Proceedings of JSPE Semestrial Meeting
2006 JSPE Spring Meeting
Displaying 1-50 of 647 articles from this issue
  • Yuki Asada, Koichi Matsuzaki, Hiromasa Suzuki, Tomoyuki Fujimori
    Session ID: A64
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Aiming at mesh segmentation applicable to reverse engineering, we propose a method to refine segmentation result of a simple and robust segmentation algorithm called variational shape approximation. In a process of reverse engineering, polygon data obtained from point group is first segmented and then parametric surface is pasted onto each region. Since variational shape approximation method only allows plane as geometric proxy of a surface region, a region representable with one or two parametric surface is often divided into several regions. Some works extend proxies to other primitive shape for refining segmentation. However, calculation cost increases to compute error metric for each shape and also smooth border can′t be obtained. In this paper, we solve these problems by using geometric features of each regions resulted from variational shape approximation and demonstrate its effectiveness with some example.
    Download PDF (509K)
  • Tomohiro Mizoguchi, Hiroaki Date, Satoshi kanai, Takeshi Kishinami
    Session ID: A65
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    We proposed segmentation method for a scanned mesh using Watershed and Region Growing method based on analytical surface fitting. In this paper, we propose an improved mesh segmentation method based on analytical surface fitting and region combinating. Our method can extract analytical surfaces including tori and their boundary edges, and recognize fillet, linear sweep, and surfaces of revolution.
    Download PDF (633K)
  • Tadashi Hiraoka, Hiromasa Suzuki, Tomoyuki Fujimori, Koichi Matsuzaki
    Session ID: A66
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    We need automatic high–quality freeform surfaces generation algorithm, in order to convert quickly the dense point sets produced by laser scanners into useful geometric models. We use the method of moving least squares(MLS) to make the density of the points constant, and to generate smooth points. Thus we can generate a high–quality freeform surface. In this paper, we present a procedure for generating constant spacing points from a set of scanned 3D points, and evaluate a MLS–surface under various conditions.
    Download PDF (181K)
  • Tetsuo Oya, Yukinori Tada, Masatake Higashi
    Session ID: A67
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    In order to remove noise on a mesh, smoothing techniques have been developed. We have proposed the smoothing method with the discrete Laplacian implemented in the diffusion process. For the feature preservation, the target model is segmented by the detected feature lines in our method. However, such treatment will lead to the problem that user must deal with many feature lines for a huge complex model. In this paper, we adopt the anisotropic diffusion flow for more efficient feature preservation. Simulation results show that the presented method worked well on a closed and an open model with preserving its features.
    Download PDF (581K)
  • Tomoyuki Fujimori, Yohei Kobayashi, Hiromasa Suzuki
    Session ID: A68
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    We propose a new technique to analyze welded machine parts. Our problem is that we cannot distinguish a welded place in CT (Computed Tomography) data. This is a principle problem of CT technology. However, analyses of those parts are heavily desired in industrial applications. Therefore, we propose a technique consist of three processes. At first, we extract medial cells from a whole voxel model. Next, we separate the model by using distance analysis and apply the medial cells extraction method once again. Finally, we implement the technique on a standard PC and demonstrate its effectiveness.
    Download PDF (812K)
  • Toshiki Miyagawa, Shoichi Kakuda, Daiki Iriya, Yoshiaki Iikawa, Hirosh ...
    Session ID: A69
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
  • Hiroaki Date, Satoshi Kanai, Takeshi Kishinami, Ichiro Nishigaki
    Session ID: A73
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    In this paper, a new mesh combining method for supporting efficient CAE is proposed. Our method consists of three steps: 1) searching the elements on the surfaces shared with different meshes, 2) inserting elements for connecting two surfaces of the input meshes, and 3) mesh simplification for the inserted elements in order to obtain conforming mesh at the shared surface. This mesh combining method is applicable to the meshes that have no–coincident elements and geometries on the surfaces to be shared, and it is reversible. This allows us to support some operations for executing CAE, such as mesh generation of the assembly models and analysis condition setting for the meshes.
    Download PDF (332K)
  • Takashi Michikawa, Takashi Kanai, Hiromasa Suzuki
    Session ID: A74
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Approximate Rigid Interpolation for tetrahedral meshes requires meshes without negating tetrahedrons. However, it is difficult to construct such meshes and this is a problem for rigid interpolation. We present a method for rigid interpolation for tetrahedral meshes with negating tetrahedrons. Our method computes pseudo rotation function for negating tetrahedrons from neighbor tetrahedrons. This enables us to compute rigid interpolation for tetrahedral meshes with negating tetrahedrons.
    Download PDF (101K)
  • Akinori Kondo, Yoshikazu Kobayasi, Kenji Shirai
    Session ID: A75
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    The cloth expression of Computer Graphics used texture mapping. As for this method, the feeling of quality of the cloth is not expressible. It proposes the method of making the cloth from the string that composes the cloth in the method of expressing the feeling of quality of the cloth. The made cloth texture of three dimensions is done and the mapping is done by the solid texture method. Moreover, the method of the mapping of the parametric surface is designed.
    Download PDF (786K)
  • Yu Sato, Yoshikazu Kobayashi, Kenji Shirai
    Session ID: A76
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    When the surface texture is designed by CAD, a genuine expression is not done by the texture where it is drawn. Then, the purpose of study is a thing to examine the method of doing the rendering that becomes shape when designing to the texture made by CAD. The geometric data which was drawn up as the expedient, by the surface texture CAD system was read, by the fact that light ray calculation is done accurately making use of Ray tracing, real rendering of texture was done.
    Download PDF (896K)
  • Shin'ichi Kawashima, Yoshikazu Kobayashi, Kenji Shirai
    Session ID: A77
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    This paper proposes development of a generation system which makes the extensive surface texture using stitching.Multiple narrow domain surface textures that measured with microscope–type surface profiling instruments change a large field of surface textures using stitching.Using measured data,it resolves the issue in which texture cant depict refined shapes finely. PC–cluster constructs a parallel processing system.The system operate in parallel. As a result, textures created with the system are more refined than conventional textures. Moreover, the execution of the generation with the parallel processing is more efficient than doing with a serial processing.
    Download PDF (629K)
  • Rei Hino
    Session ID: B06
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Production resources such as machine tools, robots, AGVs, and buffers are analyzed and modeled in order to make detail schedules in practical production manufacturing systems. This study copes with a semi–finished product stored on the production resource after finishing its operation.
    Download PDF (121K)
  • Toru Eguchi, Masashi Dougakiuchi, Fuminori Oba, Takeshi Murayama
    Session ID: B07
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    We have already proposed a dynamic job shop scheduling method using the mixture of a genetic algorithm and a priority rule. So far, we found the due date allowance plays an important role in the method. However, in the previous method, the due date allowance was not included explicitly in the performance function optimized periodically. In this report, we examine the effect of the due date allowance by including it into the performance function explicitly, and show its effect using numerical experiments.
    Download PDF (340K)
  • Yoshitaka Tanimizu, Tsuyoshi Miyamae, Yasuhiro Maeda, Tatsuhiko Sakagu ...
    Session ID: B08
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Unscheduled disruptions often occur in the actual manufacturing systems, and an initial production schedule may not satisfy the constraints due to the disruptions. A reactive scheduling method based on Genetic Algorithm (GA) was, therefore, proposed to improve the disturbed production schedule in the previous research. This research deals with a new crossover method to improve the performance of the GA based reactive scheduling process for the tardiness minimization problems and the total flow time minimization problems. The multi–objective reactive scheduling method is also proposed based on the improved reactive scheduling method in this research. The combination of the dominance of genes is discussed for the multi–objective reactive scheduling problems of the tardiness minimization and the total flow time minimization. A prototype of reactive scheduling system is developed to evaluate the effectiveness of the proposed method.
    Download PDF (55K)
  • Riku Inoue, Nobutada Fujii, Kanji Ueda
    Session ID: B09
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Process plan and production schedule are significant plans that have much effect on the productivity of factories. However, the optimality of each plan is sometimes conflictive. In order to solve the conflict, we propose a new method based on evolutionary artificial neural networks proposed in the area of evolutionary robotics, so that it can realizes simultaneous process planning and scheduling and can adapt to environment changes. The feasibilities of our proposal is confirmed in the computer simulations in which the method achieved high productivity resulting from acquiring appropriate utilization of machines and role sharing among machines.
    Download PDF (116K)
  • Hiroshi Yamada, Fumiki Tanaka, Masahiko Onosato
    Session ID: B19
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Retaining machining information is an important element for generating high quality machining information. However, existing NC data (ISO 6983) cannot be used for generating high quality machining information because they do not contain machining features, cutting tool information, workpiece information, and so on. In order to solve this problem, a method of retaining machining information (Workingstep) is proposed in this research. The Workingsteps are obtained by analyzing and transforming existing NC data to CNC data model (ISO 14649) which represents explicitly technical information, such as machining features, cutting tool information, workpiece information, and so on. In this report, the extraction of Workingteps and representation of relationship between them are described.
    Download PDF (474K)
  • Manabu Shimokobe, Tomohisa Tanaka, Yoshio Saito
    Session ID: B20
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    In the process of using CAM system, the problem that a lot of input still depends on operator′s know–how is pointed out. Therefore, in this research, the selection procedure of the processing method for form feature was made a program and it was build in existing CAM system Selecting a pocket machining as an object, the machining time and result were examined considering the differences such as ″Tool″, ″Tool path″ and ″Presence of the rough processing″. As a result, optimized condition of the pocket was systematized.
    Download PDF (1015K)
  • Satoshi Kanai, Tatsumi Tada
    Session ID: B21
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Recently, there is an increasing need for digital prototyping for efficient design and verification of the system which is composed of the mechanical, electrical and embedded software systems. In this report, a method of combining multiple simulators used in each discipline by using middleware is proposed, and its application to the electric wheel chair design is shown where the design error is hard to be found only by a simulation of one discipline.
    Download PDF (707K)
  • Taku Miyashita, Satoshi Kanai, Takeshi Kishinami, Tatsumi Tada
    Session ID: B22
    Published: March 05, 2006
    Released on J-STAGE: January 31, 2007
    CONFERENCE PROCEEDINGS FREE ACCESS
    Recently, collaborative simulation environment is needed in the mechatoronics product design. In this research, a systematic collaborative simulation environment is proposed using HLA(High Level Architecture) which is the IEEE standard of a distributed simulation. In this report, an omni–directional electric wheelchair is joined to a federation as a 1federate, and the verification of acceleration control system is done. HIL(hardware–in–the–loop) collaborative simulation using HLA can shorten the time until preparing the simulation and obtain more accurate simulation results.
    Download PDF (128K)
feedback
Top