CN114997636A - Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty - Google Patents

Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty Download PDF

Info

Publication number
CN114997636A
CN114997636A CN202210597346.9A CN202210597346A CN114997636A CN 114997636 A CN114997636 A CN 114997636A CN 202210597346 A CN202210597346 A CN 202210597346A CN 114997636 A CN114997636 A CN 114997636A
Authority
CN
China
Prior art keywords
flood
model
risk
flood control
bayesian network
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
Application number
CN202210597346.9A
Other languages
Chinese (zh)
Inventor
李洁玉
杨敏芝
钟平安
江恩慧
王远见
曲少军
马怀宝
李昆鹏
李新杰
张翎
唐凤珍
王强
颜小飞
郭秀吉
孙龙飞
王欣
刘彦晖
赵万杰
江肖鹏
杨洲
李航
刘博伦
刘刚
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.)
Yellow River Institute of Hydraulic Research
Original Assignee
Yellow River Institute of Hydraulic Research
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 Yellow River Institute of Hydraulic Research filed Critical Yellow River Institute of Hydraulic Research
Priority to CN202210597346.9A priority Critical patent/CN114997636A/en
Publication of CN114997636A publication Critical patent/CN114997636A/en
Pending legal-status Critical Current

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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • 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
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
    • Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Strategic Management (AREA)
  • Mathematical Physics (AREA)
  • Operations Research (AREA)
  • Mathematical Optimization (AREA)
  • Tourism & Hospitality (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Pure & Applied Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Quality & Reliability (AREA)
  • Probability & Statistics with Applications (AREA)
  • Game Theory and Decision Science (AREA)
  • Algebra (AREA)
  • Educational Administration (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Development Economics (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention discloses a reservoir group flood control scheduling risk analysis method considering model structure and hydrologic forecast double uncertainty, which comprises the following steps: establishing a hybrid equivalent model and a full set joint scheduling model, and providing flood risk definition considering model structure uncertainty and flood risk definition considering model structure and hydrologic prediction double uncertainty; respectively determining a Bayesian network structure of a hybrid equivalent model and a full set joint scheduling model; generating a random flood sample, carrying out Bayesian network parameter learning, and establishing a Bayesian network risk analysis model of a hybrid equivalent model and a complete set joint scheduling model; and carrying out probability reasoning based on the Bayesian network risk analysis model, and carrying out risk prediction and diagnosis. According to the method, the model structure and hydrologic forecast double uncertainties in the real-time flood control mixed dispatching of the reservoir group are considered, the influence of the dispatching model structure and hydrologic forecast errors on flood risks can be quantitatively described, and a basis is provided for flood control risk decision-making.

Description

Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty
Technical Field
The invention relates to a reservoir group flood control dispatching risk analysis method, in particular to a reservoir group flood control dispatching risk analysis method considering model structure and hydrologic forecast double uncertainties.
Background
In the real-time flood control dispatching of the reservoir group, the mixed equivalent model divides the reservoir into a significant reservoir and a non-significant reservoir according to the flood control effect, the significant reservoir is jointly dispatched, the non-significant reservoir is independently dispatched, and the decision efficiency can be improved by reducing the model structure on the premise of ensuring the dispatching effect. The reduction of the model structure may result in a reduction of flood control efficiency and an increase of flood risk.
The real-time flood control scheduling of the reservoir group is generally influenced by various uncertainty factors such as hydrologic forecast errors, curve errors of reservoir capacity, curve errors of discharge capacity, flood evolution errors and the like, so that the uncertainty of a reservoir flood control decision is caused. In the past decades, theories and methods related to risk analysis of flood control systems, such as a recurrence period method, an analytic method, a reliability analysis method, a random simulation method, a bayesian network method and the like, have been developed rapidly. Along with the gradual expansion of a flood control engineering system, flood control dispatching in a drainage basin is gradually developed into river-reservoir combined optimal dispatching and reservoir group combined optimal dispatching from single-reservoir optimal dispatching, and accordingly, research on flood control dispatching risk analysis is also subject to a development process from single-reservoir dispatching risk analysis to reservoir group dispatching risk analysis. However, the existing research aims at risk analysis of a single reservoir or a flood control system with a fixed topological structure, and flood risks caused by uncertainty factors of model structures in reservoir group scheduling are not considered.
The number of reservoirs in the reservoir group system is large, the topological structure is complex, the reservoirs are divided into significant reservoirs and non-significant reservoirs by adopting mixed equivalent model scheduling, and the solving difficulty of joint probability distribution is aggravated by dynamic combination of the two types of reservoirs.
Disclosure of Invention
The invention aims to: the invention aims to provide a reservoir group flood control scheduling risk analysis method considering dual uncertainty of model structure and hydrologic prediction, which can quantitatively describe the influence of scheduling model structure and hydrologic prediction error on flood risk and provide basis for flood control risk decision.
The technical scheme is as follows: the invention discloses a reservoir flood control dispatching risk analysis method considering model structure and hydrologic forecast double uncertainty, which comprises the following steps of:
s1, establishing a hybrid equivalent model and a full set joint scheduling model, and providing flood risk definition considering model structure uncertainty and flood risk definition considering model structure and hydrologic prediction dual uncertainty;
s2, selecting Bayesian network node random variables of the corpus joint scheduling model and the hybrid equivalent model respectively, sequencing the Bayesian network nodes of the corpus joint scheduling model and the hybrid equivalent model respectively, determining directed edges according to the direction of the result nodes pointed by the reason nodes, and determining Bayesian network structures of the hybrid equivalent model and the corpus joint scheduling model respectively;
s3, on the basis of historical flood frequency analysis, generating random flood samples by adopting a Latin hypercube sampling method, and respectively learning Bayesian network parameters of a hybrid equivalent model and a corpus joint scheduling model, so as to establish a Bayesian network risk analysis model of the hybrid equivalent model and the corpus joint scheduling model;
and S4, performing probabilistic reasoning on the Bayesian network risk analysis model respectively based on the hybrid equivalent model and the corpus joint scheduling model to obtain a flood risk considering model structure uncertainty and a flood risk considering model structure and hydrologic forecast double uncertainties, and diagnosing and analyzing factors causing downstream flood control point risks and the probability of each factor.
Further, the objective function of the hybrid equivalent model in step S1 is:
the single-base optimization scheduling adopts a maximum peak clipping criterion, and the objective function is as follows:
Figure BDA0003668635420000021
wherein, T is the number of the time segments of the scheduling period, and q (T) is the ex-warehouse flow at the T moment;
the optimal dispatching of the reservoir group aims at the minimum maximum water flow of the flood control section, and the objective function is as follows:
Figure BDA0003668635420000022
wherein q' (i, t) is the process of calculating the ex-warehouse flow of the ith warehouse in the tth time period to the public flood control point; q's' D (i, t) is a process of calculating the interval flow between the ith library and the private flood control point to the public flood control point;
Figure BDA0003668635420000023
calculating the flow of the total interval from each warehouse to the public flood control point; m is a group of * The number of reservoirs participating in the joint scheduling;
the objective function of the full-set joint scheduling model is as follows:
Figure BDA0003668635420000024
wherein M is the total quantity of the reservoirs in the reservoir group;
the constraints of both models include: the system comprises a water balance constraint, a discharge capacity constraint, a reservoir highest water level constraint, a reservoir end water level constraint and an ex-warehouse flow variation constraint.
Further, the flood risk definition considering uncertainty of the model structure and the flood risk definition considering dual uncertainty of the model structure and the hydrologic forecast in step S1 specifically include:
s11, defining flood risk of public flood control points as probability that actual flood peak flow exceeds safe discharge:
Figure BDA0003668635420000031
wherein QS is the safe discharge capacity of the public flood control point; QC (quasi-cyclic) m The peak flow rate of the flood process of the public flood control point is shown; p (QC) m >QS) is the probability that the peak flood flow of the public flood control point exceeds the safe discharge; f (QC) m ) Is QC m A probability density function of;
s12, defining the flood risk considering the uncertainty of the model structure, comparing the hybrid equivalent model with the corpus scheduling model, and representing the variation of the flood risk of the public flood control point by the following formula:
△Risk (1) ＝Risk (1) -Risk (0)
wherein Risk (1) Scheduling flood risks generated at public flood control points for the hybrid equivalent model, and calculating by using a formula in the step S11; risk (0) The flood risk generated at the public flood control point is scheduled for the complete set scheduling model and is calculated by a formula in the step S11; delta Risk (1) Considering the flood risk of model structure uncertainty for the hybrid equivalent model;
s13, taking flood peak forecasting errors into consideration, and performing flood risk analysis under the dual uncertainty factors of the mixed equivalent model real-time scheduling model structure and hydrologic forecasting; if the flood peak forecast error is epsilon, then:
Figure BDA0003668635420000032
wherein Q is m In order to be the actual peak flow of the flood,
Figure BDA0003668635420000033
forecasting the peak flow;
if the hybrid equivalent model considers the flood forecasting error, the flood Risk generated at the public flood control point is recorded as Risk (2) Then, the flood risk considering the uncertainty of the model structure and the hydrologic forecast is defined as:
△Risk (2) ＝Risk (2) -Risk (0) 。
further, step S2 includes the following steps:
s21, the flood control system comprises M reservoirs and N flood control sections, and the warehousing flow of the reservoir i is QR i The delivery flow of reservoir i is qR i The upstream interval corresponding to the section j is supplied with water QL j Flood process of section j is QC j ；
S22, respectively sequencing all M reservoirs and N flood control sections according to the principle of upstream first and downstream second and branch first and main flow second;
s23, selecting the following four random variables as nodes of a complete set joint scheduling model Bayes network: warehouse entry peak flow QR of full-set combined scheduling model reservoir i mi Ex-warehouse peak flow of reservoir i of full-set combined dispatching model
Figure BDA0003668635420000041
Flood peak flow QL of incoming water in upstream interval corresponding to flood control section j of full-set combined dispatching model mj Full-set combined dispatching model flood control section j flood peak flow
Figure BDA0003668635420000042
Selecting the following four random variables as nodes of a Bayesian network of a hybrid equivalent model: mixed equivalent model reservoir i warehousing peak flow QR mi Discharge peak flow of mixed equivalent model reservoir i
Figure BDA0003668635420000043
Flood peak flow QL of incoming water in upstream interval corresponding to flood control section j of hybrid equivalent model mj Peak flow of flood control section j of hybrid equivalent model
Figure BDA0003668635420000044
S24, analyzing the causal relationship of all random variables of the complete set combined dispatching model and the mixed equivalent model, and respectively sequencing the Bayesian network nodes of the complete set combined dispatching model and the mixed equivalent model according to the sequencing sequence of the reservoirs and the flood control sections and the principle of the random variables 'cause before and effect after';
and S25, determining directed edges according to the causal relationship of each random variable of the full-set joint scheduling model and the hybrid equivalent model and the direction of the cause node pointing to the result node, and respectively determining the Bayes network structures of the full-set joint scheduling model and the hybrid equivalent model.
Further, step S3 includes the following steps:
s31, selecting L-field historical typical flood, and estimating parameters by adopting a linear moment method to obtain probability density distribution functions of the peak flow of the flood in each reservoir; randomly sampling the probability density distribution function according to a Latin hypercube sampling method to generate Z random peak flow samples; obtaining a random real-time flood sample set containing Z-field flood from flood peak flow random samples based on a unity-fold ratio amplification method
Figure BDA0003668635420000045
i＝1,2,…,M，j＝1,2,…,N，v＝1,2,…,Z；
S32, random real-time flood sample set
Figure BDA0003668635420000046
Respectively carrying out mixed equivalent model scheduling and full set combined scheduling model scheduling on Z-field flood, and recording flood and scheduling result sample sets obtained by the mixed equivalent model as
Figure BDA0003668635420000047
Flood and scheduling result sample set obtained by full-set joint scheduling model is recorded as
Figure BDA0003668635420000048
Wherein i is 1,2, …, M, j is 1,2, …, N, v is 1,2, …, Z,
Figure BDA0003668635420000049
for the ith warehouse and the vth warehouse flood process,
Figure BDA00036686354200000410
for the ith warehouse exit flood process obtained by mixing the equivalent models,
Figure BDA00036686354200000411
for the ith warehouse exit flood process obtained by the full-set combined dispatching model,
Figure BDA00036686354200000412
for the v interval of the jth flood control point,
Figure BDA00036686354200000413
for the jth flood control point vth flood process obtained by the hybrid equivalent model,
Figure BDA00036686354200000414
obtaining a jth flood control point vth flood process for the complete set combined dispatching model;
s33, from
Figure BDA0003668635420000051
Extracting peak flow to obtain
Figure BDA0003668635420000052
Extracting peak flow to obtain
Figure BDA0003668635420000053
Wherein i is 1,2, …, M, j is 1,2, …, N, v is 1,2, …, Z,
Figure BDA0003668635420000054
for the peak flow of the ith warehouse and the vth warehouse-in flood process,
Figure BDA0003668635420000055
ith library and vth library flood process obtained for mixed equivalent modelThe peak flow of the flood is measured,
Figure BDA0003668635420000056
the flood peak flow of the ith warehouse and the vth warehouse exit flood process obtained by the full-set combined dispatching model,
Figure BDA0003668635420000057
the peak flow of flood for the nth interval of the jth flood control point,
Figure BDA0003668635420000058
the peak flow of the jth flood control point in the vth flood process obtained by the hybrid equivalent model,
Figure BDA0003668635420000059
obtaining the volume peak flow of the vth flood control point of the jth flood control process for the complete set combined scheduling model;
s34, adopting equal-width interval method to
Figure BDA00036686354200000510
And
Figure BDA00036686354200000511
carrying out discretization treatment on the four random variables; performing parameter learning by adopting a maximum likelihood estimation method, and if O observation samples are shared, marking as X ═ X 1 ,X 2 ,…,X O Then the likelihood function is:
Figure BDA00036686354200000512
the parameter estimation value is:
Figure BDA00036686354200000513
and respectively obtaining a Bayesian network risk analysis model of the hybrid equivalent model and a Bayesian network risk analysis model of the complete set joint scheduling model.
Further, step S4 includes the following steps:
(41) forecasting inference is carried out according to the Bayesian network of the hybrid equivalent model and the Bayesian network of the full-set combined scheduling model, and the forecasting warehouse entry flood peak flow and the interval incoming flood peak flow are known to form a forecasting inference sample set
Figure BDA00036686354200000514
i ═ 1,2, …, M, j ═ 1,2, …, N, v ═ 1,2, …, Z, where v ═ 1,2, … * The sequence number of the prediction samples is Z, and the total number of the prediction samples is represented by Z; reasoning the sample set SF according to the prediction m Deducing the probability Risk of flood Risk of the downstream flood control point hybrid equivalent model (1) Probability Risk of flood occurrence of complete set joint scheduling model (0) Obtaining flood Risk delta Risk caused by model structure uncertainty (1) ＝Risk (1) -Risk (0) ；
(42) Obtaining a prediction inference sample set considering the flood forecast error as epsilon
Figure BDA0003668635420000061
i is 1,2, …, M, j is 1,2, …, N, v is 1,2, … and Z, and the probability Risk of flood Risk of the downstream flood control point hybrid equivalent model is deduced (2) Obtaining flood Risk delta Risk caused by model structure and hydrologic forecast double uncertainties (2) ＝Risk (2) -Risk (0) ；
(43) And (3) supposing that the downstream flood control point exceeds the control flow, reversely reasoning out the probability density function of the flood peak flow of the father node, and diagnosing the factors causing the risk and the probability of each factor.
The invention relates to a reservoir flood control dispatching risk analysis system considering model structure and hydrologic forecast double uncertainty, which comprises:
the Bayesian network structure building module is used for respectively selecting Bayesian network node random variables of the corpus joint scheduling model and the hybrid equivalent model, respectively sequencing the Bayesian network nodes of the corpus joint scheduling model and the hybrid equivalent model, determining directed edges according to the direction from the reason node to the result node, and respectively determining the Bayesian network structures of the hybrid equivalent model and the corpus joint scheduling model;
the Bayesian network risk analysis model establishing module is used for generating random flood samples by adopting a Latin hypercube sampling method on the basis of historical flood frequency analysis, and respectively learning Bayesian network parameters of a hybrid equivalent model and a corpus joint scheduling model so as to establish Bayesian network risk analysis models of the hybrid equivalent model and the corpus joint scheduling model;
and the risk prediction and analysis module is used for performing probabilistic reasoning on the Bayesian network risk analysis model respectively based on the hybrid equivalent model and the full set joint scheduling model to obtain the flood risk considering the uncertainty of the model structure and the flood risk considering the dual uncertainties of the model structure and hydrologic forecast, and diagnosing and analyzing factors causing the risk of the downstream flood control point and the probability of each factor.
An apparatus of the present invention includes a memory and a processor, wherein:
a memory for storing a computer program capable of running on the processor;
and the processor is used for executing the steps of the reservoir group flood control dispatching risk analysis method considering the model structure and the hydrologic forecast double uncertainty when the computer program is run.
The invention relates to a storage medium, on which a computer program is stored, wherein the computer program, when executed by at least one processor, implements the steps of the above reservoir group flood control dispatching risk analysis method considering model structure and hydrologic forecast double uncertainties.
Has the advantages that: compared with the prior art, the invention has the advantages that: (1) the mixed equivalent model can improve the real-time flood control dispatching efficiency due to model structure reduction, but can cause larger flood risk, the method quantitatively describes the flood risk of considering model structure uncertainty of the mixed equivalent model in the real-time flood control dispatching of the reservoir group, and can prove the reliability of the mixed equivalent model; (2) according to the method, dual uncertainty factors of a model structure and hydrologic prediction are comprehensively considered, flood risks of a flood control system are estimated under different random flood situations, and the method has important significance for guiding real-time flood control scheduling and risk decision; (3) according to the method, under the condition that the risk of the downstream flood control point of the reservoir group is known, the reasons for risk generation are diagnosed, the probability corresponding to each reason is determined, and theoretical support can be provided for a decision maker to further carry out risk regulation.
Drawings
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is a schematic diagram of the reservoir group system structure of the present invention;
fig. 3 is a diagram of a bayesian network topology of the present invention, wherein (a) is the bayesian network topology of the corpus joint scheduling model, and (b) is the bayesian network topology of the hybrid equivalent model.
Detailed Description
The present invention will be described in detail below by way of examples with reference to the accompanying drawings.
The Bayesian network has strong capacity of solving joint probability distribution, so the invention provides a reservoir group flood control dispatching risk analysis method considering model structure and hydrologic forecast double uncertainties based on a Bayesian network theory, the risk caused by the model structure and the hydrologic forecast double uncertainties is quantitatively analyzed, the cause of the risk is clarified, and the method has important significance for guiding real-time flood control dispatching and risk decision.
As shown in fig. 1, the reservoir flood control dispatching risk analysis method considering dual uncertainties of model structure and hydrologic forecast, provided by the invention, comprises the following steps:
s1, establishing a hybrid equivalent model and a full set joint scheduling model, and providing flood risk definition considering model structure uncertainty and flood risk definition considering model structure and hydrologic prediction dual uncertainty;
in the embodiment, a five-reservoir joint-debugging flood control system in the middle and lower reaches of a yellow river is taken as an example, and the system comprises three flood control sections, namely five reservoirs of three gorges, small wave bottoms, prefectures, muddy lands and estuaries, a black stone gate, Wu\38495and a garden opening, wherein the garden opening is a public flood control point.
S11, establishing a hybrid equivalent model, wherein the objective function is as follows:
the single-base optimization scheduling adopts a maximum peak clipping criterion, and the objective function is as follows:
Figure BDA0003668635420000081
wherein T is the number of the scheduling period (h), q (T) is the delivery flow (m) at the T-th time 3 /s)。
The optimal dispatching of the reservoir group aims at the minimum maximum water flow of the flood control section, and the objective function is as follows:
Figure BDA0003668635420000082
in the formula, q' (i, t) is the process (m) of calculating the export flow of the ith time period of the ith library to the public flood control point 3 /s)；q' D (i, t) is the process (m) from the interval flow calculation between the ith base and the private flood control point to the public flood control point 3 /s)；
Figure BDA0003668635420000083
Calculating the process from the total interval flow between each warehouse and the public flood control point to the public flood control point (m) 3 /s)；M * The number of reservoirs participating in the joint scheduling.
Establishing a complete set combined scheduling model, wherein an objective function is as follows:
Figure BDA0003668635420000084
in the formula, M is the total number of the reservoirs in the reservoir group, and the total number of the reservoirs in this embodiment is 5.
The two model constraints are as follows:
and (3) water balance constraint:
Figure BDA0003668635420000085
wherein V (i, t-1) and V (i, t) are water storage amounts (m) of the initial reservoir and the final reservoir in the t period of the ith reservoir 3 ) (ii) a Q (i, t-1) and Q (i, t) are the initial and final storage flow (m) of the ith reservoir at t time period 3 S); q (i, t-1) and q (i, t) are the initial and final delivery flow (m) of the ith reservoir at t time period 3 S); Δ t is the time period length.
And (4) restriction of the drainage capacity:
q(i,t)≤q(i,Z(i,t)) (5)
wherein q (i, t) is the discharge flow (m) of the ith reservoir at time t 3 S); q (i, Z (i, t)) is the discharge capacity (m) of the ith reservoir at time t corresponding to the water level Z (i, t) 3 /s)。
Reservoir maximum water level restraint:
Figure BDA0003668635420000091
wherein Z (i, t) is the reservoir water level (m) at the moment t of the ith reservoir;
Figure BDA0003668635420000092
the maximum water level (m) is allowed for the ith reservoir at time t.
Reservoir end water level constraint:
Z i,end ＝Z i,e (7)
in the formula, Z i,end Dispatching the reservoir water level (m) calculated at the end of the term for the reservoir i; z i,e And dispatching the control water level (m) at the end of the period for the ith reservoir.
And (3) ex-warehouse flow amplitude variation constraint:
Figure BDA0003668635420000093
wherein | q (i, t) -q (i, t-1) | is the amplitude (m) of the delivery flow of the ith reservoir in the adjacent period 3 /s)；
Figure BDA0003668635420000094
Tolerance (m) for variation of delivery flow rate for adjacent time periods 3 /s)。
S12, defining flood risk of the garden opening section of the public flood control point as the probability that the actual flood peak flow exceeds the safe discharge:
Figure BDA0003668635420000095
wherein QS is the safe discharge capacity of the public flood control point; QC (quasi-cyclic) m The peak flow rate of the flood process of the public flood control point is shown; p (QC) m >QS) is the probability that the peak flood flow of the public flood control point exceeds the safe discharge; f (QC) m ) Is QC m A probability density function of;
s13, defining the flood risk considering the uncertainty of the model structure, comparing the hybrid equivalent model with the corpus scheduling model, and representing the variation of the flood risk of the public flood control point by the following formula:
△Risk (1) ＝Risk (1) -Risk (0) (10)
wherein, Risk (1) Scheduling flood risks generated at public flood control points for the hybrid equivalent model, and calculating by a formula (9); risk (0) The flood risk generated at the public flood control point is scheduled for the complete set scheduling model and is calculated by a formula (9); delta Risk (1) Considering the flood risk of model structure uncertainty for the hybrid equivalent model;
s14, taking flood peak forecasting errors into consideration, and performing flood risk analysis under the dual uncertainty factors of the mixed equivalent model real-time scheduling model structure and hydrologic forecasting; if the flood peak forecast error is epsilon, then:
Figure BDA0003668635420000101
wherein Q m In order to be the actual peak flow of the flood,
Figure BDA0003668635420000102
forecasting the peak flow;
if the hybrid equivalent model considers the flood risk generated by the flood forecast error at the public flood control point, the flood risk is recorded asRisk (2) Then, the flood risk considering the uncertainty of the model structure and the hydrologic forecast is defined as:
△Risk (2) ＝Risk (2) -Risk (0) (12)。
s2, selecting Bayesian network node random variables of the corpus joint scheduling model and the hybrid equivalent model respectively, sequencing the Bayesian network nodes of the corpus joint scheduling model and the hybrid equivalent model respectively, determining directed edges according to the direction of the result nodes pointed by the reason nodes, and determining Bayesian network structures of the hybrid equivalent model and the corpus joint scheduling model respectively;
s21, the flood control system of this embodiment includes 5 reservoirs and 3 flood control sections, as shown in fig. 2, where M is 5 and N is 3; the warehousing flow of the reservoir i is QR i The delivery flow of reservoir i is qR i The upstream interval corresponding to the section j is supplied with water in a QL form j Flood process of section j is QC j ；
S22, sequencing all 5 reservoirs and 3 flood control sections according to the principle of upstream first and downstream second and branch first and main flow second;
s23, selecting the following four random variables as nodes of a complete set joint scheduling model Bayes network: warehouse entry peak flow QR of full-set combined dispatching model reservoir i mi Warehouse-out peak flow of reservoir i of full-set combined dispatching model
Figure BDA0003668635420000103
Flood peak flow QL of incoming water in upstream interval corresponding to flood control section j of full-set combined dispatching model mj Full-set combined dispatching model flood control section j flood peak flow
Figure BDA0003668635420000104
Selecting the following four random variables as nodes of a Bayesian network of a hybrid equivalent model: warehouse entry peak flow QR of mixed equivalent model reservoir i mi And the discharge peak flow of the mixed equivalent model reservoir i
Figure BDA0003668635420000105
Mixing, etcFlood peak flow QL of incoming water in upstream interval corresponding to flood control section j of effect model mj Peak flow of flood control section j of hybrid equivalent model
Figure BDA0003668635420000106
S24, analyzing the causal relationship of all random variables of the complete set combined dispatching model and the mixed equivalent model, and respectively sequencing the Bayesian network nodes of the complete set combined dispatching model and the mixed equivalent model according to the sequencing sequence of the reservoirs and the flood control sections and the principle of the random variables 'cause before and effect after';
s25, determining directed edges according to causal relationships of random variables of the complete-set joint scheduling model and the hybrid equivalent model, and determining bayesian network structures of the complete-set joint scheduling model and the hybrid equivalent model respectively, as shown in (a) and (b) of fig. 3, in a direction from the cause node to the effect node.
S3, on the basis of historical flood frequency analysis, generating random flood samples by adopting a Latin hypercube sampling method, and respectively learning Bayesian network parameters of a hybrid equivalent model and a corpus joint scheduling model, so as to establish a Bayesian network risk analysis model of the hybrid equivalent model and the corpus joint scheduling model;
s31, selecting 20 historical typical floods, and estimating parameters by using a linear moment method to obtain probability density distribution functions of flood peak flow of each reservoir warehousing flood; randomly sampling probability density distribution functions according to a Latin Hypercube Sampling (LHS) method to generate 1000 random peak flow samples; obtaining a random real-time flood sample set containing 1000 floods from flood peak flow random samples based on a unity-fold ratio amplification method
Figure BDA0003668635420000111
i＝1,2,…,5，j＝1,2,…,3，v＝1,2,…,1000；
S32, random real-time flood sample set
Figure BDA0003668635420000112
Respectively mixing the 1000 flood fieldsModel scheduling and full set combined scheduling, flood and scheduling result sample set obtained by mixing equivalent models are recorded as
Figure BDA0003668635420000113
Flood and scheduling result sample set obtained by full-set combined scheduling model is recorded as
Figure BDA0003668635420000114
Wherein, i is 1,2, …,5, j is 1,2, …,3, v is 1,2, …,1000,
Figure BDA0003668635420000115
for the ith warehouse and the vth warehouse flood process,
Figure BDA0003668635420000116
for the ith warehouse exit flood process obtained by mixing the equivalent models,
Figure BDA0003668635420000117
for the ith warehouse exit flood process obtained by the full-set combined dispatching model,
Figure BDA0003668635420000118
for the jth flood point the v interval,
Figure BDA0003668635420000119
for the jth flood control point vth flood process obtained by the hybrid equivalent model,
Figure BDA00036686354200001110
obtaining a jth flood control point vth flood process for the complete set combined dispatching model;
s33, respectively selecting from
Figure BDA00036686354200001111
And
Figure BDA00036686354200001112
extracting peak flow to obtain
Figure BDA00036686354200001113
And
Figure BDA00036686354200001114
wherein, i is 1,2, …,5, j is 1,2, …,3, v is 1,2, …,1000,
Figure BDA00036686354200001115
for the peak flow of the ith warehouse and the vth warehouse-in flood process,
Figure BDA00036686354200001116
the flood peak flow of the ith warehouse and the vth warehouse-out flood process obtained by the mixed equivalent model,
Figure BDA00036686354200001117
the flood peak flow of the ith warehouse and the vth warehouse-out flood process obtained by the full-set combined dispatching model,
Figure BDA00036686354200001118
the peak flow of flood for the nth interval of the jth flood control point,
Figure BDA00036686354200001119
the peak flow of the jth flood control point in the vth flood process obtained by the hybrid equivalent model,
Figure BDA00036686354200001120
obtaining the volume peak flow of the vth flood control point of the jth flood control process for the complete set combined scheduling model;
s34, adopting an equal-width interval method, and taking the width of 200m 3 S, to
Figure BDA0003668635420000121
And
Figure BDA0003668635420000122
carrying out discretization treatment on the four random variables; parameter learning is performed by using a Maximum Likelihood Estimation (MLE) method, and 1000 observation samples are recorded as X ═ X 1 ,X 2 ,…,X 1000 }, the likelihood function is:
Figure BDA0003668635420000123
the parameter estimation value is:
Figure BDA0003668635420000124
and respectively obtaining a Bayesian network risk analysis model of the hybrid equivalent model and a Bayesian network risk analysis model of the complete set joint scheduling model, wherein the Bayesian network risk analysis models comprise parameters of a Bayesian network determined by parameter learning and a model structure determined in the step S2.
S4, carrying out probabilistic reasoning on the Bayesian network risk analysis model respectively based on the hybrid equivalent model and the corpus combined scheduling model to obtain flood risks considering model structure uncertainty and flood risks considering model structure and hydrologic forecast double uncertainties, and diagnosing and analyzing factors causing downstream flood control point risks and the probability of each factor;
s41, the embodiment predicts and infers the well-known forecast warehouse entry peak flow and the interval incoming flood peak flow according to the Bayes network of the hybrid equivalent model and the Bayes network of the full-set combined scheduling model to form a prediction and inference sample set
Figure BDA0003668635420000125
i ═ 1,2, …,5, j ═ 1,2, …,3, v ═ 1,2, …,100, where v is * Is the number of prediction samples. Reasoning the sample set SF according to the prediction m Deducing the probability Risk of flood Risk of the downstream flood control point hybrid equivalent model (1) Probability Risk of flood Risk of complete set combined scheduling model (0) Obtaining flood Risk delta Risk caused by model structure uncertainty (1) ＝Risk (1) -Risk (0) ；
S42, taking the flood forecast error into consideration to be 10%, and obtaining a prediction inference sample set taking the forecast error into consideration
Figure BDA0003668635420000126
Figure BDA0003668635420000126
Figure BDA0003668635420000126
1,2, …,5, 1,2, …,3, 1,2, … and 100, and deducing the probability Risk of the flood Risk of the downstream flood control point hybrid equivalent model (2) Obtaining flood Risk delta Risk caused by model structure and hydrologic forecast double uncertainties (2) ＝Risk (2) -Risk (0) ；
And S43, supposing that the downstream flood control point exceeds the control flow, reversely deducing the probability density function of the peak flow of the father node of the downstream flood control point, and diagnosing the factors causing the risk and the probability of each factor.
The invention relates to a reservoir flood control dispatching risk analysis system considering model structure and hydrologic forecast double uncertainty, which comprises:
the Bayesian network structure construction module is used for respectively selecting Bayesian network node random variables of the complete set joint scheduling model and the hybrid equivalent model, respectively sequencing the Bayesian network nodes of the complete set joint scheduling model and the hybrid equivalent model, respectively determining directed edges according to the direction from the reason nodes to the result nodes, and respectively determining the Bayesian network structures of the hybrid equivalent model and the complete set joint scheduling model;
the Bayesian network risk analysis model establishing module is used for generating random flood samples by adopting a Latin hypercube sampling method on the basis of historical flood frequency analysis, and respectively learning Bayesian network parameters of a hybrid equivalent model and a corpus joint scheduling model so as to establish Bayesian network risk analysis models of the hybrid equivalent model and the corpus joint scheduling model;
and the risk prediction and analysis module is used for performing probability reasoning on the Bayesian network risk analysis model respectively based on the hybrid equivalent model and the full set joint scheduling model to obtain the flood risk considering model structure uncertainty and the flood risk considering model structure and hydrologic forecast double uncertainties, and diagnosing and analyzing factors causing the risk of the downstream flood control point and the probability of each factor.
An apparatus of the present invention includes a memory and a processor, wherein:
a memory for storing a computer program operable on the processor;
and the processor is used for executing the steps of the reservoir group flood control dispatching risk analysis method considering the model structure and the hydrologic forecast double uncertainty when the computer program is run.
The storage medium stores a computer program, and when the computer program is executed by at least one processor, the steps of the reservoir group flood control dispatching risk analysis method considering the model structure and the hydrologic forecast double uncertainty are realized, and the technical effects consistent with the method are achieved.
In conclusion, the invention considers the model structure and hydrologic forecast double uncertainties in the real-time flood control mixed dispatching of the reservoir group, provides the flood control risk analysis method based on the Bayesian network model, can quantitatively describe the influence of the dispatching model structure and hydrologic forecast errors on flood risks, and provides a basis for flood control risk decision.

Claims (9)

1. A reservoir group flood control dispatching risk analysis method considering model structure and hydrologic forecast double uncertainties is characterized by comprising the following steps:
s1, establishing a hybrid equivalent model and a full set joint scheduling model, and providing flood risk definition considering model structure uncertainty and flood risk definition considering model structure and hydrologic prediction dual uncertainty;
s2, selecting Bayesian network node random variables of the corpus joint scheduling model and the hybrid equivalent model respectively, sequencing the Bayesian network nodes of the corpus joint scheduling model and the hybrid equivalent model respectively, determining directed edges according to the direction of the result nodes pointed by the reason nodes, and determining Bayesian network structures of the hybrid equivalent model and the corpus joint scheduling model respectively;
s3, on the basis of historical flood frequency analysis, generating random flood samples by adopting a Latin hypercube sampling method, and respectively learning Bayesian network parameters of a hybrid equivalent model and a corpus joint scheduling model, so as to establish a Bayesian network risk analysis model of the hybrid equivalent model and the corpus joint scheduling model;
and S4, performing probabilistic reasoning on the Bayesian network risk analysis model respectively based on the hybrid equivalent model and the corpus combined scheduling model to obtain the flood risk considering model structure uncertainty and the flood risk considering model structure and hydrologic forecast double uncertainties, and diagnosing and analyzing factors causing the risk of the downstream flood control point and the probability of each factor.
2. The method for risk analysis of reservoir flood control dispatch in consideration of dual uncertainties of model structure and hydrologic prediction according to claim 1, wherein the objective function of the hybrid equivalent model in step S1 is:
the single-base optimization scheduling adopts a maximum peak clipping criterion, and the objective function is as follows:
Figure FDA0003668635410000011
wherein T is the number of the periods of the scheduling period, and q (T) is the ex-warehouse flow at the T moment;
the optimal dispatching of the reservoir group aims at the minimum maximum water flow of the flood control section, and the objective function is as follows:
Figure FDA0003668635410000012
wherein q' (i, t) is the process of calculating the ex-warehouse flow of the ith warehouse in the tth time period to the public flood control point; q's' D (i, t) is a process of calculating the interval flow between the ith library and the private flood control point to the public flood control point;
Figure FDA0003668635410000013
calculating the flow of the total interval from each warehouse to the public flood control point; m * The number of reservoirs participating in the joint scheduling;
the objective function of the full-set joint scheduling model is as follows:
Figure FDA0003668635410000021
wherein M is the total quantity of the reservoirs in the reservoir group;
the constraints for both models include: the method comprises the following steps of water balance constraint, discharge capacity constraint, reservoir highest water level constraint, reservoir end water level constraint and ex-warehouse flow amplitude constraint.
3. The method for analyzing risk of flood control and dispatch of reservoir group according to claim 1, wherein the step S1 comprises defining flood risk considering uncertainty of model structure and hydrologic forecast, specifically:
s11, defining flood risk of a public flood control point as the probability that the actual peak flow exceeds the safe discharge:
Figure FDA0003668635410000022
wherein QS is the safe discharge capacity of the public flood control point; QC (quasi-cyclic) m The peak flow rate of the flood process of the public flood control point is shown; p (QC) m >QS) is the probability that the peak flood flow of the public flood control point exceeds the safe discharge; f (QC) m ) Is QC m A probability density function of;
s12, defining the flood risk considering the uncertainty of the model structure, comparing the hybrid equivalent model with the corpus scheduling model, and representing the variation of the flood risk of the public flood control point by the following formula:
△Risk (1) ＝Risk (1) -Risk (0)
wherein Risk (1) The flood risk generated at the public flood control point is scheduled for the hybrid equivalent model and is calculated by the formula in the step S11; risk (0) The flood risk generated at the public flood protection point is scheduled for the full set scheduling model, by step S11Calculating by using a middle formula; delta Risk (1) Considering the flood risk of model structure uncertainty for the hybrid equivalent model;
s13, taking flood peak forecasting errors into consideration, and performing flood risk analysis under the dual uncertainty factors of the mixed equivalent model real-time scheduling model structure and hydrologic forecasting; if the flood peak forecast error is epsilon, then:
Figure FDA0003668635410000023
wherein Q m In order to be the actual peak flow of the flood,
Figure FDA0003668635410000024
forecasting the peak flow;
if the hybrid equivalent model considers the flood Risk generated by the flood forecast error at the public flood control point, the flood Risk is recorded as Risk (2) Then, the flood risk considering the uncertainty of the model structure and the hydrologic forecast is defined as:
△Risk (2) ＝Risk (2) -Risk (0) 。
4. the method for risk analysis of reservoir flood control dispatch in accordance with claim 1, wherein step S2 comprises the following steps:
s21, the flood control system comprises M reservoirs and N flood control sections, and the warehousing flow of the reservoir i is QR i The delivery flow of reservoir i is qR i The upstream interval corresponding to the section j is supplied with water in a QL form j Flood process of section j is QC j ；
S22, respectively sequencing all M reservoirs and N flood control sections according to the principle of upstream first and downstream second and branch first and main flow second;
s23, selecting the following four random variables as nodes of a complete set joint scheduling model Bayes network: warehouse entry peak flow QR of full-set combined scheduling model reservoir i mi Warehouse-out peak flow of reservoir i of full-set combined dispatching model
Figure FDA0003668635410000031
Flood peak flow QL of incoming water in upstream interval corresponding to flood control section j of full-set combined dispatching model mj Full-set combined dispatching model flood control section j flood peak flow
Figure FDA0003668635410000032
Selecting the following four random variables as nodes of a hybrid equivalent model Bayesian network: warehouse entry peak flow QR of mixed equivalent model reservoir i mi And the discharge peak flow of the mixed equivalent model reservoir i
Figure FDA0003668635410000033
Flood peak flow QL of incoming water in upstream interval corresponding to flood control section j of hybrid equivalent model mj Peak flow of flood control section j of hybrid equivalent model
Figure FDA0003668635410000034
S24, analyzing the causal relationship of all random variables of the complete set combined dispatching model and the mixed equivalent model, and respectively sequencing the Bayesian network nodes of the complete set combined dispatching model and the mixed equivalent model according to the sequencing sequence of the reservoirs and the flood control sections and the principle of the random variables 'cause before and effect after';
and S25, determining directed edges according to the causal relationship of all random variables of the complete set joint scheduling model and the mixed equivalent model and the direction from the reason node to the result node, and respectively determining the Bayesian network structures of the complete set joint scheduling model and the mixed equivalent model.
5. The method for analyzing risk of flood control dispatching in reservoir group according to claim 1, wherein step S3 comprises the following steps:
s31, selecting L-field historical typical flood, and estimating parameters by adopting a linear moment method to obtain the probability density of the peak flow of the flood peak of each reservoir warehousing floodA distribution function; randomly sampling the probability density distribution function according to a Latin hypercube sampling method to generate Z random peak flow samples; obtaining a random real-time flood sample set containing Z-field flood from flood peak flow random samples based on a unity-fold ratio amplification method
Figure FDA0003668635410000041
S32, random real-time flood sample set
Figure FDA0003668635410000042
Respectively carrying out mixed equivalent model scheduling and full set combined scheduling model scheduling on Z field flood, and recording flood and scheduling result sample sets obtained by the mixed equivalent model as
Figure FDA0003668635410000043
Flood and scheduling result sample set obtained by full-set combined scheduling model is recorded as
Figure FDA0003668635410000044
Wherein i is 1,2, …, M, j is 1,2, …, N, v is 1,2, …, Z,
Figure FDA0003668635410000045
for the ith warehouse and the vth warehouse flood process,
Figure FDA0003668635410000046
for the ith warehouse exit flood process obtained by mixing the equivalent models,
Figure FDA0003668635410000047
for the ith warehouse exit flood process obtained by the full-set combined dispatching model,
Figure FDA0003668635410000048
for the jth flood point the v interval,
Figure FDA0003668635410000049
for the jth flood control point vth flood process obtained by the hybrid equivalent model,
Figure FDA00036686354100000410
obtaining a jth flood control point vth flood process for the complete set combined scheduling model;
s33, from
Figure FDA00036686354100000411
Extracting peak flow to obtain
Figure FDA00036686354100000412
From
Figure FDA00036686354100000413
Extracting peak flow to obtain
Figure FDA00036686354100000414
Wherein i is 1,2, …, M, j is 1,2, …, N, v is 1,2, …, Z,
Figure FDA00036686354100000415
for the peak flow of the ith warehouse and the vth warehouse-in flood process,
Figure FDA00036686354100000416
the flood peak flow of the ith warehouse and the vth warehouse-out flood process obtained by the mixed equivalent model,
Figure FDA00036686354100000417
the flood peak flow of the ith warehouse and the vth warehouse exit flood process obtained by the full-set combined dispatching model,
Figure FDA00036686354100000418
the peak flow of flood for the nth interval of the jth flood control point,
Figure FDA00036686354100000419
the peak flow of the jth flood control point in the vth flood process obtained by the hybrid equivalent model,
Figure FDA00036686354100000420
obtaining the volume peak flow of the vth flood control point of the jth flood control process for the complete set combined scheduling model;
s34, adopting equal-width interval method to
Figure FDA00036686354100000421
And
Figure FDA00036686354100000422
carrying out discretization treatment on the four random variables; performing parameter learning by adopting a maximum likelihood estimation method, and if O observation samples are shared, recording as X ═ X 1 ,X 2 ,…,X O }, the likelihood function is:
Figure FDA00036686354100000423
the parameter estimation value is:
Figure FDA0003668635410000051
and respectively obtaining a Bayesian network risk analysis model of the hybrid equivalent model and a Bayesian network risk analysis model of the complete set joint scheduling model.
6. The method for risk analysis of reservoir flood control dispatch in accordance with claim 1, wherein step S4 comprises the following steps:
(41) forecasting inference is carried out according to the Bayesian network of the hybrid equivalent model and the Bayesian network of the full-set combined scheduling model, and the forecasting warehouse entry flood peak flow and the interval incoming flood peak flow are known to form a forecasting inference sample set
Figure FDA0003668635410000052
Wherein v is * The sequence number of the prediction sample is Z, and the total number of the prediction samples is represented by Z; reasoning the sample set SF according to the prediction m Deducing the probability Risk of flood Risk of the downstream flood control point hybrid equivalent model (1) Probability Risk of flood Risk of complete set combined scheduling model (0) Obtaining flood Risk delta Risk caused by model structure uncertainty (1) ＝Risk (1) -Risk (0) ；
(42) The forecast error of the flood is considered as epsilon, and a forecast inference sample set considering the forecast error is obtained
Figure FDA0003668635410000053
Deducing probability Risk of flood Risk of downstream flood control point hybrid equivalent model (2) Obtaining flood Risk delta Risk caused by model structure and hydrologic forecast double uncertainties (2) ＝Risk (2) -Risk (0) ；
(43) And (3) supposing that the downstream flood control point exceeds the control flow, reversely reasoning out the probability density function of the flood peak flow of the father node, and diagnosing the factors causing the risk and the probability of each factor.
7. A reservoir group flood control dispatching risk analysis system considering model structure and hydrologic forecast double uncertainty is characterized by comprising:
the Bayesian network structure building module is used for respectively selecting Bayesian network node random variables of the corpus joint scheduling model and the hybrid equivalent model, respectively sequencing the Bayesian network nodes of the corpus joint scheduling model and the hybrid equivalent model, determining directed edges according to the direction from the reason node to the result node, and respectively determining the Bayesian network structures of the hybrid equivalent model and the corpus joint scheduling model;
the Bayesian network risk analysis model establishing module is used for generating random flood samples by adopting a Latin hypercube sampling method on the basis of historical flood frequency analysis, and respectively learning Bayesian network parameters of a hybrid equivalent model and a corpus joint scheduling model so as to establish Bayesian network risk analysis models of the hybrid equivalent model and the corpus joint scheduling model;
and the risk prediction and analysis module is used for performing probability reasoning on the Bayesian network risk analysis model respectively based on the hybrid equivalent model and the full set joint scheduling model to obtain the flood risk considering model structure uncertainty and the flood risk considering model structure and hydrologic forecast double uncertainties, and diagnosing and analyzing factors causing the risk of the downstream flood control point and the probability of each factor.
8. An apparatus, comprising a memory and a processor, wherein:
a memory for storing a computer program capable of running on the processor;
a processor for executing the steps of the method for risk analysis of flood control dispatch of a reservoir group taking into account the uncertainty of dual model structure and hydrologic prediction according to any of claims 1-5 when running said computer program.
9. A storage medium having stored thereon a computer program for implementing the steps of the method for risk analysis of flood control dispatch of a reservoir group taking into account the uncertainty of dual model structure and hydrologic predictions as claimed in any one of claims 1-5 when executed by at least one processor.
CN202210597346.9A 2022-05-30 2022-05-30 Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty Pending CN114997636A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210597346.9A CN114997636A (en) 2022-05-30 2022-05-30 Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210597346.9A CN114997636A (en) 2022-05-30 2022-05-30 Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty

Publications (1)

Publication Number Publication Date
CN114997636A true CN114997636A (en) 2022-09-02

Family

ID=83029668

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210597346.9A Pending CN114997636A (en) 2022-05-30 2022-05-30 Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty

Country Status (1)

Country Link
CN (1) CN114997636A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116882696A (en) * 2023-07-18 2023-10-13 黄河水利委员会黄河水利科学研究院 Reservoir group flood control safety double-layer discrimination method based on coupling fuzzy recognition and probabilistic reasoning
CN117540173A (en) * 2024-01-09 2024-02-09 长江水利委员会水文局 Flood simulation uncertainty analysis method based on Bayesian joint probability model

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116882696A (en) * 2023-07-18 2023-10-13 黄河水利委员会黄河水利科学研究院 Reservoir group flood control safety double-layer discrimination method based on coupling fuzzy recognition and probabilistic reasoning
CN116882696B (en) * 2023-07-18 2024-03-08 黄河水利委员会黄河水利科学研究院 Reservoir group flood control safety double-layer discrimination method based on coupling fuzzy recognition and probabilistic reasoning
CN117540173A (en) * 2024-01-09 2024-02-09 长江水利委员会水文局 Flood simulation uncertainty analysis method based on Bayesian joint probability model
CN117540173B (en) * 2024-01-09 2024-04-19 长江水利委员会水文局 Flood simulation uncertainty analysis method based on Bayesian joint probability model

Similar Documents

Publication Publication Date Title
Liu et al. Remaining useful life estimation for proton exchange membrane fuel cells using a hybrid method
CN109902801B (en) Flood collective forecasting method based on variational reasoning Bayesian neural network
Quan et al. A survey of computational intelligence techniques for wind power uncertainty quantification in smart grids
Chang et al. An improved neural network-based approach for short-term wind speed and power forecast
Carpinone et al. Markov chain modeling for very-short-term wind power forecasting
CN114997636A (en) Reservoir group flood control scheduling risk analysis method considering model structure and hydrologic prediction double uncertainty
Lin et al. A non-linear rainfall-runoff model using radial basis function network
Shu et al. Regional flood frequency analysis at ungauged sites using the adaptive neuro-fuzzy inference system
Faber et al. Reservoir optimization using sampling SDP with ensemble streamflow prediction (ESP) forecasts
Nandalal et al. Dynamic programming based operation of reservoirs: applicability and limits
Soltani et al. Developing operating rules for reservoirs considering the water quality issues: Application of ANFIS-based surrogate models
Zeng et al. Antithetic method‐based particle swarm optimization for a queuing network problem with fuzzy data in concrete transportation systems
Mestav et al. State estimation for unobservable distribution systems via deep neural networks
Nourani et al. Estimation of Suspended Sediment Load Using Artificial Intelligence‐Based Ensemble Model
CN108711847A (en) A kind of short-term wind power forecast method based on coding and decoding shot and long term memory network
Côté et al. Stochastic optimisation of Hydro-Quebec hydropower installations: a statistical comparison between SDP and SSDP methods
Donyaii et al. Water reservoir multiobjective optimal operation using Grey Wolf optimizer
Zhou et al. Deep learning-based rolling horizon unit commitment under hybrid uncertainties
Xu et al. A water level prediction model based on ARIMA-RNN
Azamathulla et al. Prediction of scour depth at culvert outlets using gene-expression programming
Wei et al. Two-stage pumping control model for flood mitigation in inundated urban drainage basins
Cicconet et al. Moment-SOS relaxation of the medium term hydrothermal dispatch problem
Lange et al. Probabilistic day-ahead forecast of available thermal storage capacities in residential households
Herrera et al. Optimal control of batch processes using particle swam optimisation with stacked neural network models
Abdallah et al. Combining fuzzy logic and neural networks in modeling landfill gas production

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