摘要
专家系统是人工智能应用研究最活跃和最广泛的应用领域之一。专家系统，简单地说是一种模拟人类专家解决领域问题的计算机程序系统，它将特定领域人类专家的特殊知识赋予机器，用计算机模拟专家的思维活动、推理判断，使对问题的求解达到专家的水平。专家系统在许多领域中都有成功的运用，但是随着发展的深入和要求的提高，有关知识获取以及如何处理知识的模糊性或不确定性等问题逐渐显示出来。目前专家系统的水平远远不能满足实际的需求。其根本原因在于系统的知识表示和处理的理论方法上没有取得突破。如何获得不准确或不确切的知识和关系，并在此基础上做出正确的结论就成为专家系统所需要解决的一个问题。
粗糙集理论作为人工智能领域中的一个新学术热点，能有效地处理不完整、不确定知识的表达和推理，其有效性已在许多领域的应用中得到证实。粗糙集理论无需任何先验信息，能有效地分析和处理不精确、不一致、不完整等不完备性数据，通过发现数据间隐藏的关系，揭示潜在的规律，从而提取有用信息，简化信息的处理。因此将粗糙集理论的知识获取方法和知识的约简引入到专家系统具有广泛的前景。
本文提出了一种基于粗糙集理论的专家系统模型。该模型在知识获取阶段引入知识过滤器，根据知识依赖度的变化对采集的知识进行评价和分类。该系统还在知识库构造阶段引入知识重构机制，对原有的知识库进行精简和重构。这种方法不仅消除了知识库中的冗余属性，还对属性值空间进行合理划分，对整个系统的性能有明显的改善效果。
本文还在故障诊断专家系统的基础上，引入粗糙集理论，并以信息系统属性值表为主要工具，对专家系统中的规则进行约简，并剔除不必要的属性，揭示了故障诊断信息中内在的冗余性，降低了故障诊断专家系统构成的复杂性，并建立了简化后的决策规则。
Expert System is one of the most active and the widest fields of application research in Artificial Intelligence. Expert System is a simply computer program system that simulates human experts to solve certain problems. It endows special domain knowledge of human experts to machines, reasoning and judging in thought activity that computer simulates experts, in order to achieve experts' level in solving problems. Expert system has successful application in many domains. However, with the development of technology and improvement of demands, some problems have appeared such as knowledge acquisition and how to deal with fuzzy problems or uncertainty. At present, the level of expert system cannot meet the needs of practice. Ultimate reason is that there is not breakthrough in the theory methods of system's knowledge express and management The problem that expert system must solve is that how to acquire inexact knowledge and relation, at the basis of which, a correct conclusion must be drawn.
Rough sets theory is a new learning hotspot. It can effectively deal with incomplete, uncertain knowledge express and reasoning, its validity has been proved in many fields. Rough sets theory need not any prior knowledge or information, and can analyze and dispose imprecise, inconsistent, incomplete datum. It discloses potential disciples, pick up useful information and reduce information by finding hidden relation among data. Therefore, it has a wide foreground to introduce knowledge acquisition and knowledge reducing in expert system.
In this paper a rough-set-based expert system model is proposed. This model introduces a knowledge acquisition filter during the phase of knowledge acquisition. According to the variation of the knowledge dependency, it evaluates and classifies the newly collected knowledge. This system also introduces a mechanism for knowledge reconstruction during building knowledge base to refine and restructure the primitive knowledge base building. This method presented in the paper not only removes the redundant attributes of the knowledge base, but also restructures the value space of the attributes, and
improves the performance of the whole system significantly.
Moreover, on the basis of fault diagnosis expert system, rough set theory is introduced. Knowledge representation system table is taken as a major tool to reduce the rules of expert system in which unnecessary properties are eliminated. The redundancy of fault diagnosis information is revealed. The complexity of fault diagnosis expert system's structure is also reduced. The decision-making rules are given finally.
引文
[1] Z Pawlak. Rough Sets-Theoretical Aspects of Reasoning about Data, Kluwer Academic Publishers, 1991
[2] Z Pawlak. Rough Sets Approach to Knowledge-based Decision Support, European fournal of Operational Research, 99(1997):48～57
[3] 张文修，吴伟志等．粗糙集理论与方法．北京：科学出版社，2001．12～22
[4] 史习智，王本颜等．专家系统——概念与实例．上海：上海科学技术文献出版社，1988
[5] 理查德．福西斯．专家系统原理与实例研究．北京：中国铁道出版社，1989
[6] Janson. TURBO-PROLG与专家系统．电子工业出版社，1994
[7] A. Mrozek. Rough set dependency analysis among attributes in computer implementation of expert inference models. Int. J. Man-machine studies. 1989,30:457～473.
[8] L. Plonkn. Rule-based stabilization of the inverted pendulum. Computational Intelligence. 1995, 11(2):348～356.
[9] 尹朝庆，尹皓等．人工智能与专家系统．北京：中国水利水电出版社，2001
[10] 施鸿宝等．专家系统．西安交通大学出版社，1990，3
[11] 王钰，苗夺廉，周有健．关于Rough sets理论与应用的综述．模式识别与人工智能，1996，4：337～344
[12] Pawlak Z. Rough sets. International journal of information and computer science, 1982, 11 (5): 341～356
[13] Slowisk i R. Intelligent decision support: handbook of applications of the rough sets theory. The netherlands: Kluwer Academ ic Publishers, 1992:1～235
[14] K Slow insk i. Rough sets approach to analysis of data from peritoneal lavage in acute panereaitis。Medical information, 1988, 13:143～159
[15] J Fibak. Rough sets based decision algorithms for treatment of duodenal ulcer by HSV. Bull PolAcad SciBiol, 1986, 34:227～246
[16] 郭萌．数据挖掘与数据库知识发现．模式识别与人工智能，1998，11(3)
[17] 张琦，韩祯祥，文福栓．一种基因粗糙集方法的电力系统故障诊断／警报处理的
新方法．中国电力，1998，31(4)：32～38
[18] S. K. M. Wong, W. Ziarko and R. Li Ye, "Comparison of rough-set and statistical methods in inductive learning", International Journal of Mao-Machine Studies, 1986
[19] J. R. Quinlan, C4.5: Programs for Machine Learning. Morgan Kaufmann, 1992
[20] 刘清．Rough集及Rough推理．北京：科学出版社，2001
[21] 冯博琴．实用专家系统．电子工业出版社，1992
[22] 刘健勤．进化粗糙集的逼近最优性．中南工业大学学报，2000，1
[23] 孙颖楷，张邦礼，曹龙汉，曹长修．基于粗糙集理论的人工神经网络故障诊断系统．重庆大学学报，2000，6
[24] 李卫华，汤怡群等．专家系统工具．北京：气象出版社，1987
[25] 武波等．专家系统．北京：北京理工大学出版社，2001，1
[26] 卫金茂，黄道．粗糙集理论中新的知识发现方法，华东理工大学学报，2000，4
[27] 袁小宏，赵仲生，屈梁生．粗糙集理论在机械故障诊断中的应用研究．西安交通大学学报，2001
[28] 束洪春，孙向飞，司大军．基于粗糙集理论的配电网故障诊断研究．中国电机工程学报，2001
[29] 袁保奎，郭基伟，唐国庆，卢毅．基于粗糙集理论的变压器故障分类，电力系统及其自动化学报，2001，5
[30] 李文，孙辉，陈善本．一种建立模糊模型的粗糙集方法．控制理论与应用，2001年01期
[31] 刘宜平，沈毅，童树鸿，刘志言．设备故障诊断的粗糙集方法研究．高技术通讯，2001，5
[32] 李剑，范小军，黄沛．基于粗糙集的知识理论及其应用．系统工程理论方法应用，2001
[33] 叶青，杨家本，柴跃廷．基于粗集理论的知识处理方法在专家系统中的应用．信息与控制，2001，3：193～198
[34] 曹修，孙颖楷，曹龙汉，张邦礼．基于粗糙集理论的内燃机故障诊断专家系统．重庆大学学报，2001，4：45～47
[35] 沈宝华，周志刚．故障诊断专家系统的知识处理，火力与指挥控制，1994
[36] Z. Pawlak. Rough Sets. International J. Of Fuzzy Sets and Systems 19, 1986
[37] 李桂青等．工程设计专家系统的原理与程序设计方法．气象出版社，1991
[38] 叶东毅，陈昭炯．一个改进的粗糙集属性约简算法，福州大学学报，2000
[39] 黎明，张化光，何希勤．基于粗糙集的模糊模型辨识方法．东北大学学报，2000
[40] 刘宜平，沈毅，重树鸿，刘志言．一种基于粗糙集理论的容错网络实现及其在故障诊断中的应用．电机与控制学报，2000
[41] 陈云化，叶东毅，赵士亮．基于粗糙集理论的规则提取算法的研究．福州大学学报，2001
[42] 张学民，安利平·动态数据库规则获取和更新的粗糙集方法·天津大学学报，2001
[43] 夏雨佳，李少远，席裕庚．一种基于粗糙集的信息系统决策规则提取方法．控制与决策，2001
[44] 曾黄磷，粗集理论及其应用(一)—粗糙集理论基础．四川轻化工学院学报，1996
[45] 邱丽绚．基于粗集理论的知识发现．控制与决策，1999
[46] 祝峰，王加佳，严向奎．粗集理论的现状与前景．新疆大学学报，2000
[47] 曾黄麟·基于粗糙集方法的智能专家系统，中国工程科学，2001
[48] 张启忠，罗志增．粗集理论在信息融合中的应用．机电工程，2001
[49] A. Skowron. Rough Sets in KDD. Special Invited Speaking, WCC 2000 in Beijing, Aug. 2000
[50] 谢克明，杨静．粗糙集理论及其在智能控制领域的应用前景．太原理工大学学报，1999
[51] 汪文超．故障诊断专家系统的知识获取．中国修船，1994
[52] 刘健勤．粗糙集理论及其最新进展．计算技术与自动化，1998年01期
[53] Z. Pawlak. Knowledge, Reasoning and Classificatien-A Rough Set Perspective. Bulletin of the European Association for Theoretical Computer Science 38,1989
[54] 王耀南．智能控制系统·模糊逻辑·专家系统和神经网络控制．湖南：湖南大学出版社，1996