CN114004114A - Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration - Google Patents

Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration Download PDF

Info

Publication number
CN114004114A
CN114004114A CN202111661759.0A CN202111661759A CN114004114A CN 114004114 A CN114004114 A CN 114004114A CN 202111661759 A CN202111661759 A CN 202111661759A CN 114004114 A CN114004114 A CN 114004114A
Authority
CN
China
Prior art keywords
reservoir
dam
water
burst
break
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
CN202111661759.0A
Other languages
Chinese (zh)
Inventor
赖成光
陈佩琪
王兆礼
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.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
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 South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN202111661759.0A priority Critical patent/CN114004114A/en
Publication of CN114004114A publication Critical patent/CN114004114A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Architecture (AREA)
  • Civil Engineering (AREA)
  • Structural Engineering (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Alarm Systems (AREA)

Abstract

The invention discloses a rapid simulation method for reservoir dam break flood routing based on GPU parallel acceleration, and belongs to the field of flood simulation. The method comprises the following steps: collecting basic data of a reservoir and the downstream of the reservoir; preprocessing the terrain and land utilization data; selecting a working condition and corresponding parameters thereof; simulating by using a WCA2D model version calculated by a parallel GPU; analyzing the reliability and rationality of the calculation result; the results of the WCA2D model simulation were output and processed. Compared with the traditional flood numerical simulation model, the WCA2D model of the reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration is a cellular automaton model based on weight conversion, has high simulation efficiency, supports parallel GPU calculation, and can realize rapid simulation of dam-break flood evolution.

Description

Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration
Technical Field
The invention belongs to the field of data processing and flood prediction, and particularly discloses a reservoir dam break flood evolution rapid simulation method based on GPU parallel acceleration.
Background
The reservoir dam has the functions of flood control, irrigation, power generation and the like, and plays an important role in promoting the development of social economy. With the continuous acceleration and expansion of urbanization process, some reservoirs originally in suburbs gradually or already enter urban areas, and reservoir dams in urban areas with higher concentration of lives and properties are more important vigilant objects. Although a dam break is a low-frequency and high-risk disaster, once a dam breaks down, the dam will have a catastrophic effect on the downstream area, and therefore high attention must be paid to and corresponding preventive and treatment measures must be taken. Due to improper design or construction, the dam has dangerous hidden dangers, and dam break can be generated after the hidden dangers develop to a certain degree. Besides the quality problem, the dam is broken, and the dam is broken by the ineffectiveness factors such as earthquake, war, terrorist attack and the like and artificial destruction. In such a background, it is very important to simulate and predict the destructive influence and flooding condition caused by flood discharge after dam break, and to make appropriate protective measures and emergency evacuation plans for areas that may be flooded downstream.
Since the 50 s of the last century, with the continuous improvement of computer computing capability and the further development of numerical computing technology, numerical simulation gradually became the main means for studying dam-break flood evolution. The numerical simulation comprises three modes of one-dimensional, two-dimensional and three-dimensional, the two-dimensional simulation mode is mature nowadays, the two-dimensional simulation overcomes the defect that the one-dimensional simulation can only simulate a few points, and the numerical simulation has the advantages of capability of simulating a relatively real terrain, capability of calculating plane parameter characteristics, convenience in boundary processing, good visualization effect by combining technologies such as GIS and the like, convenience in visual display of simulation results, accuracy in calculation results and the like, and is widely applied. The evolution of reservoir dam break flood can be predicted by adopting a numerical simulation method, and the two-dimensional simulation can provide richer analysis on the submerging effect compared with the one-dimensional simulation; compared with three-dimensional simulation, the two-dimensional simulation requires less calculation time and has wider research range. Therefore, the two-dimensional simulation is a common numerical simulation method for the current dam-break flood routing. Tools for implementing two-dimensional simulation include MIKE21, InfoWorks ICM, LISFLOOD-FP models, SWM models, FloodMap, second order muslc models, etc., which generally simulate the evolution process of flood by solving complete or simplified shallow water equations, but direct calculation consumes a lot of time, is difficult to satisfy timeliness of dam break flood prediction and simulation, and the commercialization nature limits the potential for generalization and secondary development like MIKE21 and InfoWorks ICM. Different from the traditional two-dimensional hydrodynamic Model, the WCA2D (Weighted Cellular Automata 2D intersection Model) Model is developed by Michele Guidolin and Albert S, Chen and the like, is a Cellular automaton Model based on weight conversion rules, has the capability of simulating a two-dimensional hydrodynamic process although inertia terms and momentum conservation are not considered, and has high simulation efficiency; the model can be used to process a variety of meshes (e.g., rectangular meshes, hexagonal meshes, or triangular meshes) and different cell neighborhood types (e.g., von-noemann neighborhoods for four cells or morgan neighborhoods for eight cells), and the GPU parallel acceleration technique of the model has also been breakthrough. Before reservoir burst such sudden disasters, the time is life, and the calculation efficiency of numerical simulation is a key factor influencing disaster defense decision.
In view of this, it is theoretically feasible to develop numerical simulation on the evolution of reservoir dam-break flood by adopting a fully open-source WCA2D model, and the timeliness and popularization requirements can be met. However, at present, no relevant technology for rapidly simulating reservoir dam-break flood routing by applying the WCA2D model is reported, so that the method for rapidly simulating reservoir dam-break flood routing based on GPU parallel acceleration is provided on the basis of the model, and is very urgent.
Disclosure of Invention
The invention aims to provide a method for realizing a two-dimensional rapid simulation method of reservoir dam break flood routing, and provides a tool which is simple to operate, high in calculation precision, high in calculation efficiency and good in visualization effect for urban dam break flood routing simulation.
The invention is realized by at least one of the following technical schemes.
A reservoir dam break flood evolution rapid simulation method based on GPU parallel acceleration comprises the following steps:
s1, collecting basic data of the reservoir and the downstream,
s2, preprocessing the terrain and land utilization data, and establishing a digital elevation model;
s3, selecting a working condition, and calculating a break opening flow process according to the reservoir capacity, the break water depth, the break opening width and the break mode corresponding to the working condition;
s4, inputting the terrain and flow process data files processed in the step S2 and the step S3 into a WCA2D model which is parallelly calculated by using a graphic manager (GPU), setting simulation parameters of the WCA2D model and then simulating;
s5, carrying out water balance and flow state analysis on the simulation result, verifying, calibrating and optimizing a numerical model, and carrying out comparative analysis on the result of the water balance and flow state analysis and an empirical value to calculate the reliability and the rationality of the result;
and S6, outputting a WCA2D model simulation result, and processing the result based on a space analysis tool of a Geographic Information System (GIS) and an R language to realize the visualization of the dam break flood evolution process.
Preferably, the basic data is not limited to reservoir capacity curves, reservoir dam related data, topographic data and land utilization data;
preferably, the output simulation result comprises the flood flow speed, the submergence water depth and the submergence range of the downstream time-sharing after the dam break and the maximum water depth of the downstream submergence range in the whole flood evolution process.
Preferably, the preprocessing is not limited to correcting terrain data and superimposing building and river data.
Preferably, step S2 includes building a digital elevation model including the shape and actual elevation of the building based on the geographic information system and combining the land utilization data and the satellite map to the terrain of the building, river, street and highway, so as to simulate the delay effect of the building on dam-breaking flood and the water-inrush effect due to the space and gap between the buildings.
Preferably, the reservoir breaking mode comprises two dam breaking modes, namely an instant dam breaking mode and a gradual dam breaking mode.
Preferably, the instant burst is divided into instant full burst and instant partial burst according to different burst widths so as to simulate the reservoir burst flood evolution process caused by the accident of strike and earthquake;
the gradual burst simulates the development process of a burst opening from small to large.
Preferably, step S3 includes the steps of:
s31, selecting corresponding parameters including a reservoir burst mode, a burst opening width, a burst water depth and a corresponding reservoir capacity according to the working condition to be simulated;
s32, calculating the maximum burst flow, wherein the maximum burst flow of the instant burst is as follows:
Figure 782192DEST_PATH_IMAGE001
in the formulaQ max、g、B 、b m 、H 0Respectively representing the maximum dam bursting flow, the gravity acceleration, the dam length, the final burst width and the burst water depth;
the dam which is gradually burst firstly breaks through osmotic deformation and then collapses instantly, and the flow of the osmotic deformation and the breakage is as follows:
Figure 577629DEST_PATH_IMAGE002
in the formula:Hcalculating a water level elevation for the reservoir water level, i.e. the reservoir;Athe cross-sectional area of the pipe through which the water flows;H P is the elevation of the central line of the pipeline;f depending on the D50 particle size, the darcy coefficient of friction, which can be calculated from the Moody curve,L the length of the pipeline along the water flow direction;Dis the diameter or width of the pipe;
s33 flood flow process of breach
Suppose that the maximum burst flow of the dam isQ maxReservoir capacity with known burst flood volumeWDuration of floodT n When it is required, thent Time of day flowQComprises the following steps:
Figure 542653DEST_PATH_IMAGE003
。
preferably, the WCA2D model is used for simulating surface water flow, and the main process comprises:
(1) acquiring the weight of water flow exchange of adjacent cells:
Figure 550579DEST_PATH_IMAGE004
Figure 738853DEST_PATH_IMAGE005
Figure 959051DEST_PATH_IMAGE006
in the formula:mrepresents the total number of cells adjacent to the central cell;
Figure 492930DEST_PATH_IMAGE007
denotes the central cell of number 0 andithe water level difference of each adjacent cellular;
Figure 898022DEST_PATH_IMAGE008
representing the corresponding volume difference;
Figure 731681DEST_PATH_IMAGE009
indicating that the water level difference is greater than the critical value
Figure 868919DEST_PATH_IMAGE010
Calculating the water quantity obtained by downstream cells;
Figure 480640DEST_PATH_IMAGE011
is shown asiWeights of adjacent cells;
Figure 361527DEST_PATH_IMAGE012
represents the weight of the central cell;
Figure 258814DEST_PATH_IMAGE013
represents the sum of the volume differences adjacent to the central cell;
Figure 820814DEST_PATH_IMAGE014
is shown and inThe minimum volume difference of the adjacent cells of the heart cell;
(2) acquiring the amount of intercellular interaction water:
Figure 173692DEST_PATH_IMAGE015
in the formula:
Figure 338830DEST_PATH_IMAGE016
to represent
Figure 362062DEST_PATH_IMAGE017
At a time downstream ofiThe exchange water quantity obtained by each cell;
Figure 298226DEST_PATH_IMAGE018
is shown in
Figure 403979DEST_PATH_IMAGE017
The amount of water transferred from the central cell at that time;
Figure 576070DEST_PATH_IMAGE019
represents the central cell area;
Figure 76977DEST_PATH_IMAGE020
representing the central cellular water depth;
Figure 222525DEST_PATH_IMAGE021
representing a weight;
Figure 305320DEST_PATH_IMAGE022
representing the difference between the water level of the central cell and the water level of the cell;
Figure 346349DEST_PATH_IMAGE023
representing the distance from the center cell midpoint to the cell midpoint;
Figure 373867DEST_PATH_IMAGE024
representing the maximum allowable flow rate of the water flow of the central cell transferred to the cell;
Figure 182992DEST_PATH_IMAGE025
represents the time step, s;
Figure 791566DEST_PATH_IMAGE026
indicating the boundary length of the cell and the central cell;
Figure 90479DEST_PATH_IMAGE027
to representtThe amount of water transferred from the central cell at that time;nis the Manning coefficient; g is the acceleration of gravity; subscript isMAll variables of (1) are for the cell with the greatest weight; all variables without superscripts are assumed at time t;
(3) calculating the water depth:
Figure 686019DEST_PATH_IMAGE028
in the formula:
Figure 677764DEST_PATH_IMAGE029
to represent
Figure 705458DEST_PATH_IMAGE017
The water depth of the central cell at all times;
Figure 808062DEST_PATH_IMAGE030
to representtThe water depth of the central cell at the moment;
Figure 70608DEST_PATH_IMAGE031
representing the amount of influent water obtained by the central cells;
Figure 498834DEST_PATH_IMAGE032
represents the amount of water flowing out of the central cell;
(4) the flow rate calculation formula is:
Figure 550842DEST_PATH_IMAGE033
in the formula:
Figure 710997DEST_PATH_IMAGE034
to represent
Figure 534990DEST_PATH_IMAGE017
The water flow of the cell at the time center is transferred to the firstiThe flow rate of the individual cells;
Figure 349098DEST_PATH_IMAGE035
to represent
Figure 425419DEST_PATH_IMAGE017
Time center cell andithe arithmetic mean of the water depth of the individual cells;
Figure 858107DEST_PATH_IMAGE036
is shown asiThe boundary length between each unit cell and the central unit cell.
Preferably, the WCA2D model is optimized by adopting an OpenCL heterogeneous algorithm, and the WCA2D model of a GPU accelerated version is obtained, so that the simulation efficiency of reservoir dam break flood routing is improved.
Preferably, the basic data is not limited to reservoir capacity curves, reservoir dam related data, topographic data and land utilization data.
Preferably, the simulation parameters include initial conditions, boundary conditions and roughness, step size, iteration number and simulation time.
Based on the ARGGIS platform, the visualization of flood routing is realized, a large amount of calculation result data are expressed in a dynamic demonstration form of dam break flood, and the visual understanding of the dynamic characteristics of the water flow from the perspective is facilitated. And the satellite map is superposed to visually and clearly present the submerging range and submerging depth of the time-sharing downstream area.
Compared with the prior art, the invention has the beneficial effects that:
the WCA2D model is a cellular automaton model based on weight conversion, the simulation efficiency is high, and the defect that the traditional two-dimensional hydrodynamic model is low in operation efficiency is overcome. In addition, the WCA2D model can realize the function of parallel GPU calculation, and the simulation efficiency can be further improved. The downstream time-sharing submerging depth, submerging range and water velocity result output by the model can be well visualized on a GIS platform, post-processing can be conveniently carried out on data, and the method has great significance for formulating an emergency plan and reducing loss. The model is a fully open source, and is beneficial to popularization and application and secondary development. The visualization of flood routing can be realized based on an ARCGIS platform, the post-processing of data is convenient, and the important significance in formulating an emergency plan and reducing loss is realized.
Drawings
FIG. 1 is a flow chart of a reservoir dam break flood evolution rapid simulation method based on GPU parallel acceleration in an embodiment of the invention;
fig. 2 is a flood flow process line diagram according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of maximum water depth of a LISFLOOD model, a WCA2D model, a WCA2D parallel CPU acceleration model, and a WCA2D parallel GPU acceleration model under a working condition 1 according to the embodiment of the present invention;
FIG. 4 is a schematic diagram of maximum water depth of a LISFLOOD model, a WCA2D model, a WCA2D parallel CPU acceleration model, and a WCA2D parallel GPU acceleration model under a working condition 2 according to the embodiment of the present invention;
FIG. 5 is a schematic diagram of maximum water depth of a LISFLOOD model, a WCA2D model, a WCA2D parallel CPU acceleration model, and a WCA2D parallel GPU acceleration model under a working condition 3 according to the embodiment of the present invention;
fig. 6 is a diagram of a second flood flow process line according to an embodiment of the present invention;
FIG. 7 is a schematic diagram of the maximum water depth of a second WCA2D parallel GPU acceleration model according to an embodiment of the present invention;
fig. 8 is a diagram of a flow process of a flood according to a third embodiment of the present invention;
FIG. 9 is a schematic diagram of the maximum water depth of three WCA2D parallel GPU acceleration models according to an embodiment of the present invention;
fig. 10 is a graph illustrating a quadplow process profile according to an embodiment of the present invention;
fig. 11 is a schematic diagram of the maximum water depth of four WCA2D parallel GPU acceleration models according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the disclosure, the invention will be described in further detail with reference to the accompanying drawings and specific embodiments. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the first embodiment, a river-head reservoir in Lianjiang, Guangdong province is taken as an example.
As shown in fig. 1, the invention discloses a reservoir dam break flood evolution rapid simulation method based on GPU parallel acceleration, comprising the following steps:
and S1, collecting basic data of the reservoir and the downstream, including but not limited to reservoir capacity curves, reservoir dam related data, topographic data and land utilization data.
In the specific embodiment, a river-head reservoir in Lianjiang city, Guangdong province is taken as an example, the dam length of the reservoir is 175m, the dam bottom elevation is 21m, the normal water storage level is 42.65m, and the corresponding reservoir capacity is 825 ten thousand m3Check flood level 41.92m, corresponding reservoir capacity 1172 km3The overtopping water level is 42.65m, and the corresponding storage capacity is 1215 ten thousand m3。
And S2, preprocessing the terrain and land utilization data, including but not limited to correcting the terrain data, superposing the building and river data, and converting the data into a format capable of being used by the model. Based on the ARCGIS, the land utilization data, satellite maps and the like are combined to accurately process original Digital Elevation Models (DEMs) on terrains such as buildings, riverways, streets, highways and the like, and the processed DEMs can accurately express the shapes and actual elevations of the buildings so as to accurately simulate the delaying effect of the buildings on dam break flood and the water inrush effect caused by spaces and gaps among the buildings.
And S3, selecting a working condition, and calculating the flow process of the burst opening through an empirical formula according to the reservoir capacity, the burst water depth, the burst opening width and the burst mode corresponding to the working condition.
S31, selecting corresponding parameters according to the condition to be simulated, wherein the parameters comprise reservoir bursting modes (including two modes of dam instant bursting and gradual bursting), burst opening width, burst water depth and corresponding reservoir storage capacity.
The reservoir bursting mode simultaneously considers two dam bursting modes of instant bursting and gradual bursting of the reservoir. The instantaneous burst can be divided into instantaneous total burst or instantaneous partial burst according to different burst widths, can simulate the reservoir burst flood evolution process caused by terrorist attack, earthquake and other emergencies, and is the simulation basis of most of dam break calculation at present. The model for gradually breaking the dam break mouth can truly simulate the development process of the break mouth from small to large, and is consistent with the actual condition of most dam breaks. Under the condition of gradual collapse, the dam is firstly damaged by osmotic deformation, and then collapses instantly when the fracture surface of the osmotic deformation is developed to a certain size, namely, the dam is firstly gradually developed into a large hole from a small hole and then collapses instantly.
In this embodiment, three working conditions are selected as shown in table 1, the width of the break opening is set to 175m, that is, the whole dam is completely broken, and the dam breaking mode is selected to be instantaneous full break. The three working conditions respectively adopt a normal water storage level, a check water level, an overtopping water level and corresponding storage capacity.
Figure 536606DEST_PATH_IMAGE037
And S32, calculating the maximum burst flow. The calculation mode of the maximum burst flow of the instant burst is as follows:
Figure 849512DEST_PATH_IMAGE001
in the formulaQ max 、g、B 、b m 、H 0Respectively represents the maximum flow (m) of dam break3(s), gravitational acceleration (m/s)2) Dam length (m), final breach width (m), and breach depth (m).
S33 flood flow process of breach
Let the flood peak be Qmax (m)3/s), reservoir capacity W (m) with known flood volume3) Flood duration tn(s) is to be solved, then the flow Q at time t is expressed by:
Figure 177244DEST_PATH_IMAGE038
finally, the process lines of the flow of the burst opening under the three working conditions are obtained and are shown in figure 2.
And S4, inputting data files such as processed terrain, flow process and the like in the WCA2D model version calculated by using the parallel GPU, setting initial conditions, boundary conditions and other simulation parameters of the model, and then simulating.
The WCA2D (Weighted Cellular Automata 2D initialization Model) Model is a Cellular automaton Model based on weight transformation rules. Different from the traditional two-dimensional simulation model, the model does not consider inertia terms and momentum conservation, so that the efficiency of simulating the two-dimensional hydrodynamic process is high. The model can be used to handle a variety of meshes and different cell neighborhood types.
The main calculation process of simulating the surface water flow by the WCA2D comprises the following steps: firstly, calculating the weight of water flow exchange of adjacent cells; calculating the amount of water interacted among the cells; thirdly, calculating the water depth; and fourthly, calculating the flow rate. The formula is as follows, and all variables without superscript are assumed at time t:
(1) the weight calculation formula is:
Figure 803836DEST_PATH_IMAGE039
Figure 838306DEST_PATH_IMAGE040
in the formula: m represents the total number of cells adjacent to the central cell;
Figure 274842DEST_PATH_IMAGE007
represents the water level difference value m between the central cellular (number 0) and the ith adjacent cellular;
Figure 773692DEST_PATH_IMAGE008
representing the corresponding volume difference, m;
Figure 386333DEST_PATH_IMAGE009
indicating a large water level differenceAt a critical value
Figure 653804DEST_PATH_IMAGE010
Calculating the water quantity obtained by downstream cells;
Figure 516368DEST_PATH_IMAGE011
representing the weight of the ith adjacent cell;
Figure 423886DEST_PATH_IMAGE012
representing the weight of the central cell.
(2) The amount of intercellular interaction water:
Figure 935157DEST_PATH_IMAGE041
Figure 463660DEST_PATH_IMAGE042
in the formula:
Figure 465768DEST_PATH_IMAGE016
to represent
Figure 40407DEST_PATH_IMAGE017
C, carrying out m-shaped cultivation on the exchange water quantity obtained by the ith cell downstream,
Figure 965156DEST_PATH_IMAGE018
is shown in
Figure 371603DEST_PATH_IMAGE017
C, carrying out m-year cultivation on the water quantity transferred from the central cells at the moment;
Figure 61122DEST_PATH_IMAGE019
represents the central cell area, m 2;
Figure 591899DEST_PATH_IMAGE020
represents the central cellular water depth, m;
Figure 444973DEST_PATH_IMAGE021
representing a weight;
Figure 502664DEST_PATH_IMAGE022
representing the difference between the water level of the central cell and the water level of the cell;
Figure 251832DEST_PATH_IMAGE023
represents the distance, m, from the center cell midpoint to the cell midpoint;
Figure 597801DEST_PATH_IMAGE024
represents the maximum allowable flow velocity, m/s, of the water flow of the central cell transferred to the cell;
Figure 661091DEST_PATH_IMAGE025
represents the time step, s;
Figure 620209DEST_PATH_IMAGE026
represents the boundary length of the cell and the central cell, m;
Figure 150066DEST_PATH_IMAGE027
representing the amount of water transferred from the central cell at time t, m; n is a Manning coefficient; g is the acceleration of gravity. All variables indexed by M are for the cell with the greatest weight.
(3) The water depth is calculated by the formula
Figure 622812DEST_PATH_IMAGE043
（5）
In the formula:
Figure 427475DEST_PATH_IMAGE029
to represent
Figure 615749DEST_PATH_IMAGE017
The water depth of the central cell at all times;
Figure 859384DEST_PATH_IMAGE030
representing the water depth of the central cell at the time t;
Figure 4714DEST_PATH_IMAGE031
representing the amount of influent water harvested by the central cell, such as rainfall, m;
Figure 365464DEST_PATH_IMAGE032
indicating the amount of water shed by the central cell, e.g., water infiltration, m.
(4) The flow velocity is calculated by the formula
Figure 647059DEST_PATH_IMAGE044
In the formula:
Figure 359799DEST_PATH_IMAGE034
to represent
Figure 726845DEST_PATH_IMAGE017
The flow velocity of the central cellular water flow transferred to the ith cellular is m/s;
Figure 861592DEST_PATH_IMAGE035
to represent
Figure 824125DEST_PATH_IMAGE017
The arithmetic mean value m of the water depths of the central cell and the ith cell at the moment;
Figure 940722DEST_PATH_IMAGE036
denotes the boundary length, m, of the ith cell and the center cell.
In this case, the WCA2D model is optimized by adopting an OpenCL heterogeneous algorithm, and the WCA2D model of a GPU accelerated version is obtained, so that the simulation efficiency of reservoir dam break flood routing is improved. OpenCL can implement communication and data transfer between heterogeneous systems (generally, cpu and GPU (graphics card)), can open up, release, and initialize memory space in a device, and can implement data transmission between a host and the device.
In order to deeply analyze the simulation efficiency of the WCA2D model based on GPU acceleration in reservoir dam break flood evolution, the simulation efficiencies of the WCA2D model based on OpenMP algorithm (CPU acceleration), the WCA2D model without acceleration and the two-dimensional hydrodynamic model LISFLOOD are compared.
And S5, carrying out water balance and flow state analysis on the numerical simulation result, comparing the result with a plurality of models, verifying, rating and optimizing the numerical models, and analyzing and calculating the reliability and rationality of the result.
In this embodiment, the results of comparing the GPU parallel acceleration version and the CPU parallel acceleration version of WCA2D, the normal version, and the two-dimensional hydrodynamic model lisflo version (with CPU acceleration) are mutually verified, and the verification results are shown in fig. 3a-d, fig. 4a-d, and fig. 5 a-d. The results show that the results of the four models are similar in the distribution of the flooding ranges and flooding depths (fig. 3a-d, fig. 4a-d, fig. 5a-d, table 2), and can be mutually corroborated. The running time of the four models is shown in table 3 (the used devices are configured as intel (r) core (tm) i7-8700 CPU), before the GPU is accelerated in parallel, the simulation efficiency of the WCA2D is worse than that of the LISFLOOD, but after the GPU is accelerated, the efficiency is greatly improved, and the running speed is faster than that of the LISFLOOD.
Table 2 LISFLOOD model, CA model and CA acceleration model, three working condition water depth area distribution (km)
Figure 290670DEST_PATH_IMAGE045
TABLE 3 simulation duration (seconds) for each condition under three models
Figure 771985DEST_PATH_IMAGE046
And S6, outputting and processing results of WCA2D model simulation, wherein the results comprise flood flow speed, submerging depth and submerging range of downstream time-sharing after dam break and the maximum depth of downstream submerging in the maximum range in the whole flood evolution process.
Based on the GIS platform, the visualization of flood routing is realized, a large amount of calculation result data are expressed in a dynamic demonstration form of dam break flood, and the visual understanding of the dynamic characteristics of the water flow from the perspective is facilitated. And the satellite map is superposed to visually and clearly present the submerging range and submerging depth of the time-sharing downstream area.
Example two, take the long and green reservoir in Lianjiang province in Guangdong province as an example. The steps in this embodiment are substantially the same as those in the first embodiment, except that:
and S1, collecting basic data of the reservoir and the downstream, including but not limited to reservoir capacity curves, reservoir dam related data, topographic data and land utilization data.
In the specific embodiment, the long and green reservoir in the city of cheap river in Guangdong province is taken as an example, the top elevation of the reservoir dam below the ridge north of the main reservoir is 50m, the check water level is 48.92m, the check reservoir capacity is 12450 ten thousand, the flood limit water level is 45m, the flood limit reservoir capacity is 7100 ten thousand, and the dam length is 525 m.
And S3, selecting a working condition, and calculating the flow process of the burst opening through an empirical formula according to the reservoir capacity, the burst water depth, the burst opening width and the burst mode corresponding to the working condition.
S31, selecting corresponding parameters according to the condition to be simulated, wherein the parameters comprise reservoir bursting modes (including two modes of dam instant bursting and gradual bursting), burst opening width, burst water depth and corresponding reservoir storage capacity.
In the present embodiment, one working condition is selected, and the burst mode is instantaneous full burst. Wherein the overtopping water level is 50.5m, the burst width of the burst opening is 414m, and the burst opening is trapezoidal when bursting to the bottom. The breach flow process is shown in FIG. 6.
Fig. 7 shows the maximum water depth and range of the flooded area in the disaster area according to the second embodiment.
Example three, take the martial reservoir of liaojiang, guangdong province as an example. The steps in this embodiment are substantially the same as those in the first embodiment, except that:
and S1, collecting basic data of the reservoir and the downstream, including but not limited to reservoir capacity curves, reservoir dam related data, topographic data and land utilization data.
In the specific embodiment, a martial reservoir in Lianjiang, Guangdong province is taken as an example, the top elevation of the reservoir dam is 42m, the check water level is 41.26m, the check reservoir capacity is 9930 ten thousand squares, the flood limit water level is 38m, the flood limit reservoir capacity is 6380 ten thousand squares, and the dam length is 150 m.
And S3, selecting a working condition, and calculating the flow process of the burst opening through an empirical formula according to the reservoir capacity, the burst water depth, the burst opening width and the burst mode corresponding to the working condition.
S31, selecting corresponding parameters according to the condition to be simulated, wherein the parameters comprise reservoir bursting modes (including two modes of dam instant bursting and gradual bursting), burst opening width, burst water depth and corresponding reservoir storage capacity.
In the embodiment, one working condition is selected, and the dam breaking mode is to break the dam gradually and break the dam instantly. Wherein the dam is gradually broken, the osmotic deformation damage is rectangular, the initial osmotic deformation damage length is 0.1m, the check flood level is 42.60m, the osmotic deformation damage height is 30.95m, the break mouth is linearly developed, and the break width is 75 m. The breach flow process is shown in FIG. 8.
The maximum flooding water depth and range of the flooded area of the disaster area in the third embodiment are shown in fig. 9.
Example four, a dragon cave reservoir, guangzhou, guangdong province, is used as an example. The steps in this embodiment are substantially the same as those in the first embodiment, except that:
and S1, collecting basic data of the reservoir and the downstream, including but not limited to reservoir capacity curves, reservoir dam related data, topographic data and land utilization data.
The present embodiment takes a dragon cave reservoir in Guangzhou city, Guangdong province as an example. Rain collecting area of reservoir is 6.36km2The normal water storage level is 64.86m (the design flood level is 64.90m in hundred years on the basis of beads, and the check flood level is 66.10m in one thousand years), and the normal storage capacity is 237 km3Total storage capacity of 250 km3. The dam length of the reservoir is 164m, the dam top elevation is 67.00m, the height of the wave wall is 0.5m, and the maximum dam height is 22.00 m.
And S3, selecting a working condition, and calculating the flow process of the burst opening through an empirical formula according to the reservoir capacity, the burst water depth, the burst opening width and the burst mode corresponding to the working condition.
S31, selecting corresponding parameters according to the condition to be simulated, wherein the parameters comprise reservoir bursting modes (including two modes of dam instant bursting and gradual bursting), burst opening width, burst water depth and corresponding reservoir storage capacity.
In the specific embodiment, the main dam is instantaneously burst due to emergencies such as earthquake, landslide and the like when the check water level of the longhole reservoir is 66.1m, the width of the burst opening is 164m of the length of the dam, and the reservoir capacity is 250 ten thousand m3. The breach flow process is shown in FIG. 10.
The maximum flooding water depth and range of the flooded area of the disaster area in the fourth embodiment are shown in fig. 11.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims (10)

1. The rapid simulation method for reservoir dam break flood routing based on GPU parallel acceleration is characterized by comprising the following steps:
s1, collecting basic data of the reservoir and the downstream of the reservoir;
s2, preprocessing the terrain and land utilization data, and establishing a digital elevation model;
s3, selecting a working condition, and calculating a break opening flow process according to the reservoir capacity, the break water depth, the break opening width and the break mode corresponding to the working condition;
s4, inputting the terrain and flow process data files processed in the step S2 and the step S3 into the WCA2D model which is calculated in parallel by using the graphic manager, setting simulation parameters of the WCA2D model and then simulating;
s5, carrying out water balance and flow state analysis on the simulation result, verifying, calibrating and optimizing a numerical model, and carrying out comparative analysis on the result of the water balance and flow state analysis and an empirical value to calculate the reliability and the rationality of the result;
and S6, outputting a WCA2D model simulation result, and processing the result based on a space analysis tool of a geographic information system and an R language to realize the visualization of the dam break flood evolution process.
2. The rapid simulation method for reservoir dam break flood routing based on GPU parallel acceleration as claimed in claim 1, wherein the preprocessing comprises correction of terrain data and superposition of building and river data.
3. The rapid simulation method for dam break flood routing of reservoir based on GPU parallel acceleration as claimed in claim 1, wherein step S2 comprises building a digital elevation model including the shape and actual elevation of the building based on the geographic information system and combining the land utilization data and the satellite map to the terrain of the building, river, street, and highway, so as to simulate the delay effect of the building on dam break flood and the water inrush effect due to the space and gap between the buildings.
4. The rapid simulation method for dam break flood routing of the reservoir based on GPU parallel acceleration as claimed in claim 1, wherein the reservoir break mode comprises two dam break modes of dam instant break and dam break gradually.
5. The rapid simulation method for dam break flood evolution of a reservoir based on GPU parallel acceleration according to claim 4, characterized in that the instant burst is divided into instant full burst and instant partial burst according to different burst widths so as to simulate the process of reservoir burst flood evolution caused by accident such as strike and earthquake;
the gradual burst simulates the development process of a burst opening from small to large.
6. The rapid simulation method for reservoir dam-break flood routing based on GPU parallel acceleration as claimed in claim 4, wherein step S3 includes the following steps:
s31, selecting corresponding parameters including a reservoir burst mode, a burst opening width, a burst water depth and a corresponding reservoir capacity according to the working condition to be simulated;
s32, calculating the maximum burst flow, wherein the maximum burst flow of the instant burst is as follows:
Figure DEST_PATH_IMAGE001
in the formulaQ max、g、B 、b m 、H 0Respectively representing the maximum dam bursting flow, the gravity acceleration, the dam length, the final burst width and the burst water depth;
the dam which is gradually burst firstly breaks through osmotic deformation and then collapses instantly, and the flow of the osmotic deformation and the breakage is as follows:
Figure 192806DEST_PATH_IMAGE002
in the formula:Hcalculating a water level elevation for the reservoir water level, i.e. the reservoir;Athe cross-sectional area of the pipe through which the water flows;H P is the elevation of the central line of the pipeline;f in order to have a coefficient of friction of darcy,L the length of the pipeline along the water flow direction;Dis the diameter or width of the pipe;
s33 flood flow process of breach
Suppose that the maximum burst flow of the dam isQ maxReservoir capacity with known burst flood volumeWDuration of floodT n When it is required, thent Time of day flowQComprises the following steps:
Figure DEST_PATH_IMAGE003
。
7. the rapid simulation method for reservoir dam-break flood evolution based on GPU parallel acceleration according to claim 1, wherein the WCA2D model is used for simulating surface water flow, and the main process comprises:
(1) acquiring the weight of water flow exchange of adjacent cells:
Figure 124595DEST_PATH_IMAGE004
Figure DEST_PATH_IMAGE005
Figure 499074DEST_PATH_IMAGE006
in the formula:mrepresents the total number of cells adjacent to the central cell;
Figure DEST_PATH_IMAGE007
denotes the central cell of number 0 andithe water level difference of each adjacent cellular;
Figure 294949DEST_PATH_IMAGE008
representing the corresponding volume difference;
Figure DEST_PATH_IMAGE009
indicating that the water level difference is greater than the critical value
Figure 517112DEST_PATH_IMAGE010
Calculating the water quantity obtained by downstream cells;
Figure DEST_PATH_IMAGE011
is shown asiWeights of adjacent cells;
Figure 818869DEST_PATH_IMAGE012
represents the weight of the central cell;
Figure DEST_PATH_IMAGE013
represents the sum of the volume differences adjacent to the central cell;
Figure 493693DEST_PATH_IMAGE014
representing the smallest volume difference of the cells adjacent to the central cell;
(2) acquiring the amount of intercellular interaction water:
Figure DEST_PATH_IMAGE015
in the formula:
Figure 437466DEST_PATH_IMAGE016
to represent
Figure DEST_PATH_IMAGE017
At a time downstream ofiThe exchange water quantity obtained by each cell;
Figure 946875DEST_PATH_IMAGE018
is shown in
Figure DEST_PATH_IMAGE019
The amount of water transferred from the central cell at that time;
Figure 153954DEST_PATH_IMAGE020
represents the central cell area;
Figure DEST_PATH_IMAGE021
representing the central cellular water depth;
Figure 378391DEST_PATH_IMAGE022
representing a weight;
Figure DEST_PATH_IMAGE023
representing the difference between the water level of the central cell and the water level of the cell;
Figure 594697DEST_PATH_IMAGE024
representing the distance from the center cell midpoint to the cell midpoint;
Figure DEST_PATH_IMAGE025
representing the maximum allowable flow rate of the water flow of the central cell transferred to the cell;
Figure 355234DEST_PATH_IMAGE026
represents the time step, s;
Figure DEST_PATH_IMAGE027
indicating the boundary length of the cell and the central cell;
Figure 66969DEST_PATH_IMAGE028
to representtThe amount of water transferred from the central cell at that time;nis the Manning coefficient; g is the acceleration of gravity; subscript isMAll variables of (1) are for the cell with the greatest weight; all variables without superscripts are assumed at time t;
(3) calculating the water depth:
Figure DEST_PATH_IMAGE029
in the formula:
Figure 53468DEST_PATH_IMAGE030
to represent
Figure 874398DEST_PATH_IMAGE017
The water depth of the central cell at all times;
Figure DEST_PATH_IMAGE031
to representtThe water depth of the central cell at the moment;
Figure 731584DEST_PATH_IMAGE032
representing the amount of influent water obtained by the central cells;
Figure DEST_PATH_IMAGE033
representing the central unit cellThe amount of water flowing out;
(4) the flow rate calculation formula is:
Figure 38323DEST_PATH_IMAGE034
in the formula:
Figure DEST_PATH_IMAGE035
to represent
Figure 986819DEST_PATH_IMAGE017
The water flow of the cell at the time center is transferred to the firstiThe flow rate of the individual cells;
Figure 948522DEST_PATH_IMAGE036
to represent
Figure 574063DEST_PATH_IMAGE017
Time center cell andithe arithmetic mean of the water depth of the individual cells;
Figure 986457DEST_PATH_IMAGE037
is shown asiThe boundary length between each unit cell and the central unit cell.
8. The rapid simulation method for reservoir dam-break flood routing based on GPU parallel acceleration as claimed in claim 1, characterized in that an OpenCL heterogeneous algorithm is adopted to optimize a WCA2D model, and a WCA2D model of GPU accelerated version is obtained, so as to improve simulation efficiency for reservoir dam-break flood routing.
9. The rapid simulation method for dam break flood routing of the reservoir based on GPU parallel acceleration as claimed in claim 1, wherein the basic data comprises reservoir capacity curve, reservoir dam related data, topographic data and land utilization data.
10. The rapid simulation method for dam break flood evolution of a reservoir based on GPU parallel acceleration according to any one of claims 1-9, characterized in that the simulation parameters comprise initial conditions, boundary conditions and roughness, step size, iteration number and simulation time.
CN202111661759.0A 2021-12-31 2021-12-31 Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration Pending CN114004114A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111661759.0A CN114004114A (en) 2021-12-31 2021-12-31 Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111661759.0A CN114004114A (en) 2021-12-31 2021-12-31 Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration

Publications (1)

Publication Number Publication Date
CN114004114A true CN114004114A (en) 2022-02-01

Family

ID=79932446

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111661759.0A Pending CN114004114A (en) 2021-12-31 2021-12-31 Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration

Country Status (1)

Country Link
CN (1) CN114004114A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117910365A (en) * 2024-03-19 2024-04-19 水利部交通运输部国家能源局南京水利科学研究院 Dam break flood evolution prediction method based on machine learning model

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109272189A (en) * 2018-08-03 2019-01-25 华南理工大学 A kind of City-scale Flooding Risk appraisal procedure based on chain structure

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109272189A (en) * 2018-08-03 2019-01-25 华南理工大学 A kind of City-scale Flooding Risk appraisal procedure based on chain structure

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
王运涛: "《基于多源信息的城市洪水模拟与弹性分析》", 《中国优秀博硕士学位论文全文数据库(博士)工程科技Ⅱ辑》 *
赖成光: "《城市地区水库溃坝洪水演进数值模拟研究》", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117910365A (en) * 2024-03-19 2024-04-19 水利部交通运输部国家能源局南京水利科学研究院 Dam break flood evolution prediction method based on machine learning model

Similar Documents

Publication Publication Date Title
CN103544390B (en) The quick evolution analog analysing method of flood due to dam-break based on cellular automata
CN114004003A (en) Reservoir dam-break flood numerical simulation method suitable for complex underlying surface of urban area
CN104156525A (en) Method for improving storm surge disaster risk predication accuracy
CN105160134B (en) The blending agent analogy method and device that oil gas flows in compact reservoir multi-dielectric
CN108021780A (en) A kind of mountain torrents dynamic emulation method based on random unstrctured grid model
CN103853934A (en) Method and system for calculating river networks model
CN110377953B (en) Storm surge fine simulation method based on parallel mode
CN104361157B (en) The evaluation method of wind environment between a kind of building
Akoh et al. High-resolution modeling of tsunami run-up flooding: a case study of flooding in Kamaishi city, Japan, induced by the 2011 Tohoku tsunami
CN115329691A (en) CFD and GIS-based super-large city wind environment simulation method
Liu et al. Numerical study on factors influencing typhoon-induced storm surge distribution in Zhanjiang Harbor
CN114492233A (en) Basin water simulation method based on webGIS platform and considering comprehensive utilization requirements
CN110990926B (en) Urban surface building hydrodynamic simulation method based on area correction rate
CN114004114A (en) Reservoir dam-break flood evolution rapid simulation method based on GPU parallel acceleration
CN115169690B (en) Dam break flood-avoiding transfer dynamic path optimization method based on improved ant colony algorithm
Zhang et al. Application of multi-agent models to urban expansion in medium and small cities: A case study in Fuyang City, Zhejiang Province, China
Belikov et al. Justification of hydrological safety conditions in residential areas using numerical modelling
CN113869804B (en) Power grid equipment risk early warning method and system under flood disaster
Wang et al. A non-uniform grid approach for high-resolution flood inundation simulation based on GPUs
CN114611780A (en) Method for calculating optimal solution for location selection of emergency refuge for levee breaking and path planning
CN114565211A (en) Risk consequence evaluation method for dam break of step reservoir group under risk transfer and superposition effects
Strasszer et al. CFD-based wind assessment for suburban buildings. The case study of aarhus university, herning campus
Wijaya et al. Combining two algorithms as a transition rules for CA-based inundation model
Ghimire et al. Application of cellular automata approach for fast flood simulation
Chen et al. Wargame Mapping and Implementation for Emergency Evacuation of Residents in Urban Waterlogging Disaster

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20220201