Training, Retraining, and Self-training Procedures for the Fuzzy Logic-Based Intellectualization of IoT&S Environments

Abstract

In the past (few) years a lot of research and development projects were proposed concerning the implementation of smart home environment, sometimes even called an intelligent environment, comforting people who use various adaptations of internet of things and services. This paper advocates an opinion that a set of electronically controlled smart things must be intellectualized using human-type reasoning. A novel approach and new algorithms for the hierarchical fuzzy training, retraining, and self-training for intellectualized home environments are proposed in this paper. Training algorithms based on fuzzy logic use top-down hierarchical analysis of home situations under consideration to conquer the curse of increasing number of rules. A successful combination of crisp algorithms for the identification of presence/absence of users in the environment with fuzzy logic-based algorithms for corresponding rules subsets development and processing enables the number of necessary rules to decrease significantly. In the paper a case is presented with the starting number of 2500 rules which later diminished approximately 5 times. For the first time, the changes in users’ wishes are taken into account during the retraining process. An entirely new ability of the system was investigated, and a fuzzy logic-based algorithm for initiating a self-training process without any a priori information is developed as well. The vitality and efficiency of the proposed methodology was tested and simulated on a specialized virtual software/hardware modeling system. The proposed and simulated algorithms are delivered for use in two industrial projects.