CN108268697A - A kind of high efficiency electric propulsion plume plasma parallel simulation method - Google Patents
A kind of high efficiency electric propulsion plume plasma parallel simulation method Download PDFInfo
- Publication number
- CN108268697A CN108268697A CN201711387842.7A CN201711387842A CN108268697A CN 108268697 A CN108268697 A CN 108268697A CN 201711387842 A CN201711387842 A CN 201711387842A CN 108268697 A CN108268697 A CN 108268697A
- Authority
- CN
- China
- Prior art keywords
- node
- vertex
- high efficiency
- electric propulsion
- simulation method
- 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
Links
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]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of high efficiency electric propulsion plume plasma parallel simulation method, including：Non-directed graph G is generated according to the adjacency matrix M of grid node；It reorders, and numbers to the vertex of non-directed graph G, obtain the sequence array of minimum bandwidth and profile；According to the sequence array, particle data structure design is carried out, realizes the solution to Large sparse equations.The derivation of the large-scale matrix and vector to heavy construction Flow Field Calculation problem is realized by the present invention, the calculating time in the numerical simulation of plasma plume is greatly reduced, improves efficiency.
Description
Technical field
The invention belongs to plasma technical field of value simulation more particularly to a kind of high efficiency electric propulsion plume plasmas
Parallel simulation method.
Background technology
By numerical simulation means, the various physical parameters of electric propulsion work plume are can be derived that, so as to calculate
Go out influence each other effect of the plume to spacecraft.However, electric propulsion space plume includes complicated continuous media stream (plume core
Heart district), several flow regimes such as Transitional stream and free molecule flow (plume external zones).Plume core space is continuous flow, outside
It encloses for Transitional stream and free molecule flow, high-altitude plume flowing is extremely complex, it is difficult to be described with a kind of method.
By taking currently used directly simulation monte carlo method as an example, directly simulating monte carlo method is imitated with limited
True molecule replace true molecular, and in a computer storage emulation molecule position coordinates, velocity component and interior energy, value with
The collision emulated between the movement of molecule and the effect on boundary and emulation molecule changes, and is emulated finally by counting in grid
The motion state of molecule realizes the simulation to actual gas flow field problem.However, due to plasma plume numerical simulation by
It is limited in its time scale and space scale, needs to expend a large amount of calculating time, therefore, it is necessary to before computational accuracy is ensured
It puts, the time is calculated to reduce using various accelerated methods.
As it can be seen that one of the most urgent problems to be solved by those skilled in the art is：A kind of high efficiency electric propulsion plume etc. is provided
Gas ions parallel simulation method.
Invention content
The technology of the present invention solves the problems, such as：Overcome the deficiencies of the prior art and provide a kind of high efficiency electric propulsion plume etc. from
Daughter parallel simulation method realizes the derivation of the large-scale matrix and vector to heavy construction Flow Field Calculation problem, significantly
The calculating time in the numerical simulation of plasma plume is reduced, improves efficiency.
In order to solve the above-mentioned technical problem, the invention discloses a kind of high efficiency electric propulsion plume plasma parallel artificials
Method, including：
Non-directed graph G is generated according to the adjacency matrix M of grid node；
It reorders, and numbers to the vertex of non-directed graph G, obtain the sequence array of minimum bandwidth and profile；
According to the sequence array, particle data structure design is carried out, realizes the solution to Large sparse equations.
In above-mentioned high efficiency electric propulsion plume plasma parallel simulation method, to the vertex of non-directed graph G into rearrangement
Sequence, and number, the sequence array of minimum bandwidth and profile is obtained, including：
Determine the border vertices set r of non-directed graph G；
An optional node assigns x1, and the hierarchical structure of non-directed graph G is generated using vertex x1 as root；
To being in the node x1 on the 1st layer marked as x1 '=1, and the sequence increased according to level is successively in each layer
Node continuous label obtains vertex number { xi ' }；
{ xi ' } is numbered according to vertex, determines the corresponding target label list { xi " } of optional node；
Each node in the vertex set r of boundary is resolved respectively, determines the corresponding target label list of each node；
According to the corresponding target label list of each node, screening obtains the node of bandwidth minimum, and construction obtains the sequence
Array.
In above-mentioned high efficiency electric propulsion plume plasma parallel simulation method, the sequence increased according to level is right successively
Node continuous label in each layer, including：
When node is located at identical layer, the node for selecting adjacent upper layer node label small is preferentially into line label.
In above-mentioned high efficiency electric propulsion plume plasma parallel simulation method, the sequence increased according to level is right successively
Node continuous label in each layer, including：
When node is located at identical layer, the node for selecting node number adjacent around node small is preferentially into line label.
In above-mentioned high efficiency electric propulsion plume plasma parallel simulation method, { xi ' } is numbered according to vertex, determines to appoint
The corresponding target label list { xi " } of node is selected, including：
{ xi ' } is numbered according to vertex, according to equation below, the corresponding target label list of optional node is calculated
{xi”}：
Xi "=N+1-xi '
Wherein, N represents node number.
The present invention has the following advantages：
(1) with the development of space technology, the configuration of spacecraft is increasingly complicated, to space plasma numerical simulation and analysis
Application demand will be more and more, and the present invention targetedly proposes a kind of high efficiency electric propulsion plume plasma parallel artificial
Method, the high efficiency electric propulsion plume plasma parallel simulation method overcome conventionally employed full particle simulation method calculating etc.
The computational efficiency problem of ion, meanwhile, the sequencing for being conducive to the finite element topological relation based on unstrctured grid is realized, can be more
Specific incorporation engineering problem is added to carry out specific spacecraft surface plasma characteristics of motion research.
(2) in the present invention, complicated electricity is pushed away using the high efficiency electric propulsion plume plasma parallel simulation method
Carry out multiple programming into plume, the optimization to OpenMP parallel programming models and calculating is realized, so that OpenMP is shared
Storage system provides a portable, telescopic model, and it is portable good, efficient to have the advantages that.
(3) in the present invention, Large Scale Sparse is solved using the high efficiency electric propulsion plume plasma parallel simulation method
Equation group has preferable Parallel Computing Performance.
Description of the drawings
Fig. 1 is a kind of bandwidth schematic diagram of the Large Scale Sparse matrix before optimization in the embodiment of the present invention；
Fig. 2 is a kind of step flow of high efficiency electric propulsion plume plasma parallel simulation method in the embodiment of the present invention
Figure；
Fig. 3 is a kind of molecule traversal loop pattern of the particle simulation program of Linked Storage Structure in the embodiment of the present invention
Schematic diagram；
Fig. 4 is the corresponding non-directed graph of a kind of grid node adjacency matrix in the embodiment of the present invention；
Fig. 5 is a kind of schematic diagram of hierarchical structure with node 4 for root generation in the embodiment of the present invention；
Fig. 6 is a kind of smporary label schematic diagram in the embodiment of the present invention；
Fig. 7 is a kind of smporary label schematic diagram in the embodiment of the present invention；
Fig. 8 is a kind of bandwidth schematic diagram of the Large Scale Sparse matrix after optimization in the embodiment of the present invention.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, it is public to the present invention below in conjunction with attached drawing
Embodiment is described in further detail.
The present invention provides a kind of high efficiency electric propulsion plume plasma parallel simulation methods, pass through algorithm optimal implementation
The solution of extensive system of linear equations.With reference to Fig. 1, the Large Scale Sparse square before a kind of optimization in the embodiment of the present invention is shown
The bandwidth schematic diagram of battle array.As it can be seen that no matter using which kind of global stiffness matrix storage mode, the bandwidth and profile of condensation matrix are to linear
The solution efficiency of equation group is all of great significance.Wherein, the node serial number of maximum bandwidth and grid each unit is closely related, right
The node serial number answered is different, and maximum bandwidth differs greatly.When the number difference of an affiliated node of grid is larger, will directly lead
Cause maximum bandwidth increase.Therefore, appropriate cell node number can be such that the nonzero element in Bulk stiffness matrix arranges as far as possible
In the closer location on diagonal both sides, the time is calculated so as to save the space occupied and reduce.For only depositing the total firm of non-zero entry
Matrix sparse storage mode entirely is spent, the non-zero entry introduced in triangle decomposition is all in bandwidth and profile, therefore, reduces band
Wide and profile can reduce the generation of filling member.On the one hand the storage capability of neutral element can be directly reduced, is avoided in excessively occupancy
It deposits；On the other hand, to substantially reduce flops, calculating speed is improved.Based on graph theory principle, certain can be pressed to the vertex of figure
Kind sequence renumbers, i.e., reorders to the adjacency matrix ranks of figure, to obtain minimum bandwidth and profile.
High efficiency electric propulsion plume plasma parallel simulation method of the present invention, by searching out a sequence number
Group so that the bandwidth of matrix and profile are smaller after being reordered with this array to matrix, and hierarchical structure is longer.The high efficiency electricity pushes away
It is based on the breadth first search of non-directed graph G boundary points, to generate longest layer into plume plasma parallel simulation method
Secondary structure is mesh calibration method：First, the border vertices r of non-directed graph G is sought, using border vertices r as root, the layer of generation non-directed graph G
Secondary structure；The sequence increased according to level generates Cuthill-Mckee sequence numbers successively to the node reordering in each layer
Group finally arranges number inverted order, generates inverse Cuthill-Mckee sequences array.
With reference to Fig. 2, a kind of high efficiency electric propulsion plume plasma parallel simulation method in the embodiment of the present invention is shown
Step flow chart.In the present embodiment, the high efficiency electric propulsion plume plasma parallel simulation method, including：
Step 101, non-directed graph G is generated according to the adjacency matrix M of grid node.
Step 102, it reorders, and numbers to the vertex of non-directed graph G, obtain the sequence number of minimum bandwidth and profile
Group.
In the present embodiment, it reorders, and numbers to the vertex of non-directed graph G, obtain the row of minimum bandwidth and profile
Ordinal number group can specifically include：Determine the border vertices set r of non-directed graph G；An optional node, assign x1, using vertex x1 as
The hierarchical structure of root generation non-directed graph G；To being in the node x1 on the 1st layer marked as x1 '=1, and increase according to level suitable
Sequence to the node continuous label in each layer, obtains vertex number { xi ' } successively；{ xi ' } is numbered according to vertex, determines optional node
Corresponding target label list { xi " }；Each node in the vertex set r of boundary is resolved respectively, determines that each node corresponds to
Target label list；According to the corresponding target label list of each node, screening obtains the node of bandwidth minimum, and construction obtains institute
State sequence array.
Preferably, in the sequence increased according to level successively to the node continuous label in each layer when, it then follows following principle：
1st, when node is located at identical layer, the node for selecting adjacent upper layer node label small is preferentially into line label.2nd, when node is located at
During identical layer, the node for selecting node number adjacent around node small is preferentially into line label.
Preferably, { xi ' } is numbered according to vertex, determines the corresponding target label list { xi " } of optional node, it specifically can be with
Including：{ xi ' } is numbered according to vertex, according to equation below, the corresponding target label list { xi " } of optional node is calculated：
Xi "=N+1-xi '
Wherein, N represents node number.
Step 103, according to the sequence array, particle data structure design is carried out, is realized to Large sparse equations
It solves.
Particle simulation method and the difference of traditional Field Flow Numerical Simulation method maximum are, generation one is needed in program
The particle of fixed number amount and the information for needing to store each particle, this just brings difficulty to data store organisation.In this implementation
In example, after the proper sequence array of optimization processing, particle data structure design can be carried out, and then realize to Large sparse equations
Solution (e.g., the large-scale matrix of heavy construction Flow Field Calculation problem and vector derivation), save calculate memory.
With reference to Fig. 3, a kind of molecule traversal of the particle simulation program of Linked Storage Structure in the embodiment of the present invention is shown
The schematic diagram of circulation pattern.Such as Fig. 3, the particle simulation program of the Linked Storage Structure is stored using the single-track link table with head node
Particle information only establishes a chained list for each particle.The particle simulation program of the Linked Storage Structure is towards particle
, it takes advantage in Particles Moving and BORDER PROCESSING, but in particle encounter and sample phase, the access of particle properties is base
In unit, so needing to rescan entire chained list, and the local chained list based on unit is set up.Molecular data only and institute
It is associated in grid cell, having cast out molecule, totally mark and its rearrangement process, traversal and the operation of molecule are based on list
Member, that is to say, that movement, the collision of molecule can be carried out with sampling in unit.Thus cast out molecule totally mark and
Its process of resequencing, has saved the calculating time.The particle simulation program of the Linked Storage Structure is convenient for node and is inserted into behaviour
Make, a pointer for being directed toward chained list tail portion is assigned in each grid cell, one molecular process of addition is also fairly simple, first
Memory space is distributed for each molecule dynamic for being inserted into, the tail portion of unit chained list where being then inserted into, using taking the lead
The double linked list storage organization of tail node and interim tail portion node.
Based on above-described embodiment, below by being further described for a concrete application scene.
(1) input parameter is determined：The adjacency matrix M [N] [N] of grid node, wherein, N is node number.
(2) non-directed graph G, such as Fig. 4 are generated according to the adjacency matrix M of grid node, shown a kind of in the embodiment of the present invention
The corresponding non-directed graph of grid node adjacency matrix.
(3) the border vertices set r of non-directed graph G, and an optional node are determined, assigns x1.
Such as Fig. 4, in the present embodiment, border vertices set r={ 1,2,3,4,5,7,8,9 }, the start node x1 of selection
=4.
(4) hierarchical structure of figure G is generated using vertex x1 as root, such as Fig. 5 shows that one kind is with node in the embodiment of the present invention
4 schematic diagram for the hierarchical structure of root generation.
(5) to being in the node x1 on the 1st layer marked as x1 '=1, the sequence increased according still further to level is successively in each layer
Node continuous label.
In the present embodiment, for same node layer, weight label is carried out according to following principle：Upper layer node mark adjacent thereto
Number smaller node elder generation label；Spending smaller node elder generation label, (" degree " refers to：Adjacent node number around node).
Wherein, xi ' can be defined as smporary label, and smporary label list is：{ xi ' }={ 9,6,3,1,2,5,4,7,8 },
Such as Fig. 6, a kind of smporary label schematic diagram in the embodiment of the present invention is shown.
(6) formula is calculated as follows in above-mentioned vertex number { xi ' } and obtains sequence array：Xi "=N+1-xi ', and
To final label list { xi " }={ Isosorbide-5-Nitrae, 7,9,8,5,6,3,2 }, such as Fig. 7, a kind of grid section in the embodiment of the present invention is shown
The label result schematic diagram again of point.
(7) remaining each point in side week vertex set r of non-directed graph G is compared each in set r using step (2)-(6)
The corresponding bandwidth of node therefrom selects bandwidth reckling.Such as Fig. 8, the big rule after a kind of optimization in the embodiment of the present invention are shown
The bandwidth schematic diagram of mould sparse matrix.By Fig. 1 and Fig. 8 comparisons it can be seen that bandwidth substantially reduces.
In conclusion the present invention targetedly proposes a kind of high efficiency electric propulsion plume plasma parallel artificial side
Method, the high efficiency electric propulsion plume plasma parallel simulation method overcome conventionally employed full particle simulation method calculate etc. from
The computational efficiency problem of son, meanwhile, the sequencing for being conducive to the finite element topological relation based on unstrctured grid is realized, can be more
Specific incorporation engineering problem carries out specific spacecraft surface plasma characteristics of motion research.
Secondly, in the present invention, using the high efficiency electric propulsion plume plasma parallel simulation method to complicated electricity
Plume is promoted to carry out multiple programming, the optimization to OpenMP parallel programming models and calculating is realized, so that OpenMP is common
It enjoys storage system and one portable, telescopic model is provided, it is portable good, efficient to have the advantages that.
In addition, it in the present invention, is solved using the high efficiency electric propulsion plume plasma parallel simulation method large-scale dilute
Equation group is dredged, there is preferable Parallel Computing Performance.
Each embodiment in this explanation is described by the way of progressive, the highlights of each of the examples are with its
The difference of his embodiment, just to refer each other for identical similar part between each embodiment.
The above, best specific embodiment only of the invention, but protection scope of the present invention is not limited thereto,
Any one skilled in the art in the technical scope disclosed by the present invention, the change or replacement that can be readily occurred in,
It should be covered by the protection scope of the present invention.
The content not being described in detail in description of the invention belongs to the known technology of professional and technical personnel in the field.
Claims (5)
1. a kind of high efficiency electric propulsion plume plasma parallel simulation method, which is characterized in that including：
Non-directed graph G is generated according to the adjacency matrix M of grid node；
It reorders, and numbers to the vertex of non-directed graph G, obtain the sequence array of minimum bandwidth and profile；
According to the sequence array, particle data structure design is carried out, realizes the solution to Large sparse equations.
2. high efficiency electric propulsion plume plasma parallel simulation method according to claim 1, which is characterized in that nothing
It reorders, and number to the vertex of figure G, obtains the sequence array of minimum bandwidth and profile, including：
Determine the border vertices set r of non-directed graph G；
An optional node assigns x1, and the hierarchical structure of non-directed graph G is generated using vertex x1 as root；
To being in the node x1 on the 1st layer marked as x1 '=1, and the sequence increased according to level is successively to the node in each layer
Continuous label obtains vertex number { xi ' }；
{ xi ' } is numbered according to vertex, determines the corresponding target label list { xi " } of optional node；
Each node in the vertex set r of boundary is resolved respectively, determines the corresponding target label list of each node；
According to the corresponding target label list of each node, screening obtains the node of bandwidth minimum, and construction obtains the sequence array.
3. high efficiency electric propulsion plume plasma parallel simulation method according to claim 2, which is characterized in that according to
The sequence that level increases successively to the node continuous label in each layer, including：
When node is located at identical layer, the node for selecting adjacent upper layer node label small is preferentially into line label.
4. high efficiency electric propulsion plume plasma parallel simulation method according to claim 2, which is characterized in that according to
The sequence that level increases successively to the node continuous label in each layer, including：
When node is located at identical layer, the node for selecting node number adjacent around node small is preferentially into line label.
5. high efficiency electric propulsion plume plasma parallel simulation method according to claim 2, which is characterized in that according to
{ xi ' } is numbered on vertex, determines the corresponding target label list { xi " } of optional node, including：
{ xi ' } is numbered according to vertex, according to equation below, the corresponding target label list { xi " } of optional node is calculated：
Xi "=N+1-xi '
Wherein, N represents node number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711387842.7A CN108268697A (en) | 2017-12-20 | 2017-12-20 | A kind of high efficiency electric propulsion plume plasma parallel simulation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711387842.7A CN108268697A (en) | 2017-12-20 | 2017-12-20 | A kind of high efficiency electric propulsion plume plasma parallel simulation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108268697A true CN108268697A (en) | 2018-07-10 |
Family
ID=62772481
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711387842.7A Pending CN108268697A (en) | 2017-12-20 | 2017-12-20 | A kind of high efficiency electric propulsion plume plasma parallel simulation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108268697A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112329201A (en) * | 2020-09-30 | 2021-02-05 | 兰州空间技术物理研究所 | Integrated ionic electric propulsion simulation model and method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101979969A (en) * | 2010-09-09 | 2011-02-23 | 南昌航空大学 | Multi-channel digital signal processor-based flow field reconstruction computing method |
| CN103440163A (en) * | 2013-09-09 | 2013-12-11 | 中国科学院近代物理研究所 | PIC (Peripheral Interface Controller)-model-based accelerator simulation method implemented by using GPU (Graphic Processing Unit) in parallel |
| US20150165550A1 (en) * | 2007-09-14 | 2015-06-18 | Robert C. Fry | Analytical laser ablation of solid samples for ICP, ICP-MS, and FAG-MS analysis |
| CN105843781A (en) * | 2015-02-02 | 2016-08-10 | 国际商业机器公司 | Matrix ordering for cache efficiency in performing large sparse matrix operations |
| CN107016180A (en) * | 2017-03-30 | 2017-08-04 | 中国石油大学(北京) | A kind of particle flow emulation mode |
-
2017
- 2017-12-20 CN CN201711387842.7A patent/CN108268697A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150165550A1 (en) * | 2007-09-14 | 2015-06-18 | Robert C. Fry | Analytical laser ablation of solid samples for ICP, ICP-MS, and FAG-MS analysis |
| CN101979969A (en) * | 2010-09-09 | 2011-02-23 | 南昌航空大学 | Multi-channel digital signal processor-based flow field reconstruction computing method |
| CN103440163A (en) * | 2013-09-09 | 2013-12-11 | 中国科学院近代物理研究所 | PIC (Peripheral Interface Controller)-model-based accelerator simulation method implemented by using GPU (Graphic Processing Unit) in parallel |
| CN105843781A (en) * | 2015-02-02 | 2016-08-10 | 国际商业机器公司 | Matrix ordering for cache efficiency in performing large sparse matrix operations |
| CN107016180A (en) * | 2017-03-30 | 2017-08-04 | 中国石油大学(北京) | A kind of particle flow emulation mode |
Non-Patent Citations (3)
| Title |
|---|
| 吴建平 等: "《稀疏线性方程组的高效求解与并行计算》", 31 May 2004 * |
| 吴建平 等: "排序对重叠区域分解型并行ILU的影响分析", 《计算机工程与应用》 * |
| 林骁雄 等: "基于非结构网格的离子推力器羽流数值模拟方法研究", 《航天器环境工程》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112329201A (en) * | 2020-09-30 | 2021-02-05 | 兰州空间技术物理研究所 | Integrated ionic electric propulsion simulation model and method |
| CN112329201B (en) * | 2020-09-30 | 2022-11-04 | 兰州空间技术物理研究所 | Integrated ionic electric propulsion simulation model and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109255829A (en) | Mixing level for ray trace | |
| CN103970960B (en) | The element-free Galerkin structural topological optimization method accelerated parallel based on GPU | |
| Fietz et al. | Optimized hybrid parallel lattice Boltzmann fluid flow simulations on complex geometries | |
| Zabelok et al. | Adaptive kinetic-fluid solvers for heterogeneous computing architectures | |
| Acharya et al. | A parallel and memory efficient algorithm for constructing the contour tree | |
| CN103345580B (en) | Based on the parallel CFD method of lattice Boltzmann method | |
| JP2013037691A (en) | System, method, and computer program product for constructing acceleration structure | |
| CN110135584A (en) | Extensive Symbolic Regression method and system based on self-adaptive parallel genetic algorithm | |
| CN102193837B (en) | Front line-based method for detecting collision of bounding volume | |
| CN102521854A (en) | Parallel flow line placing method applicable to two-dimensional flow field | |
| CN103064819A (en) | Method for utilizing microwave integrated circuit (MIC) to rapidly achieve lattice Boltzmann parallel acceleration | |
| CN106709503A (en) | Large spatial data clustering algorithm K-DBSCAN based on density | |
| CN104504468A (en) | Three-dimensional box loading method based on three-dimensional moving mode sequence and memetic algorithm | |
| CN110096838A (en) | A kind of helicopter flow field numerical value Parallel Implicit method for solving based on N-S equation | |
| Aykanat et al. | Adaptive decomposition and remapping algorithms for object-space-parallel direct volume rendering of unstructured grids | |
| CN106484532B (en) | GPGPU parallel calculating method towards SPH fluid simulation | |
| CN108268697A (en) | A kind of high efficiency electric propulsion plume plasma parallel simulation method | |
| CN106846457B (en) | Octree parallel construction method for CT slice data visual reconstruction | |
| CN116795508A (en) | Method and system for scheduling resources of tiled accelerator | |
| CN114925627B (en) | Helicopter flow field numerical simulation system and method based on graphic processor | |
| CN106780747A (en) | A kind of method that Fast Segmentation CFD calculates grid | |
| CN110135725A (en) | A kind of cable assembly sequence-planning method, device and equipment | |
| Mavriplis | Viscous flow analysis using a parallel unstructured multigrid solver | |
| Liu et al. | Massively parallel CFD simulation software: CCFD development and optimization based on Sunway TaihuLight | |
| Malakar et al. | A divide and conquer strategy for scaling weather simulations with multiple regions of interest |
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 |
Application publication date: 20180710 |
|
| RJ01 | Rejection of invention patent application after publication |