(714b) Development of a Real-Time Monitoring System of Abnormal Fire Incidents Based On the Mining of National Fire, Disaster and Meteorological Data

Prediction of fire accident is one of the important issues from all over the world. Various researches have been active to improve the sustainability. Nowaday, occurrences of fire incidents are recorded into National Fire Data System (NFDS), and   weather data go down in Korea Meteorological Data system. Their statistics are updated every day. If these big data which are accumulated on thirty or forty thousand per year apply to real time monitoring system detecting abnormal occurrences of fire, we not only do not concentrate reaction after accident but achieve proactive fire prevention and damage reduction. In this study, we develop a prediction system that considers the context related to fire occurrences on a certain day and at a specific district. We design an effective monitoring system for proactive alarm generation based on the SPC control charts. By the mining of the NFDS data collected over the past 5.5 years, data from the link through National Weather Service and fire statistics data, and considering date, season, place, weather, etc., We devide fire data recently 5.5 years to a lot of space time scale, and carry out monitoring based on fire incidence.  Most important thing is targeting of monitoring. After setting period and location, each factor is analyzed. Then we select the monitoring target using frequency of fires, casualty etc., The developed system of considering these information automatically runs the monitoring routine and launches warnings (L or H) or alarms (LL or HH) to get the attention and bring the intervention of operating personnel. In the past, almost people reacted to fixed safety regulation of experience-based and depended on their feeling. Contrastively, they are able to doing proactive action using developed system. If fire department officials use developed system, they can respond fire accident when the system senses abnormal situation and preparation is possible. So The system is expected to reduce the occurrence of fire incident.