[机械模具数控自动化专业毕业设计外文文献及翻译]【期刊】I2NC一种新型计算机数控-外文文献

I2NC: A New Type of Computer Numerical Control Junxi Bi, Tao Yu Qiang Li CIMS &Robot Center College of Mechanical Shanghai University Inner Mongolia University of Technology Shanghai City 200072,China Hohhot, Inner Mongolia , 010051,China , Abstract - This paper introduces a new type of CNC, which is named as I2NC. I2NC (Intelligent Integrated Numerical Control) is a kernel of I2MS (Intelligent integrated Manufacturing System),which includes compute-aided design, computer aided process plan, computer aided manufacturing, and CNC.I2NC provides them a unified platform oriented to CNC manufacturing, which make their work turned from distributed work to collaborative work. In this paper, we describe and analyze the conceptual framework and key technologies (such as artificial intelligence, neural networks, knowledge management, and etc.) of I2NC systematically. We also discuss STEP-NC approaches using in I2NC. Index Terms -Computer Numerical Control (CNC); Intelligent Integrated Numerical Control (I2NC);STEP-NC;Knowledge Management (KM); Artificial Intelligence (AI) I. INTRODUCTIONThere is a trend that advanced manufacturing technologies are to be developed to include numeralization, precision, uttermost, automatization, integration, latticing, intelligentization and green production technologies 1. Manufacturing enterprises are facing serious challenges and pressures from the growing globalization of the economy and the market as well as from the rapid developments of science and technology 2. Small and medium-sized enterprises especially, have to reform their traditional manufacturing methods by using advanced technologies, such as Artificial Intelligence (AI), Knowledge Management (KM), Expert System (ES), Artificial Neural Networks (ANNs), Adaptive Control Optimization (ACO) etc, in order to succeed in the increasingly intense competition for markets. Thus, it is of great importance for them to accept the concept of the intelligent integrated numerical control (I2NC). I2NC developed enhances the ability of the conventional NC to improve machining efficiency and quality, and protect the cutting tool. The capability for communicating and collaborating are improved for system integration, resource-sharing and cooperation. As the brain for industrial machinery, Computer Numerical Control (CNC) is the core element in modern intelligent integrated manufacturing system (I2MS). The structure of I2MS can be represented by a matrix the tasks and classes. A third dimension is a technology domain: CAD, CAPP, CAM, etc. This matrix represents the model for I2MS (Fig.1). CNC technology is highly complex, requiring technologies from various domains. In spite of great technological achievement, contemporary CNC still requires further improvement to satisfy the requirement of I2MS. Fig.1 The three-dimensional matrix for an I2MS In a traditional CNC manufacturing environment, the operation flow usually starts from product design via geometric models created in CAD system, and ends with the generation of machining instructions required to convert raw material into a finished product, based on geometric model. In fact this process includes CAD/CAPP/CAM/CNC/MT. Here CAPP, CAM, MT is the abbreviation respectively for Computer Aided Process Plan, Computer Aided Manufacturing, and Machine Tools. It is an executing mechanism without intelligent and it is a single direction signal workflow in current industrial manufacturing system, as it is shown in Fig. 2. CSCW (Computer Supported Collaborative Work) can be seen as Collaborative Manufacturing based on computers and network. It can be viewed as an evolution of Distributed Manufacturing, where designated individuals and organizations-both internal to a manufacturing enterprise and extended to its suppliers, customers, and partners-work together for mutual gains 3 In order to make the above workflow become double information flow and make CAD planners, CAPP planners, CAM planners, NC Programmers collaborate to work, here we provide a collaborative environment CNC system, based on STEP-NC and KM. This type of CNC can be seen as an intelligent integrated NC, so it is defined as I2NC. In this paper, a brief conception of I2NC is presented first, followed by introducing the framework of it based on STEP-NC. Some key technologies include AI, ES, ANNs, KM, etc are discussed. Sequentially successful implementation in application in roll grinding is given to illustrate how I2NC 1-4244-0466-5/06/$20.00 2006 IEEE803Proceedings of the 2006 IEEEInternational Conference on Mechatronics and AutomationJune 25 - 28, 2006, Luoyang, ChinaAuthorized licensed use limited to: Northeastern University. Downloaded on January 4, 2009 at 21:49 from IEEE Xplore. Restrictions apply.works. Finally some prospects of I2NC are also discussed in future I2MS. Fig.2 workflow in current industrial manufacturing system II. DEFINITION AND CONNOTATION OF I2NC I2NC is the abbreviation of Intelligent Integrated Numerical Control. Here intelligent and integration are two key facts in I2NC system. AI, such as ANNs, Genetic Algorithms(GA), Fuzzy Logic(FZ) , ES, ACO ,based on the operating principle of the human brain and natural biological evolution, have been received considerable and increasing interest over the past decade and have been applied to CNC machining. Most of the developments are concentrated on an intelligent control algorithm and its application to a traditional CNC machining system, so they have not achieved good results 4, 5. Various part of system being simply condensed is the idea of intelligent manufacturing model, the system is composed of these condensed intelligent body. Each condensed intelligent body is relatively isolated cell, the system is composed of these cells which can divide and collaborate to work. In order to realize the total intelligence of manufacturing system, it is essential for the intelligence of each cell device. For the intelligence of I2NC system, the direction and expression of intelligence is firstly decided, for the integration of I2NC system, the method of integration is secondly considered. The intelligence of I2NC system is shown as follow 6. 1) The intelligence of job plan 2) The ability of self-learning and improvement 3) The self-adapted man machine interface 4) The intelligent control of manufacturing process 5) The intelligent failure diagnosis Because the G and M code are employed in the traditional CNC technology, this has brought some critical problems such as incompatibility between various systems, data sharing and integration, and programming, and so on. Moreover it restricted the open ability and intelligence of CNC system, and led bottle-neck phenomena between the CNC and CAX technology. In order to solve the above-mentioned some reasons, the Europe Union passed a planning of OPTIMAL and provided the concept of STEP-NC. STEP-NC (14694) expands STEP (Standard for the Exchange of Product model data) into the field of CNC. It redefined a kind of interface between technology planning system of CAM and the control system of NC. This is helpful in decreasing machining time of traditional CAD/CAM and in eliminating post processor required in the manufacturing process, and it supports the coming intelligent manufacturing equipments. In fact STEP-NC is a CNC language which has covered the total life cycle of product. That is to say, it is a higher-level programming language for seamless integration of the CAD-CAM-CNC chain. At the present time the research and development of the standard of STEP-NC is still in the laboratory. Thus it is primary task to study and develop open I2NC system based on STEP-NC. I2NC is the abbreviation of Intelligent Integrated Numerical Control It means that abstracting some function of traditional CIMS, such as CAD, CAPP, CAM, into a series of modules, which makes the application be divided into a group of tasks frame. The tasks frames could be decomposed into detailed sub-functions. Every function modules is a practical module oriented to manufacturing. The function modules are organized and integrated into the bigger parts modules with the task frame, and these parts modules are integrated into a concrete application CNC system finally with a unified user interface. I2NC is also a kind of new philosophy and method in I2MS. III. STEP-NC STANDARD OF I2NC STEP-NC is a new and comprehensive data model for the programming language formalized as ISO 14649, which is being developed by ISO. ISO 14649 describes an interface between CAM and CNC to support the direct use of product data based on STEP (ISO10303) for creating exchangeable and workpiece-oriented data models for CNC controllers. For intelligent control, many kinds of technology must be developed, however it is important to update the new data model. The product data model of STEP-NC has covered a lot of information, such as 3D geometry model (AP-238 or AP-214), feature definition (AP-224), tolerance definition (AP-219), technology flow and so on. All information must be integrated into a overall product database though a intelligent interface. STEP-NC application protocol AP-238 is shown as Fig.3 7. AP-238 is a new input standard of CNC, it will take the replace of RS274D (ISO6983), at the same time it is a new language between clients and factory communication, IGES is also replaced. CNC system based on STEP-NC can direct read STEP file (AP-238) from CAM system, thus post processor of machining and programming from traditional CAD/CAM system has disappeared. This has been replaced by STEP-NC processor. As shown in Fig. 4, the two model styles are compared. Fig. 3 STEP-NC application protocol AP-238 804Authorized licensed use limited to: Northeastern University. Downloaded on January 4, 2009 at 21:49 from IEEE Xplore. Restrictions apply.Fig.4 Information flow of CAD-CAM-CNC Data program (file) style based on STEP-NC is shown in the following example: ISO-10303-21; HEADER; FILE_DESCRIPTION (ISO 14649-11 EXAMPLE 1, SIMPLE PROGRAM WITH A PLANAR_FACE, A POCKET, AND A ROUND_HOLE), 1); FILE_NAME (EXAMPLE1.STP,2002-02-02, (YONG TAK HYUN,JOCHEN WOLF), (WZL, RWTH-AACHEN), $,ISO 14649, $); FILE_SCHEMA(STEP_NC_SCHEMA); ENDSEC; DATA; #1= PROJECT (EXECUTE EXAMPLE1, #2,(#4),$,$,$); #2=WORKPLAN (MAIN WORKPLAN, (#10, #11, #12, # 13,#14),$,#8,$); #4=WORKPIECE (SIMPLE WORKPIECE, #6,0.010, $,$, $, (#66,#67,#68,#69); #6= MATERIAL (ST-50,STEEL, (#7); #7= PROPERTY_PARAMETER (E=200000N/M2); #8= SETUP (SETUP1, #71, #62, (#9); #9= WORKPIECE_SETUP (#4, #74, $, $, (); #10= MACHINING_WORKINGSTEP (WS FINISH PLANAR FACE1, #62, #16, #19, $); #11= MACHINING_WORKINGSTEP (WS DRILL HOLE1, #62, #17, #20, $); #12= MACHINING_WORKINGSTEP (WS REAM HOLE1, #62, #17, #21, $); #127= CUTTING_COMPONENT (100.000, $, $, $, $); ENDSEC; END-ISO-10303-21; CNC system can obtain overall description about parts and manufacturing process under the use of STEP-NC. Before the parts being machined, CNC can make some security check and optimize manufacturing process. There are many advantages about STEP-NC method data interface of STEP-NC is a neutral data description, not rely on CAM and NC machine tool, helpful in improving exchange ability and flexibility all information for product and technology are included in the file of STEP-NC, information exchange can be direct processed among CAD,CAM and CNC controller. CNC can precheck and optimize part program according to selected tool in manufacturing process. This can improve efficiency, security and adaptability. Simplifying the machining instruction STEP-NC compiler can better use new function of machine tool, the traditional post processor disappeared. Double directory data flow. E-manufacturing possibility According to ISO14649, the work flow for STEP-NC controller must include translation, planning, and control stages 8. IV THE FRAMEWORK OF I2NC The basic architecture of CNC is composed of: 1) Human-machine interface(HMI) 2) Control function 3) Data processing within CNC9 This holds true for I2NC.in the architecture presented, the user requirements are implemented in basic architecture in an extended fashion. As shown in Fig.5, it includes the following main modules: TPG(tool path generation) modules, i.e. CAD/CAPP/CAM module, which are extended HMI comprehensively covering part programming and tool path generation based on a STEP-NC data model. These modules incorporate the CAD/CAPP/CAM function into theFig. 5 A conceptual architecture of I2NC 805Authorized licensed use limited to: Northeastern University. Downloaded on January 4, 2009 at 21:49 from IEEE Xplore. Restrictions gramming system. The standard CAD data (STEP, AP203) can be converted into built-geometric modeling kernel data, by recognizing the machining features and processing these feature information. Output is stored in the machining feature database. CAPP determines the processing sequence and other processing information by using the machining knowledge. Sequentially, CAM extracts NC instruction based on STEP-NC (ISO14649). NCK/PLC control module which is related to the activities of CNC to carry out successfully the instruction given to it. NCK interprets the tool path commands and executes them by activating the servo mechanism, and PLC executes machinery commands, such as tool change and workpiece loading/unloading. The principal function is to control the machine tool so that the desired shape is produced accurately by the machining operation. Various intelligent control functions are covered here, such as adaptive control, human interaction, quality control, resource management, and self- learning. Engineering Database based on KM, which provides comprehensive data for the part programming and TPG. This database include machining feature database, machining resource database, machining process database, machining knowledge, diagnosis database and so on. V SOME TECHNOLOGIES OF I2NC Because I2NC is an intelligent and integrated NC, its key technologies not only include the technique of CNC, but also include extra integration and intelligentization technologies. Here we aim at intelligentization technologies, which include AI, ES, ANNs, and KM etc. techniques. A. AI technique using in I2NC AI refers to the capability of a machine to perform functions normally associated with human intelligence such as reasoning, learning and self-improvement. The trend towards increased use of machine intelligence in I2NC systems is clear and unlikely to be reversed. A driving force behind the development of AI techniques in I2NC system is the desire to capture specialist knowledge and lessons gained from previous experience in a form which can ensure that benefit is derived in arrange of future operation. In CNC machining, determining optimum or appropriate cutting parameters can minimize machining errors such as toll breakage, tool deflection and tool wear, thus yielding high productivity or minimum cost. Features of AI that have been applied in CNC system are illustrated in Fig. 6. B. Knowledge-based Expert system Knowledge-based expert system(KBES) are intelligent computer programs that capture knowledge of a particular domain and mimic the problem-solving strategies of human experts to provide recommendation.10-12 A body of knowledge can be manipulated in such a way as to perform a function which would normally require human intelligence. A structure of KBES for machining control is shown Fig.7.it Fig.6 AI techniques applied in CNC contains three modules: a knowledge base, an inference engine, and sensor data acquisition and processing module. The reference engine drives the system and the interfaces with the knowledge base and hence supplies advice to the user and explanation to justify the system line of reasoning. The knowledge base can provide near-optimal machining control with experimental data. C ANN technique using in I2NC ANNs have been successfully used in many process control applications. There ability to approximate arbitary nonlinear vector functions, combined with dynamic elements has yielded a powerful tool for modeling nonlinear dynamical systems. ANNs have being successfully applied in manufacturing automation. In CNC system, ANNs are mainlyFig.7 A KBES conceptual scheme for machining control 806Authorized licensed use limited to: Northeastern University. Downloaded on January 4, 2009 at 21:49 from IEEE Xplore. Restrictions apply.used in parameter selection and optimization of machining process. ANNs differ in various ways from conventional expert systems to traditional computing. ANNs have the following main advantages: Knowledge can be obtained by learning from training example. Knowledge is stored in a distributed fashion ANNs can provide solutions to problem which have not previously been experienced as long as the solutions lie within the same domain as training data. The another advantage of using ANNs in this situation is that such a system can be easily incorporated into PC based CNC and form a quick reference for operators as problems of machining parameters are experienced. These characteristics make ANNs potentially useful for modeling machining processes. For building machining knowledge base, a two-layer feedforward Neural net with one hidden layer is shown in Fig. 8.The user inputs the machining conditions. Then, the input information is encoded into a form that ANNs can recognize. The function of the network is to map input vectors to output vectors. The output module uses the output of the ANNs to identify a suitable outcome in the machining database. D. Knowledge management (KM) Here, knowledge management (KM) namely Knowledge engineering, which is used in processing some process technology knowledge, such as machine preparation, adjustment,