【機械類畢業(yè)論文中英文對照文獻翻譯】基于智能代理的分布式數(shù)字控制仿真系統(tǒng),機械類畢業(yè)論文中英文對照文獻翻譯,機械類,畢業(yè)論文,中英文,對照,對比,比照,文獻,翻譯,基于,智能,代理,分布式,數(shù)字控制,仿真,系統(tǒng) Proceedings of the Fourth International Conference on Machine Learning and Cybernetics, Guangzhou, 18-21 August 2005 INTELLIGENT AGENT-ENABLED DISTRIBUTED NUMERICAL CONTROL SIMULATION SYSTEM YA-BO LUO, DING-FANG CHEN School of Mechanical and Electrical Engineering, Wuhan University of Technology, Wuhan 430070, China E-MAIL: luoyabo1973@sohu.com, dfchen@public.wh.hb.cn Abstract: NC simulation is an important component of virtual manufacturing, which enables testing, debugging and optimizing of NC programs in advance. However, development of web-based distributed NC machining system is facing difficulties of intelligent control supporting real-time rendering over web and the remote simulation of cutting process. First, this paper addresses the multi-agents framework, which employs VRML to model the models of cutter and workpiece, and Java to control the process of milling simulation, taking the EAI as the bridge. Second, the novel programming methodologies for the realization of 3 key majority agents, namely, time the interpolation agent, NC codes flow listener agent, and the dynamical removal simulation agent, are detailed represented. Keywords: Multi-agents system; net manufacturing; virtual manufacturing; distributed simulation 1. Introduction A virtual manufacturing (VM) system is an integrated computer-based model that represents the physical and logical schema and the behavior of a real manufacturing system [1]. Pre-estimating and optimizing the design before it is putted on the product line is the main objective of VM. Virtual NC machining is an important component of VM, which enable testing, debugging and optimizing of NC programs. NC verification has always been the majority part of NC machining [2]. Traditional method for NC verification is to test the programming using real machine tools with cheap material, obviously which disadvantage is expensive, time taking and low efficient. Net manufacturing is a fresh concept proposed in recently years, which integrates the internal and external resources of enterprises group. The core of net manufacturing is resource sharing, which is a good approach to lower the cost of enterprises group. The rapid development of intelligent simulation and net technologies makes it possible to develop a web-based virtual NC machining system. A web-based virtual NC machining system can be distributed on web to benefit many users without having to improve the cost of NC verification. However, development of web-based virtual NC machining system is facing 2 difficulties [3]. The first is the intelligent framework, which should support real-time rendering and remote control over web. The second is the simulation of cutting process involving the workpiece dynamical modeling. Taking milling simulation as case study, this research employs multi-agents technology realized through virtual reality modeling language (VRML) and Jave external authoring interface (EAI) to overcome the above difficulties. 2. Intelligent agents framework for remote simulation The intelligent framework for 3D simulation over web comprises of technologies framework and files framework. The distributed NC milling simulation deals with 3D graphic technology over web. Currently the typical 3D rendering packages on web are Java 3D and VRML. The Java 3D API is a hierarchy of Java classes that serve as the interface to a sophisticated three-dimensional graphics and sound rendering system. Java 3D provides high-level constructs to create and manipulate 3D geometry, and to build the structures used to render that geometry [4]. As the details of rendering are handled automatically, developers can efficiently create precise virtual universes on web using this Java 3D package taking advantage of Java threads. Though the Java 3D is capable of automatically optimizing for improved rendering performance, however, it is difficult to find a function to dynamically model the work-piece, which should have capability to be modified in real time. A dynamic model constructed by the elementary API should meet complex data structure and uncertain data amount. Compared to the Java 3D, VRML, which allows to create virtual worlds networked via the Internet and hyperlinked with the World Wide Web, is a better approach to create work-piece model supporting remote 3Dsimulation. Interaction and internetworking can be specified using VRML without being dependent on special gear like head-mounted devices (HMD). It is the intention of its designers to develop VRML as the standard language for interactive simulation within the World Wide Web. Ease to use is an advantage of VRML. VRML is used in the same way to build 3-D worlds on the World Wide Web as HTML is used to build standard Web pages. For work-piece modeling, the crucial advantage is it contains some various dynamical shape nodes comprising of fields to describe how an object is rendered, which will be represented late. The above 2 advantages decide on that it is advisable to take the VRML as the tool for remote 3D work-piece modeling. A main disadvantage of VRML is that a virtual scene constructed with only VRML has poor interaction capability. Only navigation, moving object and viewpoint changing are agreed. Dynamic controlling the profiles of VRML models to realize intelligent simulation is difficult. The EAI is a good solution to the difficulty. EAI allows you to control the contents of a VRML browser window embedded in a web page from a Java applet on the same page. It does this with a browser plug-in interface that allows embedded objects on web page to communicate with each other. Consequently, the technologies framework is built by employing VRML to model the models of cutter and work-piece, and Java to control the process of intelligent milling simulation, taking the EAI as the agent maker. Current there are some browser agree EAI, such as the Netscape Communicator 4.04 for Macintosh, Netscape Communicator 4.04 for Win32, and the Netscape Navigator 3.01S for IRIX. Some plug-ins can be employed to run VRML files, such as Cosmo Player 2.1 or later for Macintosh. This research employs Netscape Communicator 4.04 + Cosmo Software Cosmo Player 2.1 to run the system. The files framework is showed as in Figure 1. VRML files comprise of sta_model.wrl used to build static models such as the cutter model and dyn_model.wrl used to build dynamic models such as the workpiece model. Users visit and manipulate the virtual scene by EAI and applet through the navigator and cosmo player. VirtualNC.html is the interface between virtual scene and users. 3. Intelligent agents for cutting simulation To make a NC milling system to run as if they were on the machine tool itself, showing machine motion and material removal is crucial. Currently, many 3D modeling packages are available such as the OpenGL, and the Direct3D soft packages. However, when a model constructed with a current 3D modeling package is immersed in a virtual environment, its profile function cannot be modified in real time. This research employs the VRML ElevationGrid node attached with the metal light effect to implement a dynamical work-piece with variable sizes. The xDimension, zDimension, and the height fields address the key fields of VRML ElevationGrid node in favor of dynamical modeling. The xDimension and zDimension fields indicate the number of elements of the grid height array in the X and Z directions. Overlap checks is fundamental for implementing the real-time workpiece material removal process. When an overlap is detected along with the inputted NC codes, the heights of these vertices of dynamical model, which is built in the dyn_model.wrl, inside the circle of the cutter, which is modeled in the sta_model.wrl, will be set equal to the height of the cutter. The realization of milling simulation contains 3 key majority agents modules demonstrated as followings. (1) Time interpolation agent DEF TS_XZ TimeSensor{ cycleInterval 32 loop FALSE } ROUTE CLOCK.cycleTime TO SCRIPT.cycleTime ROUTE Trsf_Y.translation_changed TO SCRIPT.Ytranslation ROUTE Trsf_XZ.translation_changedTO SCRIPT.XZtranslation ROUTE SCRIPT.height TO EGRID.set_height (2) NC codes flow listener agent public void ValueChanged extends Applet implements ActionListener, EventOutObserver { TheTextField.setText(String.valueOf(TheScrollbar.getValue())); Output.append("Set Scale Value:" + " x=" + ScaleValue[0] + " y=" + ScaleValue[1]+ " z=" + ScaleValue[2] + "\n"); Output.append("Set Trans Value:" + " x=" + XZTransValue[0] + " y=" + YTransValue[1]+ " z=" + XZTransValue[2] + "\n"); set_XZtranslation.setValue(XZTransValue); set_Ytranslation.setValue(YTransValue); GcodeInput=new TextArea(init_input, val); add(GcodeInput); } (3) Dynamical removal simulation agent set_height_changing(){ if( (((var_float+XZPosition[0])-(x*Long))*(( var_float+XZPosition[1])-(x*Long))+((var_float+XZPosition[1])-(z*z_Long))*(( var_float+XZPosition[1])-(z*Long)))<=(R*R_cut) ) height.setValue( heightFields ); } 4. conclusions A distributed intelligent milling system has been developed by the above key technologies combining the latest graphics, net with intelligent agent technologies. It enables users to see, manipulate, and share a virtual NC milling machine on Internet. This system detects all NC program errors, which can be immediately corrected online. Since this system runs on standard Windows platforms, and no exterior attachment is required to users, thus it is ease to use. Acknowledgements This paper is supported by the Hubei province Natural Science Fund for young elitist (No.2005ABB023), Wuhan city dawn plan (No.20055003059). References [1] K. L. Lee, and S. D. Noh, “Virtual manufacturing system-a test bed of engineering activities”, Annuals of the CIRP, Vol 46, No.1, pp.347-350, 2002. [2] H. Narita, K. Shirase, H. Wakamatsu, A. Tsumaya and E. Arai, “Real-time cutting simulation system of a milling operation for autonomous and intelligent machine tools”, International Journal of Production Research, Vol 40, No.15, pp.3791-3806, 2002. [3] Y. B. Luo, S. K. Ong, D. F. Chen, A. Y. C. Nee, “An Internet-enabled image- and model-based virtual machining system”, International Journal of Production Research, Vol 40, No.10, pp.2269-2289, 2002. [4] L. H. Wang, R. Sams, M. Verner and F. F. Xi, “Integrating Java 3D model and sensor data for remote monitoring and control”, Robotics & Computer-Integrated Manufacting, Vol 19, No.1/2, pp.13-20, 2002.