gaussian process can be used for both classification and regression. Because the reservoir water level prediction problem is a regression problem, the application of the Gaussian process to the regression problem is mainly introduced in the invention, and the basic form is as follows:
yi＝f(xi)+∈i，i＝1，2，...，n，
wherein c represents independent and identically distributed Gaussian noise, and the hidden variable f assumes a Gaussian process prior. In practice, since the data is finite, f is often a multi-dimensional Gaussian distribution that can be expressed as
Wherein mugIs the mean value. For the sake of simplifying the expression, it is often assumed that μg0. K, represents the covariance matrix, equation (K)g)i，j＝cov(f(xi)，f(xj))＝k(xi，xj(ii) a θ) represents a kernel function. Due to epsiloniIs a Gaussian distribution which is independently and identically distributed, and therefore has
Wherein
Represents the variance of the noise and I represents the unit matrix. The following equation can be obtained by simple derivation:
therein are provided with
In practical applications, in order to learn the hyper-parameter σgAnd θ, generally using a maximum marginal likelihood function, minimizes the negative log marginal likelihood, i.e., equivalent to minimizing the following equation:
after learning the hyper-parameter sigmagAnd θ can be followed by a prediction. For arbitrary x*∈RDThe predicted distribution is as follows
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
Claims (10)
1. A reservoir water level model prediction method is characterized in that: the method comprises the following steps:
step 1: collecting information data of a selected reservoir;
step 2: selecting a moment t, and regressing according to the acquired information quantity to obtain a plurality of water levels of different positions at the moment t;
and step 3: taking an average value of a plurality of water levels at the time t;
and 4, step 4: and predicting the water level amount after a certain time through the obtained water level average value and the change amount.
2. The reservoir water level model prediction method of claim 1, characterized by: and the information data of the reservoir in the step 1 are the increase amount and the decrease amount of the reservoir.
3. The reservoir water level model prediction method of claim 2, characterized in that: the increase of the reservoir includes rainfall and inflow.
4. The reservoir water level model prediction method of claim 2, characterized in that: the reduced amount of the reservoir comprises evaporation amount, industrial water supply amount, domestic water supply amount, agricultural water supply amount and reservoir outflow amount.
5. The reservoir water level model prediction method of claim 1, characterized by: the calculation formula of the reservoir water level at the time t is as follows:
RWLt＝RWLt-1+ΔtRF+ΔtRI-ΔtEVA-ΔtIWS-ΔtDWS-ΔtAWS-ΔtRO，
wherein RW Lt, RW Lt-1Respectively represents the reservoir water level at t and t-1 times, deltatRF、ΔtRI、ΔtEV A，ΔtIW S、ΔtDW S、ΔtAW S、ΔtRO represents rainfall, reservoir inflow, evaporation, industrial water, domestic water, agricultural water and reservoir outflow in the time from t-1 to t and delta t, respectively.
6. The reservoir water level model prediction method of claim 1, characterized by: the variation in step 4 includes rainfall, temperature, humidity, season, and open gate data.
7. The reservoir water level model prediction method of claim 6, characterized by: the formula of the predicted water level is as follows:
RWLt～RWLt-1+ΔtRF+ΔtTEM+ΔtHUM+ΔtSEA，
wherein RWLtIndicating the reservoir level at time t.
8. The reservoir water level model prediction method of claim 6, characterized by: the information quantity regression can adopt any one of multiple linear regression, random forest, Gaussian process, compressed sensing, neural network and deep learning algorithm.
9. The reservoir water level model prediction method of claim 8, characterized by: the algorithm of the multiple linear regression algorithm comprises the following steps:
giving a random sample (Y)i，Xi1，…，Xip) I 1, …, n, a linear regression model assuming the regression sub-YiAnd the regressive quantity Xi1The relationship between is that in addition to the effect of X, there are other variables present; adding an error term epsilon in the form of a random variablei(ii) a To capture except Xi1，…，XipAny effect on Y; the expression form is as follows:
Yi＝β0+β1Xi1+β2Xi2+...+βpXip+εi，i＝1，...，n。
10. the reservoir water level model prediction method of claim 8, characterized by: the random forest algorithm comprises the following steps:
representing the number of training cases by N, and representing the number of features by M; inputting a characteristic number M for determining a decision result of a node on a decision tree, wherein M is far less than M; sampling N times from N training cases in a mode of sampling with a back sampling to form a training set, and using the non-sampled cases as predictions to evaluate errors of the training sets; for each node, randomly selecting m characteristics, and determining the decision of each node on the decision tree based on the characteristics; calculating the optimal splitting mode according to the m characteristics; each tree grows completely and can not be pruned, and pruning is carried out after a normal tree classifier is built.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011611776.9A CN112734091A (en) | 2020-12-30 | 2020-12-30 | Reservoir water level model prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011611776.9A CN112734091A (en) | 2020-12-30 | 2020-12-30 | Reservoir water level model prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112734091A true CN112734091A (en) | 2021-04-30 |
Family
ID=75610310
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011611776.9A Pending CN112734091A (en) | 2020-12-30 | 2020-12-30 | Reservoir water level model prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112734091A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113538858A (en) * | 2021-07-15 | 2021-10-22 | 华能澜沧江水电股份有限公司 | Hydropower station drainage area flood prevention early warning device and early warning method |
| CN114021836A (en) * | 2021-11-16 | 2022-02-08 | 电子科技大学 | Multivariable reservoir water inflow amount prediction system based on different-angle fusion, training method and application |
| CN114254833A (en) * | 2021-12-25 | 2022-03-29 | 福建中锐网络股份有限公司 | Reservoir water level prediction and scheduling method based on multiple linear regression and meteorological data |
| CN116341769A (en) * | 2023-05-30 | 2023-06-27 | 山东科技大学 | Neural network residual water level forecasting method based on Bayesian super-parameter optimization |
| CN116882215A (en) * | 2023-09-07 | 2023-10-13 | 北京国信华源科技有限公司 | Multi-element self-excitation early warning method |
| CN117689501A (en) * | 2024-02-03 | 2024-03-12 | 深圳市广汇源环境水务有限公司 | Reservoir flood prevention dynamic forecasting and early warning method and system platform |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105843942A (en) * | 2016-04-01 | 2016-08-10 | 浙江大学城市学院 | Urban flood prevention decision support system based on big data technique |
| CN109978235A (en) * | 2019-03-04 | 2019-07-05 | 宁波市气象服务中心 | A kind of flooded water level prediction method of product based on sample learning |
| CN110991687A (en) * | 2019-09-26 | 2020-04-10 | 深圳市东深电子股份有限公司 | Water resource scheduling optimization method based on empirical model |
| CN111753461A (en) * | 2020-05-12 | 2020-10-09 | 中山大学 | Tidal water level correction method, target residual water level acquisition method, device and equipment |
| CN111811575A (en) * | 2020-07-17 | 2020-10-23 | 河南省南阳水文水资源勘测局 | Hydrology equipment remote monitoring system |
-
2020
- 2020-12-30 CN CN202011611776.9A patent/CN112734091A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105843942A (en) * | 2016-04-01 | 2016-08-10 | 浙江大学城市学院 | Urban flood prevention decision support system based on big data technique |
| CN109978235A (en) * | 2019-03-04 | 2019-07-05 | 宁波市气象服务中心 | A kind of flooded water level prediction method of product based on sample learning |
| CN110991687A (en) * | 2019-09-26 | 2020-04-10 | 深圳市东深电子股份有限公司 | Water resource scheduling optimization method based on empirical model |
| CN111753461A (en) * | 2020-05-12 | 2020-10-09 | 中山大学 | Tidal water level correction method, target residual water level acquisition method, device and equipment |
| CN111811575A (en) * | 2020-07-17 | 2020-10-23 | 河南省南阳水文水资源勘测局 | Hydrology equipment remote monitoring system |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113538858A (en) * | 2021-07-15 | 2021-10-22 | 华能澜沧江水电股份有限公司 | Hydropower station drainage area flood prevention early warning device and early warning method |
| CN114021836A (en) * | 2021-11-16 | 2022-02-08 | 电子科技大学 | Multivariable reservoir water inflow amount prediction system based on different-angle fusion, training method and application |
| CN114021836B (en) * | 2021-11-16 | 2023-05-16 | 电子科技大学 | Multi-variable reservoir water inflow prediction system based on different angle fusion, training method and application |
| CN114254833A (en) * | 2021-12-25 | 2022-03-29 | 福建中锐网络股份有限公司 | Reservoir water level prediction and scheduling method based on multiple linear regression and meteorological data |
| CN116341769A (en) * | 2023-05-30 | 2023-06-27 | 山东科技大学 | Neural network residual water level forecasting method based on Bayesian super-parameter optimization |
| CN116882215A (en) * | 2023-09-07 | 2023-10-13 | 北京国信华源科技有限公司 | Multi-element self-excitation early warning method |
| CN116882215B (en) * | 2023-09-07 | 2023-12-29 | 北京国信华源科技有限公司 | Multi-element self-excitation early warning method |
| CN117689501A (en) * | 2024-02-03 | 2024-03-12 | 深圳市广汇源环境水务有限公司 | Reservoir flood prevention dynamic forecasting and early warning method and system platform |
| CN117689501B (en) * | 2024-02-03 | 2024-05-14 | 深圳市广汇源环境水务有限公司 | Reservoir flood prevention dynamic forecasting and early warning method and system platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112734091A (en) | Reservoir water level model prediction method | |
| Fan et al. | Comparison of Support Vector Machine and Extreme Gradient Boosting for predicting daily global solar radiation using temperature and precipitation in humid subtropical climates: A case study in China | |
| Fan et al. | Empirical and machine learning models for predicting daily global solar radiation from sunshine duration: A review and case study in China | |
| Dong et al. | Enhancing short-term forecasting of daily precipitation using numerical weather prediction bias correcting with XGBoost in different regions of China | |
| Schoof et al. | Downscaling temperature and precipitation: A comparison of regression‐based methods and artificial neural networks | |
| Robertson et al. | A Bayesian approach to predictor selection for seasonal streamflow forecasting | |
| Wang et al. | Spatio-temporal changes of precipitation and temperature over the Pearl River basin based on CMIP5 multi-model ensemble | |
| Huang et al. | Fuzzy neural network and LLE algorithm for forecasting precipitation in tropical cyclones: comparisons with interpolation method by ECMWF and stepwise regression method | |
| CN114707688A (en) | Photovoltaic power ultra-short-term prediction method based on satellite cloud chart and space-time neural network | |
| Coutinho et al. | Application of artificial neural networks (ANNs) in the gap filling of meteorological time series | |
| Guo et al. | Prediction of daily crop reference evapotranspiration (ET0) values through a least-squares support vector machine model | |
| Ellenburg et al. | The role of evapotranspiration in streamflow modeling–An analysis using entropy | |
| Rust et al. | Mapping weather-type influence on Senegal precipitation based on a spatial–temporal statistical model | |
| Lund et al. | Temperature trends in the United States | |
| Ghamariadyan et al. | Monthly rainfall forecasting using temperature and climate indices through a hybrid method in Queensland, Australia | |
| Chen et al. | Sub-daily soil moisture estimate using dynamic Bayesian model averaging | |
| Shiri | Prediction vs. estimation of dewpoint temperature: assessing GEP, MARS and RF models | |
| Shakeri et al. | Projection of important climate variables in large cities under the CMIP5–RCP scenarios using SDSM and fuzzy downscaling models | |
| Pei et al. | Analysis of spring drought in Northeast China from the perspective of atmosphere, snow cover, and soil | |
| Wu et al. | A novel bayesian additive regression trees ensemble model based on linear regression and nonlinear regression for torrential rain forecasting | |
| Bazrafshan et al. | Trivariate risk analysis of meteorological drought in Iran under climate change scenarios | |
| Khan et al. | Evaluation of statistical downscaling models using pattern and dependence structure in the monsoon‐dominated region of Pakistan | |
| Ender et al. | Extreme value modeling of precipitation in case studies for China | |
| Nyongesa et al. | Non-homogeneous hidden Markov model for downscaling of short rains occurrence in Kenya | |
| Joshi et al. | Comparison of direct statistical and indirect statistical-deterministic frameworks in downscaling river low-flow indices |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210430 |
|
| RJ01 | Rejection of invention patent application after publication |