对基于网络的专家系统开发平台的研究
文献类型：学位论文
| 作者 | 王文胜 |
| 学位类别 | 工学硕士 |
| 答辩日期 | 2002-05-01 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 中国科学院自动化研究所 |
| 导师 | 高东杰 |
| 关键词 | 专家系统 推理机 推理机控制机制 粗集理论 知识获取 Expert System Rough Set Reasoning Knowledge Acquisition Control Method of Reasoning for Multiple Tasks |
| 学位专业 | 控制理论与控制工程 |
| 中文摘要 | 专家系统是人工智能中的一个重要的研究领域，它已经成功的被运用到许多 的行业当中去，随着网络技术，智能决策技术的发展，将专家系统，网络技术， 智能决策技术结合起来给研究新一代的专家系统提供了动力。 本文以国家“九五”科技攻关项目“钢铁企业综合自动化关键技术研究与示 范工程”中的子课题“人工智能技术软件工程环境开发研究”中的专题“专家系 统开发工具”为背景，结合网络技术和智能决策系统的研究，介绍了基于网络的 专家系统工具的设想，并对在基于网络的专家系统工具开发过程中遇到的一些理 论和实际的问题进行了研究和开发，介绍了多领域知识库、多任务推理机等模块， 其主要内容包括： 1．在分析了专家系统和专家系统开发工具结构的基础上，介绍了基于网络的 专家系统开发工具NESDP的总体结构设计； 2．NESDP系统的详细设计。介绍了知识获取，多领域知识库，推理机，推 理机控制机制和数据通讯等关键技术问题；从理论的角度，详细讨论了 NESDP系统所涉及到的一些理论；从实际开发的角度，详细讨论了NESDP 中各个模块，包括类模块和功能模块： 3．对NESDP系统中的推理机控制机制和知识获取问题作了更进一步的讨 论，将等分时算法，抢占式调度方式和多级反馈轮转算法运用到推理机控 制机制问题中；提出了基于粗集理论的知识获取方法； 4．主要创新的地方：将等分时算法运用到推理机控制机制当中：采用域标记 的方法来区分同一个知识库中不同领域的知识；提出利用粗集理论来实现 知识的自动获取。 |
| 英文摘要 | Expert System is one of the most important AI fields, and it is successfully applied in every walk of life. With the development of Internet and Intelligent Decision Support System, a combination of Internet, Intelligent Decision Support System and Expert System urges the new generation of Expert System to develop. The research background of this dissertation is the Tool of Expert System, which is one of the subsystems of Research & Development of AI Technology Software Engineering, which is one of the subsystems of the National Key 95 project, Research and Exemplifying Project of Key Integration Automation Technologies in Steel Industry. The dissertation designs the Networking based Expert System Development Platform (NESDP) based on research of Networking Intelligent Decision Support System. There are four main parts: 1. After analyzing the traditional Expert System and tools of Expert System, we introduce the frame of Networking-based Expert System Development Platform. 2. The detailed design of NESDP. The knowledge acquisition, implementation technique for the knowledge base of multiple domains, control method of reasoning for multiple tasks, the data communication technique, database technique and so on are especially discussed in this part. From the theoretical point of view, we discuss some theories related to NESDP; from the practical point of view, we discuss some important modules, including class modules and functional modules. 3. In this part we make some further discussion about control method of reasoning for multiple tasks and knowledge acquisition. Equal Time-Sharing algorithm, Round Robin with Multiple Feedback algorithm and Privative algorithm are applied in control method of reasoning for multiple tasks; and Rough Set theory is applied in knowledge acquisition. 4. The main innovative points: Equal Time-Sharing algorithm is applied in control method of reasoning for multiple tasks, Rough Set theory is applied in knowledge acquisition, the method of domain sign is applied to divide different domain knowledge in one knowledge base. |
| 语种 | 中文 |
| 其他标识符 | 637 |
| 源URL | [http://ir.ia.ac.cn/handle/173211/6785]  |
| 专题 | 毕业生_硕士学位论文 |
| 推荐引用方式 GB/T 7714 | 王文胜. 对基于网络的专家系统开发平台的研究[D]. 中国科学院自动化研究所. 中国科学院研究生院. 2002. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。