KR-102964587-B1 - Intelligent Wireless Fire Detection and Response System and Method Using the Same

KR102964587B1KR 102964587 B1KR102964587 B1KR 102964587B1KR-102964587-B1

Abstract

An intelligent fire detection and response system for detecting and responding to fire wirelessly according to an embodiment of the present invention comprises: a detector that collects detection data of an installed area; a reliability score assigning unit that calculates a reliability score for each detector based on the data collected from the detector and assigns an overall reliability score by synthesizing the reliability scores of each detector; a server that determines whether a fire has occurred and the fire stage based on the overall reliability score and the detection data; and an alarm stage calculating unit that calculates an alarm level for each fire stage according to the judgment result of the server. The reliability score assigning unit is characterized by being configured to dynamically adjust to assign weights based on the past detection data and failure history of each detector, determine whether a specific detector is abnormal by analyzing the correlation between detectors by comparing it with detection data of other detectors installed within a predefined distance range, and collect work environment information input by an administrator and reflect a correction coefficient corresponding to the work environment information in the reliability score.

Inventors

정문숙
김영재

Assignees

주식회사 금강씨엔텍

Dates

Publication Date: 20260513
Application Date: 20251021

Claims (10)

As an intelligent fire detection and response system for detecting and responding to fires based on wireless technology, A detector that collects detection data from an installed area; A reliability score assignment unit that calculates a reliability score for each detector based on data collected from the above detectors, and assigns an overall reliability score by combining the calculated reliability scores of each detector; A server that determines whether a fire has occurred and the fire stage based on the above total reliability score and the above detection data; and It includes an alarm level calculation unit that calculates the fire level alarm level according to the judgment result of the above server, and The above reliability score assignment unit is, (i) Dynamically adjust to assign weights based on the past detection data and failure history of each detector, and (ii) Determining whether a specific detector is abnormal by analyzing the correlation between detectors by comparing the detection data of other detectors installed within a predefined distance range, and (iii) An intelligent wireless fire detection and response system characterized by collecting work environment information entered by a manager and configuring a correction coefficient corresponding to the work environment information to be reflected in the reliability score.
In Article 1, The above server is, Based on the above total reliability score and the above detection data, the detection cycle of each detector is determined, and An intelligent wireless fire detection and response system characterized by being configured to control fire response facilities within a building according to the fire stage derived from the above-mentioned alarm stage calculation unit.
In Article 1, An intelligent wireless fire detection and response system characterized by further including an AI environment learning unit that communicates with the above-mentioned server and performs learning to distinguish between fire and non-fire events by utilizing data collected from the above-mentioned detector together with data from a non-fire event library stored in an external cloud.
In Paragraph 3, The above AI environment learning unit is, A dynamic threshold setting unit that learns environmental data of the installed zone and automatically adjusts the threshold based on the normal state for each space; and An intelligent wireless fire detection and response system characterized by including a signature analysis unit that analyzes detection data collected from the above-mentioned detector to identify the signature of a non-fire event and stores or updates the identified signature in a non-fire event library.
In Article 1, The above server is, An evacuation path calculation module that calculates an evacuation route based on the result of determining whether a fire has occurred and the fire stage, and location information collected from detectors within the building, and Controls a guidance information generation module that automatically generates guidance text corresponding to calculated evacuation routes and fire stages, and The above system is, An intelligent wireless fire detection and response system characterized by further including a user interface that receives guidance information generated by the guidance information generation module above and outputs it to a user.
In Article 5, It further includes a digital twin construction unit that creates a virtual space based on building structural information and firefighting equipment layout information, and visualizes the fire situation by reflecting real-time data collected from the detectors into the virtual space. The above digital twin construction unit is, An intelligent wireless fire detection and response system characterized by being configured to support fire response decision-making by displaying to firefighters flame spread paths, temperature distributions, expected locations of rescue targets, and the placement of fire extinguishing equipment.
As a method for detecting and responding to fires based on wireless technology, A step of collecting detection data from detectors in the installed area; A step of calculating a reliability score for each detector using the above-mentioned collected detection data, and assigning an overall reliability score by combining the calculated reliability scores; A step of determining whether a fire has occurred and the fire stage by performing cross-validation based on the above total reliability score and detection data; A step of determining the response level according to the determined fire stage and controlling fire response facilities within the building; and It includes the step of calculating an evacuation route based on the determined fire stage and location information within the building, and providing guidance information corresponding to the calculated route and fire stage through a user interface. The step of assigning the above overall confidence score is, A step of dynamically assigning weights to each detector based on past detection data and failure history; A step of analyzing the correlation between detectors by comparing the detection data of other detectors within a predefined distance range; and An intelligent wireless fire detection and response method characterized by including a step of correcting the reliability score by reflecting work environment information entered by an administrator.
In Article 7, The step of determining the response level according to the fire stage determined above is, If the fire stage is determined to be Stage 1, the step of saving the generated notification record and registering it as a maintenance target, When the fire stage is determined to be Level 2, a step of providing a notification so that the manager can recognize the alarm and verifying the on-site situation through an integrated video device, and An intelligent wireless fire detection and response method characterized by including a step of issuing a general alarm while automatically controlling facilities within the building when the fire stage is determined to be stage 3.
In Article 7, An intelligent wireless fire detection and response method characterized by further including the step of identifying non-fire events that may be mistaken for fire by referring together detection data collected from the above-mentioned detector and data from a non-fire event library stored in an external cloud, and improving the accuracy of fire determination by reflecting the identified results in a learning process.
In Article 9, The step of identifying the above non-fire event is, Extract feature information by analyzing patterns of collected detection data, and The above feature information is compared with existing patterns stored in the above non-fire event library to determine whether a match exists, and An intelligent wireless fire detection and response method characterized by including the step of labeling the new pattern and updating the non-fire event library when classified as a new pattern that is not matched.

Description

Intelligent Wireless Fire Detection and Response System and Method Using the Same The present invention relates to the field of fire safety management technology. Specifically, it concerns a fire detection system and method that collects data from detectors installed via a wireless network and intelligently determines whether a fire has occurred by calculating a reliability score based on the collected data. In particular, it relates to an intelligent wireless fire detection and response system and method that enhances the accuracy of fire determination by comprehensively reflecting historical information for each detector, inter-correlation, and work environment information, and further minimizes detector malfunctions and enables efficient fire response by providing integrated alarm issuance, equipment control, and evacuation guidance according to the fire stage. Fire safety management technology has evolved with the goal of minimizing casualties in large buildings or multi-use facilities. Existing technologies have been centered around detection devices utilizing temperature, smoke, and gas sensors, as well as alarm systems or basic evacuation guidance systems that operate based on these sensors. These technologies have played a crucial role in shortening initial response times by rapidly detecting a fire and guiding evacuees to the location of emergency exits or escape directions. Furthermore, some systems operate in conjunction with firefighting equipment, such as door controls or sprinklers, to suppress the spread of fire and support evacuation. However, conventional fire response and evacuation guidance technologies have often remained limited to relying on single sensor detection signals or operating based on fixed threshold criteria. Consequently, alarms frequently malfunctioned in non-fire situations, and in actual fire scenarios, these systems failed to account for spatial characteristics or environmental conditions, resulting in reduced accuracy and efficiency in response. Therefore, there is a need for a next-generation fire safety management system capable of quantitatively evaluating the reliability of detection data, distinguishing non-fire events through AI learning, and supporting firefighting response in real time by integrating advanced technologies such as digital twins. In this regard, Korean Registered Patent No. 10-2312310 (Prior Art 1) discloses a system capable of guiding evacuation routes in the event of a fire. This document includes a configuration in which local devices installed in each zone within a facility detect whether a fire has occurred and the location of evacuees through fire detection sensors and occupancy detection sensors, and a relay and management module receives this information to determine the fire location and occupancy status, and then establishes an evacuation plan. However, this document does not disclose advanced technical features such as a structure that makes a comprehensive judgment by scoring the reliability of the detectors, or an AI-based environment learning function that reduces malfunctions by learning and classifying non-fire events. In addition, Korean Published Patent No. 2020-0044213 (Prior Art 2) discloses an intelligent safety management system capable of managing fire and earthquake situations based on the Internet of Things. This document includes a configuration comprising a door operating member that controls a building door, a display member that indicates the status thereof, a fire detection member that detects a fire, a human body detection member that detects a human body, a fire extinguishing member that sprays fire extinguishing substances, and a fire management server that controls said members. In particular, it presents a system structure that considers evacuation and fire extinguishing simultaneously through a control logic that automatically closes the door when a fire is detected and opens the door when a human body is detected. However, this document also does not disclose an AI-based environment learning function that comprehensively analyzes detection data by scoring its reliability, or minimizes malfunctions by learning and distinguishing non-fire events. The aforementioned prior art technologies commonly aim to ensure rapid evacuation and safety in the event of a fire, and are characterized by providing response functions such as evacuation routes and door control through detection sensors and network-based control. However, since these technologies are each limited to specific functional aspects, such as guiding evacuation routes or controlling doors and fire extinguishing equipment, functions for comprehensively calculating the reliability of detection data or minimizing malfunctions by learning from non-fire events are not considered. Furthermore, prior art technologies do not provide advanced features such as real-time visualization of fire scenes using digital twins, precise determination of response levels for eac