Reservoir flood control scheduling scheme reservoir area submerging risk map construction method

Abstract

The invention discloses a reservoir flood control scheduling scheme reservoir area submerging risk map construction method. The method comprises the following steps: acquiring and sorting basic data; a reservoir area real-time water surface line simulation model with multiple reservoir entering points is constructed; generating combined working conditions of reservoir main and branch reservoir inflow and flood control dispatching control parameters; simulating a real-time water surface line of the reservoir area under each combined working condition and counting reservoir area submerging indexes; establishing a multivariate joint probability distribution function of reservoir main and branch reservoir inflow, flood control dispatching control parameters and reservoir area submerging indexes; and constructing a conditional probability transformation model, and making a reservoir area submerging risk map. According to the method, the relevance among the reservoir main and branch reservoir inflow, flood control scheduling scheme control parameters and reservoir area submerging indexes is fully utilized, the uncertainty of traditional reservoir area water surface line prediction is reduced, the time cost of complex hydrological hydrodynamic calculation is reduced, and the robustness and timeliness of reservoir real-time flood control scheduling decision making are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of reservoir scheduling risk assessment and relates to a method for constructing a reservoir area flooding risk map for a reservoir flood control scheduling plan. Background Art

[0002] Reservoirs that undertake downstream flood control tasks use the reserved flood control storage capacity in the reservoir area when floods come, and implement flood control scheduling to intercept and stagger flood peaks to ensure downstream flood control safety. In the reservoir design stage, according to the design flood standard for reservoir inundation treatment and the sediment deposition years, severe flood processes have been selected, and reservoir flood control calculations and backwater calculations have been carried out. Based on certain normal reservoir water level conditions, the wind and wave climbing value, ship wave influence, and the prediction results of bank collapse and landslide around the reservoir have been considered to determine the scope of reservoir inundation impact treatment, and the boundaries of reservoir area resident migration and land acquisition have been measured and demarcated. In the reservoir construction stage, the design results are used to organize the implementation of the relocation of rural and town settlements, industrial and mining enterprises within the inundation impact range, and the expropriation of farmland, gardens, forests, and grasslands to ensure the safe and effective performance of the reservoir flood control function. Therefore, under normal circumstances, the reservoir needs to reduce the water level in front of the dam to the flood control limit water level before the flood in the flood season. At this time, if the flood magnitude does not exceed the design flood standard for reservoir inundation treatment, there is generally no risk of inundation in the reservoir area.

[0003] Whether the reservoir area is flooded and to what extent during the flood control operation of the reservoir mainly depends on the combined working conditions of flood control operation control parameters such as the inflow, the reservoir discharge, and the water level in front of the dam. At present, under the dual pressure of climate change and human activities, the increase in extreme climate events and changes in underlying surface conditions have increased the probability that the inflow exceeds the land acquisition standard (such as once in five years) or even the resident migration standard (such as once in 20 years), and the reservoir area may face certain flooding losses; in addition, in the event of continuous compound floods, the reservoir continues to intercept floods, the reservoir water level continues to rise, and the water level in front of the dam is high when the next round of flood peaks occurs, which may also cause temporary flooding of the reservoir area. In the real-time operation process, the reservoir operation department needs to evaluate the impact of different flood control operation schemes on project safety, downstream flooding and reservoir flooding according to the inflow, make scientific and rapid decisions on the reservoir storage and release strategy in the face of the period, reasonably control the discharge flow and the water level in front of the dam, and take into account the upstream and downstream flood control pressures on the premise of ensuring the safety of the project, so as to avoid or reduce the loss of life and property of the people as much as possible.

[0004] The traditional method for assessing the flooding risk of reservoir flood control schemes is to construct a high-precision hydrological and hydrodynamic model of the reservoir area, simulate the change process of the water surface line of the reservoir area under different flood control schemes under assumed water conditions, compare it with the demarcated boundaries of resident migration and land acquisition, statistically analyze the flooding impact range and degree indicators, and estimate the flooding losses in the reservoir area. However, due to the uncertainty of rainfall, flood forecasts and the model itself, the accuracy of water surface line prediction is subject to certain constraints; at the same time, reservoirs generally have multiple inlet points, the relationship between the evolution of reservoir floods and the distribution along the way is complex, the calculation scale is large and the time is long, which affects the decision-making efficiency. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method for constructing a reservoir area flooding risk map for a reservoir flood control scheduling scheme. By offline acquiring a large number of numerical simulation experimental samples of reservoir water surface lines under different combinations of inflow flows and flood control scheduling scheme control parameters, and applying multidimensional joint distribution theory, a probability conversion model between reservoir inflow and outflow, flood control scheduling control parameters such as water level in front of the dam, and corresponding flooding indicators is constructed to evaluate the possibility and severity of reservoir area flooding caused by a preset flood control scheduling scheme under certain water inflow conditions, thereby constructing a reservoir area flooding risk map for a reservoir flood control scheduling scheme, so as to reduce the uncertainty of reservoir water surface line prediction, reduce the time cost of complex hydrological and hydrodynamic calculations, and improve the robustness and timeliness of real-time reservoir flood control scheduling decisions.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for constructing a reservoir flood control scheduling plan reservoir area flooding risk map, comprising the following steps: Step 1: Basic data collection and collation; Step 2, constructing a real-time water surface simulation model of a reservoir area with multiple inlet points; Step 3, generating the combined operating conditions of the inflow flow of the main and branch reservoirs and the flood control dispatching control parameters; Step 4, using the model constructed in step 2, traverse the real-time water surface lines of the reservoir area under each combination of working conditions generated in step 3, and calculate the flooding index of the reservoir area under each combination of working conditions; Step 5, establishing a multivariate joint probability distribution function of the inflow flow of the main and tributary reservoirs, flood control dispatching control parameters and reservoir area inundation index; Step 6: Construct a conditional probability conversion model between the inflow volume of the main and tributary reservoirs, flood control dispatch control parameters and reservoir area inundation indicators, and produce a reservoir area inundation risk map.

[0007] In step 1, the basic data include the characteristic water level of the reservoir, the water level-reservoir capacity curve, the discharge capacity curve, the cross-section topography along the river channel in the reservoir area and the mileage from the dam, the location and elevation of the land acquisition line boundary stakes, the location and elevation of the resident migration line boundary stakes, the statistics of the number of physical objects submerged at different elevations in the reservoir area and the minimum discharge flow reservoir characteristic data; the long series of synchronous flood data and frequency distribution statistical parameters of the control stations of the main and tributary inflows into the reservoir area, the water level and flow process of the control hydrological stations in the reservoir area for the actual floods measured since the construction of the reservoir, and the discharge flow of the corresponding reservoir flood control dispatching process, the hydrological and dispatching operation data of the water level in front of the dam.

[0008] In step 2, constructing a real-time water surface simulation model of a reservoir area with multiple inlet points includes the following steps: Step 2-1, collect S The sections along the river in the reservoir area are numbered from downstream to upstream, and the terrain and mileage data of each section are read to divide the river sections. Section 1 is generally the section where the dam is located, and Section 2 is the section where the dam is located. S Section No. is generally the section where the main inflow point is located; Step 2-2, establish a one-dimensional hydrodynamic model of the reservoir area, use the water level and flow process of the reservoir area control hydrological station measured in flood events since the reservoir was built and the corresponding reservoir flood control dispatch process and dam front water level data to determine the roughness of each section by river section; Step 2-3, initialize the water level and flow rate of each section. S The flow rate of section No. 1 is taken as the upper boundary, and the water level or flow rate of section No. 2 is taken as the lower boundary. The water level is calculated section by section to obtain the real-time water surface line of the reservoir area.

[0009] In step 2 and step 3, a one-dimensional steady flow or non-steady flow hydrodynamic model of the reservoir area is selected based on the characteristics of flood evolution, river channel topographic conditions, and available data.

[0010] In step 3, generating the combined operating conditions of the inflow flow of the main and branch reservoirs and the flood control dispatch control parameters includes the following steps: Step 3-1: Based on the synchronous flood data and frequency distribution statistical parameters of the main and branch inflow control stations in the reservoir area, the master-slave method is used to randomly generate N Combination of flow rates from main and branch banks into the reservoir ,in Main stream i Group inbound traffic, For the k Tributary i Group inbound traffic; Step 3-2: Increase the flood control limit water level of the reservoir Z FCL To high flood level Z FSL The water level value interval is based on the step size ΔZ Discretize and use Bootstrap resampling technology to generate N Water level before dam Z i ( i =1,2,..., N ), and for each group Z i , at the minimum discharge flow of the reservoir Water level before dam Z i Corresponding discharge capacity A set of downstream flows is randomly generated within the value range ; Step 3-3, N Inflow of main and branch reservoirs and flood control dispatch control parameter scheme Perform random combination without replacement and obtain N Combination conditions .

[0011] In step 3-2, in order to reduce sampling errors and to be as close as possible to flood control operation practice, .

[0012] In step 4, the reservoir area inundation index includes the land acquisition line inundation index and the resident migration line inundation index. By comparing the real-time water surface line of the reservoir area with the land acquisition line or the resident migration line, a set of indicators U={ U 1, U 2,..., U J}.

[0013] In step 5, the N The boundary conditions for calculating the water surface line in the reservoir area and the corresponding statistical samples of the inundation index are combined. Based on the maximum entropy principle, the probability distribution of the random variable joint information entropy reaching the maximum value is solved while keeping the main statistical characteristics of the random variable system unchanged, and the flow rate Q of the main and tributary inflows into the reservoir and the water level in front of the dam are used as the probability distribution. Z , Downflow The joint probability distribution density function of the random variable system composed of the reservoir area flooding index U ; The acquisition of the joint probability distribution density function of the random variable system is described as the following mathematical programming problem: Objective function: ; Where: is the joint information entropy of the random variable system X; f(x) is the joint probability distribution function of the random variable system; R is a Euclidean space with the same dimension as the number of system variables; Constraints: ; Where: g m (x) and EX [ g m (x)] are respectively m =1,2,..., M The function of a known random variable system X and the corresponding mathematical expectation value; g m (x) Using low-order origin moments, sample moments or moment product functions of each random variable to maintain the statistical characteristics of the random variable system expectation, variance, skewness coefficient and correlation; In order to ensure the normalization and standardization of the probability density function, the following constraints must also be met: ; By setting the number of constraints reasonably, the Lagrange multiplier method is used to solve the above constrained maximum optimization problem to estimate f (x) shape, scale and position parameters, and then get f (x).

[0014] In step 6, according to the predicted reservoir inflow size and the real-time water level in front of the dam during the period, a reservoir area flooding risk value map or a reservoir area flooding risk warning map is selected for construction; when the predicted reservoir inflow is large or the water level in front of the dam is high, and the reservoir area flooding risk is difficult to avoid, the reservoir area flooding risk value map is selected for construction, including the following steps: According to the downstream flood control demand during the period, the forecast value of the inflow flow of the main and tributary reservoirs Q=q and the preset flood control dispatch control parameters are given. Z = z , ,right Perform bounded multiple integration to obtain the conditional probability distribution density function of the reservoir area flooding index during the period ; Combining recognized standards and manual experience, the severity level of risk consequences corresponding to different reservoir flooding index thresholds is divided; the conditional probability distribution density function Perform continuous spatial multiple integration to obtain the cumulative probability of different reservoir flooding index thresholds and divide the risk occurrence possibility level; comprehensively consider the risk occurrence possibility and consequence severity level corresponding to different reservoir flooding index thresholds, calculate the risk value, and obtain the reservoir flooding risk value map; When the inflow is forecast to be small or the water level in front of the dam is not high, and the risk of reservoir flooding can be avoided or reduced, the construction of reservoir flooding warning map is selected, including the following steps: According to the reservoir area flooding risk control target, given the reservoir area flooding index U=u during the facing period, By performing bounded multiple integration, we can obtain the conditional probability distribution density function of the combined operating conditions of the main and branch reservoir inflow and flood control dispatching control parameters during the period. ; Based on the downstream flood control needs during the period, it is assumed that the reservoir discharge flow , calculate the flow rate of different main and branch reservoirs into the reservoir under the condition of Q=q The water level in front of the dam with a cumulative probability of not being lower than a certain acceptable risk level Z The integral upper bound of is taken as the critical warning water level to avoid or alleviate reservoir flooding, and the reservoir flooding risk warning map is obtained.

[0015] The main beneficial effects of the present invention are: The correlation between the inflow volume of the main and tributary reservoirs, the control parameters of the flood control scheduling scheme and the reservoir area inundation index is fully utilized, and a joint probability distribution and conditional probability conversion model is constructed, which is conducive to reducing the uncertainty of inputs, models, etc. in the process of calculating the reservoir area inundation index based on the inflow volume of the main and tributary reservoirs and the control parameters of the flood control scheduling scheme, and can reversely obtain the critical control conditions that can avoid or alleviate reservoir area flooding, thereby improving the robustness and flexibility of flood control scheduling decisions.

[0016] By acquiring a large number of real-time reservoir water surface simulation samples offline, which are computationally cumbersome and time-consuming, and constructing a reservoir flooding risk map to integrate and quickly provide reservoir flooding risk information under different combined working conditions, precious flood control scheduling decision-making time is saved.

[0017] The method for obtaining the joint probability distribution density function of a random variable system based on the maximum entropy principle does not need to assume a priori distribution in advance. It can optimize the parameters of the joint distribution as a whole by using only various statistical characteristics of the original simulation data, and has a strong statistical basis. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0019] Figure 1 It is a flow chart of the present invention.

[0020] Figure 2 It is a schematic diagram comparing the outer envelope line of the reservoir water surface line during the flood season with the land acquisition line and the resident migration line of the present invention.

[0021] Figure 3 It is a schematic diagram of the simulation results of the reservoir flood process of the present invention.

[0022] Figure 4 It is a schematic diagram of the reservoir area flooding risk value spectrum of the present invention.

[0023] Figure 5 It is a schematic diagram of the reservoir area flooding risk warning map of the present invention. DETAILED DESCRIPTION

[0024] like Figure 1~Figure 5 A method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme includes the following steps: Step 1: Collect and organize basic data, including characteristic water level of the reservoir, water level-reservoir capacity curve, discharge capacity curve, topography of sections along the river in the reservoir area and mileage from the dam, location and elevation of land acquisition line boundary stakes, location and elevation of resident migration line boundary stakes, statistics on the number of physical objects submerged at different elevations in the reservoir area, minimum discharge flow and other reservoir characteristic data; synchronous flood data and frequency distribution statistical parameters of the main series of control stations of the main and tributary inflows into the reservoir area, water level and flow process of the control hydrological stations in the reservoir area in the flood field measured since the construction of the reservoir, and the discharge flow of the entire process of flood control dispatch of the corresponding reservoir, water level in front of the dam and other hydrological and dispatching operation data.

[0025] Among them: the cross-sectional topography and mileage data along the river channel in the reservoir area refer to the latest measured cross-sectional position along the reservoir area, the distance from the starting point of each node in the cross-sectional area, the riverbed elevation, etc.; the statistical data on the number of physical objects submerged at different elevations in the reservoir area can be obtained by visiting the reservoir area on the spot after the construction of the reservoir, investigating and compiling, or by interpreting satellite and aerial photos of the reservoir area.

[0026] Step 2: construct a real-time water surface simulation model of a reservoir area with multiple inlet points. Specifically, this step includes the following sub-steps: Step 2.1, collect S The sections along the river in the reservoir area are numbered from downstream to upstream, and the terrain and mileage data of each section are read to divide the river sections. Section 1 is generally the section where the dam is located, and Section 2 is the section where the dam is located. S Section No. is generally the section where the main stream flows into the reservoir.

[0027] Step 2.2: divide the reservoir area into K The tributaries are connected to the river channel as concentrated lateral inflows in turn, and the hydraulic characteristics of the sections of each tributary entering the reservoir are generalized according to the parameters of the nearest sections upstream and downstream of the access point.

[0028] Step 2.3, establish a one-dimensional hydrodynamic model of the reservoir area, and use the water level and flow process of the reservoir area's controlling hydrological station measured in flood events since the reservoir was built, as well as the corresponding reservoir flood control dispatching process and dam-front water level data to determine the roughness of each section by river section.

[0029] Specifically, we can choose to establish a one-dimensional steady flow or non-steady flow hydrodynamic model for the reservoir area based on the characteristics of flood evolution, river topographic conditions, data availability, etc.

[0030] In order to establish a one-dimensional steady-flow hydrodynamic model of the reservoir area, the reservoir cross-section is usually used to represent the hydraulic characteristics along the process, the inflow and outflow flow process and the water level process in front of the dam are simplified, and the local head loss and flow velocity non-uniform coefficient are ignored. The reservoir hydrodynamic model is constructed based on the constant non-uniform gradually varying flow theory.

[0031] In order to establish a one-dimensional unsteady flow dynamics model in the reservoir area, the Preissmann four-point implicit difference format is usually used to discretize the Saint-Venant equations. The pursuit method is used to solve the equations, which can realize the continuous simulation of the water surface line in the reservoir area.

[0032] Step 2.4, initialize the water level and flow rate of each section. S The flow rate of section No. 1 is taken as the upper boundary, and the water level or flow rate of section No. 2 is taken as the lower boundary. The water level is calculated section by section to obtain the real-time water surface line of the reservoir area.

[0033] Step 3: Generate the combined operating conditions of the inflow flow of the main and branch reservoirs and the flood control dispatch control parameters. Specifically, this step includes the following sub-steps: Step 3.1: Based on the synchronous flood data and frequency distribution statistical parameters of the main and branch inflow control stations in the reservoir area, the master-slave method is used to randomly generate N Combination of flow rates from main and branch banks into the reservoir ,in Main stream i Group inbound traffic, For the k Tributary i Group inbound traffic.

[0034] Specifically, the master station-slave station method generally takes the main inflow reservoir control station with a large control basin area and a significant impact on the reservoir backwater as the master station, and the branch inflow reservoir stations as slave stations. First, the master station flood is randomly simulated, and then the slave station flood is indirectly generated from the master station flood by maintaining the main statistical characteristics of the slave station flood and the zero-order cross-correlation between the master and slave station floods in space.

[0035] Step 3.2: Increase the flood control limit water level of the reservoir Z FCL To high flood level Z FSL The water level value interval is based on the step size Δ Z Discretize and use Bootstrap resampling technology to generate N Water level before dam Z i ( i =1,2,..., N ), and for each group Z i , at the minimum discharge flow of the reservoir Water level before damZ i Corresponding discharge capacity A set of downstream flows is randomly generated within the value range .

[0036] Specifically, when the water level in front of the reservoir dam is [ Z FCL , Z FSL ] interval, the reservoir can play a role in flood control for the downstream through flood control scheduling; Bootstrap resampling technology is used to [ Z FCL , Z FSL ], the feasible values ​​of the dam front water level within the interval are sampled with equal probability and replacement to achieve random generation of the dam front water level; in order to reduce sampling errors and be as close to flood control scheduling practice as possible, it is required .

[0037] Step 3.3, N Inflow of main and branch reservoirs and flood control dispatch control parameter scheme Perform random combination without replacement and obtain N Combination conditions .

[0038] Step 4, using the model constructed in step 2, traverse the real-time water surface lines of the reservoir area under each combination of working conditions generated in step 3, and calculate the flooding index of the reservoir area under each combination of working conditions.

[0039] Specifically, the reservoir area inundation index includes the land acquisition line inundation index and the resident migration line inundation index. By comparing the real-time water surface line of the reservoir area with the land acquisition line or the resident migration line, combined with the statistics of the number of physical objects inundated at different elevations in the reservoir area, a set of indicators U={ U 1, U 2,..., U J}.

[0040] in: U j ( j =1,2,..., J ) includes but is not limited to water surface line calculation quantitative indicators such as inundation mileage, average inundation depth, maximum inundation superelevation under various combined working conditions, or physical quantitative indicators such as population affected by the inundation area, inundated cultivated land area, economic losses, etc.

[0041] Step 5: The data obtained from steps 3 and 4 above NThe boundary conditions for calculating the water surface line of the reservoir area and the corresponding statistical samples of the inundation index are combined. Based on the maximum entropy principle, the probability distribution of the random variable joint information entropy reaching the maximum value is solved while keeping the main statistical characteristics of the random variable system unchanged, as the inflow flow Q of the main and branch reservoirs and the water level in front of the dam in the present invention. Z , Downflow The joint probability distribution density function of the random variable system composed of the reservoir area flooding index U Specifically, the acquisition of the joint probability distribution density function of the random variable system can be described as the following mathematical programming problem: Objective function: ; Where: is the joint information entropy of the random variable system X; f (x) is the joint probability distribution function of the random variable system; R is the Euclidean space with the same dimension as the number of system variables.

[0042] Constraints: ; Where: g m (x) and EX [ g m (x)] are respectively m =1,2,..., M The function of a known random variable system X and the corresponding mathematical expectation value. g m (x) Use low-order origin moments, sample moments or moment product functions of each random variable to maintain the statistical characteristics of the random variable system expectation, variance, skewness coefficient, correlation, etc.

[0043] For example, in combination with the random variable system of the present invention, some typical constraints can be described as: ; ; ; ; In particular, in order to ensure the normalization and standardization of the probability density function, the following constraints must be met: .

[0044] By setting the number of constraints reasonably, the Lagrange multiplier method is used to solve the above constrained maximum optimization problem to estimate f (x) shape, scale and position parameters, and then get f (x).

[0045] Specifically, the Lagrange multiplier Description f (x) is in the following form: ,in .

[0046] Generally, the more constraints there are, the higher the accuracy of the joint distribution construction will be. However, the corresponding number of distribution parameters (Lagrange multipliers) that need to be estimated will also be greater, and the number of samples required to estimate the parameters will also be greater.

[0047] Step 6: Construct a conditional probability conversion model between the inflow volume of the main and tributary reservoirs, flood control dispatch control parameters and reservoir area inundation indicators, and produce a reservoir area inundation risk map.

[0048] Specifically, we can choose to construct a reservoir flooding risk value map or a reservoir flooding risk warning map based on the predicted reservoir inflow size and the real-time water level in front of the dam during the period. When the forecast inflow is large or the water level in front of the dam is high, and the risk of reservoir flooding is difficult to avoid, you can choose to build a reservoir flooding risk value map, including the following steps: according to the downstream flood control needs during the period, the forecast value of the inflow flow of the main and tributary reservoirs Q=q and the preset flood control dispatch control parameters are given Z = z , ,right Perform bounded multiple integration to obtain the conditional probability distribution density function of the reservoir area flooding index during the period ; Combining recognized standards and manual experience, the severity level of risk consequences corresponding to different reservoir flooding index thresholds is divided; the conditional probability distribution density function Continuous spatial multiple integration is performed to obtain the cumulative probability of different reservoir area flooding index thresholds and divide the risk possibility level; the risk possibility and consequence severity level corresponding to different reservoir area flooding index thresholds are comprehensively considered to calculate the risk value and obtain the reservoir area flooding risk value map.

[0049] For example, Table 1 lists a risk value level classification standard for reference.

[0050] Table 1 Risk value level standard

[0051] For example, Table 2 and Table 3 list a reference standard for grading the likelihood of risk occurrence and the severity of risk consequences.

[0052] Table 2 Risk probability level standards grade 1 2 3 4 5 Qualitative description Very low, almost impossible Low, unlikely to occur Occurrence High, may occur Very high, frequent Probability interval 0.00001～0.001 0.001～0.1 0.1～0.5 0.5～0.9 0.9～1.0 Table 3 Risk consequence severity level standards grade Population affected by flooding Flooded cultivated land area Economic losses 1 Less than 50,000 people Less than 100,000 mu Less than 5 million yuan 2 50,000-100,000 people 100,000 to 200,000 mu 5 million to 10 million yuan 3 100,000 to 200,000 people 200,000 to 500,000 mu 10 million to 50 million yuan 4 200,000 to 500,000 people 500,000 to 1 million mu 50 million to 100 million yuan 5 More than 500,000 people More than 1 million mu More than 100 million yuan When the predicted inflow is not large or the water level in front of the dam is not high, and the risk of reservoir flooding may be avoided or reduced, a reservoir flooding warning map can be constructed, including the following steps: according to the reservoir flooding risk control target, given the reservoir flooding index U=u during the period, By performing bounded multiple integration, we can obtain the conditional probability distribution density function of the combined operating conditions of the main and branch reservoir inflow and flood control dispatching control parameters during the period. ; Based on the downstream flood control needs during the period, it is assumed that the reservoir discharge flow , calculate the flow rate of different main and branch reservoirs into the reservoir under the condition of Q=q The water level in front of the dam with a cumulative probability of not being lower than a certain acceptable risk level Z The integral upper bound of is taken as the critical warning water level to avoid or alleviate reservoir flooding, and the reservoir flooding risk warning map is obtained.

[0053] Specifically, in real-time scheduling, based on the inflow forecast of the main and tributary reservoirs, if the water level in front of the dam is close to or higher than the corresponding critical warning water levels at all levels, an early warning must be issued in a timely manner and certain risk countermeasures must be taken, including close monitoring, taking measures to pre-lower the water level, and switching to flood control scheduling.

[0054] Taking a control reservoir A on the main stream of the Yangtze River as an example, a flood risk map of the reservoir area of ​​its flood control scheduling scheme is constructed. The specific implementation method is as follows: According to step 1, collect and organize the basic data of Reservoir A. The flood control limit water level of Reservoir A is 145m, the normal water storage level is 175m, the flood control high water level is 175m, and the flood control storage capacity is 22.15 billion m 3 ; The land acquisition standard is once in five years, and the backwater end is the DTB section; the resident migration standard is once in 20 years, and the backwater end is at DZT; CT is the main inflow control station, and the main tributary inflow control station is WL. The controlling hydrological stations along the reservoir area also include YJW, MD, WJT, CS, WG, ZJYZ, ST, QXC, ZX, WX, FJ, WS, BD, ZG, etc.

[0055] During the flood season of 2020, the A Reservoir experienced the largest flood inflow since its construction. The A Reservoir played an important role in flood control. The reservoir water level continued to rise, and the flood control situation in the reservoir area was once tense. Figure 2 This is a schematic diagram comparing the outer envelope of the reservoir water surface line, the land acquisition line, and the resident migration line of Reservoir A during the flood season in 2020. As can be seen from the figure, in some periods, the real-time water surface line of the reservoir area began to exceed the land acquisition line at the section above the QXC station, and some sections even slightly exceeded the resident migration line, causing temporary flooding of the reservoir area.

[0056] According to step 2, a real-time water surface simulation model of the reservoir area of ​​Reservoir A with multiple inlet points is constructed. Based on the latest measured 319 cross-section terrains from the main inlet station CT to the dam front, a one-dimensional non-steady flow dynamic model of the reservoir area with 1 main stream (the inlet control station is CT) and 1 main tributary (the inlet control station is WL), with a total of 2 inlet points, is constructed. A total of 6 measured flood processes during the flood season since 2012 and the corresponding flood control dispatch discharge and dam front water level change process are used as the upper and lower boundaries to calibrate the model parameters. Figure 3 This is a schematic diagram of the simulation results of the flood process at the QXC station in the reservoir area from July to August 2020. As can be seen from the figure, the water level process of the QXC station simulated by the constructed real-time water surface line model of the reservoir area can be well consistent with the measured water level process.

[0057] According to step 3, generate the combined operating conditions of the inflow flow of the main and branch waters of reservoir A and the flood control dispatch control parameters. N = 100,000 sets of main and branch inflow flow combinations ; with Δ Z =0.01m Discrete water level interval of reservoir A [145m,175m].

[0058] According to step 4, the real-time water surface line of reservoir A under each combination of working conditions is simulated, and the reservoir flooding index under each combination of working conditions is calculated. H As a representative flooding indicator, it refers to the maximum value of the difference between the backwater elevation of each section of any group of simulated reservoir water surface lines and the backwater elevation obtained by interpolating the line connecting the upstream and downstream boundary stakes closest to the land acquisition line or the resident migration line at the section.

[0059] Table 4 lists the maximum inundation height index of the reservoir water surface line relative to the resident migration line under some combined working conditions. It can be seen from Table 4 that the sensitive inundation river section in the reservoir area of ​​A is mainly located between the ST and DZT sections, and the maximum inundation height usually occurs in the CS section.

[0060] Table 4 Maximum inundation height index of the reservoir water surface line relative to the resident migration line under some combined working conditions

[0061] According to step 5, based on the principle of joint information entropy and maximum entropy, the inflow flow of the main stream CT of reservoir A, the main tributary WL, the water level in front of the dam, the discharge flow and the maximum inundation height are established. H The multivariate joint probability distribution function .

[0062] According to step 6, construct the inflow of the main stream CT of reservoir A, the main tributary WL, the water level in front of the dam, the discharge flow and the maximum inundation heightH The conditional probability conversion model between the two is used to produce the reservoir area flooding risk value map and the reservoir area flooding risk warning map.

[0063] In order to simplify and intuitively illustrate the construction effect of the flood risk value map of the reservoir area A, the embodiment of the present invention first fixes the facing period Q out = q out , Q WL = q WL Boundaries, focus on Q CT and Z These two factors have the most significant impact on the backwater of the reservoir area. Different working conditions are set q CT and z , respectively ; Combined with the mapping relationship between the maximum flooding super-elevation amplitudes of the CS section and the number of submerged objects at different elevations in the reservoir area, the severity of the flooding risk consequences corresponding to different maximum flooding super-elevation thresholds is divided; The cumulative probability corresponding to different maximum flooding super-high thresholds is calculated by integration, and the possibility of flooding risk is divided; the risk value is calculated to obtain the flooding risk value map of the reservoir area during the period faced by Reservoir A, such as Figure 4 shown.

[0064] In order to simplify and intuitively illustrate the construction effect of the flood risk warning map of the reservoir area of ​​A, the embodiment of the present invention only takes the control target of avoiding flooding of any section of the reservoir area during the period as an example, that is, given H =0, we get ; Also assuming different Q out = q out , Q WL = q WL , the trial is different q CT Under the conditions The water level in front of the dam with a cumulative probability of not being lower than a certain acceptable risk level Z The integral upper bound of is used as the critical warning water level to avoid or reduce the flooding of the reservoir area, and the flooding risk warning map of the reservoir area in the period A is obtained, as shown in Figure 5 shown.

[0065] The above analysis shows that by adopting the method described in the present invention, a large number of numerical simulation experimental samples of reservoir water surface lines can be obtained offline, and the correlation between the inflow flow of the main and tributary reservoirs, the control parameters of the flood control scheduling scheme and the reservoir inundation index can be fully utilized. The deterministic conversion process of calculating the real-time water surface line of the reservoir area based on the inflow and outflow flow of the reservoir and the water level boundary in front of the dam and then statistically analyzing the inundation index is replaced by a conditional probability conversion process based on multivariate joint probability distribution, which is beneficial to reducing the uncertainty of inputs, models, etc. in the calculation process, and can reversely obtain the critical conditions for controlling the risk of reservoir inundation, providing richer decision-making information, saving the time cost of complex hydrodynamic calculations, and improving the robustness and timeliness of flood control scheduling.

[0066] The above-mentioned embodiments are only preferred technical solutions of the present invention and should not be regarded as limitations of the present invention. The embodiments and features in the embodiments of the present application can be arbitrarily combined with each other without conflict. The protection scope of the present invention shall be the technical solutions recorded in the claims, including the equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme, characterized in that: The steps include: Step 1: basic data collection and collation; Step 2, constructing a real-time water surface simulation model of a reservoir area with multiple inlet points; Step 3, generating the combined operating conditions of the inflow flow of the main and branch reservoirs and the flood control dispatching control parameters; Step 4, using the model constructed in step 2, traverse the real-time water surface lines of the reservoir area under each combination of working conditions generated in step 3, and calculate the flooding index of the reservoir area under each combination of working conditions; Step 5, establishing a multivariate joint probability distribution function of the inflow flow of the main and tributary reservoirs, flood control dispatching control parameters and reservoir area inundation index; Step 6: Construct a conditional probability conversion model between the inflow volume of the main and tributary reservoirs, flood control dispatch control parameters and reservoir area inundation indicators, and produce a reservoir area inundation risk map.

2. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 1 is characterized in that: In step 1, the basic data include the characteristic water level of the reservoir, the water level-reservoir capacity curve, the discharge capacity curve, the cross-section topography along the river channel in the reservoir area and the mileage from the dam, the location and elevation of the land acquisition line boundary stakes, the location and elevation of the resident migration line boundary stakes, the statistics of the number of physical objects submerged at different elevations in the reservoir area and the minimum discharge flow reservoir characteristic data; the long series of synchronous flood data and frequency distribution statistical parameters of the control stations of the main and tributary inflows into the reservoir area, the water level and flow process of the control hydrological stations in the reservoir area for the actual floods measured since the construction of the reservoir, and the discharge flow of the corresponding reservoir flood control dispatching process, and the hydrological and dispatching operation data of the water level in front of the dam.

3. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 1 is characterized in that: In step 2, constructing a real-time water surface simulation model of a reservoir area with multiple inlet points includes the following steps: Step 2-1, collect S The sections along the river in the reservoir area are numbered from downstream to upstream, and the terrain and mileage data of each section are read to divide the river sections. Section 1 is generally the section where the dam is located, and Section 2 is the section where the dam is located. S Section No. is generally the section where the main inflow point is located; Step 2-2, establish a one-dimensional hydrodynamic model of the reservoir area, use the water level and flow process of the reservoir area control hydrological station measured in flood events since the reservoir was built and the corresponding reservoir flood control dispatch process and dam front water level data to determine the roughness of each section by river section; Step 2-3, initialize the water level and flow rate of each section. S The flow rate of section No. 1 is taken as the upper boundary, and the water level or flow rate of section No. 2 is taken as the lower boundary. The water level is calculated section by section to obtain the real-time water surface line of the reservoir area.

4. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 1 is characterized in that: In step 2 and step 3, according to the characteristics of flood evolution, river channel topography and available data, a one-dimensional steady flow or non-steady flow hydrodynamic model of the reservoir area is selected.

5. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 1 is characterized in that: In step 3, generating the combined operating conditions of the inflow flow of the main and branch reservoirs and the flood control dispatch control parameters includes the following steps: Step 3-1: Based on the synchronous flood data and frequency distribution statistical parameters of the main and branch inflow control stations in the reservoir area, the master-slave method is used to randomly generate N Combination of flow rates from main and branch banks into the reservoir ,in For the mainstream i Group inbound traffic, For the k Tributary i Group inbound traffic; Step 3-2: Increase the flood control limit water level of the reservoir Z FCL To high flood level Z FSL The water level value interval is based on the step size Δ Z Discretize and use Bootstrap resampling technology to generate N Water level before dam Z i ( i =1,2,..., N ), and for each group Z i , at the minimum discharge flow of the reservoir Water level before dam Z i Corresponding discharge capacity A set of downstream flows is randomly generated within the value range ; Step 3-3, N Inflow of main and branch reservoirs and flood control dispatch control parameter scheme Perform random combination without replacement and obtain N Combination conditions .

6. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 5 is characterized in that: In step 3-2, in order to reduce sampling errors and to be as close as possible to flood control dispatch practice, .

7. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 1 is characterized in that: In step 4, the reservoir area inundation index includes the land acquisition line inundation index and the resident migration line inundation index. By comparing the real-time water surface line of the reservoir area with the land acquisition line or the resident migration line, a set of indicators U={ U 1, U 2,..., U J }.

8. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 1 is characterized in that: In step 5, the N The boundary conditions for calculating the water surface line in the reservoir area and the corresponding statistical samples of the inundation index are combined. Based on the maximum entropy principle, the probability distribution of the random variable joint information entropy reaching the maximum value is solved while keeping the main statistical characteristics of the random variable system unchanged, and the flow rate Q of the main and tributary inflows into the reservoir and the water level in front of the dam are used as the probability distribution. Z , Downflow The joint probability distribution density function of the random variable system composed of the reservoir area flooding index U ; The acquisition of the joint probability distribution density function of the random variable system is described as the following mathematical programming problem: Objective function: ; Where: is the joint information entropy of the random variable system X; f (x) is the joint probability distribution function of the random variable system; R is a Euclidean space with the same dimension as the number of system variables; Constraints: ; Where: g m (x) and EX [ g m (x)] are respectively m =1,2,..., M The function of a known random variable system X and the corresponding mathematical expectation value; g m (x) Using low-order origin moments, sample moments or moment product functions of each random variable to maintain the statistical characteristics of the random variable system expectation, variance, skewness coefficient and correlation; In order to ensure the normalization and standardization of the probability density function, the following constraints must also be met: ； By setting the number of constraints reasonably, the Lagrange multiplier method is used to solve the above constrained maximum optimization problem to estimate f (x) shape, scale and position parameters, and then get f (x).

9. The method for constructing a reservoir flood risk map for a reservoir flood control scheduling scheme according to claim 1 is characterized in that: In step 6, according to the predicted reservoir inflow size and the real-time water level in front of the dam during the period, a reservoir area flooding risk value map or a reservoir area flooding risk warning map is selected for construction; when the predicted reservoir inflow is large or the water level in front of the dam is high, and the reservoir area flooding risk is difficult to avoid, the reservoir area flooding risk value map is selected for construction, including the following steps: According to the downstream flood control demand during the period, the forecast value of the inflow flow of the main and tributary reservoirs Q=q and the preset flood control dispatch control parameters are given. Z = z , ,right Perform bounded multiple integration to obtain the conditional probability distribution density function of the reservoir area flooding index during the period ; Combining recognized standards and manual experience, the severity level of risk consequences corresponding to different reservoir flooding index thresholds is divided; the conditional probability distribution density function Perform continuous spatial multiple integration to obtain the cumulative probability of different reservoir flooding index thresholds and divide the risk occurrence possibility level; comprehensively consider the risk occurrence possibility and consequence severity level corresponding to different reservoir flooding index thresholds, calculate the risk value, and obtain the reservoir flooding risk value map; When the inflow is forecast to be small or the water level in front of the dam is not high, and the risk of reservoir flooding can be avoided or reduced, the construction of reservoir flooding warning map is selected, including the following steps: According to the reservoir area flooding risk control target, given the reservoir area flooding index U=u during the facing period, By performing bounded multiple integration, we can obtain the conditional probability distribution density function of the combined operating conditions of the main and branch reservoir inflow and flood control dispatching control parameters during the period. ; Based on the downstream flood control needs during the period, it is assumed that the reservoir discharge flow , calculate the flow rate of different main and branch reservoirs into the reservoir under the condition of Q=q The water level in front of the dam with a cumulative probability of not being lower than a certain acceptable risk level Z The integral upper bound of is taken as the critical warning water level to avoid or alleviate reservoir flooding, and the reservoir flooding risk warning map is obtained.

Images