CN112926786B - Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation - Google Patents

Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation Download PDF

Info

Publication number
CN112926786B
CN112926786B CN202110262237.7A CN202110262237A CN112926786B CN 112926786 B CN112926786 B CN 112926786B CN 202110262237 A CN202110262237 A CN 202110262237A CN 112926786 B CN112926786 B CN 112926786B
Authority
CN
China
Prior art keywords
water level
lake
association rule
stage
level
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.)
Active
Application number
CN202110262237.7A
Other languages
Chinese (zh)
Other versions
CN112926786A (en
Inventor
何建兵
蔡梅
龚李莉
王元元
潘明祥
石亚东
陆志华
刘增贤
钱旭
韦婷婷
李勇涛
李敏
李蓓
向美焘
白君瑞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Water Resources Development Research Center Of Taihu Basin Authority
Original Assignee
Water Resources Development Research Center Of Taihu Basin Authority
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Water Resources Development Research Center Of Taihu Basin Authority filed Critical Water Resources Development Research Center Of Taihu Basin Authority
Priority to CN202110262237.7A priority Critical patent/CN112926786B/en
Publication of CN112926786A publication Critical patent/CN112926786A/en
Application granted granted Critical
Publication of CN112926786B publication Critical patent/CN112926786B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G06N5/022—Knowledge engineering; Knowledge acquisition
    • G06N5/025—Extracting rules from data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling

Abstract

The invention discloses a shallow lake target water level reverse prediction method based on an association rule model and numerical simulation, which comprises the following steps of: (1) determining a reverse prediction stage and water level target; (2) mining the lake water level response rule of the later stage of the pre-precipitation water level by adopting a method combining multi-dimensional scene construction and a digital-analog test; (3) determining a water level safety threshold; (4) constructing an association rule model, and mining association rules at the early stage of the pre-reduction water level; (5) and (5) identifying a water conservancy project regulation and control mode in the pre-reduction period in the future scene according to the association rule obtained by mining in the step (4) and the lake water level response rule in the post-reduction water level period obtained by mining in the step (2). The invention integrates the advantages of the association rule model and the numerical simulation technology, overcomes the core difficulty that the traditional technical means depends on the rainfall forecast data in the medium-long term water level simulation forecast, and provides a technical means with practical operability for the shallow lake water level regulation and control based on a specific target.

Description

Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation
Technical Field
The invention relates to a shallow lake target water level reverse prediction method and system based on an association rule model and numerical simulation, and relates to the field of water conservancy technology and data mining.
Background
The large shallow lake is a regulation and storage center for flood water and water resources in river network areas, and the lake water level is an important indicator of the drought and waterlogging situation of a drainage basin. The method scientifically predicts the lake level to guide the joint scheduling of the complex river network hydraulic engineering group so as to regulate and control the lake level, and has very important practical significance for guaranteeing the water safety of plain river network areas, particularly the flood control safety. In recent years, with frequent extreme weather and climate events such as climate change, heavy rainfall and the like, deep study of lake water levels and accumulation of flood control scheduling practices, determination of lake water level targets before flood and medium-long term prediction (more than 15 days, depending on more accurate rainfall forecast period, the same applies hereinafter) have become necessary requirements for river basin flood control scheduling.
At present, reservoir is mainly used as a research object in water level regulation and control research related to basin flood management, the reservoir water level is mainly determined by water from a few upstream warehousing rivers and the amount of discharged water at a dam site, and the reservoir water level regulation and control target is realized by regulating the amount of discharged water through reservoir control engineering. The numerical simulation technology is one of the commonly used technologies in lake level prediction and regulation research, but the difficulty of long-term water level prediction in shallow lakes is that accurate rainfall data of a long time scale in the future cannot be obtained in advance, so that the water level simulation prediction cannot be directly performed by using the traditional numerical simulation technology, and the actual water inflow and outflow quantity is related to the branch inflow and outflow quantity and the regulation and control capacity of the lake-in and outflow control engineering, so that the shallow lake level prediction and regulation and control mechanism in the plain river network area is more complex and variable compared with a reservoir, and is the core difficulty of the long-term prediction in the shallow lake level and the key technology which needs to be broken through urgently.
Although attempts of pre-flood water level reduction are carried out on a few shallow lakes in plain areas in recent years, the traditional prediction and regulation methods are adopted, and the requirements on whether the water level of the pre-flood water level reduction lake can be achieved to a specific target or not are greatly uncertain mainly depending on empirical judgment and multiple trial and error.
Disclosure of Invention
The purpose of the invention is as follows: the invention provides a shallow lake target water level reverse prediction method and system based on an association rule model and numerical simulation aiming at the core difficulty of a shallow lake water level prediction technology.
The technical scheme is as follows: in a first aspect, a shallow lake target water level reverse prediction method based on an association rule model and numerical simulation is provided, and the method comprises the following steps:
(1) determining a reverse prediction stage and a water level target;
(2) mining the lake water level response rule under the multiple driving factors at the later stage of the pre-precipitation water level by adopting a method combining multi-dimensional scene construction and a digital-analog test;
(3) determining a water level safety threshold;
(4) constructing an association rule model, and excavating an association rule of 'basin rainfall-lake water level' in the early stage of the pre-rainfall water level;
(5) and determining a water conservancy project regulation and control mode in the pre-reduction period in the future scene according to the association rule obtained by mining in the step (4) and the lake water level response rule in the post-reduction water level period obtained by mining in the step (2).
In some implementations of the first aspect, step (1) further includes:
(1-1) dividing a reverse prediction stage: dividing the prediction time interval into a pre-reduction water level early stage and a pre-reduction water level later stage, wherein the pre-reduction water level early stage comprises a pre-reduction early stage starting time T0Pre-decreasing the early-stage termination time Ta(ii) a The later stage of the pre-reducing water level comprises the termination time T of the early stage of the pre-reducingaConsistent pre-reduction late start time, and end time Tb；
(1-2) determining a water level target: the pre-reduction early-stage starting time T0The ending time of the early stage of pre-reduction or the starting time T of the late stage of pre-reductionaEnd time TbCorresponding water levels are respectively Z0、Za、Zb；ZaThe threshold and lake level targets are respectively recorded as thre _ Za、obje_Zb。
In some implementations of the first aspect, step (2) further includes:
(2-1) constructing a post-stage multi-dimensional scene set S of the pre-reduction water level: choosing obje _ ZbThe key driving factor X is screened out by hydrologic, meteorological and manual intervention1,X2,…Xi,…XnN is a natural number, the sequence comprising Za(ii) a Based on the key driving factor, a multi-dimensional scene set S is constructed, specifically as follows:
Figure BDA0002970503390000021
construction of the incremental sequence Δ x for each drive factorij；
Xij＝Xi1+Δxij i＝1,2,…,n；j＝1,2,..,mn
Defining a matrix containing a sequence of all drive factors as a scene, thereby generating M multi-dimensional scenes:
Figure BDA0002970503390000022
(2-2) establishing a river network hydrodynamic model, wherein the basic principle of the hydrodynamic model is as follows:
equation of continuity
Figure BDA0002970503390000031
Equation of force
Figure BDA0002970503390000032
In the formula: x is the distance, m; t is time, s; a is the water passing area, m2(ii) a Q is a cross-sectional flow, m3S; z is water level, m; alpha is a momentum correction coefficient; r is hydraulic radius, m; q. q.sLFor side inflow, m2S, inflow is positive and outflow is negative; v. ofxThe velocity of the inflow in the direction of the water flow, m/s, v if the side inflow is perpendicular to the main flowx＝0；
(2-3) excavating a lake water level response rule under multiple driving factors at the later stage of the pre-precipitation water level by adopting a digital-analog test method: inputting the driving factor sequence under each scene into a hydrodynamic model, and obtaining Z under each scene through model numerical simulationbSequence, digging and analyzing lake level ZbResponse law Z for each driving factorb＝f(X1,X2,…,Xn)。
In some realizations of the first aspect, step (3) is performed while satisfying Zbk≤obje_ZbIn the scenario of (1), take ZakMean value as ZaIs safe threshold thre _ Za：
Figure BDA0002970503390000033
Wherein M' is Zbk≤obje_ZbThe number of scenes.
In some implementations of the first aspect, step (4) further includes:
(4-1), discretization rainfall and water level data: according to Z0Historical data and rainfall data P of early rainfall basin of rainfall forecast levelaDistribution characteristics, procedure(3) Determined thre _ ZaIs a reaction of Z0、PaAnd ZaCarrying out discretization treatment to form a plurality of groups;
(4-2) constructing a 'river basin rainfall-lake water level' association rule model: x, Y is a non-empty set of items, which is in the form of
Figure BDA0002970503390000034
(wherein X ∈ I,
Figure BDA0002970503390000035
and is
Figure BDA0002970503390000036
) The logical implication relationship of (a) is called an association rule, (b) X is called a front-part or a prerequisite of the association rule, and (c) Y is called a back-part or a result of the association rule;
digging a 'basin rainfall-lake water level' association rule in the early period of the pre-rainfall water level by constructing an association model; according to the association rule obtained by mining, enabling Z to be realizedaControl at thre _ ZaThe early-stage hydraulic engineering regulation and control suggestion of the pre-precipitation water level is provided; the association rule model relates to support degree and confidence degree as follows:
Figure BDA0002970503390000037
Figure BDA0002970503390000038
in the formula: d is a transaction set;
Figure BDA0002970503390000041
the support degree of the association rule;
Figure BDA0002970503390000042
is the confidence of the association rule.
In some realizations of the first aspect, step (5) is further:
(5-1) based on the association rule obtained by mining in the step (4), according to the future situation, the lake level z'0And watershed rainfall data P'aPredicting future scene z'aAnd through z'aAnd thre _ ZaComparing, recognizing and determining T0～TaRegulation and control mode of hydraulic engineering, if z'a≤thre_ZaRegular scheduling is implemented in hydraulic engineering, otherwise, the lake level is pre-reduced in advance by hydraulic engineering scheduling to enable z'a≤thre_Za：
Z′aa＝min{Z′a,thre_Za}
In formula (II), Z'aaFor T in future scenariosaActual water level corresponding to the moment;
(5-2) according to ZbIdentifying and determining T according to response rule of each key driving factora～TbAnd (3) an optimal hydraulic engineering regulation and control mode, bringing the actual scene into the scene set in the step (2), and continuously optimizing the steps through information feedback and training.
In a second aspect, a shallow lake target water level reverse prediction system based on an association rule model and numerical simulation is provided, and the system includes: a first module (front module) for determining a reverse prediction staging, water level target; a second module (digital-analog analysis module) for excavating lake water level response rules under multiple driving factors at the later stage of the pre-precipitation water level; a third module for determining a water level safety threshold (threshold analysis module); a fourth module (association rule mining module) for mining the association rule of 'basin rainfall-lake water level' in the early stage of the rainfall precipitation level; and a fifth module (decision support module) for determining a water and utility engineering regulation and control mode in the future scene in the pre-reduction period according to the association rule obtained by mining of the fourth module and the lake level response rule in the post-reduction stage of the pre-reduction water level obtained by mining of the second module, as shown in fig. 2.
The first module (pre-module) is further configured to partition the backward prediction partition: dividing the prediction time interval into a pre-reduction water level early stage and a pre-reduction water level later stage, wherein the pre-reduction water level early stage comprises a pre-reduction early stage starting time T0Pre-decreasing the early-stage termination time Ta(ii) a The later stage of the pre-reducing water level comprises the termination time T of the early stage of the pre-reducingaConsistent pre-reduction late start time, and end time Tb；
Determining a water level target: the pre-reduction early-stage starting time T0The ending time of the early stage of pre-reduction or the starting time T of the late stage of pre-reductionaEnd time TbCorresponding water levels are respectively Z0、Za、Zb；ZaThe threshold and lake level targets are respectively recorded as thre _ Za、obje_Zb。
The second module (digital-to-analog analysis module) is further configured to construct a predicted level late multi-dimensional scene set S: choosing obje _ ZbThe key driving factor X is screened out by hydrologic, meteorological and manual intervention1,X2,…Xi,…XnN is a natural number, the sequence comprising Za(ii) a Based on the key driving factor, a multi-dimensional scene set S is constructed, specifically as follows:
Figure BDA0002970503390000051
construction of the incremental sequence Δ x for each drive factorij；
Xij＝Xi1+Δxij i＝1,2,…,n；j＝1,2,..,mn
Defining a matrix containing a sequence of all drive factors as a scene, thereby generating M multi-dimensional scenes:
Figure BDA0002970503390000052
establishing a river network hydrodynamic model, wherein the basic principle of the hydrodynamic model is as follows:
equation of continuity
Figure BDA0002970503390000053
Equation of force
Figure BDA0002970503390000054
In the formula: x is the distance, m; t is time, s; a is the water passing area, m2(ii) a Q is a cross-sectional flow, m3S; z is water level, m; alpha is a momentum correction coefficient; r is hydraulic radius, m; q. q.sLFor side inflow, m2S, inflow is positive and outflow is negative; v. ofxThe velocity of the inflow in the direction of the water flow, m/s, v if the side inflow is perpendicular to the main flowx＝0；
Adopting a digital-analog test method to excavate the lake water level response rule under the multiple driving factors at the later stage of the pre-precipitation water level: inputting the driving factor sequence under each scene into a hydrodynamic model, and obtaining Z under each scene through model numerical simulationbSequence, digging and analyzing lake level ZbResponse law Z for each driving factorb＝f(X1,X2,…,Xn)。
The third module (threshold analysis module) further satisfies Zbk≤obje_ZbIn the scenario of (1), take ZakMean value as ZaIs safe threshold thre _ Za：
Figure BDA0002970503390000055
Wherein M' is Zbk≤obje_ZbThe number of scenes.
The fourth module (association rule mining module) is further used for discretizing rainfall and water level data: according to Z0Historical data and rainfall data P of early rainfall basin of rainfall forecast levelaDistribution characteristics, thre _ Z determined by the third moduleaIs a reaction of Z0、PaAnd ZaCarrying out discretization treatment to form a plurality of groups;
constructing a 'drainage basin rainfall-lake water level' association rule model: x, Y is a non-empty set of items, which is in the form of
Figure BDA0002970503390000061
(wherein X ∈ I,
Figure BDA0002970503390000062
and is
Figure BDA0002970503390000063
) The logical implication relationship of (a) is called an association rule, (b) X is called a front-part or a prerequisite of the association rule, and (c) Y is called a back-part or a result of the association rule;
digging a 'basin rainfall-lake water level' association rule in the early period of the pre-rainfall water level by constructing an association model; according to the association rule obtained by mining, enabling Z to be realizedaControl at thre _ ZaThe early-stage hydraulic engineering regulation and control suggestion of the pre-precipitation water level is provided; the association rule model relates to support degree and confidence degree as follows:
Figure BDA0002970503390000064
Figure BDA0002970503390000065
in the formula: d is a transaction set;
Figure BDA0002970503390000066
the support degree of the association rule;
Figure BDA0002970503390000067
is the confidence of the association rule.
The fifth module (decision support module) is further used for mining the association rule obtained by the fourth module according to the lake level z 'under the future situation'0And watershed rainfall data P'aPredicting future scene z'aAnd through z'aAnd thre _ ZaComparing, recognizing and determining T0～TaRegulation and control mode of hydraulic engineering, if z'a≤thre_ZaIf not, the water conservancy project implements conventional scheduling, otherwise, the water is passedPre-reducing the lake water level in advance by engineering scheduling to ensure that z'a≤thre_Za：
Z′aa＝min{Z′a,thre_Za}
In formula (II), Z'aaFor T in future scenariosaActual water level corresponding to the moment;
according to ZbIdentifying and determining T according to response rule of each key driving factora～TbAnd the optimal hydraulic engineering regulation and control mode is realized, the actual scene is brought into the scene set of the second module, and the steps are continuously optimized through information feedback and training.
In a third aspect, an apparatus for inversely predicting a target water level of a shallow lake based on an association rule model and numerical simulation is provided, the apparatus comprising: a processor, and a memory storing computer program instructions; the processor, when reading and executing the computer program instructions, implements the method of the first aspect or some realizations of the first aspect.
In a fourth aspect, there is provided a computer storage medium having computer program instructions stored thereon that, when executed by a processor, implement the method of the first aspect or some of the realizations of the first aspect.
Has the advantages that: the invention provides a shallow lake target water level reverse prediction method and system based on an association rule model and numerical simulation, which integrate the advantages of the association rule model and the numerical simulation technology, and when the association rule model is used for water level prediction, the prediction of lake water level change on medium and long term scales is realized mainly based on historical data and independent of rainfall data required by water level prediction; the numerical simulation technology is mainly used for performing simulation prediction on the lake level process on a short time scale (depending on a more accurate rainfall forecast period, such as 14 days and 7 days). The invention overcomes the core difficulty that the traditional technical means depends on the rainfall forecast data in the long-time scale water level simulation forecast, and provides a technical means with practical operability for shallow lake water level prediction and regulation based on a specific target.
Drawings
FIG. 1 is a technical flow diagram of the present invention.
Fig. 2 is a system architecture diagram of the present invention.
Fig. 3 is a schematic topology structure diagram of a river network and a hydraulic structure according to an embodiment of the present invention.
Fig. 4 is a schematic diagram of a hydrodynamic model of a river network according to an embodiment of the present invention.
Fig. 5 is a water level process of a third lake tai according to an embodiment of the present invention.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the present invention. It will be apparent, however, to one skilled in the art, that the present invention may be practiced without one or more of these specific details. In other instances, well-known features have not been described in order to avoid obscuring the invention.
The applicant believes that although attempts of pre-flood water level reduction are made in a few shallow lakes in plain areas in recent years, the traditional prediction and regulation methods are adopted, and the requirements on whether the water level of the pre-flood water level reduction lake can be achieved to a specific target or not are uncertain mainly depending on empirical judgment and multiple trial and error.
Therefore, for the defects of the prior art, it is necessary to provide a method for reversely predicting and regulating the water level of a shallow lake for a specific water level target under the condition that the rainfall data forecasted for a medium-long term cannot be accurately obtained.
The first embodiment is as follows:
in this embodiment, taking the lake tai as an example, the method and the system for reversely predicting the target water level of the shallow lake based on the association rule model and the numerical simulation provided by the present invention are applied to mine the lake water level response rule with multiple driving factors and determine the water level safety threshold, and the method and the system are specifically characterized in that:
step 1, determining a reverse prediction stage and a water level target:
(1-1) inverse prediction staging: according to the rainfall of the Taihu basin in 1988-2017 and the historical data of the Taihu water level, the average water of the Taihu lake for many yearsThe number of the plants shows a more obvious ascending trend from the beginning of the year to the beginning of 4 months. The 1 st to 15 th days from 4 th month are the flood front flood season of the drainage basin, so the 1 st to 1 st months from 1 st month are taken as the research period of the reverse prediction of the water level of the Taihu lake, and the current rainfall forecast period limit is considered, and the research period is divided into a pre-precipitation water level front period and a pre-precipitation water level rear period, wherein the pre-precipitation water level front period is 1 st to 1 st 15 days from 1 st month, and the pre-precipitation water level rear period is 3 th to 16 th to 4 th 1 day. The water levels of the Taihu lake in 1 month and 1 day, the Taihu lake in 3 months and 15 days and the Taihu lake in 4 months and 1 day are respectively recorded as Z1/1、Z3/16、Z4/1。
(1-2) water level target determination: the target of pre-reducing the water level of the Taihu lake in 1 day in 4 months is recorded as obje _ Z4/1And the safe threshold of the water level of the Taihu lake in 3 and 15 months is recorded as thre _ Z3/16. The current water level of the Taihu lake for flood control in 4 months and 1 day is 3.10m, and in the embodiment, the water level of the Taihu lake for 4 months and 1 day is controlled to be not more than 3.10m as the target of pre-reducing the water level, i.e., obje _ Z4/1＝3.1。
Step 2, digging a lake water level response rule under multiple driving factors at the later stage of the pre-precipitation water level by adopting a method combining multi-dimensional scene construction and a digital-analog test:
(2-1) constructing a post-stage multi-dimensional scene set S of the pre-reduction water level: choosing obje _ Z4/1The key driving factors are screened out, including hydrology, meteorology, manual intervention and the like, to obtain the water level Z of the Taihu lake in 3 months and 16 days3/16And the expected rainfall PP at the later stage of pre-reductionbAnd a hydraulic engineering regulation and control mode set U. Based on the key driving factors, a multi-dimensional scene set S is constructed as follows:
Figure BDA0002970503390000081
Za＝{Za1,Za2,…,Zan}
PPb＝{PPb1,PPb2,…,PPbn}
U＝{U1,U2,…,Un}
in the formula: zaiThe water level is the possible initial water level in the later stage of the pre-reduction water level and is in the unit of m; PP (polypropylene)biForecasting the rainfall of the drainage basin in unit mm, wherein the rainfall is different in the later stage of the rainfall level; u shapeiThe method is different in the later stage of the pre-precipitation water level.
When the hydraulic engineering regulation and control mode set U is constructed, the method mainly aims at two most main lake-entering and lake-exiting river channels of the Taihu lake, namely the Hopoppy river engineering and the Taipu river engineering. For different Taihu lake bottom water and expected rainfall, the pre-descending difficulty is considered comprehensively, and on the basis of conventional scheduling of the Homing Yu river engineering and Taipu river engineering, a regulation and control strategy set of the Homing Yu river and Taipu river engineering based on the purpose of pre-descending the Taihu lake water level in advance is constructed to form a hydraulic engineering regulation and control mode set, as shown in Table 1. It should be noted that, when constructing the set U, the actual possibility is considered, and for example, the combinations such as "proper drainage" of the hope-Yu-river project, and "regular scheduling" of the Taipu-river project are less likely to actually occur, and therefore, such combinations are not considered.
TABLE 1 Hydraulic engineering regulation and control mode set based on the purpose of pre-lowering Taihu lake water level in advance
Figure BDA0002970503390000082
Figure BDA0002970503390000091
(2-2) establishing a river network hydrodynamic model based on a water yield and water quality digital model platform of the lake Taihu basin, which comprises the following steps:
generalization of a production and convergence module: the generalized hydrological calculation unit mainly relates to a rainfall station weight scheme, evaporation station information, underlying surface information, a convergence time interval distribution coefficient and other parameters related to production convergence calculation;
generalization of river network: introducing a riverway center line shapefile, extracting a riverway center line, deriving a model, and checking a river network topological relation;
and (3) section generalization: standardizing actually measured or generalized river channel section data to form a dat file in a standard format, enabling river channel large section data to correspond to river channel positions in a river channel center line shapefile one by one, introducing the standardized section data file, and generating a one-dimensional river channel;
and (3) lake generalization: in this embodiment, lakes are generalized to zero-dimensional regulation nodes, collected lake shape files and lake bottom elevation data are arranged into standard dat-format general terrain files, and a model is introduced, wherein the lakes have a regulation function, so that the relationship between water level (or depth) -area (or volume) of each lake needs to be set, and the zero-dimensional lakes are connected with a river network through the nodes or weir gates;
generalization of hydraulic buildings and scheduling conditions: the hydraulic structure (including culvert, gate, pump station, etc.) is generalized into the relation key element in the model, its flow rate of overflowing adopts hydrodynamic method to simulate, set up the node upstream and downstream of the building, connect with generalized river course through the node (as shown in fig. 2), water level and flow rate between the nodes depend on weir flow formula and hydraulic structure operation mode, the hydraulic structure operation mode is generalized into a series of logical statements;
and (3) generalization of boundary conditions: the boundary condition is mainly the external tide level boundary, and the characteristic tide level data of the actually measured tide level station is interpolated into the integral tide level and further interpolated into the external boundary gates corresponding to all the generalized riverways.
The hydrodynamic model is shown schematically in FIG. 4.
(2-3) carrying out a Taihu lake water level regulation digital-analog test under the multi-dimensional scenes in the later stage of the pre-precipitation water level, inputting the driving factor sequence under each scene into a hydrodynamic model, and obtaining Z under each scene through model numerical simulation4/1And (3) excavating a lake water level response rule under multiple driving factors at the later stage of the pre-precipitation water level, wherein the model test results under partial scenes are shown in a table 2.
TABLE 2 Taihu lake Water level response results in partial scenarios (not all listed)
Figure BDA0002970503390000092
Figure BDA0002970503390000101
Step 3, determining a water level safety threshold value: one of the key problems in the invention is to reasonably determine the water level threshold thre _ Z of the Taihu lake of 3 months and 15 days3/16. At the point of satisfying Z4/1，kIn the scene less than or equal to 3.1, take Z4/1，kMean value as safety threshold thre _ Z3/16. According to Z in step 24/1Response law, thre _ Z3/16＝3.1。
Example two:
in this embodiment, on the basis of the first embodiment, a shallow lake target water level inverse prediction method and system based on an association rule model and numerical simulation provided by the present invention are further applied to mine a "basin rainfall-lake water level" association rule in a pre-precipitation water level period, and with reference to the accompanying drawings, the specific steps are as follows:
(4-1) discretization of rainfall and water level data: when the association rule model is constructed, firstly, the historical water level and rainfall data are subjected to discretization processing. Will Z1/1Discretization to "Z1/1≤3.0m”、“3.0＜Z1/1≤3.2m”、“Z1/1>3.2 m' three classes, Z3/16Discretization to "Z3/16＜3.1m”、“Z3/16More than or equal to 3.1 m. According to the long series rainfall data of Taihu basin, PaThe lower quartile (25%), the median (50%), and the upper quartile (75%) were 145mm, 171mm, and 216mm, respectively, and thus P was assignedaDiscretized into three classes, respectively "Pa≤145mm”、“145mm＜PaLess than or equal to 216mm and Pa>216mm”。
(4-2) constructing a 'basin rainfall-lake water level' association rule model: in this embodiment, the association rule model adopts a GRI algorithm, sets the minimum support degree to 10%, sets the minimum confidence degree to 50%, and obtains 7 association rules by mining, as shown in table 3.
TABLE 3Z1/1-Pa-Z3/16Association rule extraction and specification
Figure BDA0002970503390000102
Figure BDA0002970503390000111
According to the association rule of Table 1, for Z3/16Less than or equal to 3.1, and adopting a corresponding hydraulic engineering regulation and control mode for main hydraulic engineering of the Taihu lake, as shown in Table 4.
TABLE 4 forecast water level early stage ring Taihu lake main hydraulic engineering scheduling suggestion
Figure BDA0002970503390000112
Example three:
example three:
in this embodiment, based on the second embodiment, the shallow lake target water level inverse prediction method and system based on the association rule model and the numerical simulation provided by the present invention are further applied to predict and regulate the middle and long term lake water levels, and with reference to the accompanying drawings, the specific steps are as follows:
based on the association rule obtained by mining in the second step (4) of the embodiment, according to the future situation, the lake level Z'1/1And watershed rainfall data P'aPredicting future scene Z'3/16And through Z'3/16Comparing with 3.1m, identifying and determining a hydraulic engineering regulation and control mode of 1-3-16 months from 1 month, if Z'3/16And (3) performing conventional scheduling by hydraulic engineering, otherwise, pre-reducing the lake level in advance by the hydraulic engineering scheduling to enable the lake level in Taihu lake in 16 days in 3 months to be below 3.1 m. Z in this example'1/1＝3.17m，P′aIn anticipation of exceeding 180mm, it is recommended in Table 4 according to example II that a stepwise pre-descent schedule is implemented for the major entrances and exits of the Taihu lake from 1 st to 3 st 15 th, and it is verified that Z is performed in this manner3/16＝3.1m。
Further according to Z4/1And identifying and determining the optimal hydraulic engineering regulation and control mode in 16-4 months and 1 day 3 months. In this example, Z3/16The rainfall is forecast to be 52.2mm in 3.1m and 16-31 months, and according to the table 2 in the first embodiment, the Taihu lake is mainly accessed from the exit of the lake in 16-4 months and 1 day 3 and 3 months4(u4) Performing scheduling, verified in this way Z4/13.09m (as shown in fig. 5), the target water level requirement is met.
The above specific implementation manner and embodiments are specific support for the technical idea of the shallow lake target water level inverse prediction method based on the association rule model and the numerical simulation, and the protection scope of the present invention cannot be limited thereby, and any equivalent changes or equivalent changes made on the basis of the technical scheme according to the technical idea of the present invention still belong to the protection scope of the technical scheme of the present invention.

Claims (3)

1. A shallow lake target water level reverse prediction method based on an association rule model and numerical simulation is characterized by comprising the following steps:
(1) determining a reverse prediction stage and a water level target, wherein the reverse prediction stage is divided into a pre-precipitation water level early stage and a pre-precipitation water level later stage;
(1-1) dividing a reverse prediction stage: the early stage of the pre-reducing water level comprises the initial time of the early stage of the pre-reducing water levelT 0 Early-stage termination time of pre-precipitation water levelT a (ii) a The later stage of the pre-reducing water level comprises the termination time of the early stage of the pre-reducing water levelT a Consistent pre-precipitation level late start time, and end timeT b ；
(1-2) determining a water level target: preliminary water level early-stage starting timeT 0 Early-stage termination time of pre-precipitation water levelT a Late end time of pre-lowering water levelT b Corresponding lake levels are respectivelyZ 0 、Z a 、Z b (ii) a Lake water levelZ a The safety threshold value and the lake water level target value are respectively recorded asthre_Z a 、obje_Z b ；
(2) Mining the lake water level response rule under the multiple driving factors at the later stage of the pre-precipitation water level by adopting a method combining multi-dimensional scene construction and a digital-analog test;
(2-1) constructing a pre-reduction water level later-stage multi-dimensional scene setS: selecting a lake water level target valueobje_Z b The key driving factors are screened out by hydrologic, meteorological and manual interventionX 1 ,X 2 ,…X i ,…X n ，nThe sequence comprising the end time of the pre-precipitation period being a natural numberT a Corresponding lake water levelZ a (ii) a Constructing a multi-dimensional scene set based on key driving factorsSThe method comprises the following steps:
Figure 461965DEST_PATH_IMAGE002
construction of an incremental sequence for each drive factor
Figure DEST_PATH_IMAGE003
；
Figure DEST_PATH_IMAGE005
Defining a matrix comprising a sequence of all drive factors as a scene, thereby generatingMA plurality of multi-dimensional scenarios:
Figure DEST_PATH_IMAGE007
(2-2) establishing a river network hydrodynamic model, wherein the basic principle of the hydrodynamic model is as follows:
equation of continuity
Figure DEST_PATH_IMAGE009
Equation of force
Figure DEST_PATH_IMAGE011
In the formula: x is the distance, m; t is time, s; a is the water passing area, m2(ii) a Q is a cross-sectional flow, m3S; z is water level, m; alpha is a momentum correction coefficient; r is hydraulic radius, m; q. q.sLFor side inflow, m2S, inflow is positive and outflow is negative; v. ofxThe velocity of the inflow in the direction of the water flow, m/s, v if the side inflow is perpendicular to the main flowx=0；
The process of establishing the river network hydrodynamic model comprises the following steps:
generalization of a production and convergence module: the generalized hydrological calculation unit relates to a rainfall station weight scheme, evaporation station information, underlying surface information, a confluence time period distribution coefficient and parameters related to production confluence calculation;
generalization of river network: introducing a riverway center line shapefile, extracting the riverway center line, deriving a model, and checking the river
A network topology relationship;
and (3) section generalization: standardizing actually measured or generalized river channel section data to form a dat file in a standard format, and obtaining the river channel
The large section data correspond to the river channel positions in a river channel center line shapefile one by one, and a standardized section data file is introduced to generate a one-dimensional river channel;
and (3) lake generalization: generalizing lakes into zero-dimensional regulation nodes, arranging collected lake shape files and lake bottom elevation data into standard dat format general terrain files, introducing a model, setting a relation between water level and area of each lake due to regulation of the lakes, and establishing a connection between the zero-dimensional lakes and a river network through the nodes or weir gates;
generalization of hydraulic buildings and scheduling conditions: the hydraulic structure is generalized to be a link element in the model, the flow of the hydraulic structure is simulated by a hydrodynamic method, nodes are arranged at the upstream and the downstream of the structure and are connected with a generalized river channel through the nodes, the water level and the flow among the nodes depend on a weir flow formula and the operation mode of the hydraulic structure, and the operation mode of the hydraulic structure is generalized to be a series of logic statements;
and (3) generalization of boundary conditions: the boundary condition is mainly external tide level boundary, and the characteristic tide level data of the actually measured tide level station is interpolated
The tide level is an integral point, and the tide level is further interpolated to outer boundary gates corresponding to all generalized riverways;
(2-3) excavating a lake water level response rule under multiple driving factors at the later stage of the pre-precipitation water level by adopting a digital-analog test method: inputting the driving factor sequence under each scene into a hydrodynamic model, and obtaining the lake water level under each scene through model numerical simulationZ b Sequence, digging and analyzing lake water levelZ b Response law to each driving factor
Figure 76748DEST_PATH_IMAGE012
；
(3) Determining a water level safety threshold value when meeting
Figure DEST_PATH_IMAGE013
In the scenario of (1), getZ ak Mean value asZ a Safety threshold ofthre_Z a ：
Figure DEST_PATH_IMAGE015
In the formula (I), the compound is shown in the specification,
Figure 107765DEST_PATH_IMAGE016
to satisfy
Figure 258123DEST_PATH_IMAGE013
The number of scenes;
(4) constructing an association rule model, and excavating an association rule of 'basin rainfall-lake water level' in the early stage of the pre-rainfall water level;
(4-1), discretization rainfall and water level data: according to the early-stage starting time T of the pre-precipitation water level0Corresponding water levelZ 0 Historical data and rainfall data of early rainfall basin of rainfall forecast levelP a Distribution characteristics, safety threshold determined in step (3)thre_Z a The early-stage initial time T of the pre-precipitation water level0Corresponding water levelZ 0 Early-stage rainfall level river basin historical rainfall dataP a And the initial time T of the later period of the pre-lowering water levelaCorresponding water levelZ a Carrying out discretization treatment to form a plurality of groups;
(4-2) constructing a 'river basin rainfall-lake water level' association rule model: x, Y is a non-empty set of items, which is in the form of
Figure DEST_PATH_IMAGE017
The logical implication relationship of (a) is called an association rule, (b) X is called a front-piece or prerequisite of the association rule, (c) Y is called a back-piece or result of the association rule,
Figure 72627DEST_PATH_IMAGE018
；
digging a 'basin rainfall-lake water level' association rule in the early period of the pre-rainfall water level by constructing an association model; according to the association rule obtained by mining, the initial time T of the later period of the pre-reducing water level is realizedaCorresponding water levelZ a Control at safe thresholdthre_Z a The early-stage hydraulic engineering regulation and control suggestion of the pre-precipitation water level is provided; the association rule model relates to support degree and confidence degree as follows:
Figure 599423DEST_PATH_IMAGE020
Figure 471564DEST_PATH_IMAGE022
in the formula: d is a transaction set;
Figure 543556DEST_PATH_IMAGE024
the support degree of the association rule;
Figure 625782DEST_PATH_IMAGE026
is the confidence of the association rule;
(5) determining a water conservancy project regulation and control mode in a pre-reduction period in a future scene according to the association rule obtained by mining in the step (4) and the lake water level response rule in the post-reduction water level period obtained by mining in the step (2);
(5-1) based on the association rule obtained by mining in the step (4), according to the lake water level in the future situation
Figure DEST_PATH_IMAGE027
River basin rainfall data
Figure 500809DEST_PATH_IMAGE028
Predicting lake water level of future situation
Figure DEST_PATH_IMAGE029
And passing the lake level of the future scene
Figure 306086DEST_PATH_IMAGE029
And a safety thresholdthre_Z a Comparison, identification and determinationT 0 ~T a The hydraulic engineering regulation and control mode is that if the lake level of the future situation is in, the hydraulic engineering implements conventional scheduling, otherwise, the lake level of the future situation is pre-reduced in advance through the hydraulic engineering scheduling, so that the lake level of the future situation is in
Figure 329405DEST_PATH_IMAGE030
：
Figure 774293DEST_PATH_IMAGE032
In the formula (I), the compound is shown in the specification,
Figure DEST_PATH_IMAGE033
for the future situationT a Actual water level corresponding to the moment;
(5-2) lake water level according to the late end time of the pre-reduction water levelZ b Identifying and determining the response rule of each key driving factorT a ~T b And (3) an optimal hydraulic engineering regulation and control mode, bringing the actual scene into the scene set in the step (2), and continuously optimizing the steps through information feedback and training.
2. An apparatus for inversely predicting a target water level of a shallow lake based on an association rule model and numerical simulation, the apparatus comprising:
a processor and a memory storing computer program instructions;
the processor reads and executes the computer program instructions to implement the method of claim 1.
3. A computer-readable storage medium having computer program instructions stored thereon which, when executed by a processor, implement the method of claim 1.
CN202110262237.7A 2021-03-10 2021-03-10 Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation Active CN112926786B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110262237.7A CN112926786B (en) 2021-03-10 2021-03-10 Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110262237.7A CN112926786B (en) 2021-03-10 2021-03-10 Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation

Publications (2)

Publication Number Publication Date
CN112926786A CN112926786A (en) 2021-06-08
CN112926786B true CN112926786B (en) 2022-01-04

Family

ID=76172470

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110262237.7A Active CN112926786B (en) 2021-03-10 2021-03-10 Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation

Country Status (1)

Country Link
CN (1) CN112926786B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115829420B (en) * 2023-02-14 2023-05-02 清华四川能源互联网研究院 Method for judging steady-state conversion threshold value of shallow lake

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108921345A (en) * 2018-06-28 2018-11-30 杭州市水文水资源监测总站 The river flood-peak stage Real-time Forecasting Method that precipitation forecast is coupled with support vector machines
CN109948235A (en) * 2019-03-14 2019-06-28 南京大学 Water resource scheduling and precision configuration method
CN111898303A (en) * 2020-08-05 2020-11-06 苏州大圜科技有限公司 River basin water level and waterlogging forecasting method based on weather forecasting and hydrodynamic simulation

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108921345A (en) * 2018-06-28 2018-11-30 杭州市水文水资源监测总站 The river flood-peak stage Real-time Forecasting Method that precipitation forecast is coupled with support vector machines
CN109948235A (en) * 2019-03-14 2019-06-28 南京大学 Water resource scheduling and precision configuration method
CN111898303A (en) * 2020-08-05 2020-11-06 苏州大圜科技有限公司 River basin water level and waterlogging forecasting method based on weather forecasting and hydrodynamic simulation

Also Published As

Publication number Publication date
CN112926786A (en) 2021-06-08

Similar Documents

Publication Publication Date Title
Shokri et al. Reservoir operation for simultaneously meeting water demand and sediment flushing: Stochastic dynamic programming approach with two uncertainties
CN110298076A (en) A kind of urban waterlogging intelligent modeling and analysis method based on GIS and SWMM
JP6716328B2 (en) Inundation risk diagnostic device, inundation risk diagnostic method, control device, and computer program
JP2007205001A (en) Discharge forecasting apparatus
CN111046574A (en) Flood control and drainage calculation method for lake and river gate pump system in plain lake region
CN110219277A (en) The dam bursting flood analysis system of dam system
CN111199298B (en) Flood forecasting method and system based on neural network
CN114219131A (en) Watershed runoff prediction method based on LSTM
CN112052545B (en) Urban surface runoff and pipe network confluence coupling method based on cellular automaton
CN112926786B (en) Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation
Haktanir et al. Fifteen-stage operation of gated spillways for flood routing management through artificial reservoirs
Safavi et al. Conjunctive use of surface and ground water using fuzzy neural network and genetic algorithm
CN108876047B (en) Research method based on GALSS model sand transportation contribution rate
Paudel Role of sediment in the design and management of irrigation canals: Sunsari Morang Irrigation Scheme, Nepal
CN115730739A (en) Urban waterlogging prediction method based on LSTM neural network
CN108959816B (en) Detection method for high arch dam construction early-middle diversion risk rate
Mahmoud et al. Hybrid meta-heuristic adaptive fuzzy inference systems in rockfill dam multi-objective shape optimization
CN110348639A (en) A kind of coal mine roof plate gushing water danger classes prediction technique
CN117852850B (en) Flood storage area flood scheduling method and system based on virtual river channel replacement gate drainage calculation
Kyaw et al. Optimization–based solution for reducing water scarcity in the Greater Chao Phraya River Basin, Thailand: through re–operating the Bhumibol and Sirikit Reservoirs using non–linear programming solver
Kokkinos et al. Development and evaluation of an adaptive neuro fuzzy inference system for the calculation of soil water recharge in a watershed
JPH08180109A (en) River supervising and facility operating method and operator training system
Li et al. Study on auxiliary operation control of machine learning in multiobjective complex drainage system
CN117993495A (en) Method and system for constructing flood-delivery safety knowledge graph by combining machine learning with traditional mechanism model
Hii et al. Review of Reservoir Sediment Inflow Estimation for Teriang Reservoir, Malaysia

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