PROCEDURES AND SYSTEMS FOR ENHANCING THE DETECTION AND REDUCTION OF FALSE FIRE ALERTS IN BUILDINGS USING ARTIFICIAL NEURAL NETWORK (ANN) TECHNOLOGY TO ASSESS THE OPERATIONAL STATUS OF APARTMENT ELECTRICAL SYSTEMS

Abstract

The invention relates to a process for enhancing the detection capability and reducing false alarms in a building's automatic fire alarm system using artificial neural network (ANN) technology for analyzing the total load consumption current in each apartment of the building.The process includes the following steps: Step 1: collect data from the measuring equipment system and sensors with a frequency of 1 second and a delay of less than 1.5 seconds; Step 2: send data to a computer with an AI module, and this data is calculated and processed to identify abnormal load currents; Step 3: if a warning signal appears from a smoke sensor (8), the system will determine the location (address) of the apartment with the smoke sensor sending the warning; Step 3: send the apartment address to the artificial neural network ANN to get the results from the processing of the ANN network for I, u, cosφ of the apartment with this address from before to that time (within 1h); Step 4: the AI ​​module, which is the above-mentioned ANN network, performs analysis, combines and compares the results with other independent signals such as smoke sensor (8), co sensor (7), temperature and humidity sensor (6), and gives appropriate warnings.The invention also relates to a system for detecting and reducing false fire alarms in buildings using artificial neural network (ANN) technology for analyzing load current, voltage, and total power consumption of each household, including: computer (1); automatic fire alarm (smoke alarm) system FDAS (10); temperature and humidity sensors (6), smoke sensors (7); AI module integrated in the computer (1); Hubdata data receiver (2) receives sensor information (3); fire alarm transmission device (11) to the server; wireless / wired router (12); fire alarm light and siren (13), this system performs the above process.