A Novel Distributed Knowledge Reasoning Model

Abstract

This paper proposes a novel model and device based on distributed knowledge reasoning, for predicting the environmental risk degree for an abnormal event (e.g., fire) in the Internet of Things (IoT) environment, which includes: (i) determining the predictive environmental information of the current node based on its historical environmental information and the predefined time-series prediction algorithm; (ii) determining the probability distribution function (PDF) of the environmental information obtained at each time corresponding to the historical environmental information; (iii) determining the deviated environment information based on the current environmental information and the probability distribution function mentioned above; and (iv) determining the environmental risk degree of the current node, based on the current environmental information, the predictive environmental information and the deviated environmental information. The wireless node of in the proposed model has the ability of perceiving and reasoning for the occurrence of an abnormal event. According to the current environmental information, the predictive environmental information and the deviated environmental information, the environmental risk degree of the current node could be determined jointly.