一种针对物联网智能系统的规则冲突检测方法  

A Rule Conflict Detection Approach for Intelligent System of Internet of Things

作　　者：杨波 郭浩然[2] 冯俊辉[2] 李戈 金芝 Yang Bo;Guo Haoran;Feng Junhui;Li Ge;Jin Zhi(School of Information Science&Technology,Beijing Forestry University,Beijing 100083;School of Information Science&Technology,North China University of Technology,Beijing 100144;Engineering Research Center for Forestry Oriented Intelligent Information Processing,National Forestry and Grassland Administration(Beijing Forestry University),Beijing 100083;School of Computer Science,Peking University,Beijing 100871;Key Laboratory of High Confidence Software Technologies(Peking University),Ministry of Education,Beijing 100871)

机构地区：[1]北京林业大学信息学院,北京100083 [2]北方工业大学信息学院,北京100144 [3]国家林业和草原局林业智能信息处理工程技术研究中心(北京林业大学),北京100083 [4]北京大学计算机学院,北京100871 [5]高可信软件技术教育部重点实验室(北京大学),北京100871

出　　处：《计算机研究与发展》2023年第3期592-605,共14页Journal of Computer Research and Development

基　　金：中央高校基本科研业务费专项资金(BLX202003);国家自然科学基金项目(62072007,62192733,61832009,62192730,61772078,61502011);北京林业大学热点追踪项目(BLRD202124)。

摘　　要：物联网系统架构的核心是逻辑控制器,逻辑控制器中使用规则控制业务逻辑,减少物联网系统的开发、维护成本,提高物联网设备的灵活性.但随着物联网系统的规模扩大,规则间的关系变得复杂,从而可能产生规则冲突.为避免规则发生冲突,一些研究者提出了规则冲突检测方法.但是,以往的规则冲突检测方法还存在规则冲突类型分析不全面、检测结果准确性较低的问题.为此提出一种针对物联网系统控制逻辑的形式化规则冲突检测方法(formal rule conflict detection,FRCD).该方法首先形式化定义规则及规则冲突,其中将规则定义为控制主体、动作、触发条件、符号的组合;然后根据规则对系统的影响以及规则的结构特征,总结出7类规则冲突类型;最后设计规则冲突检测的算法,并介绍规则冲突检测的详细过程.在2个物联网系统上开展实验,与已有的3种典型的物联网规则冲突检测方法进行对比.这3种方法分别是基于用户、触发器、环境实体和作动器的冲突检测方法(user,triggers,environment entities,and actuators,UTEA)、基于Web语义的策略冲突检测方法(semantic Web-based policy interaction detection with rules,SPIDER)和半形式化的冲突检测方法(identifying requirements interactions using semiformal,IRIS).实验结果显示,FRCD规则冲突检测方法效果更好.The core of the Internet of things(IoT)system architecture is the logic controller.The logic controller uses rules to control the business logic,which reduces the development and maintenance costs of the IoT system and improves the flexibility of the IoT devices.As the scale of the IoT system expands,the relationship between the rules becomes complicated.This may cause rule conflicts.In response to this problem,some researchers have proposed some detection methods for rule conflicts.However,the existing rule conflict detection methods still have some problems,such as incomplete analysis of rule conflict types and low accuracy of detection results.For these reasons,a formal rule conflict detection(FRCD)method for the control logic of the IoT intelligent system is proposed.This method formalizes the structure of rules,and defines rules as a combination of control subjects,actions,trigger conditions,and symbols.Then according to the influence of the rules on the system and the structural characteristics of the rules,7 types of rule conflicts are summarized.Finally,an algorithm for rule conflict detection is designed,and the detailed process of rule conflict detection is introduced.We carry out experiments on two IoT systems and compare them with three typical IoT rule conflict detection methods.These three methods are the formal rule model conflict detection method based user,triggers,environment entities,and actuators(UTEA),semantic web-based policy interaction detection with rules(SPIDER),identifying requirements interactions using semiformal(IRIS).The experimental results show that the formal rule conflict detection method in this paper is more effective.

关 键 词：物联网 智能系统 规则冲突 冲突检测 规则形式化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]