Turkish Journal of Electrical Engineering and Computer Sciences




Fires generally occur due to human carelessness and the change in environmental conditions. The uncontrolled fire results in death incidents of humans and animals as well as severe threats to the ecosystem. The preservation of the natural environment is important. The wireless sensor networks, widely used in different monitoring applications, is used in this work. For fire detection, we use flame, smoke, temperature, humidity, and light intensity sensors in our proposed network node which is low-cost, reduced-size, and power-efficient. The experiments are performed in a well-controlled real-time environment. The proposed node transmits the sensed data to the central node. The central node then transfers the data gathered from all the nodes to the control station using an air interface. To decide whether there is an incident of fire or not, and to have an idea on fire intensity, we combine multiple attributes sensed from a single node using Bayesian approach due to its simplicity and resemblance with human reasoning. In the experimental setup, the conditions for fire with different intensity are generated and the results confirm the validity of the proposed approach in terms of accuracy and less false alarms.


Wireless sensor networks, microcontrollers, fires, classification algorithms, gas detectors

First Page


Last Page