基于粗糙集的智能家居规则提取方法  

Rough set based rules extraction method for smart home

作　　者：霍磊[1] 王志良[1] 

机构地区：[1]北京科技大学自动化学院,北京100083

出　　处：《现代电子技术》2016年第5期38-41,46,共5页Modern Electronics Technique

基　　金：国家973项目(2011CB505402);国家自然科学基金项目(61170117);省部产学研结合项目(2011A090200008)专项资金资助

摘　　要：智能家居的关键在于在不需要用户干预的条件下,针对不同的用户生成个性化的服务规则。用户习惯的主观性造成了服务规则不能预先定义,需要在家居系统运行过程中提取。粗糙集是一种处理不完备信息的工具,适合在大量复杂数据中发现潜在规律。把智能家居看成是一个状态驱动的环境,建立状态控制的家居控制系统平台,同时提出了一种基于粗糙集的智能家居个性化规则提取方法,把家庭历史状态记录看作挖掘对象,对状态值进行离散化,通过属性约简,找出影响决策的关键因素,再进行值约简,得出决策规则。智能家居实验室的实验表明,该种方法具有实用价值。The key issue of smart home is to generate the personalized service rules for different uses without user interven- tion. The service rules can＇t be predefined due to the subjectivity of user habit, so they should be extracted in the running pro- cess of smart home system. The rough set is a tool to deal with the incomplete information, and suitable for mining the potential rules in mass complex data. The smart home is regarded as a state-driven environment to establish the smart home control sys- tem platform with controlled state. A personalization rules extraction method for smart home based on rough set is proposed, in which the history status records of the family are taken as the mining object to discrete the state values. The key factors affecting on the decision were found out by attribute reduction, and then the value is reduced to obtain the decision rules. The experimen- tal results from smart home laboratory show this method has practical value.

关 键 词：粗糙集 智能家居 规则提取 属性约简 数据挖掘 

分 类 号：TN92-34[电子电信—通信与信息系统] TP273[电子电信—信息与通信工程]