Application of the hottest KBE Technology in UG

  • Detail

The application of KBE Technology in UG 1. Introduction the development of CAD technology has experienced the following stages: in the 1970s, the use of CAD technology to generate digital drawings improved the efficiency of drawing engineering drawings, but did not really play a role in aided design. In the 1980s, parametric modeling technology used feature-based CAD technology to establish a fully parametric driven three-dimensional model, and on this basis, carried out finite element analysis, motion analysis, assembly interference inspection and NC automatic programming for the overall design and components, so as to ensure that the design meets the actual engineering needs. But at this time, CAD technology is only a part of the product development process

in order to make designers concentrate on innovative work, CAD system should help designers get rid of repetitive work. Therefore, the next generation of mechanical CAD system should use computers to extend the design ability of human experts with creative thinking as the core, and realize the automation of the design process as much as possible. This is the real design tool. The design process of human experts depends on the mastery, processing and application of knowledge by experts. Without knowledge, analysis, judgment and decision-making cannot be carried out. Therefore, design automation is the automatic processing of knowledge. At present, KBE (knowledge based Engineering) technology is the most promising solution to this problem

kbe system puts forward a scheme to solve the problems existing in the traditional CAD system, such as the embodiment of the design principle, whether the constraints conflict, how to evaluate the product in the design stage, whether the design and manufacturing is feasible, and whether the designed final product meets the appearance requirements, and provides an optimization scheme at the same time. The purpose of KBE system is to make the product information can be applied in the whole life cycle, so as to obtain the optimal scheme

II. The concept of KBE

kbe is a computer integrated processing technology that provides the best solution to engineering problems and tasks through the drive and reproduction of knowledge. Due to the openness of KBE technology, so far, there is no recognized and complete definition. The KBE center of Coventry University in the UK believes that KBE system is a computer system that stores and processes knowledge related to product models and is based on product models; Washington University believes that KBE is a design methodology, which will be closely combined with the next generation CAD technology. It uses heuristic design rules and will cover component, assembly and system development. KBE system stores product models that contain geometric and non geometric information and engineering guidelines that describe how products are designed, analyzed, and manufactured

to sum up, the connotation of KBE can be summarized as follows: KBE is the inheritance, integration, innovation and management of domain expert knowledge, and the integration of CAX technology and AI technology

in order to make KBE system the center of engineering process, it must be applied to the relevant design principles of each stage of the product. In order to effectively use KBE system from design analysis to manufacturing, KBE system is required to be flexible, open, reusable and customizable. KBE system is deeply embedded in CAX system, rather than simply adding to an existing system. Therefore, KBE system should have the following basic functions

(1) use of knowledge - convenient access to knowledge databases

knowledge databases exist in many forms, which can be persistent databases or spreadsheets. KBE system can easily access these knowledge databases, so the openness of KBE system is very important

(2) knowledge representation - create and apply knowledge rules

knowledge rules are a representation of knowledge. Knowledge representation is a set of conventions to describe the world. It is the symbolization of knowledge. This method can easily express human knowledge into a knowledge structure that computers can receive and process. Knowledge representation should consider not only the storage of knowledge, but also the use of knowledge. Knowledge representation is divided into rules, predicate logic and framework. Rules refer to the text description of a behavior, which corresponds to a certain input, can reference various forms of knowledge base, and produce certain results according to the corresponding logical conditions, such as feature modeling, parameter calculation, selection of specified types, or warning when violating constraints

knowledge rules can be established at any time in the project life cycle. The form of rules is flexible and can be applied to many disciplines. Therefore, rules can be used at all stages of the product life cycle

knowledge rules are always relevant to their related objects. For example, when knowledge rules are applied to establish a geometry, the rules and geometry remain relevant during its existence. The designer can use this rule to generate this geometry, or generate a new geometry. If there is no such correlation, using this rule again can only regenerate another geometry, and the information of the first geometry is completely irrelevant to this rule, so the whole life cycle becomes a strictly serial process, and it is impossible to realize concurrent engineering

(3) knowledge acquisition - knowledge reproduction mechanism

kbe system has a "self generated" knowledge reproduction mechanism. This is the main difference between it and traditional expert system. Traditional expert systems rely on knowledge engineers to obtain expert knowledge manually, which is inefficient. The main reason is that experts in the professional field lack understanding of the system and do not know how to provide comprehensive knowledge; However, system engineers do not have a deep understanding of the field of expert knowledge and cannot effectively go deep into the essence of the problem, which leads to the low application value of expert system. KBE system provides a means to multiply knowledge, broadens the way to obtain knowledge, and allows users to use the tools provided by KBE to add their own unique knowledge, so as to expand the system

kbe system automates repeated design and engineering tasks, shortens product development time, and integrates design, analysis, and manufacturing to achieve parallel work. Using KBE to build a model can combine geometric modeling and analysis, realize multidisciplinary optimization, accurately evaluate the feasibility, apply standards and practical experience to improve the quality of products, digitally acquire and reuse knowledge information such as design practice and process experience, so as to improve the efficiency of automation process

As an integrated cad/cam/cae software, UG is committed to the whole product development process from conceptual design to functional analysis, engineering drawing generation, NC code generation and processing. UG provides a knowledge driven automation solution, which fully integrates KBE system with CAX software system. KDa is a system that can record and reuse engineering knowledge and use it to drive, establish, select and assemble corresponding geometric models. This solution includes ug/kf (knowledge fusion) and a series of process wizards

in the past, KBE systems were often used alone and could not be embedded into existing engineering systems, and could not be related to existing applications. Ug/kf solves this problem. Because ug/kf is completely embedded, users do not need to know what KBE is and what ug/kf is. As long as they use familiar applications, they can reuse all kinds of knowledge. Therefore, ug/kf is for most engineers. This means that engineers and designers can devote themselves to product development without converting data between KBE system and product development system

ug/kf language is an object-oriented language, which is based on intent language. Intent language is an industry recognized rule-based knowledge programming language, which is mainly described by classes, and its applications can be executed without compilation. Users can browse and modify KBE programs with a general texter, Thus, the company continues to maintain the "growth momentum in the second half of 2013", which makes the program open and enables the engineering knowledge to be updated, supplemented and maintained at any time.

UG's kDa scheme uses "rules" To represent the relationship between different geometric parameters and engineering attributes of products. Users use KF language to establish their own rules, and use rules to easily add knowledge. Since this language is declarative rather than procedural, there is no sequence of rules. Ug/kf system automatically judges the execution order according to the relationship between rules. These rules are used to calculate the influence of engineering parameters on product geometric parameters, so as to drive the final geometric model. In addition, this language can access external knowledge sources, such as databases or spreadsheets, and provides interfaces with other application modules such as analysis and optimization modules. At the same time, UG provides a wealth of KF classes. KF class refers to the set of KF rules that can implement a task. Users can also use KF language to create their own classes

ug provides a visual tool - knowledge fusion navigator. Users can use the navigator to create, acquire and reuse knowledge rules to drive the generation of geometry. Users can directly add sub rules (establish instances of classes), sub rules (parameters in instances), add attributes (add user-defined attributes to instances) and reference other attributes through the visual aggregation 400+ industry elite interface

the rules created by users themselves must be reusable, which is the basic requirement of KBE system. KF technology connects rules with UG objects and keeps this correlation. Users use rules to control UG objects. In this way, the reuse of related rules is realized when creating objects

ug adopts generative and adoptive mechanisms to integrate knowledge with CAX system. In particular, the absorption mechanism solves the problem of summarizing and reverse seeking knowledge from existing mature products and practices. Using adoption, you can establish rules for an existing UG object to make it related to rules. Because it is too cumbersome to establish complex modeling with KF language, geometric modeling can be carried out interactively. After the object is created, the system automatically generates the corresponding rules by using adoption. The parameters of this object are defined as the attributes of the rules. Rules control objects through attributes. Therefore, by using adoption, g/kf can integrate the advantages of efficient modeling of CAD system with KBE. Users can establish and apply rules more quickly and conveniently

next, use the above tools to create a block feature and generate corresponding rules for it

the steps to establish a UG object using the generation mechanism are as follows:

(1) first, create a text file a with KF language to realize the geometric modeling of block features

#! UG/KBE 17.0

DefClass: MyBox (ug_base_part);

(number parameter) my_ height: 84;

(number parameter) my_ width: 36;

(number parameter) my_ length: my_ width:/12;

(child) block1:


class, ug_block;

length, my_length:;

width, my_width:;

height, my_height:;

color, 3;


(2) then open UG and specify the path of a file in the preset

(3) select knowledge fusion navigation under the View menu. The KF navigator displays an object tree, and each object node represents an instance of a class and the attributes of a class instance. Now select add child rules from the right-click menu to pop up the add child rules dialog box. Add my in the name column_ b

Copyright © 2011 JIN SHI