Multi Sensor Fire Detection in Low Voltage Electrical Panel Using Modular Fuzzy Logic

Fire in a building is one of the disasters that is very damaging and very serious for human safety, it is known that DKI Jakarta has the highest fire cases with an average of two incidents per day with the biggest cause coming from electricity. The use of multiple sensors, known as multi-sensors, has demonstrated its superiority in carrying out early detection of fire so that the development of technology is currently focused on the use of better decision making algorithms. This study uses many sensors consisting of flame sensors, carbon monoxide sensors, temperature sensors and smoke sensors as an early detection of fire with fuzzy logic decision making algorithms in modeling low voltage electrical panels, presented in the form of a comparison between conventional type fuzzy logic with fuzzy logic modular type. Sensor data processing from the simulation of smoke to fire through Arduino Mega, then simulations using Simulink on MATLAB, the simulation results showed a reduction in the number of rules from 81 to 43 in the modular type, while the probability output at the 5th second was higher and at the 21st second lower than the conventional type.