CN102609990B - Massive-scene gradually-updating algorithm facing complex three dimensional CAD (Computer-Aided Design) model - Google Patents
Massive-scene gradually-updating algorithm facing complex three dimensional CAD (Computer-Aided Design) model Download PDFInfo
- Publication number
- CN102609990B CN102609990B CN201210009810.4A CN201210009810A CN102609990B CN 102609990 B CN102609990 B CN 102609990B CN 201210009810 A CN201210009810 A CN 201210009810A CN 102609990 B CN102609990 B CN 102609990B
- Authority
- CN
- China
- Prior art keywords
- model
- scene
- simplification
- magnanimity
- lod
- 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.)
- Expired - Fee Related
Links
Landscapes
- Processing Or Creating Images (AREA)
Abstract
The invention relates to a massive-scene gradually-updating algorithm facing a complex three dimensional CAD (Computer-Aided Design) model. The algorithm comprises the following steps of: using different files to respectively save scene tree data, model-inputting geometric data and simplified LOD (Levels of Detail) model data; using a double-layer BSP (Binary Space Partition) scene tree structure to fast update the whole scene space index; and adopting a parallel online structure technology of massive three-dimensional scene LOD models based on multiple GPU (Graphics Processing Unit) hardware platforms of a PC (Personal Computer) to fast structure the LOD. The algorithm has the advantages that the scene updating speed can be quickened, and the purpose of online interactive rendering is achieved.
Description
Technical field
The invention belongs to computer graphics and multimedia and technical field of virtual reality, particularly relate to a kind of gradual update algorithm of magnanimity scene towards complex three-dimensional cad model.
Background technology
In industries such as machine-building (as aircraft manufacturing, high ferro locomotive manufacturing and shipbuilding), so more and more that to employ complex product Informationization Manufacture technology, particularly utilization rate Virtual Prototype Technique in the life cycle of product design, manufacture.
Virtual Prototype Technique is a kind of digitalized design method based on product computer simulation model, and to simulate from vision, the sense of hearing, sense of touch and function and behavior be really the technology of product.Complex Products Virtual Prototype Technology merges advanced modeling/emulation technology, modern information technologies, advanced design manufacturing technology and modern management technology further in each field CAX (as CAD, CAM, CAE etc.)/DFX (as DFA, DFM etc.) technical foundation, these technology are applied to complex product Life cycle, total system, and integrated management is carried out to them and emphasizes virtual, to analyze from the aspect of system, the systematized engineering design of one of Simulation of Complex product and management method.The exploitation key of complex product Life cycle virtual prototype is exactly the emulation of complex product three-dimensional model and the visual edit technology under policy environment.And the model data amount that complex model self exists is large, structural topology is complicated etc., and feature all brings very large challenge to computer software and hardware.The simplation visualizing technology of current complex three-dimensional model mainly contains two kinds:
One is the drafting complexity reducing model scene by exporting sensitive technology.The major technique of this respect has: the data assignment that visibility culling, simplification, level of detail, buffer memory are consistent and spatial index and online updating etc. thereof.
Two is adopt parallel drawing technology, utilizes computer cluster to replace single computing machine to process model.Current parallel drawing technology is mainly divided into sort-first (Data distribution8 type drawing parallel graphic), sort-middle (render instruction level walks abreast), sort-last (rendered pixel level walks abreast) three major types.
More existing large-scale and complex scenes browing systems at present, as by the GigaWalk of card university, the FarVoxels of Italy CRS4 Visual calculation group, the iWalk of Princeton University, the vLOD of Hope Jin Si university, Saarland university exploitation based on the extensive cad model drawing system of shared drive formula multi-CPU architecture and the AdaptiveVoxels system etc. of Zhejiang University.
Although magnanimity three-dimensional model rendering technique obtains pay attention to widely, its Theories and methods is still not mature enough, greatly limit its application.First, existing magnanimity model scene tree is all only applicable to static scene.Scene tree must be recalculated for needing the scene carrying out dynamically updating.Secondly, existing external memory File Technology is not supported to carry out incremental update preservation to contextual data, and thus the renewal of scene can not persistence.Finally, lack the online simplification technology to more new scene, thus can not support the drafting of more new scene.
Summary of the invention
Technique effect of the present invention can overcome above-mentioned defect, provides a kind of gradual update algorithm of magnanimity scene towards complex three-dimensional cad model.Its online drafting function aspects after magnanimity three-dimensional CAD data scene update achieves breakthrough, forms verification system.
For achieving the above object, the present invention adopts following technical scheme: it comprises the following steps:
(1) system architecture supporting scene online updating is built;
(2) by different files saving scenario tree data respectively, mode input geometric data and simplification LOD model data;
(3) upgrade whole scene space index fast, make it remain valid in rendering accelerating algorithm;
(4) the magnanimity three-dimensional scenic LOD model parallel on-line constructing technology of Based PC many GPU hardware platform.
The present invention is intended to study the gradual update algorithm of magnanimity scene towards complex three-dimensional cad model.Based on scene tree renewal technology, the incremental update of research uniformly subdivision index and the simplification problem of LOD model parallel computation, be devoted to solve efficiency that external memory file data upgrades, the reconstruction and optimization problem of spatial index when upgrading and online LOD simplify Mission Scheduling.Adopt block-based external memory file structure and corresponding update algorithm, adaptive spatial index update algorithm and the simplification task scheduling algorithm based on priority, final online drafting after magnanimity three-dimensional CAD data scene update functionally makes a breakthrough, and forms verification system.
Mentality of designing of the present invention is: for the magnanimity three-dimensional scenic after renewal, for avoiding the whole uniformly subdivision index of reconstruct and LOD model, by reducing the renewal set of uniformly subdivision index node and the LOD node set of simplification may being needed, reduce the calculated amount of scene update, rapid voxel LOD Model Simplification Method is obtained in conjunction with based on many GPU, accelerate scene renewal speed, reach the object that online interaction is drawn.
In described step (1), scenes in parallel update algorithm flow process aspect, inserts spatial index dynamic subdivision algorithm after out-of-core algorithm, inserts parallel shortcut calculation after visibility culling algorithm; Hardware structure aspect, the parallel simplification deployment scheme that the present invention adopts is: the many GPU based on a PC are parallel simplifies framework.
In described step (2), comprise one and support the out-of-core algorithm that magnanimity three-dimensional model upgrades, step is as follows:
A external memory Divide File is the block of fixed size by (), be then sent in these blocks of files by the geometric data in enclosure body after object subdivision;
B () opens up enough large file, calculate an existing collection of objects nearest with newly adding object distance;
C this set is divided into several subsets according to the blocks of files at its geometric data place by (), set up according to distance the relation of object and these blocks of files of newly adding, and the closest blocks of files of relation of choosing is inserted and newly added object.
In described step (3), when magnanimity three-dimensional scenic upgrades, use the double-deck BSP scene tree structure in conjunction with coarseness scenario B SP and particulate object BSP; When carrying out object editor, object BSP structure remains unchanged, and only upgrades the scenario B SP of coarseness; When drawing, fine-grained object BSP ensures the efficiency of visibility culling algorithm, adopts lazy update strategy to upgrade scenario B SP, the BSP node that reconstruct was lost efficacy.
In described step (4), the magnanimity modeling rendering system based on parallel simplification is that traditional magnanimity modeling rendering system and spatial index dynamic subdivision and parallel simplification combine.The data of scene tree in internal memory are revised when wherein spatial index dynamically updates, parallel shortcut calculation collects the potential visible renewal node of current view point, these nodes are formed potential simplification with its father node in LOD hierarchical structure together with child node collect (PSS, Potentially Simplification Set).Simplify hardware platform according to many GPU are parallel, need optimum distribution to simplify task multiple simplification in end.
Innovation of the present invention is mainly reflected in:
(1) break through the thinking only considering static magnanimity modeling rendering in the past, propose the magnanimity three-dimensional scenic renewal technology based on blocks of files data store strategy;
(2) be different from the method for existing reconstruct whole scene space subdivision index, propose the method refusing to upgrade magnanimity model space index.Accelerating keeping spatial index, on the basis of operation validity and speed, to reduce the calculated amount of spatial index reconstruct, decreasing the LOD node quantity needing to upgrade simultaneously;
(3) based on many GPU hardware environment, online parallel construction voxel LOD model, the magnanimity three-dimensional model supporting input responsive upgrades to be drawn.
Beneficial effect of the present invention is: the update algorithm of magnanimity cad model editor scene tree of the present invention is applicable to the drafting of dynamic scene, and preserves Data Update with the external memory file layout that supported data upgrades; Uniformly subdivision index delta renewal technology of the present invention can the application on site demand of satisfying magnanimity three-dimensional scenic; Parallel simplification technology based on GPU sampling of the present invention carrys out the LOD model algorithm of more new scene, finally can support that the renewal of magnanimity model scene is drawn.
Accompanying drawing explanation
Fig. 1 is scenes in parallel update algorithm flow process of the present invention;
Fig. 2 is the parallel simplification system hardware framework of Based PC of the present invention;
Fig. 3 is potential simplification collection (PSS) of the present invention.
Embodiment
The gradual update algorithm of magnanimity scene towards complex three-dimensional cad model of the present invention, comprises the following steps:
(1) system architecture supporting scene online updating is built.
Magnanimity modeling rendering system based on parallel simplification is the combination of traditional magnanimity modeling rendering system and spatial index dynamic subdivision and parallel simplification technology, wherein spatial index dynamic subdivision needs the data of amendment scene tree in internal memory, therefore insertion algorithm flow process after out-of-core algorithm, and parallel shortcut calculation needs could work after obtaining potential simplification collection (PSS), therefore insertion algorithm flow process after visibility culling algorithm, scenes in parallel update algorithm flow process as shown in Figure 1.
In hardware structure, the parallel simplification deployment scheme that the present invention proposes is: the many GPU based on a PC are parallel simplifies framework.On current main-stream PC platform, load the GPU graphics process hardware of more than 2 or 2, thus make full use of the powerful ability in sampling of GPU to carry out parallel LOD Construction of A Model, hardware is disposed as shown in Figure 2.
(2) be different from single massive data files structure, the present invention distinguishes saving scenario data by different files, mode input geometric data and simplification LOD model data.
Mode input geometric data and simplification LOD model data are mass data, are the renewal rewards theory supporting magnanimity scene, when expressing scene tree, use the space structure of two BSP.First each model BSP in scene is split into enclosure body, then to enclosure bodies all in scene cluster generate whole scene BSP tree.Supporting the out-of-core algorithm that magnanimity three-dimensional model upgrades for one that adopts the present invention to propose, improving the buffering hit rate when carrying out drawing data access.
Support that the out-of-core algorithm that magnanimity three-dimensional model upgrades is:
A external memory Divide File is the block of fixed size by (), be then sent in these blocks of files by the geometric data in enclosure body after object subdivision;
B (), according to priori (estimating the data volume of scene) to scene, need open up enough large file in advance, join in the process of scene by new object, calculate an existing collection of objects nearest with newly adding object distance;
C this set is divided into several subsets according to the blocks of files at its geometric data place by (), set up according to distance the relation of object and these blocks of files of newly adding, and the closest blocks of files of relation of choosing is inserted and newly added object.Thus guarantee in the ever-increasing process of model, also close apart from the arrangement of close object geometry on hard disk;
D () equally also there will be the situation of Delete Objects when scene update, at this moment again external memory file fast in there will be fragment, thus after the renewal rewards theory carrying out some, we need carry out defragmentation to external memory file.With the class of algorithms of new family object seemingly, each object in block will be organized when carrying out defragmentation and take out, and regard it as and newly add object, according to interpolation algorithm, it be added in external memory file again.
(3) upgrade whole scene space index fast, make it remain valid in rendering accelerating algorithm.
When magnanimity three-dimensional scenic upgrades, the uniformly subdivision index being used for carrying out acceleration drawing can lose efficacy, and we use the double-deck BSP scene tree structure in conjunction with coarseness scenario B SP and particulate object BSP.When carrying out object editor, object BSP structure remains unchanged, and only upgrades the scenario B SP of coarseness.When drawing, fine-grained object BSP ensure that the efficiency of visibility culling algorithm, and when upgrading scenario B SP, we adopt lazy update strategy simultaneously, only reconstruct the BSP node of inefficacy.
Spatial index node is divided into three kinds of state: lazy (subdivision is treated in expression), vacant (inclusion dignity sheet number is less than setting value) and confirmed (through reconstruct).When adding object, first the enclosure body of the object be added is put in existing uniformly subdivision structure, each enclosure body judges whether from top to bottom in the bounding box of current uniformly subdivision node from root node, if not, then put it in the enclosure body queue of the father node of current node, and father node is labeled as lazy.If then sent in child's node and done same judgement, by that analogy, until joined in leaf node, if the quantity that leaf node comprises enclosure body has exceeded subdivision setting value, then mark it for lazy, and add corresponding leaf editor node in the set of the leaf editor node of leaf node.From scene during case-deleted models, generate enclosure body by object subdivision and find corresponding BSP leafy node, disconnect contacting of itself and leafy node, if BSP leafy node was originally masked as confirmed, then leafy node is labeled as vacant, represent that it is what have vacant position, be lazy if original, keep lazy constant.When mobility model, in moving process, do not revise the enclosure body of object and the index relative of uniformly subdivision node.After movement, first each enclosure body is judged whether also in original BSP leaf node, if still, then do nothing, if gone out original BSP leafy node, then first remove the relation with original BSP leafy node, and it is upwards pushed away, until shift onto in a node within its bounding box, then, be labeled as lazy.If root node bounding box do not hold mobile after model, then to put it in scene tree in an independent enclosure body queue, and be lazy by whole uniformly subdivision structure tag.
After in BSP structure, the state of node is determined, system, along with specific operation (as draw) is when traversal uniformly subdivision structure, adaptively to upgrade uniformly subdivision structure.When traveling through BSP structure, all lazy nodes run into are reconstructed and are denoted as confirmed.When carrying out spatial index reconstruct to lazy node, we use the BSP generating algorithm based on SAH, to expect to obtain having the space index structure having acceleration most.
(4) the magnanimity three-dimensional scenic LOD model parallel on-line constructing technology of Based PC many GPU hardware platform.
Magnanimity modeling rendering system based on parallel simplification is that traditional magnanimity modeling rendering system and spatial index dynamic subdivision and parallel simplification combine.Wherein spatial index dynamically updates needs the data of amendment scene tree in internal memory, parallel shortcut calculation collects the potential visible renewal node of current view point, these nodes are formed potential simplification with its father node in LOD hierarchical structure together with child node collect (PSS, PotentiallySimplification Set) (as shown in Figure 3).Simplify hardware platform according to many GPU are parallel, need optimum distribution to simplify task multiple simplification in end.After the uniformly subdivision topology update of scene is complete, range decrease on the basis of renewal node, according to visibility judge such as the ken cone rejecting, occlusion cullings that current view point expands, the current quantity needing to upgrade LOD model node is reduced from the degree of depth, just can obtain potential simplification collection (PSS), then we carry out simplify processes to the LOD model node in PSS.On the basis of the existing drawing system based on observability, the reconstruction calculations of scene LOD model after adopting a unified simplification framework to process to upgrade.Set up in the parallel simplification task scheduling engine of system and have to simplify a task queue, each simplification task agent contains: uniformly subdivision subtree root node to be simplified, need the error being simplified model, and one simplifies task priority mark, simplify task queue according to the arrangement of simplification task priority ascending order.Operationally, the LOD model that do not complete in visible oBject (namely drawn in former frame object) queue is endowed the highest priority.Then be the LOD model do not completed in PVS queue, the simplification priority of the model in PSS queue is minimum.While in utilization, this simplifies queue, we still judge with lru algorithm the service efficiency simplifying task result.Based on this thought, in simplification queue, first we simplify those and be modified time longer model, and for the model be frequently modified recently, then wait until its finally amendment result again it is simplified after determining.When carrying out simplification task matching, the parallel task scheduling engine that simplifies takes out the highest simplification task of a priority at every turn from task queue, is distributed to idle simplification end.
Each simplification end (namely on GPU), we used rapid voxel LOD Construction of A Model algorithm (AdaptiveVoxels) based on GPU sampling, within the shortest time, construct the voxel LOD model meeting specification error.
Claims (1)
1., towards the gradual update method of magnanimity scene of complex three-dimensional cad model, it is characterized in that comprising the steps:
(1) system architecture supporting scene online updating is built;
(2) by different files saving scenario tree data, mode input geometric data and simplification LOD model data respectively;
(3) upgrade whole scene space index fast, make it remain valid in rendering accelerating algorithm;
(4) the online building method of magnanimity three-dimensional scenic LOD model of Based PC many GPU hardware platform, support to export responsive magnanimity three-dimensional model interactive editor and real-time update drafting, described step (4) is specific as follows:
A. the magnanimity modeling rendering method based on parallel simplification is existing magnanimity modeling rendering method and spatial index dynamic subdivision and the combination simplified that walks abreast, and simplify hardware platform according to many GPU are parallel, in multiple simplification end, optimum allocation simplifies task;
B. after the uniformly subdivision topology update of scene is complete, range reduces and upgrades node, the ken cone expanded according to current view point is rejected, these visibility judging methods of occlusion culling, the current quantity needing to upgrade LOD model node is reduced from the degree of depth, obtain potential simplification collection PSS, then simplify processes is carried out to the LOD model node in PSS;
C. on the basis of the existing drawing system based on observability, in the parallel simplification task scheduling engine of system, set up one simplify task queue, each simplification task agent contains: simplify uniformly subdivision subtree root node, need the error being simplified model, and one simplifies task priority mark, simplify task queue according to the arrangement of simplification task priority ascending order;
D. operationally priority is from high to low: the model in the LOD model that not completing in visible oBject queue does not complete in LOD model, PVS queue, PSS queue; Meanwhile, judge with lru algorithm the service efficiency simplifying task result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210009810.4A CN102609990B (en) | 2012-01-05 | 2012-01-05 | Massive-scene gradually-updating algorithm facing complex three dimensional CAD (Computer-Aided Design) model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210009810.4A CN102609990B (en) | 2012-01-05 | 2012-01-05 | Massive-scene gradually-updating algorithm facing complex three dimensional CAD (Computer-Aided Design) model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102609990A CN102609990A (en) | 2012-07-25 |
| CN102609990B true CN102609990B (en) | 2015-04-22 |
Family
ID=46527330
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210009810.4A Expired - Fee Related CN102609990B (en) | 2012-01-05 | 2012-01-05 | Massive-scene gradually-updating algorithm facing complex three dimensional CAD (Computer-Aided Design) model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102609990B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103713938A (en) * | 2013-12-17 | 2014-04-09 | 江苏名通信息科技有限公司 | Multi-graphics-processing-unit (GPU) cooperative computing method based on Open MP under virtual environment |
| CN103942306B (en) * | 2014-04-18 | 2017-04-19 | 重庆市勘测院 | Three-dimensional city model self-adaption scheduling method |
| KR102244619B1 (en) * | 2014-09-30 | 2021-04-26 | 삼성전자 주식회사 | Method for generating and traverse acceleration structure |
| CN105677743B (en) * | 2015-12-30 | 2019-06-28 | 南京邮电大学 | Environmental pollution by water three-dimensional fast display method and system |
| CN106023315B (en) * | 2016-05-10 | 2019-07-12 | 国网电力科学研究院武汉南瑞有限责任公司 | Three-dimensional entity model fine granularity multiresolution stage division in Distributed Virtual Reality System |
| CN107745785B (en) * | 2017-09-27 | 2019-05-03 | 中国船级社 | A kind of best Calculation of stowage on board method of loading instrument of bulk freighter and oil carrier |
| CN108765538B (en) * | 2018-06-16 | 2022-03-22 | 桂林理工大学 | Method for hierarchical rendering of OSGB data based on CAD platform |
| CN110853152B (en) * | 2019-11-14 | 2024-01-30 | 上海未高科技有限公司 | Subdivision cutting loading method for three-dimensional oversized scene |
| CN112337093B (en) * | 2021-01-08 | 2021-05-25 | 成都完美时空网络技术有限公司 | Virtual object clustering method and device, storage medium and electronic device |
| CN114708140B (en) * | 2022-06-07 | 2022-09-13 | 中国电建集团华东勘测设计研究院有限公司 | Real-time interactive sectioning visualization method for numerical calculation model based on unit |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101593361A (en) * | 2009-06-25 | 2009-12-02 | 北京航空航天大学 | A kind of large-scale terrain rendering system based on double-layer nested grid |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6346956B2 (en) * | 1996-09-30 | 2002-02-12 | Sony Corporation | Three-dimensional virtual reality space display processing apparatus, a three-dimensional virtual reality space display processing method, and an information providing medium |
-
2012
- 2012-01-05 CN CN201210009810.4A patent/CN102609990B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101593361A (en) * | 2009-06-25 | 2009-12-02 | 北京航空航天大学 | A kind of large-scale terrain rendering system based on double-layer nested grid |
Non-Patent Citations (1)
| Title |
|---|
| 海量三维模型的交互绘制;田丰林;《中国博士学位论文全文数据库》;20110815(第8期);27-77 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102609990A (en) | 2012-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102609990B (en) | Massive-scene gradually-updating algorithm facing complex three dimensional CAD (Computer-Aided Design) model | |
| US11087052B2 (en) | Automatic partitioning of a 3D scene into a plurality of zones processed by a computing resource | |
| Kim et al. | Feature-based simplification of boundary representation models using sequential iterative volume decomposition | |
| CA2919229C (en) | Temperature coupling algorithm for hybrid thermal lattice boltzmann method | |
| CN105261066B (en) | A kind of three-dimensional geographic information system real-time rendering multithreading distribution and control method | |
| CN108563808A (en) | The design method of heterogeneous reconfigurable figure computation accelerator system based on FPGA | |
| US20110282641A1 (en) | Method and system for real-time particle simulation | |
| JP2014191826A (en) | Multi-component computational fluid dynamic simulation | |
| CN103021016B (en) | Magnanimity three-dimensional geographic information data processing method in a kind of render process | |
| CN102385762A (en) | Modelica integrated three-dimensional scene simulation system | |
| Zheng et al. | An improved local remeshing algorithm for moving boundary problems | |
| US8935129B1 (en) | System and method for simplifying a graph'S topology and persevering the graph'S semantics | |
| Chen et al. | Improvements in the reliability and element quality of parallel tetrahedral mesh generation | |
| CN103714192A (en) | Adaptive R-tree based large-data-volume three-dimensional railway design model rendering method | |
| CN105243238A (en) | Integrated quick product iteration forming device and method | |
| Mavriplis | Viscous flow analysis using a parallel unstructured multigrid solver | |
| Qi et al. | An enhanced sweep and prune algorithm for multi-body continuous collision detection | |
| CN113378279A (en) | Building modeling method, building modeling system, electronic equipment and storage medium | |
| Aldrich et al. | Collision-Driven Volumetric Deformation on the GPU. | |
| Dai et al. | An open platform of shape design optimization for shell structure | |
| CN113392348A (en) | BIM-based tunnel main body structural steel IFC2x3 data visualization method | |
| Ghisi et al. | Impact of tetrahedralization on parallel conforming octree mesh generation | |
| CN105335520A (en) | Data processing method based on subway integrated automation system and processor | |
| Jie et al. | LOD methods of large-scale urban building models by GPU accelerating | |
| CN110543690B (en) | Parameter changing method for digital prototype of complex product model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150422 Termination date: 20170105 |