KR20210028949A - System and Method for Detcting Fire of in advance - Google Patents

Abstract

The present invention relates to a fire detection monitoring system which provides real-time fire detection monitoring and step-by-step countermeasures, and to a method thereof. The system of the present invention includes an arc detector for detecting an arc phenomenon, and a server which receives arc data detected by the arc detector in real time, predicts the arc data, and provides a countermeasure step by step according to a predefined dangerous level.

Description

System and Method for Detcting Fire of in Advance

The present invention relates to a fire detection monitoring system, and more particularly, to a fire detection monitoring system and method capable of detecting a fire occurrence sign in advance through analysis of predictive data for arc data measured in real time.

Most of the electrical fires can be attributed to short circuits of wires or arcing caused by poor contact (ie, spark discharge, sparks generated between wires, etc.). Looking at the nationwide ignition factors by year, over 30% of them are ignition due to electrical factors, and among them, fires due to arcing are about 80%, and the proportion of fires due to arc is large.

As such, the proportion of fires caused by arcs is high, but the response to this is still insufficient. For example, it is not obligatory to install a sensing device that detects arcing. Currently, an earth leakage circuit breaker or an overload circuit breaker is installed, but these devices do not have a function to cut off power by detecting an arc phenomenon. And even though a device for detecting arc phenomenon is installed, a system for monitoring and analyzing fire caused by arc in real time is inadequate.

It is not an exaggeration to say that there was no fundamental plan to prevent the occurrence of fires in advance due to these problems. Therefore, after the fire occurred, the necessary countermeasures to extinguish the fire had to be taken, which in turn always resulted in a problem leading to enormous personal and property damage.

Conventionally, most fire related systems compare and analyze acquired image data and pre-collected data at a predetermined location, such as Korean Patent Registration No. 10-1894738 (registered on August 29, 2018), and display an alarm for fire alarm. It was about.

Accordingly, an object of the present invention is to solve the above problems, by detecting an arc phenomenon that is one of the causes of an electric fire through real-time monitoring, analyzing and predicting the predicted data of the arc data to detect the fire occurrence symptoms in advance. It is to provide a fire pre-detection monitoring system and method that can cope with it.

Another object of the present invention is to increase the accuracy of pre-prediction of fire occurrence symptoms by extracting only a real arc signal that causes an electric fire during an arc phenomenon.

The present invention for achieving the above object, the arc detector for detecting the arc phenomenon; And a server that receives the arc data detected by the arc detector in real time, predicts the arc data, and provides a step-by-step response plan according to a predefined dangerous water level level.

According to the present embodiment, the arc detector transmits the arc data including unique information for identifying the arc detector to the server through a hypertext transfer protocol (HTTP) method.

According to the present embodiment, the server is a collection unit for collecting arc data including a real arc signal, a signal irrelevant to an electric fire, and a noise component, which causes an electric fire from the arc detector, the A prediction data generation unit that extracts only the real arc signal from the arc data and generates prediction data that predicts the amount of change of the extracted real arc signal, a monitoring unit that monitors the dangerous water level through analysis of the prediction data, and the risk It is composed of a notification/blocking unit and a report/removal service unit that provide step-by-step countermeasures according to the water level.

According to the present embodiment, the prediction data generation unit generates prediction data for the real arc signal using a Kalman filter, and calculates the prediction data by the following equation.

. 여기서,

는 k-1 시점의 측정값을 토대로 한 k 시점의 상태,

는 해당 시간에서 이전 상태에 기반한 상태 전이 행렬,

는 사용자 입력에 의한 상태 전이 행렬,

는 사용자 입력값을 말한다.

. here,

Is the state at point k based on the measured value at point k-1,

Is the state transition matrix based on the previous state at that time,

Is the state transition matrix by user input,

Refers to the user input value.

According to this embodiment, the server visualizes and provides the arc data.

According to another feature of the present invention, the plurality of arc detectors each detecting an arc phenomenon; Receiving, by a server, arc data according to the detection result through a communication method; Extracting, by the server, a real arc signal using a Kalman filter; Generating predicted data that predicts the amount of change of the real arc signal; And providing, by the server, a risk level level through analysis of the predicted data, and providing a step-by-step response plan according to the risk level level.

According to the present embodiment, the communication method is HTTP (hypertext transfer protocol), and the arc data includes and transmits detected arc information and unique information for identifying the arc detector.

According to the present embodiment, the predicted data is calculated by the following equation, and the state transition matrices F and B are learned by performing prediction using the current value and the change amount of the current value as the state vector x to generate the predicted data.

. 여기서,

는 k-1 시점의 측정값을 토대로 한 k 시점의 상태,

는 해당 시간에서 이전 상태에 기반한 상태 전이 행렬,

는 사용자 입력에 의한 상태 전이 행렬,

는 사용자 입력값을 말한다.

. here,

Is the state at point k based on the measured value at point k-1,

Is the state transition matrix based on the previous state at that time,

Is the state transition matrix by user input,

Refers to the user input value.

According to the fire pre-detection monitoring system and method of the present invention as described above, through detection of arc data in real time and prediction through a Kalman filter, it is possible to detect a fire occurrence sign due to an arc phenomenon in advance, thereby preventing the fire itself. Even if a fire occurs, it is possible to quickly respond to fire, so there is an effect of remarkably reducing human and property damage.

In addition, the present invention extracts a real arc signal, which can be a direct cause of an electric fire, and predicts the amount of change in the extracted real arc signal, thereby monitoring and predicting accurate fire occurrence symptoms.

1 is a block diagram showing a fire pre-detection monitoring system according to an embodiment of the present invention
2 is an exemplary graph showing a spectrum change state when an arc signal is generated according to the present invention
3 is an exemplary graph showing a simulation result of arc data and predicted data measured according to the monitoring result of the arc phenomenon of the present invention
4 is a graph illustrating prediction of a change amount of a real arc signal according to the present invention
5 is a flow chart showing a fire pre-detection monitoring method according to an embodiment of the present invention

Since the present invention can apply various transformations and have various embodiments, specific embodiments are illustrated in the drawings and will be described in detail in the detailed description. However, this is not intended to be limited to a specific embodiment of the present invention, it should be understood to include all conversions, equivalents, or substitutes included in the spirit and scope of the present invention. In describing the present invention, when it is determined that a detailed description of a related known technology may obscure the subject matter of the present invention, a detailed description thereof will be omitted.

Terms such as first and second may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another component.

The terms used in the present invention are used only to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present application, terms such as "comprise" or "have" are intended to designate the presence of features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, but one or more other features. It is to be understood that the presence or addition of elements or numbers, steps, actions, components, parts, or combinations thereof does not preclude in advance.

Spatially relative terms such as below (below, beneath, lower) and above (above, upper) facilitate the correlation between one device or components and other devices or components as shown in the drawing. Can be used to describe. Spatially relative terms should be understood as terms including different directions of the device during use or operation in addition to the directions shown in the drawings. For example, when an element shown in the figure is turned over, an element described below (beneath) another element may be placed above the other element. Accordingly, below, which is an exemplary term, may include both directions below and above. The device may be oriented in other directions, and thus spatially relative terms may be interpreted according to the orientation.

Therefore, the spirit of the present invention is limited to the described embodiments and should not be defined, and all things that are equivalent or equivalent to the claims as well as the claims to be described later fall within the scope of the inventive concept. .

Hereinafter, the present invention will be described in more detail based on the embodiments shown in the accompanying drawings. The present invention proposes an invention capable of generating revenue by detecting the signs of fire in advance through big data analysis while monitoring the arc phenomenon of an electric distribution box in real time, and providing relevant data for knowing the signs of fire to specific customers. will be.

As shown in FIG. 1, an arc detector 100 for detecting an arc occurrence is installed at a predetermined location. As a place where the arc detector 100 is installed, it may be a distribution box that may cause damage due to electrical cutoff in the event of a fire, or a power pole to which various lines are connected. In addition, a circuit breaker (not shown) is installed together with the arc detector 100. The circuit breaker is to cut off the current when a fire sign is detected.

A server 200 for storing and managing arc data measured in real time from the arc detector 100 is provided. The arc detector 100 and the server 200 are connected through an Internet-based hypertext transfer protocol (HTTP) communication protocol, and the arc detector 100 transmits arc data to the server 200 in an HTTP method. Here, since not only one arc detector 100 is installed, but a plurality of arc detectors are distributed over several areas, the arc detector 100 is included in the arc data so that the arc detector 100 that is installed at a specific location and transmits arc data can be identified. It would be natural that the unique information assigned to each 100) should be included. Examples of unique information may be a specific area or building, street information, serial number, and the like.

The server 200 includes a collection unit 210 that collects arc data measured by the arc detector 200 and a prediction data generation unit 220 that generates prediction data of the arc data.

In this embodiment, the collection unit 210 also collects unnecessary signals and noise components included in the arc data. That is, the arc data includes a real arc signal that causes an electric fire, a signal unrelated to an electric fire (unnecessary signal), and a noise component, and all these signals are collected. Here, the unnecessary signal and noise are irrelevant to the electric fire as described above, but the size and fluctuation of the signal tends to appear very severe immediately before the fire occurs (part'A' in FIG. 2).

In this embodiment, the prediction data generation unit 220 may generate prediction data in a current state using a Kalman filter. At this time, only the real arc signal is used as the predicted data. That is, when the Kalman filter is used, unnecessary signals and noise components are removed, and only real arc signals are extracted.

In addition, the following Equation 1 is used to predict arc data using the Kalman filter.

는 k-1 시점의 측정값을 토대로 한 k 시점의 상태,

는 해당 시간에서 이전 상태에 기반한 상태 전이 행렬,

는 사용자 입력에 의한 상태 전이 행렬,

는 사용자 입력값을 말한다. here,

Is the state at point k based on the measured value at point k-1,

Is the state transition matrix based on the previous state at that time,

Is the state transition matrix by user input,

Refers to the user input value.

이고,

이다. Also,

ego,

to be.

By using this, the state transition matrices F and B are learned by repeatedly performing prediction by using the current value i and the amount of change d of the current value as the state vector x, so that the measured arc data can be more accurately predicted. .

In addition, the server 200 according to the present embodiment includes a monitoring unit 230 that monitors the signs of fire through the analysis of the predicted data. The monitoring unit 230 performs monitoring according to different standards depending on where the arc detector 100 is installed. This is because, for example, when comparing a general home and a factory, the corresponding standard according to the arc phenomenon may be different because the current capacity supplied is different. The criteria according to these arc data are shown in the following [Table 1].

Relief step Warning stage Risk stage Fire occurrence stage Moni
Turing Arc signal
(example)

standard Level within 50% of the capacity of quality analysis and monitoring devices Continuous for 1 hour at 80% or more of the capacity of the quality analysis and monitoring device or 1 minute at 10mA or more Quality analysis and monitoring lasts 15 to 30 minutes at 100% or more of device capacity or 1 minute at 15 mA or more Quality analysis and monitoring lasts 10 to 20 minutes at more than 120% of the device capacity or 1 minute at 20 mA or more prediction Big Data and Kalman Filter Grasp the phenomenon Arc signal prediction
Similar pattern analysis Arc signal prediction
Similar pattern analysis Response Web/Mobile Alert notification transmission Alert notification transmission, guard dispatch Fire alarm transmission, automatic 119 report

Here, [Table 1] is only an example, and it may be changed as much as possible depending on the site. In other words, in [Table 1], it can be regarded as a warning stage in case of sustaining for 1 hour at 80% or more of capacity or 1 minute at 10mA or more, but these standards cannot be applied equally to homes and factories.

In addition, the server 200 includes a notification/blocking unit 240 that outputs an inspection status or an alarm sound or cuts off a current. The notification/blocking unit 240 includes an inspection notification function that notifies an inspection notification when abnormal data occurs or inspection notification information at predetermined intervals, a danger notification function that generates an alarm sound according to a dangerous water level, and a fire place in real time. It performs a location notification function that notifies you, and a current cutoff service function that cuts off the current by running a circuit breaker after predicting a fire.

In addition, the server 200 also includes a report/dispatch service unit 250 that notifies the manager or worker of the contents according to the level of danger and automatically reports it to a related organization (eg, fire department, etc.), and requests dispatch service. .

In addition, the server 200 is interconnected via a wired/wireless network with a computer 10 and various mobile devices 20 that can be checked by an administrator/worker, and a fire station and a company providing dispatch services. Accordingly, the customer may use various services provided by the server 100 through the computer 100 or the mobile device 20.

Meanwhile, the above-described collection unit 210, prediction data generation unit 220, monitoring unit 230, notification/blocking unit 240, report/dispatch service unit 240, etc. are included in the server 200. Although it is described that it is configured, it is natural that only some or all of the configurations can be designed and configured separately from the server.

Next, we will look at the fire detection and monitoring method.

In order to detect fire in advance of the present invention, the arc detector 100 must be installed at a predetermined location such as a distribution box. The installed arc detector 100 may be a detector having a different capacity depending on the installed location, for example, a home or a factory, a multi-use facility, and the like.

In addition, a circuit breaker for blocking the current flow is also installed together with the arc detector 100. In other words, the heat generated by the arc (spark or electric flame) is expanded to the surrounding combustible material to cause a fire. This is because the occurrence of fire can be prevented by blocking the arc generation itself.

From such a configuration, the arc detectors 100 installed at predetermined locations detect arc occurrence states such as flames or sparks (s100). In addition, the detected arc data is transmitted in real time to the server 200 in an Internet-based HTTP method, and the collection unit 210 collects and stores arc data transmitted by the plurality of arc detectors 100 (s102). The stored arc data are classified for each arc detector and processed into a db format, and will be managed together with the predicted data described below.

At this time, the collection unit 210 collects all of a real arc signal, a signal irrelevant to an electric fire (unnecessary signal), and a noise component. That is, as shown in (a), which shows the change in the normal (normal state) spectrum, as shown in Fig. 2, there are few unnecessary signals and noise components, but when an arc occurs, unnecessary signals and noise components such as'A' in (b). The back is amplified and appears. Collecting all of these signals.

In addition, the prediction data generation unit 220 generates prediction data for the collected arc data (s104). At this time, since the arc data contains unnecessary signals and noise components, it is necessary to remove it and extract only the real arc signal, and predict the amount of change in the extracted real arc signal to generate predicted data.

The predicted data is predicted using a Kalman filter, and the predicted data may be extracted through the above-described [Equation 1].

An exemplary graph showing simulation results for the measured arc data and predicted data is shown in FIG. 3. Measurement data (actual value) and predicted data can be checked with respect to the graph of FIG. 3, and such graphs may be numericalized by time and provided as shown in [Table 2] below.

In this way, a predicted value can be generated by applying a Kalman filter to the actual measured value of the real arc signal. For reference, the relative error (%) is used to check the accuracy of the prediction.

Then, as shown in FIG. 4, the amount of change in the real arc signal can be predicted. For example, it can be predicted that the slope increases at a level (normal state) where the slope is almost '0', and then the slope increases rapidly, such as'B'.

The monitoring unit 230 analyzes the pattern of the predicted data and compares it with the reference range described in [Table 1]. That is, the pattern of the predicted data is to determine which step is the'safety step', the'warning step', the'dangerous step', and the fire generating step' (s106). In other words, the risk level level can be determined according to the analyzed pattern of the generated prediction data, where the prediction similarity pattern analysis analyzes the pattern at the time of the previous arc alarm or the fire occurrence pattern through big data analysis, and predicts It is to check whether there are cases similar to the pattern of the data.

Thereafter, when the risk level level is confirmed by the pattern analysis result of the predicted data, the server 200 performs a notification service and a response action corresponding to the identified risk level level (s108). Specifically, the notification/blocking unit 240 and the report/dispatch service unit 250 are in charge of these notification services and countermeasures, and for example, the predicted data is in the warning stage "quality analysis and more than 80% of the capacity of the monitoring device. In the case of continuing for 1 hour or 1 minute at 10 mA or more, the notification/blocking unit 240 transmits an alarm notification to the manager (worker) according to the monitoring result, and generates an alarm sound.

If a fire occurs, the notification/blocking unit 240 generates an alarm sound, and the report/dispatch service unit 250 transmits a fire occurrence notification and reports it to the fire department. In this case, the actual fire may not have occurred, but there may be signs of a fire, or it may have been before people directly visually check the fire condition and report it.

As another example of the countermeasure, it is possible for the server 200 to remotely control a circuit breaker installed together with the arc detection paper 100 to block the current flow. In this way, it is possible to prevent the occurrence of an arc due to an abnormal current or an overcurrent, and an advantage of initially blocking the occurrence of fire is also expected.

In addition, since the target of the notification service and the countermeasure has already been identified through the unique information included in the arc data, the most appropriate countermeasures can be selected based on this. That is, the fire can be notified directly to the nearest fire station, or the facility manager can be notified of a fire alarm or an abnormality.

As described above, the present invention can shorten the response time in that it is possible to detect a fire occurrence sign in advance according to the dangerous water level level identified through the pattern analysis of the predicted data. In fact, if the fire is dispatched more quickly due to the pre-detection of the signs of the fire, and the initial extinguishing is possible, the fire can be prevented in advance, and even if a fire has occurred, it will be possible to block the spread of a larger fire.

Meanwhile, according to the present invention, it is possible to generate revenue by managing arc data that detects an arc phenomenon and predicted data thereof.

Alerts or alarm sounds are delivered through real-time monitored arc data and predicted data for specific customers, or if necessary, when a fire sign is detected, a manager can be dispatched directly, or a dispatch service can be performed in connection with the fire department. can do.

In addition, it allows the customer to check the arc data using the computer 10 through the web or the mobile device 20 through the wireless network, and the customer may directly request the server 200 manager for dispatch. Moreover, the server 200 visualizes the arc data and the dangerous water level in real time so that the customer can check it, so that the current state can be easily checked.

In addition, when real-time monitoring of arc data is not normally performed due to damage or damage to the arc detector 100, a notification for replacement of the arc detector 100 may be transmitted to the customer, and replacement service of the arc detector 100 Notification information can be provided.

As described above, the present invention can cope with the risk of fire occurrence step by step according to the danger level level according to the predicted data using the Kalman filter for the arc data measured in real time. It can be seen that it can provide reporting and dispatch services.

Although it has been described with reference to the illustrated embodiments of the present invention as described above, these are only exemplary, and those of ordinary skill in the art to which the present invention pertains, without departing from the gist and scope of the present invention, various It will be apparent that variations, modifications and other equivalent embodiments are possible. Therefore, the true technical protection scope of the present invention should be determined by the technical spirit of the appended claims.

That is, the present embodiment describes a configuration example in which a fire occurrence symptom is detected in advance and an appropriate countermeasure is provided according to real-time arc data monitoring, but the present invention uses predictive data that can know the fire occurrence symptom in a multi-use facility or factory. It can be sufficiently applied to a management system that provides a variety of services for each situation, while providing them to specific customers such as homes, households, etc., and generating profits from service use.

100: arc detector
200: server
210: collection unit
220: prediction data generation unit
230: monitoring unit
240: notification/blocking unit
250: Report/Response Service Unit

Claims (8)

An arc detector that detects an arc phenomenon; And
And a server that receives the arc data detected by the arc detector in real time, predicts the arc data, and provides a step-by-step response plan according to a predefined danger level level. The method of claim 1,
The arc detector,
Fire pre-detection monitoring system for transmitting the arc data including unique information for identifying an arc detector to the server in a hypertext transfer protocol (HTTP) method. The method of claim 1,
The server,
A collection unit for collecting arc data including a real arc signal, a signal irrelevant to the electric fire, and a noise component from the arc detector;
A prediction data generation unit that extracts only a real arc signal from the arc data and generates prediction data by predicting a change amount of the extracted real arc signal;
A monitoring unit monitoring the dangerous water level through analysis of the predicted data; And
A fire pre-detection monitoring system comprising a notification/blocking unit and a report/dispatch service unit providing step-by-step response measures according to the dangerous water level. The method of claim 3,
The predictive data generation unit generates predictive data for the real arc signal using a Kalman filter, and calculates the preliminary fire detection monitoring system by the following equation.

.
here,

Is the state at point k based on the measured value at point k-1,

Is the state transition matrix based on the previous state at that time,

Is the state transition matrix by user input,

Refers to user input The method of claim 1,
The server,
A fire pre-detection monitoring system that visualizes and provides the arc data. Each detecting an arc phenomenon by a plurality of arc detectors;
Receiving, by a server, arc data according to the detection result through a communication method;
Extracting, by the server, a real arc signal using a Kalman filter;
Generating predicted data that predicts the amount of change of the real arc signal; And
And providing, by the server, a risk level level through analysis of the predicted data, and providing a step-by-step response plan according to the risk level level. The method of claim 6,
The communication method is HTTP (hypertext transfer protocol),
A fire pre-detection monitoring method in which the arc data includes detected arc information and unique information for identifying an arc detector. The method of claim 6,
The predicted data is calculated by the following equation,
A fire pre-detection monitoring method in which prediction data is generated by learning the state transition matrices F and B by performing prediction using the current value and the amount of change in the current value as the state vector x.

. here,

Is the state at point k based on the measured value at point k-1,

Is the state transition matrix based on the previous state at that time,

Is the state transition matrix by user input,

Refers to user input

Images