CN116720409B - Electromagnetic scattering field calculation method for moving time-varying dispersion medium target - Google Patents
Electromagnetic scattering field calculation method for moving time-varying dispersion medium target Download PDFInfo
- Publication number
- CN116720409B CN116720409B CN202310754740.3A CN202310754740A CN116720409B CN 116720409 B CN116720409 B CN 116720409B CN 202310754740 A CN202310754740 A CN 202310754740A CN 116720409 B CN116720409 B CN 116720409B
- Authority
- CN
- China
- Prior art keywords
- coordinate system
- boundary
- slave
- parameters
- fdtd
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 61
- 239000002612 dispersion medium Substances 0.000 title claims abstract description 19
- 238000000034 method Methods 0.000 claims abstract description 186
- 230000008569 process Effects 0.000 claims abstract description 160
- 230000033001 locomotion Effects 0.000 claims abstract description 62
- 230000009466 transformation Effects 0.000 claims abstract description 19
- 230000005672 electromagnetic field Effects 0.000 claims abstract description 16
- 230000005684 electric field Effects 0.000 claims description 27
- 230000006870 function Effects 0.000 claims description 24
- 238000010521 absorption reaction Methods 0.000 claims description 22
- 239000013598 vector Substances 0.000 claims description 20
- 239000002609 medium Substances 0.000 claims description 19
- 238000004891 communication Methods 0.000 claims description 18
- 238000005192 partition Methods 0.000 claims description 11
- 230000011218 segmentation Effects 0.000 claims description 9
- 108020001568 subdomains Proteins 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 6
- 230000010287 polarization Effects 0.000 claims description 5
- 230000002123 temporal effect Effects 0.000 claims 1
- 230000001131 transforming effect Effects 0.000 abstract 2
- 238000006243 chemical reaction Methods 0.000 abstract 1
- 238000004422 calculation algorithm Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 4
- 238000012546 transfer Methods 0.000 description 4
- 238000009795 derivation Methods 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 238000013213 extrapolation Methods 0.000 description 2
- 235000011437 Amygdalus communis Nutrition 0.000 description 1
- 241000220304 Prunus dulcis Species 0.000 description 1
- 235000020224 almond Nutrition 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000004020 conductor Substances 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011112 process operation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
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
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/54—Interprogram communication
- G06F9/542—Event management; Broadcasting; Multicasting; Notifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/28—Fuselage, exterior or interior
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Automation & Control Theory (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Aviation & Aerospace Engineering (AREA)
- Multimedia (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an electromagnetic scattering field calculation method of a moving time-varying dispersion medium target, which comprises the following steps: firstly, target modeling, namely reading a dispersion medium model file in a laboratory coordinate system by a main process and reading corresponding data from a screen; transforming the time and space components by Lorentz transformation to obtain the time and space components in a motion coordinate system; dividing the FDTD region and establishing a virtual topological structure; the main process introduces an incident wave source after Lorentz conversion and transmits parameters to the auxiliary process; each slave process carries out iterative operation of an electromagnetic field by using an FDTD method and processes a special boundary; after the calculation of each slave process is finished, the calculation result is sent to the master process; and finally, obtaining electromagnetic field data in the motion coordinate system, and then, transforming the electromagnetic field in the motion coordinate system into a laboratory coordinate system by utilizing inverse Lorentz transformation to obtain the electromagnetic field data in the laboratory coordinate system. The invention realizes the rapid calculation of the electromagnetic scattering field of the motion time-varying dispersion medium.
Description
Technical Field
The invention belongs to the field of electromagnetic field calculation, and particularly relates to an electromagnetic scattering field calculation method of a moving time-varying dispersion medium target based on a parallel Finite Difference Time Domain (FDTD) algorithm formed by combining an MPI library with the FDTD algorithm.
Background
With the continuous increase of the speed of aircrafts, electromagnetic scattering research on high-speed moving objects is attracting more and more attention. Meanwhile, due to the influence of factors such as the flight attitude of the aircraft, a turbulent flow field and the like, a time-varying plasma sheath is often generated around the aircraft. The time-varying plasma sheath can affect the original electromagnetic scattering characteristics of the target, so that the detection and identification of the target are difficult. Therefore, on the basis of electromagnetic transmission and scattering mechanisms of a high-speed moving dispersion medium target, the method for researching the electromagnetic calculation of the time-varying dispersion medium target in a high-speed moving state is of great significance.
Currently, electromagnetic computing methods for high-speed moving conductor and media targets begin to emerge, such as the Lorentz-FDTD method, the Relativistic Boundary Conditions (RBC) method. However, most of the previous computing methods have studied metal or media targets that move at high speed and consume excessive computing time and computing resources (memory). Therefore, for calculation of electromagnetic field values of a high-speed moving dispersive medium, it is highly necessary to design a technique capable of shortening the operation time and improving the resource utilization.
Patent document 1 discloses a method of calculating an electromagnetic fringe field of a three-dimensional high-speed translational target, which proposes a method capable of calculating an electromagnetic fringe field value of an almond body target of high-speed translational under three-dimensional conditions, which is capable of calculating an electromagnetic fringe field value of a complex target of high-speed translational, however, the patent document is not involved in improvement of program operation efficiency.
Reference to the literature
Patent document 1 chinese invention patent application, publication No.: CN107944113a, publication date: 2018.04.20.
disclosure of Invention
The invention aims to provide an electromagnetic scattering field calculation method of a motion time-varying dispersion medium target, which is based on a parallel Finite Difference Time Division (FDTD) algorithm formed by combining an MPI library with an FDTD algorithm, and can realize the rapid calculation of the electromagnetic scattering field of the motion time-varying dispersion medium, greatly shorten the running time and improve the resource utilization rate.
In order to achieve the above purpose, the invention adopts the following technical scheme:
an electromagnetic scattering field calculation method of a moving time-varying dispersion medium target comprises the following steps:
step 1, modeling a dispersive medium target in a laboratory coordinate system K to obtain a modeling program; outputting a model file after the modeling program is operated; in the main process, MPI initialization is firstly carried out; then reading in model file parameters by the main process, and sequentially reading in operation parameters required by the main process operation from a screen;
step 2, in the main process, combining a target space grid component in the model file parameter and a time component in the operation parameter, and obtaining space and time components in a motion coordinate system K' by using a Lorentz transformation formula;
step 3, initializing parameters in the main process in a laboratory coordinate system K' to obtain main process initialization parameters;
step 4, according to the number of partitions in three directions in the operation parameters read in the step 1, carrying out three-dimensional region segmentation on a given FDTD calculation region in a motion coordinate system K' to obtain each subdomain; establishing a virtual topological structure by utilizing an MPI library function, and equally distributing the grid numbers in the length, width and height directions contained in each subdomain to each slave process; then entering MPI, adding all slave processes into the same FDTD group, and starting each slave process;
step 5, the main process transmits the parameters in the steps 2 to 4 to each auxiliary process through the broadcasting function in the MPI library;
step 6, firstly entering MPI from the process, and adding FDTD group; then each slave process receives the parameters transmitted by the master process; initializing parameters of each slave process respectively to obtain initialization parameters of each slave process;
step 7, transmitting boundary information of the connection boundary in the main process initialization parameter from the main process to each sub-domain, and forming a complete connection boundary in each sub-domain; determining an incident wave in a laboratory coordinate system K, defining amplitude and frequency, and obtaining the frequency and the amplitude of the incident wave in a motion coordinate system K' by Lorentz transformation;
each slave process judges the logic variable of the connection boundary according to the initialization parameters of each slave process, and judges whether the grid range of the subdomain corresponding to the process contains the connection boundary or not; if the connection boundary is included, introducing an identical transformed incident wave into a motion coordinate system K' in each sub-field including the connection boundary, if the connection boundary is not included, performing magnetic field communication between adjacent sub-fields, and then performing electric field iterative update under the FDTD method;
step 8, in the motion coordinate system K', after the step 7 is completed, each slave process performs electromagnetic field iterative operation under the FDTD method by using the initialization parameters of each slave process, and performs operation processing on a connection boundary, an absorption boundary and an output boundary;
after the calculation of each slave process is finished, the slave process sends the calculation result to the master process;
step 9, the main process receives far field data in all sub-domains containing output boundaries in the secondary process, and performs superposition and derivation processing on the far field data to obtain a total time domain electromagnetic scattering field value in a motion coordinate system K';
and (3) performing inverse Lorentz transformation on the electromagnetic field in the obtained motion coordinate system K' to obtain far-region scattered field data of K in the laboratory coordinate system, outputting a calculation result, exiting the FDTD group, exiting the MPI, and ending the timing of the program.
The invention has the following advantages:
as described above, the invention relates to an electromagnetic scattering field calculation method of a moving time-varying dispersion medium target, which is based on a parallel Finite Difference Time Division (FDTD) algorithm formed by combining an MPI library with an FDTD algorithm, and can be used for calculating the electromagnetic scattering field of the moving time-varying dispersion medium target, so that the running time of a program can be greatly shortened. The method can realize the rapid calculation of the electromagnetic scattering field of the time-varying motion dispersion medium, and is suitable for processing the electromagnetic scattering field of more three-dimensional unevenly distributed motion plasma targets. The invention greatly shortens the running time, improves the resource utilization rate and greatly expands the application range.
Drawings
FIG. 1 is a schematic diagram of incident wave direction, scattering direction and polarization direction in an embodiment of the present invention.
Fig. 2 is a schematic diagram of two coordinate systems for Lorentz transformation according to an embodiment of the present invention.
Fig. 3 is a schematic diagram of parallel FDTD computing region segmentation in an embodiment of the present invention.
Fig. 4 is a time domain waveform of a far field scattered field of a moving plasma sphere in accordance with an embodiment of the present invention.
Fig. 5 is a radar cross-section (RCS) diagram of a moving plasma sphere in an embodiment of the invention.
Fig. 6 is a main process flow chart of an electromagnetic fringe field calculation method of a moving time-varying dispersive medium target in an embodiment of the present invention.
Fig. 7 is a flow chart of a slave process of the electromagnetic fringe field calculation method of the moving time-varying dispersive medium target in the embodiment of the invention.
Detailed Description
The invention is described in further detail below with reference to the attached drawings and detailed description:
the embodiment describes an electromagnetic scattering field calculation method of a moving time-varying dispersion medium target, which is realized based on a parallel Finite Difference Time Division (FDTD) algorithm formed by combining an MPI library with the FDTD algorithm.
As shown in fig. 6 and 7, the electromagnetic scattering field calculation method of the moving time-varying dispersion medium target includes the following steps:
and step 1, modeling a dispersive medium target in a laboratory coordinate system K to obtain a modeling program.
Outputting a model file after the modeling program is operated; in the main process, MPI initialization is firstly carried out; and then reading in the model file parameters by the main process, and sequentially reading in the operation parameters required by the operation of the main process.
Step 1.1, modeling a dispersion medium target model in a laboratory coordinate system K through Fortran language, namely a modeling program, and outputting a model file object.
The model file includes the spatial grid components dx, dy, dz of the target, the region boundary Itmin, itmax, jtmin, jtmax, ktmin, ktmax of the target model, the number of media MediaNumber and media parameters for the main program to read.
dx, dy, dz denote the magnitudes of the spatial grid components in the x, y, z directions, respectively.
Where dx=dy=dz=δ, δ represents the size of the spatial grid component.
Itmin and Itmax respectively represent the minimum value and the maximum value of the boundary of the x-direction model; jtmin, jtmax represent the minimum and maximum values in the y-direction, respectively; ktmin, ktmax represent the minimum and maximum value, respectively, in the z-direction.
The medium parameters of each medium are four, including dielectric coefficient epsilon, magnetic conductivity mu, conductivity sigma and magnetic conductivity sigma m 。
i. j and k represent the grid numbers in the x, y and z directions, respectively.
i∈(Itmin,Itmax)，j∈(Jtmin,Jtmax)，k∈(Ktmin,Ktmax)。
Step 1.2. The main process reads in parameters in the model file object. Sct.
And step 1.3, reading corresponding parameters required by the running of the main program from the screen in turn.
The read-in parameters are specifically as follows:
the number of partitions of the region is calculated in three directions of x, y and z: nubdomainx, nubdomainy, nubdomainz.
II. coefficient delta/(c x dt) of the laboratory coordinate system K satisfying the Courant stability condition:
here, 2 is entered (this data is typically. Gtoreq.2).
c is the speed of light, and according to the input coefficient, dt in K in the laboratory coordinate system, i.e., dt=δ/(2*c) can be obtained.
III, incidence angle: θ i ,Scattering angle: θ s ,/>Polarization ofAngle: alpha, as shown in figure 1.
As shown in FIG. 1, θ i Is the included angle between the direction of the incident wave and the positive half axis of the z-axis in the space,is the included angle between the projection of the incident wave direction on the xoy plane and the positive half axis of the x axis. Here is provided with->
θ s Is the included angle between the scattered wave direction and the positive half axis of the z axis in space,is the included angle between the projection of the scattered wave direction on the xoy plane and the positive half axis of the x axis. />
Alpha represents the polarization angle of the incident wave, typically taken as 0 deg. and 90 deg..
For an incident wave, α=0° represents that the incident wave electric field vector is parallel to the incident plane, i.e., horizontally polarized; α=90° represents that the scattered wave electric field vector is perpendicular to the plane of incidence, i.e. vertically polarized. Here, α=0°.
IV, total time iteration step number: timestop. Here, let timestop=5000.
Since the dispersive medium (plasma) is time-varying, parameters of the plasma, such as plasma strike frequency and plasma frequency, need to be set after the time iteration begins, each slave process uses FDTD for field iterative update.
And 2, in the main process, combining the target space grid component in the model file parameter and the time component in the operation parameter, and obtaining the space and time components in the motion coordinate system K' by using a Lorentz transformation formula.
When the Lorentz transformation is used for analyzing the electromagnetic scattering characteristics of a moving object, two coordinate systems exist: one is a laboratory coordinate system K that remains stationary with the ground, and the other is a motion coordinate system K' that remains relatively stationary with the moving object.
As shown in FIG. 2, the object has the same velocity v as the motion coordinate system K ', and the motion angle θ' v ,
Here v=0.1c, where the motion angle isThus, according to relativistic covariances, the electromagnetic scattering problem of moving plasma in the laboratory coordinate system K can be transformed into the moving coordinate system K' for solution.
From the spatial step in the three directions x, y, z in the laboratory coordinate system K in step 1.1: dx, dy, dz and the time component Δt in step 1.3, transformation relationship of the spatial and time components of the laboratory coordinate system K and the motion coordinate system K':
wherein dx, dy, dz represent spatial components in three directions of x, y, z in the laboratory coordinate system K; dx ', dy', dz 'represent spatial components in three directions x, y, z in the motion coordinate system K'.
dt 'represents the time component in the motion coordinate system K'.
v x ,v y ,v z The components of the movement velocity v of the target in the x, y and z directions are respectively.
β=v/c, c is the speed of light in vacuum.
Unit vector representing scattering direction, < >>A unit vector representing the speed.
Wherein (1)>The unit vectors in the x, y and z directions are respectively represented.
Wherein θ v The included angle between the motion speed and the positive half axis of the z axis in the laboratory coordinate system K is shown;and the included angle between the motion direction in the laboratory coordinate system K and the positive half axis of the x-axis after being projected to the xoy plane is shown.
And 3, initializing parameters in the main process in a laboratory coordinate system K' to obtain main process initialization parameters.
The main process initialization parameters comprise boundary information of a connection boundary, an output boundary and an absorption boundary, an incident wave origin, a cut-off point and maximum and minimum time from the output boundary to a remote receiving point.
And 4, carrying out three-dimensional region segmentation on a given FDTD calculation region in a motion coordinate system K' according to the number of the partitions in three directions in the read operation parameters in the step 1, so as to obtain all subdomains. Establishing a virtual topological structure by utilizing an MPI library function, and equally distributing the grid numbers in the length, width and height directions contained in each subdomain to each slave process; then the MPI is entered, all slave processes are added into the same FDTD group, and each slave process is started.
The step 4 specifically comprises the following steps:
step 4.1. The number of partitions in three directions according to the input in step 1.3: nsubdomainx, nsubdomainy, nsubdomainz, performing three-dimensional region segmentation on the given FDTD calculation region to obtain each subdomain.
Step 4.2. Virtual topology is built using MPI library functions as shown in FIG. 3.
The number of each sub-field is similar to the matrix element, namely (m, n, p) is used for representing a certain sub-field, each sub-field is provided with a unique number, and the position of each sub-field in the total calculation area can be determined according to the number.
Wherein m, n and p are numbers in three directions of x, y and z respectively.
And equally distributing the grid numbers in the corresponding length, width and height directions contained in each subdomain to each slave process.
In this embodiment, automatic segmentation is adopted, and boundary information of each sub-domain can be obtained after segmentation.
The following functions in the MPI library were used in this example:
(1)MPI_Cart_create(comm_old,ndims,dims,periods,reorder,comm_cart)。
the function is a function for creating a Cartesian topology, the system constructs a topology meeting the conditions in the comm_old communication domain according to the structure in the dims array, and obtains the communication domain comm_cart with a new Cartesian topology.
(2) MPI_COMM_rank (COMM, node, ierr). The function is used to determine a process ranking in the communicator of the cartesian location. Wherein the node returns the current process sequence number and takes the values of 0,1,2 and … N-1.
(3)MPI_COMM_SIZE(comm,np,ierr)。
The function is used to determine the number of processes involved in the communicator or the total number of processes available. Where np returns the total number of processes.
(4) MPI_Cart_shift (comm, direction, disp, rank_source, rank_dest). The function can enable each slave process to determine the specific position information of the slave process in the topological structure, and the numbering conditions of the front, rear, left, right, upper and lower processes are clear.
Step 4.3, adding all slave processes into the same FDTD group;
the Nubdomainx Nubdomainy X Nubdomainz slave processes are started.
And 5, the master process transmits the parameters in the steps 2 to 4 to each slave process through the broadcasting function in the MPI library.
The parameters passed by the master process to the slave process mainly include two types:
the first is the same parameters for all processes, such as the space grid size, time step and the like in the step 2; the other is that the parameter values are different for different processes, such as boundary information of each sub-field in step 4.
In this embodiment, global communication is adopted for communication between the master process and the slave processes, and the function MPI_BCAST of one-to-many communication is utilized, and the function can transfer relevant parameters in the master process to each slave process.
The broadcasting function, i.e., mpi_bcast (buffer, count, datatype, root, comm), has the following roles:
in the designated comm communication domain, data of the number of data types in the buffer of the root process are sent to all processes in the communication domain, and other processes do not need to execute receiving operation.
And step 6, initializing parameters of each slave process respectively to obtain the initialized parameters of each slave process.
The initialization parameters in the slave process include:
the FDTD calculates coefficients of an iterative formula, judges whether a connection boundary, an absorption boundary and an output boundary are logic variables in a grid range of subdomains included in each slave process, and outputs distances from each point of the boundary to a receiving point.
Step 7, transmitting boundary information of the connection boundary in the main process initialization parameter from the main process to each sub-domain, and forming a complete connection boundary in each sub-domain; each slave process judges the logic variable of the connection boundary according to the initialization parameters of each slave process, and judges whether the grid range of the subdomain corresponding to the process contains the connection boundary or not; if the connection boundary is included, the completely same transformed incident wave is introduced into the motion coordinate system K' in each sub-field including the connection boundary, and if the connection boundary is not included, magnetic field communication is carried out between adjacent sub-fields, and electric field iterative update under the FDTD method is carried out.
In the step 7, the incident wave in the laboratory coordinate system K is determined to be Gaussian pulse, and the amplitude of the incident wave in the laboratory coordinate system K is defined as E 0 The amplitude of the incident wave in the motion coordinate system K' after Lorentz transformation is calculated according to formula (5): e's' 0 =0.9045V/m. Meanwhile, the expression of the gaussian pulse in the motion coordinate system K' after Lorentz transformation is shown as (6).
Wherein E is 0 Respectively representing the amplitude of incident waves in a laboratory coordinate system K; e's' 0 Respectively representing the amplitude of the incident wave in the motion coordinate system K'.
β=v/c, v is the speed of movement of the target, and c is the speed of light in vacuum.
Wherein v is x ,v y ,v z The components of the movement velocity v of the target in the x, y and z directions are respectively; />The unit vectors in the x, y and z directions are respectively represented.
Representing the magnetic field in the laboratory coordinate system K.
Psi is the angle between the electric field vector and the velocity vector, i.e Representing the unit vector of the electric field.
E′ i (t ') is Gaussian pulse incident wave in motion coordinate system K', t 0 Is the moment when the peak of the Gaussian pulse occurs, τ is the pulse width of the Gaussian pulse, where t is 0 ＝τ。
Unit vector representing incident wave, +.>Is a unit vector of velocity.
Step 8, in the motion coordinate system K', after the step 7 is completed, each slave process performs electromagnetic field iterative operation under the FDTD method by using the initialization parameters of each slave process, and performs operation processing on a connection boundary, an absorption boundary and an output boundary;
after the calculation of each slave process is finished, the slave process sends the calculation result to the master process.
The step 8 specifically comprises the following steps:
and 8.1, performing magnetic field communication between adjacent subfields in a motion coordinate system K' after the step 7 is completed, and performing iterative calculation of electric field intensity by utilizing magnetic field data.
Here, an ADE-FDTD method is used to implement iterative computation of the time-varying plasma electric field intensity.
Since the FDTD calculation region is divided into subfields, the connection boundary, the absorption boundary, and the output boundary are all divided into some subfields, and thus it is necessary to process the special boundaries of the connection boundary, the absorption boundary, and the output boundary. If the connection boundary is not included, a determination is made as to whether or not there is an absorption boundary.
Then according to the logic variable of the connection boundary determined from the process initialization parameters, each slave process determines whether the sub-domain contains the connection boundary, if so, an incident wave magnetic field component is added into the sub-domain containing the connection boundary to process the connection boundary of the electric field; if the connection boundary is not included, a determination is made as to whether or not there is an absorption boundary.
Step 8.2, judging whether the sub-domain contains the absorption boundary or not according to the logic variable of the absorption boundary from the process initialization parameters by each slave process; if the absorption boundary is included, calculating the absorption boundary; if the absorption boundary is not included, a determination is made as to whether or not there is an output boundary.
Step 8.3, judging whether the sub-domain contains the output boundary according to the logic variable of the output boundary from the process initialization parameters by each slave process; if the output boundary is not included, electric field communication is carried out between adjacent subfields, and FDTD iterative calculation of magnetic field intensity is carried out by utilizing electric field data; if the output boundary is included, performing far-field extrapolation on the output boundary to obtain far-field data in all sub-domains including the output boundary, performing electric field communication between adjacent sub-domains, and performing FDTD iterative calculation of magnetic field intensity by using the electric field data.
For the calculation of the far field on the extrapolation boundary, specifically:
(1) After the FDTD total calculation area is divided into subdomains, the output boundaries are distributed in different subdomains; for the output boundary, the FDTD electromagnetic field iteration of each step needs to calculate the contribution of the equivalent electromagnetic current on the output boundary at the moment to the far field.
(2) Since the contributions of the points on the output boundary to the far zone are in a linear superposition relationship, the contributions of the equivalent electromagnetic currents on the output boundary to the far zone field are calculated on the sub-fields containing the output boundary.
And 8.4, after the magnetic field value is obtained through iterative calculation in the step 8.3, judging whether the sub-domain contains the connection boundary according to the logic variable of the connection boundary from the process initialization parameter.
If the connection boundary is included, an incident wave electric field component is added into the subdomain containing the connection boundary to process the connection boundary of the magnetic field; if the connection boundary is not included, judging whether the total time iteration step number is ended.
And 8.5, finally, judging whether the total time step is iterated to be ended, and if the time iteration is not ended, repeating the steps 7 to 8 until the time iteration is ended. After the calculation of all the slave processes is completed, the slave processes send the calculation results to the master process. At this point, each slave process exits the FDTD group and exits the MPI.
In step 8, the MPI_SENSRECV function and the MPI_BARRER function in the MPI library are mainly used for message transfer between subfields. Since the positions of the electric field and magnetic field components in the Yee grid cross each other by half a grid, when dividing the subfields, adjacent subfields overlap by half a grid, so that the magnetic field (electric field) of the adjacent subfields needs to be in the upper half a time step before the field (magnetic field) iteration of the subfield process. This requires data transfer between adjacent subfields, and synchronous message transfer between processes, mainly using the mpi_send function.
At the same time, the MPI_BARRER function is also used to synchronize the processes. After each process in the FDTD group completes the task of the sub-domain, an MPI_BARRER function is called, then the process is in a waiting state, and all processes in the FDTD group synchronously enter the next working task after calling the MPI_BARRER function.
And 9, receiving far field data in all sub-domains containing output boundaries in the slave process by the master process, and carrying out superposition and derivation processing on the far field data to obtain a total time domain electromagnetic scattering field value in a motion coordinate system K'.
And (3) performing inverse Lorentz transformation on the electromagnetic field in the obtained motion coordinate system K' to obtain far-field scattered field data of K in the laboratory coordinate system, and outputting the far-field scattered field data of K in the laboratory coordinate system, namely outputting the result to a file scate. Dat far-field scattered field, and outputting the data of an excitation source to a source. Dat file. At this point, the main process exits the FDTD group, exits the MPI, and the program ends the timer. The electromagnetic field in the motion coordinate system K' is subjected to inverse Lorentz transformation, and the electromagnetic field value is transformed into a laboratory coordinate system K by the formula:
wherein,respectively representing the electric and magnetic fields in the laboratory coordinate system K,/for>Representing the electric and magnetic fields in the motion coordinate system K', respectively; />
Wherein v is x ,v y ,v z The components of the velocity of the object in the x, y and z directions are respectively.
Beta = v/c, v is the targetThe speed of movement, c, is the speed of light in vacuum.
As shown in fig. 4, after obtaining the scan.dat data file, the Origin drawing software is opened, and the data is imported into book1, so as to obtain the far-field fringe field time domain waveform.
As shown in fig. 5, the obtained remote field scattering electric field file scate. Dat and excitation source file source. Dat are imported into a program for calculating the target RCS, the field values are subjected to an FFT algorithm to obtain the target RCS file RCS. Dat, and then the RCS data is imported into book2, thereby obtaining the target RCS. Meanwhile, the present embodiment records the running time of the different FDTD calculation region decomposition schemes, respectively, as shown in the following table 1.
| Partition(s) | 1×1×1 | 1×1×4 | 1×4×1 | 2×2×1 |
| Time | 529.4700 | 150.1528 | 148.4874 | 149.1687 |
| Speed-up ratio | 1 | 3.5262 | 3.5657 | 3.5494 |
As can be seen from table 1, the run time of the program is also different for different partitioning schemes, but compared to a 1 x 1 partition, the running time of the parallel FDTD method is greatly shortened, so that the running time of a program can be greatly shortened, and the resource utilization efficiency is improved. When the parallel FDTD method is used for carrying out time-varying motion dispersion medium electromagnetic scattering calculation, a mode of synchronously coordinating and calculating by a plurality of computers can be adopted, so that the running time and the running efficiency of a program can be further improved.
The method of the invention includes, but is not limited to, time-varying moving plasma spheres, plasma cylinders, cones, etc., and may also include, for example, complex shaped dispersive medium objects, such as combinations of cones and hemispheres, and missile models with non-uniform tail flames, etc. The method of the invention greatly shortens the operation time, improves the utilization efficiency of resources and greatly expands the application range.
The foregoing description is, of course, merely illustrative of preferred embodiments of the present invention, and it should be understood that the present invention is not limited to the above-described embodiments, but is intended to cover all modifications, equivalents and alternatives falling within the spirit and scope of the present invention as defined by the appended claims.
Claims (8)
1. The electromagnetic scattering field calculation method of the moving time-varying dispersion medium target is characterized by comprising the following steps of:
step 1, modeling a dispersive medium target in a laboratory coordinate system K to obtain a modeling program;
outputting a model file after the modeling program is operated; in the main process, MPI initialization is firstly carried out, then the main process reads in model file parameters, and the running parameters required by the running of the main process are sequentially read in from a screen;
the operation parameters required by the main process in operation comprise the number of partitions of the region in three directions, time components, incidence angles, polarization angles, scattering angles and total time steps in a laboratory coordinate system K;
step 2, in the main process, combining a target space grid component in the model file parameter and a time component in the operation parameter, and obtaining space and time components in a motion coordinate system K' by using a Lorentz transformation formula;
step 3, initializing parameters in the main process in a motion coordinate system K' to obtain main process initialization parameters; the main process initialization parameters comprise boundary information of a connection boundary, an output boundary and an absorption boundary, an incident wave origin, a cut-off point and maximum and minimum time from the output boundary to a remote receiving point;
step 4, according to the number of partitions in three directions in the operation parameters read in the step 1, carrying out three-dimensional region segmentation on a given FDTD calculation region in a motion coordinate system K' to obtain each subdomain; establishing a virtual topological structure by utilizing an MPI library function, and equally distributing the grid numbers in the length, width and height directions contained in each subdomain to each slave process;
then entering MPI, adding all slave processes into the same FDTD group, and starting each slave process;
step 5, the main process transmits the parameters in the steps 2 to 4 to each auxiliary process through the broadcasting function in the MPI library;
step 6, firstly entering MPI from the process, and adding FDTD group; then each slave process receives the parameters transmitted by the master process, and each slave process is initialized to obtain the initialized parameters of each slave process; the initialization parameters of the slave process comprise coefficients of an FDTD calculation iteration formula, and whether a connection boundary, an absorption boundary and an output boundary are in logic variables in a grid range of subdomains contained in each slave process or not and the distances from each point of the output boundary to a receiving point are judged;
step 7, transmitting boundary information of the connection boundary in the main process initialization parameter from the main process to each sub-domain, and forming a complete connection boundary in each sub-domain; determining an incident wave in a laboratory coordinate system K, defining amplitude and frequency, and obtaining the frequency and the amplitude of the incident wave in a motion coordinate system K' by Lorentz transformation;
step 8, in a motion coordinate system K', each slave process performs electromagnetic field iterative operation under the FDTD method by using the initialization parameters of each slave process, and performs operation processing on a connection boundary, an absorption boundary and an output boundary;
after the calculation of each slave process is finished, the slave process sends the calculation result to the master process;
and 9, receiving far field data in all sub-domains containing output boundaries in the slave process by the master process, superposing and deriving the far field data to obtain a total time domain electromagnetic scattering field value in a motion coordinate system K', obtaining the far field scattering field data in a laboratory coordinate system K through inverse Lorentz transformation, and outputting the result.
2. The method of electromagnetic fringe field calculations for a moving time-varying dispersive medium object of claim 1,
the step 1 specifically comprises the following steps:
step 1.1, modeling a dispersive medium target in a laboratory coordinate system K through Fortran language to obtain a modeling program; outputting a model file object. Sct after the modeling program is operated;
the model file parameters include target space grid components dx, dy, dz, target region boundaries Itmin, itmax, jtmin, jtmax, ktmin, ktmax, media number MediaNumber and media parameters;
dx, dy and dz respectively represent the sizes of the space grids in the x, y and z directions;
itmin and Itmax respectively represent the minimum value and the maximum value of the x direction; jtmin, jtmax represent the minimum and maximum values in the y-direction, respectively; ktmin, ktmax represent the minimum and maximum value, respectively, in the z-direction;
the dielectric parameters of each medium are four, namely dielectric coefficient epsilon, magnetic conductivity mu, conductivity sigma and magnetic conductivity sigma m ；
Step 1.2, the main process reads in model file parameters in the model file object. Sct;
and step 1.3, sequentially reading the operation parameters required by the operation of the main process, wherein the operation parameters comprise the number of partitions of the calculated area in three directions, the time component, the incidence angle, the polarization angle, the scattering angle and the total time step in a laboratory coordinate system K.
3. The method of electromagnetic fringe field calculations for a moving time-varying dispersive medium object of claim 1,
in the step 2, the transformation relationship between the spatial and temporal components of the laboratory coordinate system K and the motion coordinate system K' is as follows:
wherein dx, dy, dz represent spatial components in three directions of x, y, z in the laboratory coordinate system K; dx ', dy', dz ', represent spatial components in three directions of x, y and z in a motion coordinate system K';
dt represents the time component in the laboratory coordinate system K, dt 'represents the time component in the motion coordinate system K';
v x ,v y ,v z the components of the movement velocity v of the target in the x, y and z directions are respectively;
beta = v/c, c is vacuumIs the speed of light in (a);
unit vector representing scattering direction, < >>A unit vector representing a speed;
wherein (1)>The unit vectors in the x, y and z directions are respectively represented;
wherein θ v The included angle between the motion speed and the positive half axis of the z axis in the laboratory coordinate system K is shown;and the included angle between the motion direction in the laboratory coordinate system K and the positive half axis of the x-axis after being projected to the xoy plane is shown.
4. The method of electromagnetic fringe field calculations for a moving time-varying dispersive medium object of claim 1,
the step 4 specifically comprises the following steps:
step 4.1, carrying out three-dimensional region segmentation on the FDTD calculation region according to the number of the partitions in three directions;
step 4.2, establishing a virtual topological structure by utilizing an MPI library function, and representing the number of a certain subdomain after the FDTD calculation region is subjected to three-dimensional region segmentation by using (m, n, p); wherein m, n and p are numbers in three directions of x, y and z respectively;
equally distributing the grid numbers in the three directions of the corresponding length, width and height of each subdomain to each slave process;
step 4.3, adding all slave processes into the same FDTD group;
starting Nubdomainx X Nubdomainy X Nubdomainz slave processes; wherein Nsubdomainx, nsubdomainy, nsubdomainz represents the number of partitions in the x, y, z directions, respectively.
5. The method of electromagnetic fringe field calculations for a moving time-varying dispersive medium object of claim 1,
in the step 5, global communication is adopted for communication between the master process and the slave process; the related parameters in the master process are transferred to the respective slave processes using the function mpi_bcast of one-to-many communication.
6. The method of electromagnetic fringe field calculations for a moving time-varying dispersive medium object of claim 1,
in the step 7, the incident wave in the laboratory coordinate system K is determined, the amplitude and the frequency are defined, and the frequency and the amplitude of the incident wave in the motion coordinate system K' after Lorentz transformation are as follows:
wherein omega i And E is 0 Representing the frequency and amplitude of the incident wave in the laboratory coordinate system K; omega i ' and E 0 'represents the frequency and amplitude of the incident wave in the motion coordinate system K';unit vector representing incident wave, +.>Is a unit vector of velocity;
β=v/c, v being the speed of movement of the target, c being the speed of light in vacuum;
wherein v is x ,v y ,v z The components of the movement velocity v of the target in the x, y and z directions are respectively; />The unit vectors in the x, y and z directions are respectively represented;
representing the magnetic field in the laboratory coordinate system K;
psi is the angle between the electric field vector and the velocity vector, i.e Representing the unit vector of the electric field.
7. The method of electromagnetic fringe field calculations for a moving time-varying dispersive medium object of claim 1,
the step 8 specifically comprises the following steps:
step 8.1, after the step 7 is completed, performing magnetic field communication between adjacent subfields, performing iterative calculation of electric field intensity by utilizing magnetic field data, and performing iterative calculation of time-varying plasma electric field intensity by adopting an ADE-FDTD method; then, according to the logic variable of the connection boundary determined from the process initialization parameters, each slave process determines whether the sub-domain contains the connection boundary;
if the connection boundary is included, adding an incident wave magnetic field component into the subdomain containing the connection boundary to process the connection boundary of the electric field; if the connection boundary is not included, judging whether an absorption boundary exists or not;
step 8.2, judging whether the sub-domain contains the absorption boundary or not according to the logic variable of the absorption boundary from the process initialization parameters by each slave process; if the absorption boundary is included, calculating the absorption boundary; if the absorption boundary is not included, judging whether an output boundary exists or not;
step 8.3, judging whether the sub-domain contains the output boundary according to the logic variable of the output boundary from the process initialization parameters by each slave process; if the output boundary is not included, electric field communication is carried out between adjacent subfields, and FDTD iterative calculation of magnetic field intensity is carried out by utilizing electric field data; if the output boundary is included, the far field is extrapolated on the output boundary, after the far field data in all the sub-domains including the output boundary are obtained, electric field communication is carried out between adjacent sub-domains, and FDTD iterative calculation of the magnetic field intensity is carried out by utilizing the electric field data;
step 8.4, after the magnetic field value is obtained through iterative computation in the step 8.3, judging logic variables of the connection boundaries according to the process initialization parameters, and judging whether the connection boundaries are contained in the sub-domain or not according to each process; if the connection boundary is included, an incident wave electric field component is added into the subdomain containing the connection boundary to process the connection boundary of the magnetic field; if the connection boundary is not included, judging whether the total time iteration step number is finished;
step 8.5, if the time iteration is not finished, repeating the steps 7 to 8 until the time iteration is finished; after the calculation of all the slave processes is completed, the slave processes send calculation results to the master process; and then exits the FDTD group and exits the MPI.
8. The method of electromagnetic fringe field calculations for a moving time-varying dispersive medium object of claim 1,
in the step 9, the electromagnetic field in the motion coordinate system K' is subjected to inverse Lorentz transformation, and the electromagnetic field value is transformed into the laboratory coordinate system K by the formula:
wherein,respectively representing the electric and magnetic fields in the laboratory coordinate system K,/for>Representing the electric and magnetic fields in the motion coordinate system K', respectively; />
Wherein v is x ,v y ,v z The components of the movement speed of the target in the x, y and z directions are respectively;
β=v/c, v is the speed of movement of the target, and c is the speed of light in vacuum.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310754740.3A CN116720409B (en) | 2023-06-21 | 2023-06-21 | Electromagnetic scattering field calculation method for moving time-varying dispersion medium target |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310754740.3A CN116720409B (en) | 2023-06-21 | 2023-06-21 | Electromagnetic scattering field calculation method for moving time-varying dispersion medium target |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116720409A CN116720409A (en) | 2023-09-08 |
| CN116720409B true CN116720409B (en) | 2024-03-26 |
Family
ID=87871361
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310754740.3A Active CN116720409B (en) | 2023-06-21 | 2023-06-21 | Electromagnetic scattering field calculation method for moving time-varying dispersion medium target |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116720409B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101770542A (en) * | 2010-02-25 | 2010-07-07 | 中国科学院上海光学精密机械研究所 | Method for simulating electromagnetic wave propagation by cluster computer |
| CN107944113A (en) * | 2017-11-17 | 2018-04-20 | 西北工业大学 | A kind of method for calculating three-dimensional high-speed translation Electromagnetic Scattering of Target field |
| CN111339688A (en) * | 2020-03-27 | 2020-06-26 | 南京工程学院 | Method for solving rocket simulation model time domain equation based on big data parallel algorithm |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12052518B2 (en) * | 2020-07-23 | 2024-07-30 | University Of Utah Research Foundation | Multi-modal computational imaging via metasurfaces |
| CN114781220A (en) * | 2022-04-27 | 2022-07-22 | 安徽大学 | Electromagnetic field simulation method based on sub-grid and single-step ADI-FDTD |
-
2023
- 2023-06-21 CN CN202310754740.3A patent/CN116720409B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101770542A (en) * | 2010-02-25 | 2010-07-07 | 中国科学院上海光学精密机械研究所 | Method for simulating electromagnetic wave propagation by cluster computer |
| CN107944113A (en) * | 2017-11-17 | 2018-04-20 | 西北工业大学 | A kind of method for calculating three-dimensional high-speed translation Electromagnetic Scattering of Target field |
| CN111339688A (en) * | 2020-03-27 | 2020-06-26 | 南京工程学院 | Method for solving rocket simulation model time domain equation based on big data parallel algorithm |
Non-Patent Citations (3)
| Title |
|---|
| Lorentz变换普遍形式的推导;肖恩蓉;《南京理工大学学报(自然科学版)》;19870830(第04期);第18-23页 * |
| 并行FDTD方法在海面及其上方漂浮目标复合电磁散射中的应用;贾春刚 等;《电波科学学报》;20160229;第31卷(第01期);第116-122页 * |
| 电各向异性介质FDTD并行算法的研究;杨利霞 等;《电波科学学报》;20060425;第21卷(第01期);第43-48页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116720409A (en) | 2023-09-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tezduyar et al. | Flow simulation and high performance computing | |
| Jamshidi et al. | Real time UAV path planning by parallel grey wolf optimization with align coefficient on CAN bus | |
| Peterkin et al. | A virtual prototyping environment for directed-energy concepts | |
| Desell et al. | An analysis of massively distributed evolutionary algorithms | |
| CN114528716B (en) | Efficient numerical simulation method applied to multi-scale electromagnetic wave problem analysis | |
| He et al. | Solving electromagnetic scattering problems with tens of billions of unknowns using GPU accelerated massively parallel MLFMA | |
| Wang et al. | Simulations of ion thruster plume–spacecraft interactions on parallel supercomputer | |
| WO2020068244A2 (en) | General scattered field simulator | |
| CN117540586B (en) | Multidisciplinary multilevel coupling simulation method taking hyperstimulation as core | |
| CN116720409B (en) | Electromagnetic scattering field calculation method for moving time-varying dispersion medium target | |
| Zhao et al. | Parallel hybrid method of HOMoM–MLFMA for analysis of large antenna arrays on an electrically large platform | |
| Araki et al. | Dynamic load balancing with over decomposition in plasma plume simulations | |
| US5933794A (en) | Scalable parallel processing apparatus and method for performing fast multipole scattering calculations | |
| Yang et al. | Analysis of the calculation of a plasma sheath using the parallel SO‐DGTD method | |
| Gao et al. | Parallel shooting and bouncing ray method on GPU clusters for analysis of electromagnetic scattering | |
| CN110021339B (en) | Cluster parallel computing acceleration method based on protein folding calculation protein structure | |
| Hong et al. | Easy-hardware-implementation MMPF for maneuvering target tracking: Algorithm and architecture | |
| Yamaguchi et al. | Fast finite element analysis method using multiple GPUs for crustal deformation and its application to stochastic inversion analysis with geometry uncertainty | |
| Weatherill et al. | Towards fully parallel aerospace simulations on unstructured meshes | |
| Shrestha et al. | Multi-level domain-decomposition strategy for solving the eikonal equation with the fast-sweeping method | |
| Su et al. | Parallel integral equation based nonoverlapping DDM for solving electrically-large scattering problems in half space | |
| Namburu et al. | Scalable electromagnetic simulation environment | |
| CN111353261B (en) | GPU optimization method for immersion boundary high-precision statistical dynamic fluid simulation | |
| Ahuja et al. | A Parallel Finite-Volume Runge–Kutta Algorithm for Electromagnetic Scattering | |
| Wang et al. | A fast RCS simulation method for dynamic target based on Kriging interpolation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |