A Rule Verification System for Smart Buildings

Abstract

As an important component of Internet of Things (IoT), wireless sensor-actuator networks can significantly improve the practicality and flexibility of smart building systems. In smart building systems, services, stored as rules, are achieved by rule analyzing and executing. However, the irrational contents of rules and the conflicts among rules may bring confusion and maloperation in the system. To verify the correctness of rules, we propose a lightweight rule verification and resolution framework which mainly include: 1) a rule verification system for content anomaly detection and rule conflict detection by using domain knowledge and probability analysis and 2) a quick resolution strategy for rule conflicts based on conflict-scenario-analysis. This framework can balance the verification quality with speed and guarantee that the rule system performs appropriately. Moreover, in order to reduce the comparison in rule verification, we apply a tree structure to store the building structure hierarchically and bind every rule to one node in the location tree. This way, the position information of a rule can be extracted, and the verification efficiency can be improved. The experimental results show that our proposed framework and strategies can perform efficiently and flexibly in a smart building system.