Design and management of process knowledge base in

2022-07-26
  • Detail

Design and management of process knowledge base in CAPP system. At first, the research on CAPP focused on the automation of process design, and then on the auxiliary role of computer in process design. Now people are more and more aware of the importance of process knowledge in process design

The CAPP system can not only make scientific decisions by using the experience and knowledge of process personnel and various process data, and automatically generate process procedures in JTG E50 ⑵ 006 "test procedures for geosynthetics in Highway Engineering", but also automatically calculate process dimensions, draw process diagrams, select cutting parameters and optimize process design results, so as to design process procedures with good consistency and high quality. Domestic manufacturing enterprises have accumulated a wealth of process design knowledge and a large number of manufacturing resource data through learning foreign advanced manufacturing processes and equipment technologies. However, most of these valuable process knowledge and manufacturing resource data are still in the minds of some designers or scattered paper documents, which is not conducive to enterprises' rapid search and reuse of these valuable process design knowledge and manufacturing resource information, It restricts the development speed and management level of China's manufacturing industry, so it is urgent to solve the problem of industry common technology and process knowledge management. How to make use of the existing and mature process knowledge to ensure the stability and reliability of the process method is the main problem faced by every process worker and software designer. This paper introduces the design and management method of the process knowledge base in CAPP system

2 knowledge representation method

selecting the knowledge representation method used by the system should not only consider the performance requirements for knowledge representation, but also consider the characteristics of knowledge involved in manufacturing process preparation, so as to improve the efficiency of knowledge processing. According to the requirements of expert tools, the characteristics of product manufacturing process knowledge and the requirements of system programming and expansion, object-oriented representation and production representation are selected to describe the knowledge

2.1 object-oriented representation

object-oriented representation applies object-oriented ideas and methods to knowledge representation. The so-called object-oriented means that when people understand and analyze problems, they can decompose problems into some objects and the combination and connection of objects. In the object-oriented knowledge system, the knowledge of an object constitutes the static attributes of the object. The knowledge processing methods and various operations of an object describe the intelligent behavior of the object. Object oriented representation can be used to represent process content knowledge, process equipment knowledge, processing equipment knowledge, heat treatment knowledge, etc

2.2 production representation

production representation is also called production rule representation, which is usually used to represent knowledge with causal relationship. Because the knowledge base of production rules is a rule base composed of a group of independent knowledge, and the rules communicate with each other through the facts in the knowledge base, although changing one rule will affect the characteristics of the system, it will not affect other rules of the system. The knowledge base constructed with production rule representation is easy to add, modify and delete, so it can conveniently realize the function of updating information

3 design of knowledge base

3.1 classification of knowledge base

the reasoning process of expert system needs to match the conditions of rules in the rule base with the knowledge in the knowledge base and the data in the dynamic database, which requires a lot of knowledge and data. The system can not only obtain data from the database, but also add new knowledge and new rules summarized by experts to the database to update the obsolete data in the database. According to the form of knowledge in the expert system, knowledge can be divided into: the fact knowledge about the object is the narrow knowledge, and the knowledge about the method is the commonly referred to rules. The above two forms of knowledge are stored in the knowledge base and the rule knowledge base respectively. The classification structure of the knowledge base is shown in Figure 1

Figure 1 Structure of process knowledge base

process flow inventory stores the logical relationship between the processing sequence of process nodes (i.e. operations) on the product processing process flow chart. The process rule library stores several rules about process processing, such as a series of rules for extracting information during process generation and rules for generating manufacturing serial numbers. These rules are a set of perfect rules that have been analyzed and comprehensively designed during the detailed design of the system, and can deal with various situations that may occur during the operation of the system

the process content knowledge base stores the attribute knowledge of process node objects, including the type of process node, input items and expert prompt information, as well as the specific content of equivalent process node, superior process node and process (i.e. work step content). The process equipment knowledge base and the processing equipment knowledge base store the relevant knowledge and information of the process equipment and processing equipment used in the processing process respectively. The heat treatment knowledge base stores the heat treatment related knowledge of processing technology. In addition, other knowledge bases such as standard parts knowledge base store relevant knowledge respectively. This knowledge is the experience summary of process experts. It provides modification interface during system design, which can easily add new knowledge and rules of experts, and update or delete obsolete data. Knowledge base needs a certain storage carrier. The knowledge base of expert system takes relational database as its storage carrier

3.2 design of rule base

the function of rule base is to summarize process design rules, including processing methods of typical geometric elements, machine tool selection rules, dimensional accuracy selection rules, logical judgment principles of process sequencing and related processing A. dynamometer structure: the type of force measuring piston is installed in the force measuring cylinder. The database is used to store machining data, including machining allowance, tool (die) parameters, cutting parameters, auxiliary tool code, measuring tool code, machine tool parameters and number, tooling code, man hours, etc. The source of these data can be created by the user according to the product characteristics of the enterprise and the environment of manufacturing resources, or based on the existing database

because process design itself is a multi parameter, multi constraint, experience dependent and complex thinking creation process, its knowledge structure is very complex. A knowledge expression method of multi-level, multi expression mode and organic collection is proposed here. That is, the above process rules and processing data knowledge are collected and arranged in a hierarchical manner. The first layer is part family feature acquisition; The second layer is the process knowledge base of processing method and process selection; The third layer is the manufacturing resource library for machine tool selection, processing type, tooling and fixture selection; The fourth layer is processing data, processing hours and other process databases. The low-level knowledge is expressed by database; For high-level knowledge such as machining sequence, tooling equipment, cutting parameters, process design, etc., the framework, production, logic and process integrated expression modes are used. Target driven mode (reverse reasoning strategy) is not suitable for process reasoning, but data-driven mode (forward reasoning strategy) is suitable, that is, starting from the blank of the part (at this time, the process specification is empty), adopting data-driven strategy and introducing heuristic knowledge for multi-level search and hierarchical reasoning. The knowledge base formed in this way not only has the logic principle, but also has the creation function, that is, it has the function of inferring high-level knowledge from low-level knowledge. The knowledge expression form of process decision module mainly adopts production rules. Production rule is a rule that is judged according to a group of sentences composed of conditions and conclusions, and organized from top to bottom according to the order and corresponding conditions. This rule is more in line with the thinking mode of experts. Its general form is: if (condition 1) and (condition 2) and (condition n) the existence of then (conclusion) holds

in order to solve the conflict, redundancy and insufficient expression of process rules, it is made up by the expression of propositional logic. For example, when searching a process rule, a high-frequency quenching operation exists according to the production condition, and there must be an operation related to quenching and tempering in the previous operation arrangement. Therefore, there is a "" relationship between the high-frequency quenching operation and the operation related to quenching and tempering

4 knowledge base management

4.1 knowledge management

having knowledge is an important sign that expert systems are different from other computer software systems, and the quality and quantity of knowledge are the key factors that determine the performance of expert systems. Knowledge management mainly includes knowledge acquisition, knowledge query, knowledge modification, knowledge consistency maintenance, etc. In the expert system, the knowledge management module of the expert system is constructed by using the object-oriented method through the friendly human-computer interface. The experience of the process experts is transformed into the knowledge and rules that the system can understand, and connected with the knowledge base. It is convenient to achieve the acquisition of knowledge and rules, knowledge query, knowledge modification, knowledge maintenance and so on. Under the guidance of man-machine interface, users can easily add, browse, modify and delete knowledge and rules in the knowledge base without knowing the syntax required by production rules, so as to enrich and perfect the knowledge base, so as to improve the flexibility and practicality of the software

4.2 knowledge acquisition

the basic task of knowledge acquisition is to acquire knowledge for the expert system and establish a sound, complete and effective rule base to meet the needs of solving domain problems. Knowledge base acquisition provides a way to continuously expand the content of the knowledge base. According to the knowledge obtained from experts or relevant materials, after certain sorting, it can be input into the knowledge base through the knowledge acquisition module. The knowledge acquisition module is responsible for detecting the integrity and consistency of knowledge in the process of acquiring knowledge. In the knowledge acquisition module, the operation of several rule objects or knowledge objects is used to sort out the knowledge, detect the integrity and consistency, and input the knowledge into the corresponding knowledge base or rule base

4.3 knowledge query, modification and maintenance

like knowledge acquisition, knowledge query, modification and maintenance are also carried out through rule objects or knowledge objects. Knowledge query provides the query of rules and rule elements. Rules can be queried by rule number, rule name and rule elements contained in the rule; The rule element can be queried according to the rule element name, rule code and natural language description. When inconsistencies are found in the process of knowledge maintenance, knowledge query can quickly locate the knowledge to facilitate knowledge modification. Knowledge modification provides a mechanism for correcting error rules when errors are found in consistency check. At the same time, the consistency of the modified knowledge is detected. The modified knowledge can be stored in the corresponding database only after passing the detection. The knowledge in the knowledge base may be accidentally damaged, making the knowledge in the knowledge base inconsistent. In order to ensure the consistency and integrity of the knowledge base, it is necessary to regularly maintain the consistency and integrity of the knowledge

is it non-linear and at what level is the non-linear degree?

5 application method of process knowledge in process design

in the knowledge-based CAPP system, the process design of parts and components is the core of the whole system. Its main feature is that in knowledge-based process planning, process designers can query and quote various process knowledge in the process knowledge base, which is the current process

Copyright © 2011 JIN SHI