Adaptive neuro-fuzzy inference system–based fire detection system performance improvement using wired and wireless communication

Abstract

Conventional fire detection/suppression systems have several limitations, including their slow response time, high costs associated with wastewater handling, and limited area of coverage. This work develops a model of an Adaptive Neuro-Fuzzy Inference System (ANFIS)-based fire suppression and detection system within a comprehensive set of simulation environments. The design of this simulation integrates ANFIS control technologies with a digital display, enabling the real-time monitoring of both wired and wireless communications. The structure of the system is based on two major components: (1) a layout structure that illustrates the spatial relationships between temperature and smoke detector controllers, ANFIS decision units, and alarm interfaces; and (2) a sprinkler structure that describes the fluid dynamics of fire suppression systems using computational fluid dynamics (CFD). The sprinkler systems utilize programmable variable-speed control to adjust the water flow and alarm levels, both of which are based on the ANFIS outputs, which are graded according to temperature and smoke intensity. The use of graded ANFIS outputs will both reduce operational costs in the simulation and enhance the efficiency of the delivered water during simulations. The results of our simulations demonstrate that the proposed system model produces highly faithful representations of fire through various fire events, exhibits superior suppression performance compared to other systems, and efficiently integrates hydraulic technology (CFD) and data-driven approaches (ANFIS) into a single fire monitoring and response system.