Approach to prototyping a technical device for fire detection based on the principles of artificial intelligence
Abstract
The aim of the study is to justify the developed prototype of a technical device that uses machine learning methods for early detection of fires in order to minimise damage and save human lives. The authors formulated a hypothesis about the expediency of expanding the frequency range within which the fire automation system detects the primary signs of fire in order to increase the efficiency of its operation. To test the hypothesis, a prototype has been developed, based on the typical structure of the existing fleet of fire automation systems, including an additional unit operating on the principles of artificial intelligence. The article describes an additional unit working on the principles of fuzzy logic and convolutional neural networks to analyse the data coming from video cameras and fire detectors in order to determine the fire in the early stages of ignition.