Publication: Enhancement of The Fire Detection Devices' Functionality For The Prediction of Fire Possibility Via Machine Learning Algorithms
Early fire alarm devices were made singular and then as a network to detect fire in a better way. Although recent advances in the design of fire alarm devices have led to better fire detection, current devices cannot deal with sudden or large fires. The best way to be prepared for this type of incident is to predict the possibility of fire. The device we have presented is in the form of a connected network of smoke and dangerous gas detection devices centered on a single device that forms the Master-Slave architecture, and the sub-nodes of this network are primarily responsible for collecting The dataset of the entire building environment and the main node (master) uses this dataset and machine learning algorithms to investigate the possibility of fire. In this article, we first investigate the mechanism of the fire alarm network, and then we consider two supervised and unsupervised scenarios and implement three machine learning models, and compare the output accuracy of each of these models. Finally, we come to the conclusion that the presence of these algorithms in the fire alarm device has clearly improved the ability to detect the possibility of fire. The device we have presented is in the form of a connected network of smoke and dangerous gas detection devices centered on a single device that forms the Master-Slave architecture, and the sub-nodes of this network are primarily responsible for collecting the dataset of the entire building environment and the main node (master) uses this dataset and machine learning algorithms to investigate the possibility of fire.
Certification: https://www.linkedin.com/feed/update/urn:li:activity:7060821286897233920/