CN103729483A - Device and method for simulating and predicting mountain torrent disasters - Google Patents

Abstract

The invention relates to a hydraulics parameter monitoring device for numerical simulation of mountain torrent disasters and a method for simulating and predicting the evolution process of the mountain torrent disasters. The method comprises the steps that hydrologic data such as rainfall and water level of a monitored zone are collected through a mountain torrent simulation parameter monitoring node device and transmitted to a remote simulation computing terminal in real time, the evolution process of the mountain torrent disasters is calculated circularly on the computing terminal through a grid outflow correcting method, and therefore simulation and prediction with high reliability are achieved. The core calculation steps are that firstly, numerical value discretization treatment is conducted on a mass continuity equation; when the value of the depth of water of a grid is negative, whether M and N are inflow or outflow is judged firstly, and the inflow Qin and the outflow Qout of the grid are calculated; then correction coefficients are calculated; finally, calculation is conducted circularly and repeatedly until all the grids with the water depth of the negative values are collected, and the next time step is conducted. According to the method, the difficult problems that simulating calculation of the mountain torrent disasters is prone to divergence due to the steep terrain in a drainage basin and the rapid change of water flow in a traditional method are solved.

Description

Mountain flood simulation and forecast device and method

One, technical field

The present invention is a kind of parameter monitor device for mountain flood numerical simulation and mountain torrents evolution process simulating and predicting method, utilize the hydrographic datas such as mountain torrents analog parameter monitoring node device acquisition monitoring district rainfall, water level, flow, be sent in real time long-range analog computation terminal, at computing terminal, adopt institute's grid of carrying to flow out revised law and analyze data, finally prediction generates the simulation process of mountain torrents evolution.Institute's extracting method has solved in classic method because basin landform is precipitous, discharge changes rapidly, cause the negative depth of water of in the moving boundary processing procedure compared with under computation time step, the simple deal with return to zero of discharge being brought, thus the problem of quality and momentum nonconservation before and after in the simulation process causing.Make the mass conservation in mountain torrents analog computation process, reduced the conservative errors of momentum, guaranteed simulation precision under precipitous landform dynamic boundary condition and the stability of evaluation, solved the difficult problem that the calculating of mountain flood simulation under precipitous landform is dispersed.This invention belongs to Computer Simulation and numerical simulation field.

Two, background technology

The development of computational hydraulics can be concluded displaying with the evolution of its fundamental equation.What first with differential equation form, explain is the balance differential equation (Euler, 1755) of fluid, and it has represented unit mass flow body upper volume power and area equilibrium of forces relation under stationary state.Recognize after fluid viscosity, Navier and Stokes (1843,1845) have set up the equation of motion of real fluid, i.e. N-S equation, and it shows classical hydraulic completing.Within 1883, Reynolds has found that by the experiment of system liquid has laminar flow and two kinds of fluidised forms of turbulent flow; Prandtl in 1904 have proposed the concept in boundary layer, indicate that hydraulics has entered modern hydraulic developing stage.So for many years, in to continuous the improving of equation, the understanding of people's convection cell inwardness and slight movement rule thereof is also increasing.Because the turbulent fluctuation problem of fluid is very complicated, though carried out a large amount of research and formed the turbulent theory of each family, arrive far away the stage of basic solution.

Because fundamental equation is nonlinear partial differential equation on mathematics, all standards very of the fairly simple and type of form, are therefore extremely suitable for numerical evaluation, are about to fluid mechanics physical problem and are converted into the just numerical solution problem of boundary values of partial differential equation.

The discrete method of commonly using in numerical evaluation comprises method of characteristic, finite element method, finite volume method and method of finite difference.In the process of establishing of mathematical model, although the numerical discretization method difference adopting, but there is identical feature, first zoning is divided into many fritters or grid, then on these grid cells, the differential equation is separated into algebraic equation, again the algebraic equation on grid is merged into overall Algebraic Equation set, finally under certain first boundary value condition, solve equation of n th order n group, thereby try to achieve the physical quantity of each node in zoning.Method of finite difference, finite element method and finite volume method are the discrete numerical solutions of the differential equation, and this several method is not complete difference, exist some contacts between them.Finite element method is different with method of finite difference starting point in the thinking of discrete equation, and its key distinction is:

1, method of finite difference is that point is approximate, adopt the value on discrete grid node to carry out approximate expression continuous function, and that finite element method is segmentation (or burst, piecemeal) is approximate, in unit, approximate solution is to resolve continuously, and between unit, approximate solution is continuous.

What 2, limit that first method obtains is an abundant smooth approximate solution, at unit interior derivative, exists.And difference method generally can not guarantee the slickness of separating.

3, first method has no particular limits the division of computing unit, processes flexibly, and particularly, when processing the problem of complex boundary, this advantage is more outstanding.

With regard to discrete method, finite volume method can be regarded the intermediate of Finite Element and method of finite difference as.Finite Element must the Changing Pattern (be interpolating function) of assumed function value between net point, and set it as approximate solution.A method of finite difference considering gridding point superior function value φ and do not consider the situation of change of φ between net point.Finite volume method only seek must nodal value, this and method of finite difference are similar; But finite volume method, when seeking to control the integration of volume, must supposition must be worth the distribution between net point, and this thinks similar with Finite Element again.In finite volume method, an interpolating function integration for computer control volume, after drawing discrete equation, just can be without interpolating function; If needed, can take different interpolating functions to items different in the differential equation.

Different numeric formats by discrete the differential equation be different Algebraic Equation set, even if employing same numerical method but is selected different forms, to the same differential equation, also can obtain multiple different Algebraic Equation set.But select different numeric formats, if select certain specific form, also may by discrete the same differential equation be identical Algebraic Equation set.Therefore, say in a sense, during various numeric format, come round.For example, in finite volume method, while calculating the normal direction flux at control volume interface place, if take the mean value of adjacent control volume centre of form place flux, if the computing grid adopting is rectangular node, be equivalent to the centered finite difference form of second order, if adopt triangle or quadrilateral mesh, be equivalent to again the golden finite element scheme of Linear Triangular shape or bilinearity tetragonal gal the Liao Dynasty.

Consider on the whole, in processing problem, different algorithms has relative merits separately.Finite difference ratio juris is fairly simple, if border is uncomplicated, mathematical reasoning and program composition are all fairly simple, and it is fewer to calculate EMS memory occupation, and computing velocity is fast, and in the more intense problem of non-constancy, application is many.But finite difference method dirigibility when processing complex boundary is poor, and computational accuracy is not high, on complex boundary, usually there is false mobile phenomenon, there is sometimes water yield nonconservation phenomenon.Finite element method and finite volume method be mathematical reasoning and just complexity of program composition comparatively speaking, and because the grid of these two kinds of methods can be the arbitrary messes such as triangle or quadrilateral, although this network is easy to process the problem of border and landform more complicated, but need the position relationship between data structure definition computing unit, committed memory is larger, and computing velocity is relatively slow.Finite volume method physical concept is clear, when solving control volume interface normal direction flux, is easy to process discontinuous problem, and the water yield and momentum conservation, can be used to process the problems such as current interruption.But discretely on irregular grid be difficult to adopt high-precision discrete scheme, the theoretical analysis that discrete scheme is carried out to system is more difficult.

With certain form to fundamental equation carry out discrete after, the groundwork of numerical simulation is solving linear algebric equation group.Method for solving can be divided into direct method and indirect method (process of iteration).Often all to separate Large Scale Sparse matrix (Sparse Matrix).During direct method solves, determinant (Cramer) operand is very large, elimination method (matrix is turned to upper triangular matrix) and LU factorization (matrix being turned to the product of lower triangle battle array and upper triangular matrix) calculated amount is also larger, therefore conventional chasing method solves.Indirect method is asked

Solution is an iteration for explicit scheme, conventional simple some iteration (Jacobi iteration), improvement iteration (Gauss-Seidel iteration) and over relaxation method acceleration point iteration (Successive OverRelaxation); For implicit schemes, it is row iteration, the simple row iteration of Xiang Yingyou, improvement row iteration, line overrelaxation method are accelerated row iteration (SLOR) and alternating direction iteration (ADI iteration), also has in addition SIP technique (SIP), every time iteration can make information spread all over universe, solves speed faster.

For flood simulation, the people such as Jin W L have proposed to use the nonlinear convection item of the discrete equation of momentum of Second-order Up-wind form and the overbank process in second order leap-frog form Discrete Linear convective term simulation Nakdong basin.The people such as Roger A F propose a kind of edge fitting numerical model of MacCormack+TVD form, by the depth of water detecting under each time step, whether arrive withered critical value judgement moving boundary scope.Denmark DHI water resource and Environmental Studies institute have adopted implicit expression alternating direction algorithm development Hydrodynamic Simulation software Mike21.In addition, also have the representative numeric formats such as Simple method and CIP method.

But in flush flood and debris flow disaster Two Dimensional Numerical Simulation, because basin landform is precipitous, discharge changes rapidly, classic method often causes the negative depth of water to the simple deal with return to zero of discharge in the moving boundary processing procedure compared with under computation time step, cause quality and momentum nonconservation in simulation process, finally cause unstable even calculating of evaluation to disperse and can not get result.At present, also there is no a kind of method that is applicable to the mountain torrents Two Dimensional Numerical Simulation under precipitous landform.

The advantage of this invention is: based on monitoring terminal Real-time Collection hydrographic data, carry out grid and flow out corrected Calculation, can solve in classic method because basin landform is precipitous, discharge changes rapidly, the problem of the quality in the simulation process causing and momentum nonconservation.Institute's extracting method makes the mass conservation in mountain torrents analog computation process, reduced the conservative errors of momentum, guarantee simulation precision under precipitous landform dynamic boundary condition and the stability of evaluation, solved the difficult problem that the calculating of mountain flood simulation under precipitous landform is dispersed.

Three, summary of the invention

Mountain torrents analog parameter monitoring device of the present invention and Forecasting Methodology, is characterized in that adopting the grid that the present invention proposes to flow out modification method correction calculation grid discharge.Solved in classic method because basin landform is precipitous, discharge changes rapidly, cause the negative depth of water of in the moving boundary processing procedure compared with under computation time step, the simple deal with return to zero of discharge being brought, thus the problem of quality and momentum nonconservation before and after in the simulation process causing.Make the mass conservation in mountain torrents analog computation process, reduced the conservative errors of momentum, guaranteed simulation precision under precipitous landform dynamic boundary condition and the stability of evaluation, solved the difficult problem that the calculating of mountain flood simulation under precipitous landform is dispersed.

Described mountain torrents analog parameter monitoring device and Forecasting Methodology, it is characterized in that described mountain torrents analog parameter monitoring node device comprises velocimeter module, rain sensor module, Temperature Humidity Sensor module, evaporation gauge module, water-level gauge module and level shifting circuit, wherein the output terminal of velocimeter module, rain sensor module, Temperature Humidity Sensor module connects respectively the input end of microcontroller, and the output terminal of evaporation gauge module, water-level gauge module is connected in series respectively after analog-to-digital conversion module by level shifting circuit and microprocessor communication.

Described mountain torrents analog parameter monitoring device and Forecasting Methodology, it is characterized in that described each mountain torrents analog parameter monitoring node device comprises solar-electricity source unit, data acquisition unit, input-output unit, Versatile Interface Unit, microprocessor unit and communication unit, wherein data acquisition unit consists of sensor assembly serial connection analog-to-digital conversion module, microprocessor unit consists of microcontroller serial connection storer, communication unit is connected and composed by Big Dipper communication small station and antenna electric, analog-to-digital conversion module, input-output unit is connected with microcontroller by Versatile Interface Unit respectively, the input end of the output termination microcontroller of analog-to-digital conversion module, solar-electricity source unit is given respectively data acquisition unit, input-output unit, Versatile Interface Unit, microprocessor unit and communication unit power supply.

The present invention can reach following beneficial effect:

(1) the present invention has adopted the grid outflow revised law of original creation to carry out mountain torrents evolution numerical simulation.Because the landform of mountain torrents spot is precipitous and uneven, therefore the bounds of flood, flow velocity and the depth of water all sharply change.In moving boundary in the past calculates, often there is the situation of water surface absolute altitude lower than ground elevation, the depth of water after namely calculating becomes negative value.Cause quality can not guarantee conservation, bad stability, even calculates and disperses and can not get result.Therefore, by grid discharge is circulated, revise to eliminate the negative depth of water, the conservation of ensuring the quality of products, improves numerical evaluation stability.

(2) the present invention possesses the reconfigurable of function and emat sensor more flexibly.The present invention adopts general-purpose interface in the design of sensor node, by toggle switch identification sensor type, makes five kinds of hydrographic information sensors can realize plug and play in using at the scene.

(3) possess can routing mode in the present invention, and each sensor node can form message according to sensor type in the present invention, and the optimal path of search uploading data detects connection status and reports in time error message.

(4) what in the design, adopt is solar cell for supplying power, and there is accumulator as reserve battery, and circuit itself adopts ultralow Consumption, the maximum short-distance wireless communication part of wherein consuming energy adopts dormancy and the mode waken up reaches lowest power consumption and extends life-span of monitoring device node.

Four, accompanying drawing explanation:

Accompanying drawing 1: grid flows out modification method calculation process;

Accompanying drawing 2: mountain torrents analog parameter monitoring node workflow diagram;

Accompanying drawing 3: mountain torrents analog parameter monitoring node processor unit circuit diagram;

Accompanying drawing 4: mountain torrents analog parameter monitoring node sensor and interface unit circuit thereof;

Accompanying drawing 5: mountain torrents analog parameter monitoring node power supply unit circuit diagram.

Five, specific embodiments:

Accompanying drawing 1 has shown that grid flows out the calculation process of modification method, comprises the steps:

The first step, carries out numerical discretization processing to mass continuity equation, is expressed as follows:

$h_{i+1/2,j+1/2}^{n+3}(\mathrm{\Δx}\·\mathrm{\Δy})=h_{i+1/2,j+1/2}^{n+1}(\mathrm{\Δx}\·\mathrm{\Δy})+2\mathrm{\Δt}[(M_{i,j+1/2}^{n+2}-M_{i+1,j+\mathrm{1,2}}^{n+2})\mathrm{\Δy}+(N_{i+1/2,j}^{n+2}-N_{i+1/2,j+1}^{n+2})\mathrm{\Δx}]$

That is, the water yield+2 Δ t in the water yield=n+1 moment grid (i, j) in n+3 moment grid (i, j) (unit interval flows into the water yield of the water yield-unit interval outflow grid (i, j) of grid (i, j))

Second step, when negative value appears in the grid depth of water, first judges M, and N flows into or flows out, and the influx Qin of computing grid and discharge Qout:

$h_{i+1/2,j+1/2}^{n+3}=h_{i+1/2,j+\mathrm{1,2}}^{n+1}+\frac{Q_{\mathrm{in}}-{\mathrm{\λQ}}_{\mathrm{out}}}{\mathrm{\Δx}\·\mathrm{\Δy}}=0$

The 3rd step, then use following formula to obtain correction factor:

$\mathrm{\λ}=(h_{i+1/2,j+1/2}^{n+1}(\mathrm{\Δx}\·\mathrm{\Δy})+Q_{\mathrm{in}})/Q_{\mathrm{out}}$

The 4th step, but, because the influx of adjacent mesh reduces, may cause that the depth of water of adjacent mesh becomes negative.In this case, need continuous double counting, until all negative depth of water grids are revised, then proceed to next time step.

Accompanying drawing 2 has shown mountain torrents analog parameter monitoring node workflow, the designed mountain torrents analog parameter monitoring node of the present invention is characterized in that requiring access water level, quantity of precipitation, evaporation capacity, flow, five kinds of sensors of humiture according to monitoring, is deployed on the gauge point in planning region.Described each intelligent sensor node comprises solar-electricity source unit, data acquisition unit, input-output unit, Versatile Interface Unit, microprocessor unit and communication unit, wherein data acquisition unit consists of sensor assembly serial connection analog-to-digital conversion module, microprocessor unit consists of microcontroller serial connection storer, communication unit is connected and composed by Big Dipper communication small station and antenna electric, analog-to-digital conversion module, input-output unit is connected with microcontroller by Versatile Interface Unit respectively, the input end of the output termination microcontroller of analog-to-digital conversion module, solar-electricity source unit is given respectively data acquisition unit, input-output unit, Versatile Interface Unit, microprocessor unit and communication unit power supply.Input-output unit comprises some input-output device such as LED, LCD, keyboard.

Accompanying drawing 3 has shown mountain torrents analog parameter monitoring node processor unit circuit diagram.That flush bonding processor adopts is the CC2430 (U1) of TI company.Inductance L 1, L2, L3, capacitor C 9 and Antenna form the peripheral circuit of transmitter unit, by pin 32,33,34, are connected with U1.Five toggle switchs (SW1) are for carrying out the selection of node sensors configured state, and each is corresponding Temperature Humidity Sensor, rain sensor, velocimeter respectively.The configuration status of water-level gauge and evaporation gauge, 0 represents " having ", 1 represents " nothing ", and the corresponding pin 1,2,16,17,18 of U1 need to be configured to input pull-up state.Two pilot lamp of LED1 and LED2 are connected to 11,12 pins of U1, for the state indication of debug process.

Accompanying drawing 4 has shown mountain torrents analog parameter monitoring node sensor and interface unit circuit thereof.Module (1) is the TTL-RS232 level shifting circuit of microprocessor U1, SP3232E chip and DB9 connector, consists of.Transmitting terminal TXD is connected with 13 pins with 14 of U1 respectively with receiving end RXD.CC2430 has two groups of UART interfaces, by a slice SP3232E chip and 2 DB9 connectors, can realize 2 groups of RS232 interfaces.The LS20B type rotary propeller type velocimeter of module (2) and the JDZ05-1 type rain sensor of module (3) are pulse switch amount output type sensor, its output terminal is connected with 8 with the pin 6 of U1 respectively, output low level when the tongue tube of sensor is closed, this negative edge pulse will trigger the external interrupt of microprocessor, and completes once counting.Module (4) is integrated digital hygro sensor DHT11, and it is connected with the 9th pin of U1 by pin 2, adopts a kind of agreement of unibus to send the data of temperature and humidity, and its sampling interval is by microprocessor control.Module (5) and module (6) represent respectively remote measurement evaporation gauge and WQC-1 type air bubble type water gauge, and the two all adopts 12V power supply, and by RS232 interface and microprocessor communication.

It is characterized in that described mountain torrents analog parameter monitoring node sensor assembly comprises velocimeter module, rain sensor module, Temperature Humidity Sensor module, evaporation gauge module, water-level gauge module and level shifting circuit, wherein the output terminal of velocimeter module, rain sensor module, Temperature Humidity Sensor module connects respectively the input end of microcontroller, and the output terminal of evaporation gauge module, water-level gauge module is connected in series respectively after analog-to-digital conversion module by level shifting circuit and microprocessor communication.

Accompanying drawing 5 has shown mountain torrents analog parameter monitoring node power supply unit circuit diagram.In module (1), solar panel (SBT) charges to lithium battery (BT1) by charging intelligent management chip CN3063 (U3), and provides 5V Voltage-output at J1 interface.J1 is connected with the 5V power input interface (J3) in module (2) by power lead.Module (2) provides power supply and the switch control thereof of 3.3V for processor unit, and linear voltage regulator SPX1117-3.3 is that 3.3V voltage is for CC2430 and DHT11 by the 5V voltage transitions of input.Module (3) provides the power supply of 12V for remote measurement evaporation gauge and WQC-1 type air bubble type water gauge, 5V voltage is introduced from the J2 interface of module (1) by J4 interface, by MC34063 chip (U5), boost to 12V, and exported by interface J5.

The present invention can work under unattended state, and possesses highly reliable, routable communication system.Node power unit adopts solar cell for supplying power, and solar cell for supplying power unit is by solar panel, and the part such as solar charging/discharging control circuit, accumulator forms.Wherein solar panel is that sun power is converted to electric energy; thereby drive node work or electric energy is sent in accumulator and is stored in case of need; the effect of solar charging/discharging control circuit be monitoring whole power supply unit duty, and to accumulator overcharge, over.

Claims (4)

1. the parameter monitor device for mountain flood numerical simulation and simulating and predicting method, utilize the hydrographic datas such as mountain torrents analog parameter monitoring node device acquisition monitoring district rainfall, water level, flow, evaporation capacity, humiture, be sent in real time long-range analog computation terminal, at computing terminal, adopt grid to flow out correction cycle calculations method mountain torrents evolution process is carried out to high reliability simulation and forecast.Core calculations step is first mass continuity equation to be carried out to numerical discretization processing; When there is negative value in the grid depth of water, first judge M, N flows into or flows out, and the influx Qin of computing grid and discharge Qout; Then solve correction factor; Finally, because the influx of adjacent mesh reduces, may cause that the depth of water of adjacent mesh becomes negative.In this case, need continuous double counting, until all negative depth of water grids are revised, then proceed to next time step.

2. mountain torrents analog parameter monitoring node device according to claim 1, it is characterized in that: each monitoring node device comprises velocimeter module, rain sensor module, Temperature Humidity Sensor module, evaporation gauge module, water-level gauge module and level shifting circuit, wherein the output terminal of velocimeter module, rain sensor module, Temperature Humidity Sensor module connects respectively the input end of microcontroller, and the output terminal of evaporation gauge module, water-level gauge module is connected in series respectively after analog-to-digital conversion module by level shifting circuit and microprocessor communication.Sensor kind can need to be configured according to monitoring.

3. mountain torrents analog parameter monitoring node device according to claim 1, it is characterized in that: each monitoring node device comprises solar-electricity source unit, data acquisition unit, input-output unit, Versatile Interface Unit, microprocessor unit and communication unit, wherein data acquisition unit consists of sensor assembly serial connection analog-to-digital conversion module, microprocessor unit consists of microcontroller serial connection storer, communication unit is connected and composed by Big Dipper communication small station and antenna electric, for transmitting Monitoring Data to remote terminal computing machine.Analog-to-digital conversion module, input-output unit are connected with microcontroller by Versatile Interface Unit respectively, the input end of the output termination microcontroller of analog-to-digital conversion module, solar-electricity source unit is given respectively data acquisition unit, input-output unit, Versatile Interface Unit, microprocessor unit and communication unit power supply.Input-output unit comprises some input-output device such as LED, LCD, keyboard.

4. grid according to claim 1 flows out the calculation process of modification method, it is characterized in that its workflow is as follows:

The first step, carries out numerical discretization processing to mass continuity equation, is expressed as follows:

$h_{i+1/2,j+1/2}^{n+3}(\mathrm{\Δx}\·\mathrm{\Δy})=h_{i+1/2,j+1/2}^{n+1}(\mathrm{\Δx}\·\mathrm{\Δy})+2\mathrm{\Δt}[(M_{i,j+1/2}^{n+2}-M_{i+1,j+\mathrm{1,2}}^{n+2})\mathrm{\Δy}+(N_{i+1/2,j}^{n+2}-N_{i+1/2,j+1}^{n+2})\mathrm{\Δx}]$

That is, the water yield+2 Δ t in the water yield=n+1 moment grid (i, j) in n+3 moment grid (i, j) (unit interval flows into the water yield of the water yield-unit interval outflow grid (i, j) of grid (i, j))

Second step, when negative value appears in the grid depth of water, first judges M, and N flows into or flows out, and the influx Qin of computing grid and discharge Qout:

$h_{i+1/2,j+1/2}^{n+3}=h_{i+1/2,j+\mathrm{1,2}}^{n+1}+\frac{Q_{\mathrm{in}}-{\mathrm{\λQ}}_{\mathrm{out}}}{\mathrm{\Δx}\·\mathrm{\Δy}}=0$

The 3rd step, then use following formula to obtain correction factor:

$\mathrm{\λ}=(h_{i+1/2,j+1/2}^{n+1}(\mathrm{\Δx}\·\mathrm{\Δy})+Q_{\mathrm{in}})/Q_{\mathrm{out}}$

The 4th step, but, because the influx of adjacent mesh reduces, may cause that the depth of water of adjacent mesh becomes negative.In this case, need continuous double counting, until all negative depth of water grids are revised, then proceed to next time step.

Images