Uzman sistem ve otomotiv sektöründeki bir uygulama

dc.contributor.advisor Tümkor, Serdar
dc.contributor.author Burgut, Hüseyin
dc.contributor.authorID 83032
dc.contributor.department Makine Mühendisliği tr_TR
dc.date.accessioned 2023-03-16T06:04:55Z
dc.date.available 2023-03-16T06:04:55Z
dc.date.issued 1999
dc.description Tez (Yüksek Lisans) -- İstanbul Teknik Üniversitesi, Sosyal Bilimler Enstitüsü, 1999 tr_TR
dc.description.abstract Uzman Sistem ve Otomotiv Sektöründeki Bir Uygulaması adlı çalışma 9 bölümden oluşmaktadır. Birinci Bölüm olan Giriş kısmında bu konuda yapılan çalışmalar hakkında bilgiler ve genel olarak yapılan çalışmanın tanıtımı bulunmaktadır. Problemleri çözmek maksadıyla, bilgisayar yardımıyla biriktirilmiş tecrübelerden yararlanan, insan zekasının taklidi programlar Yapay Zeka (Al) olarak adlandırılmaktadır. Yapay Zeka'nın literatürdeki tanımları ve uygulama alanları Bölüm 2'de açıklanmıştır. Yapay Zeka'nın önemli bir uygulama sahası olan Uzman Sistemler (ES), uzman insanların hareketlerinin bilgilerinin biriktirildiği, karar vermeye yönelik programlardır. Uzman Sistemler'in geçmişi pek uzaklara gitmemesine rağmen uygulamaya yönelik olduklarından kullanım sahası oldukça geniştir. Bölüm 3, 4, 6, 7, 8'de Uzman Sistemler başlığı altında Uzman Sistem ifade şekilleri, Uzman Sistemin tarihçesi, Uzman Sistem' in uygulama alanları ve örnekleri, Uzman Sistem geliştirme programlan, teknikleri, karakteristikleri, yapısı, avantajları, eksik yönleri ve Türkiye'deki kullanımı hakkında bilgiler yeralmaktadır. Bilginin bilgisayar sistemleri ile birleştirildiği mühendislik disiplini olan bilgi mühendisliği alanındaki bilgi toplama süreci ve bilgi mühendisi hakkındaki bilgiler Bölüm 5'te bulunmaktadır. Son kısım olan dokuzuncu bölümde Hazırlanan Uzman Sistem hakkında bilgiler verilmiştir. Burada Uzman Sistemde kullanılan Bilgi Bankalarından, Sonuç Dosyalarından ve Karar Mekanizmalarından sözedilmektedir. Elde edilen sonuçlar grafikler halinde takdim edilmiştir. tr_TR
dc.description.abstract To understand the internal structure and workings of Expert Systems (ESs), we must first develop on understanding of basic Artificial Intelligence (AI) techniques. AI represents a large body of concepts and techniques that has been develop by many researchers since the late 1950s. During this period, many definitions of AI have appeared, but it is diffucult to find a universal definition of AI. The literature is full of different definitions of AI. But for our purposes of this text we will start with the following general definitions. " AI is the computer-based solution of complex problems throug the application of processes that are analogous to the human reasoning process." The recently years, some result of AI research have indicated that many concepts, procedures and techniques developed in AI laboratories have great commercial value and a new industry, dediceted to the commercialization of the most promising aspects of AI has been established Figure 1 shows some commercial offspring of the AI research conducted during the last several years. The five most active areas of commercialization are natural language, robotics, improved human interfaces, exploratory programming and EXPERT SYSTEMS. Artificial Intelligence Natural Language Improved Human Interfaces Robotics Exploratory Programming! EXPERT SYSTEM Figure 1 The five most active areas of commercialization An Expert System (ES) is a computer applications that solves complicated problems that would otherwise require extensive human expertise. Expert Systems have been in development since the 1960s and the first appeared in the early 1970s. Commercial interest increased in the early 1980s and as a result, so have efforts to development of the mid-1980s has been the introduction of Expert System building tools and environments that do not require AI scientist. These tools and environments can often be used by subject domain experts. Most of the tools can be used by software engineers with niinimal training and exposure. İt will be important for all software engineers to learn about ESs and how to build them in the 1990s. Internally, an ideal ES can be characterized as including the following.. Extensive spesific knowledge from the domain of interest.. Application of search techniques.. Symbolic processing.. An ability to explain its own reasoning.. Capacity to infer new knowledge from existing knowledge. Expert Systems have expertise, symbolic reasoning, depth, self-knowledge. Expertise;. Exhibit expert performance. Have high level skill. Have adequate robustness Sybolic reasoning;. Represent knowledge symbolically. Reformulate symbolic knowledge Depth;. Handle difficult problem domains. Use complex rules Self-knowledge;. Examine its own reasoning. Explain its operation ESs can be designed for spesific hardware and software configurations, or they can be software systems that are designed to run on a general purpose computer. The knowledge the ES uses is made up of either rules or experience information about the behavior of the elements of particular subject domain. ESs today support many problem-solving activities. Future ESs will support even more. ESs have been built to model the problem-solving strategies of human experts. Because different human experts use different problem-solving techniques, XI the ESs modelled after the strategies of the human experts use a variety of problem- solving approaches. ESs aren't comprehensively and commonly discused in Turkey yet. Only one voice in this subject comes from universities. Recently, the increased of post graduate and doctoral studies about Expert Systems technology results from the interest on this technology. However, the use of ESs in industry doesn't exceed several individual efforts. Also there are no ESs doesn't come into existence yet. The major problems for the development of ESs technology in Turkey are as follow.. High cast (software, hardware and labor).. High development time.. Being unknown of this technology and the knowledge engineering concept.. Insufficiencies in the software sector of Turkey. Overtime these problems will be exceed but necessory steps must be taken before it is to late. Knowledge Engineering; Knowledge engineering is the process of acquiring specific-domain knowledge and building it into the knowledge base. Figure 2 illustrates this process. Domain Expert Knowledge Source "Z. Knowledge Engineer Expert System Knowledge Based Update System Representation Transformation Figure 2 Knowledge engineering Knowledge engineer is the person who acquires the knowledge from the domain expert and transports it to the knowledge base. Because the ES requires that knowledge in the knowledge base be stored in accordance with the systems knowledge representation of the knowledge as a part of the transportation process. To acquire the necessary knowledge, the knowledge engineer must first establish an overall understanding of the domain, form a mental dictionary of the domains essential vocabulary. He or She must then distill succinat knowledge from the information provided by the expert. Xll Expert System Development Life Cycle; Expert System development life cycle consits of the six major stages that are following.. Problem selection and identification: Determing problem characteristics.. Conceptualization and prototype construction: Finding concepts to represent knowledge and prototype construction.. Implementation: Formulating rules that embody knowkedge.. Testing and avaluation: Validating rules that embody knowkedge.. Maintenance and updating: Adding knowledge and suitable technology. Architecture of Expert Systems; ESs use a wide variety of spesific system architectures. Extensive research is currently in progress to investigate various aspects of ES architectures and considerabledabete remains. In spite of significant differences most of the architectures have several general components in common. The typical components are user, user interface, knowledge update facility, explanation facility, knowledge base, inference engine.. User is tester, tutor, pupil, and customer.. User Interface accepts information from the system and convert it to a form that can be understood by the user.. Knowledge update facility is used to perform such update.. Explanation facility consits of an identification for each steps in the reasoning process and a justification for each step.. Knowledge base is containing knowledge about a problem area.. Inference engine; manupulating the stored knowledge to produce solutions to problems. It includes forward chaining and backward chaining. There are several knowledge representation. The two important knowledge representations are production rules and frames. The production rule is the most widely used form of knowledge representation in Expert Systems. A production rule has two parts that IF and THEN. IF states a condition and THEN expresses a corresponding conclusion. ES Building Tools; ES building tools are classified in a number of different ways. If you focus on the overall knowledge representation techniques available in a tool, then you can divide the current crop of commercially available tools into five general types * Inductive tools * Simple rule - based tools * Structered rule - based tools * Hybrid tools Xlll *Domain specific tools Selecting the correct tool is important decision in the development of an ESs. Building Expert Systems; ES tools are available for several reasons; * They provide rich software development environments that would assist in development of any software * They include specific assistance for rapid prototyping * It provides a solid basis for quick capture of knowledge and rapid system development; and it eliminates the labor required to build the basic software * In many cases an E.S tool can provide extensive assistance in some specific area of system development. At that time, there are several tools such as programing languages that is Fortran, Pascal, C, and AI languages. The major ES programming languages are LISP and PROLOG. AI scientists use PROLOG and LISP. LISP and PROLOG are used to create ES. Comparing human and Expert Systems shows in table 1 The areas of application of Expert Systems; There are a lot of expert systems reported to be in use today, and their number is rapidly increaising. Finding examples is, consequently, not difficult. We can look at the main areas and make an arbitrary breakdown into;. Diagnostic aids: The classical systems are those Which help diagnosis, expecially medical diagnosis. Aids to design and manufacture: A second group of systems is those which help with design and construction. xiv . Teaching aids: This is a wide field for an ES. Its two essential characteristics of dialoque and intelligent behaviour.. Problem solving: It includes four application areas.. Recognition of forms. Robotics. Games. Automatic demonstration of theorems APPLICATION; Companies activating in industry are trying to develop various approaches and management models in order to get best solutions in relative environment, which includes hard and increased competitive market conditions. Main purpose of this afford is to continue for a long term working activity by increasing the power of competition. It is a bit further understood day by day that the most important factor of taking the first place in sector is the usages of acknowledge and data's. Nowadays, main reason for companies to survive, are how to get information, how to use it and the capability of practising such an information in business world. By the help of that perspective, the usage of this kind of existing information, improving it and collecting period of this data, have an important role. Expert Systems have been improved for master thesis against lack of or up to late information flow between top manager who are responsible for making decisions, again, some difficulties in sharing and transmission of information among departments such as Purchasing Department, Engineering Department, Stock Control and Manufacturing Department, Spare Department. As a result of this, a system, which is called Shaft System, is taken consideration in improved Expert Systems and which transport circular movements or revolutions just after having it from gearbox to differential gear as a longitudinal movement. This Shaft System applies to those models that are represented by MD23, MD27 (Midibus) and NPR66, NKRWTDE66 (light lorries or trucks) In these improved Expert Systems for the complex data's (among Purchasing department, Spare parts Dept., Engineering Dept., Quality Control Dept.) which belongs to parts of vehicle to be manufactured. Getting data's from each department by the help of Expert Systems without visiting them separately, will be helpful for decision makers and mechanism, to realise that, data banks for each department have been formed available yet. Information of shaft system and department data's that are forming the main structure of system placed in those data banks. Information of parts which are included in shaft system, have been combined entirely and submitted to software users by forming separate sheets and tables and those data's or information are permanently upgraded. XV An example sheet is given in the table 2 Getting all information have been realised by linking between conclusion files and data banks. Table 2 Purchasing knowledge A linking facility is shown as an example below Figure 3 Linking -..L'Cj r. It can also be leaded to necessary information by the help of a graphical approach Sevklyat Durumu JF jr ^ * J? c? ^ Figure 4 Graphical approach i ESerlesl XVI Conclusion files have been prepared for users so as to get desired information for each department independently without going to every department to have all necessary information. There are conclusion files for each department that in these files, it is made all desired information being listed in tables by extracting part numbers from pull down menu. Through the prepared data banks and conclusion files it can be gotten desired information in a short period and even people who are not familiar with subjects can easily reach the information through the prepared expert system. Two additional auxiliary decision mechanisms have been formed apart from data banks and conclusion files, these are explained below separately. Decision Mechanism 1; In this system, if the user would like to see the final situation of a material in production stock area such as in the state of raw - material or in the state of finish good, program user can reach the conclusion section in prepared sheets. And tables by choosing part number from pull - down menu which includes warming indicator for the moment the part is chosen, part number and the part name will be listed in the table. In the meantime, there is a conclusion section which is formed by Decision Mechanism I, this conclusion section can be monitored by the indication of the comparison between the amount of minimum level of stock and the amount of warehouse stock which has been made by Decision Mechanism I. Two different conclusions occure on Decision Mechanism I system, these are explained below; 1- IF the amount of warehouse stock < Minimum level of stock, then check the delivery of the firm and according to the deficit balance and have the part made. 2- IF the amount of warehouse stock > Minimum level of stock, then check the delivery program of other parts Falling possibility through the critical position of the supplied, intermediate part has been prevented during the final assembly operation of the vehicle, by the help of DM I and continuity of the production has been realised without being faced with any problem. During the controlling period of the stock amount of current part, a macro program has been prepared. so macro If '1' clause is acceptable, then press the switch |Check the Balance program will start up and relative message will be sent to the responsible person through the Lotus Notes program. Decision Mechanism II has been used for determination of suppliers regarding new parts to be applied for the current production assembly line, at the same time, in the xvii purpose of determination of firms orders of mentioned parts to be used for mass production. The content of Decision Mechanism II within the year according to the previous months; consists of a comparison of supplier performance against average performance so that suppliers can be prefered according to this criteria. Purchasing department is a department, which has a great strategic importance for suppliers; therefore it is compulsory for a company to eliminate one man factor in terms of any failure possibility during operation. The elimination of self-decision in working conditions and getting rid of the personality in terms of all decision making levels, indicate the main purpose of Decision Mechanism II. Another function of Decision Mechanism II is availability for a flexible production assembly program additionally; responsible manager who is usually purchasing manager can decide amount of orders in current parts to be increased/decreased from a certain supplier to another one. All these facilities can be done through prepared macro, purchasing manager can give his message to responsible departments by using buttons. Search for new suppliers |give| an order of a new part Since then, two different conclusions on Decision Mechanism II have taken placed " You can give completely new order concerning current or new parts, you can extract your order from a supplier balance and send it to a new supplier's balance". Otherwise, command as follows " Send the order of this part to other suppliers and search for new suppliers alternatively". en_US
dc.description.degree Yüksek Lisans tr_TR
dc.identifier.uri http://hdl.handle.net/11527/23935
dc.language.iso tr
dc.publisher Fen Bilimleri Enstitüsü tr_TR
dc.rights Kurumsal arşive yüklenen tüm eserler telif hakkı ile korunmaktadır. Bunlar, bu kaynak üzerinden herhangi bir amaçla görüntülenebilir, ancak yazılı izin alınmadan herhangi bir biçimde yeniden oluşturulması veya dağıtılması yasaklanmıştır. tr_TR
dc.rights All works uploaded to the institutional repository are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. en_US
dc.subject Otomotiv sektörü tr_TR
dc.subject uzman sistemler tr_TR
dc.subject Yapay zeka tr_TR
dc.subject Automotive sector en_US
dc.subject Expert systems en_US
dc.subject Artificial intelligence en_US
dc.title Uzman sistem ve otomotiv sektöründeki bir uygulama
dc.title.alternative Expert system and an application in the automotive sector
dc.type Tez tr_TR
Dosyalar
Orijinal seri
Şimdi gösteriliyor 1 - 1 / 1
thumbnail.default.alt
Ad:
83032.pdf
Boyut:
4.41 MB
Format:
Adobe Portable Document Format
Açıklama
Lisanslı seri
Şimdi gösteriliyor 1 - 1 / 1
thumbnail.default.placeholder
Ad:
license.txt
Boyut:
3.16 KB
Format:
Plain Text
Açıklama