在当今信息爆炸的时代，如何从大量的、杂乱无章的、强干扰的海量数据中挖掘潜在的、有利用价值的信息，这对人类的信息处理能力提出了前所未有的挑战。由此产生了人工智能一个崭新领域——数据挖掘（DM）和数据库知识发现（KDD）。 在DM和KDD诸多方法中，粗糙集理论（RST）是一种较为有效的方法，它是一种新的软计算(soft computing)数学工具，可以处理不精确、不一致与不完整等各种不完备的数据，利用它可以从数据中获取有用的知识。RST从产生至今，已在许多方面获得了成功应用。但在实际应用中，随着数据样本规模的增大，就很难（甚至不可能）用人工的方法获取知识（规则集），所以有必要开发一种基于粗糙集算法的从数据样本中提取规则的软件，从而使RST在实际中得到更好应用。人可以对复杂系统进行有效的控制，利用RST可以从人的控制行为数据中提取人的控制行为规则。本文的目的就是开发一种基于RST从人的控制行为数据中自动提取控制规则的软件，并通过实际的控制系统进行检验规则有效性。 为了完成基于RST控制规则自动获取软件，首先分析了RST知识获取算法，得到软件系统目标；完成系统总体设计，其中包括建立决策信息表、决策信息表离散化、离散决策信息表约简、建立原始决策表、原始决策表完备化功能模块划分以及各功能模块间流转数据的存储即数据库设计；完成人机界面设计和各功能模块详细设计；使用VC++和SQL Server 2000编程实现各模块功能，完成控制规则自动获取。 为了验证软件提取规则的有效性，本文采用由变频器、交流电动机、水泵和单水箱组成的单容水箱控制系统进行验证。人通过调节变频器的控制电压使水箱的液位保持恒定，首先使用PCL-812PG采集卡采集人的控制行为数据；然后使用软件从数据中提取控制规则；提取规则设计仿人智能控制器并用它控制单容水箱系统。实验结果表明，仿人智能控制器达到一定控制指标，证明软件提取规则是有效的。

Nowadays, it is the era of information explosion. How to mine latent and valuable information from massive,chaotic,and strong jamming data which is unprecedented challenge to humanity’s information-handling capacity.Therefore , new fields of artificial intelligence are produced that are Data Mining (DM) and database knowledge discovery (KDD). Among many ways of DM and KDD, rough set theory (RST) is a effective way, which is a kind of new mathematical tool of soft computing. Inaccurate, inconsistent ,incomplete, and other incomplete Data can be processed by RST.Some useful knowledge can be extracted from a data sample applying RST. From there so far, RST has been successfully applied in many areas. However, with the data sample size enlargement , it will be very difficult (even impossible) to use artificial methods to acquire knowledge (rule set) in practical application,, so it is necessary to develop software method based on RST to extracte Rules from the data sample , so that RST can be well applied in practice.human being can make effective control to a complicated system and it is possible to use RST to withdraw control rules from person’s control behavioural data.The purpose of this paper is to develope a software based on RST which can automaticly get control rules from person’s control behavioural data ,and the control rules are validated by actual control system. In order to complete the software , firstly, algorithm of RST is analyzed to obtain the goal of the software system.Then integrated system design is completed,which include system functional module division and database design used to store the data transfered between modules, the divided system include building, discretizating decision information table, reduction of discrete decision information table, taking and completizating the original ecision table. Human-computer interface and detailed modules design are completed. VC + + and SQL Server 2000 are used to realize the modules function and gain control rules automatically. In order to confirm the validity of the rules gained through the software, the control system of a single water-tank composed by inverter, AC motors, pumps and single water tank is adopted. a operator can maintain water level constant by adjusting converter’s control voltage. firstly, person’s control behavioural data is acquired by PCL-812PG. then,the software is used to withdraw the control rules from the data;thirdly,according to the rules,a humanoid intelligent controller is designed, then the system is controled by the controller. Experimental results show the controller can achieve certain guide posts and it is proved that the rules gained from software are effective.