KR20230071565A - System for disaster observation monitoring and method thereof - Google Patents

Abstract

The present invention relates to a disaster observation and monitoring system and a method thereof, and more specifically to the disaster observation and monitoring system and the method thereof in which a computer can automatically check for abnormalities in an operation of sensors that collect weather data to predict disasters through analysis of weather data collected on a regular basis, not when a disaster occurs. The disaster observation and monitoring system includes a sensor unit; a data processing unit; and a central processing unit.

Description

Disaster observation monitoring system and method {System for disaster observation monitoring and method thereof}

The present invention relates to a disaster observation and monitoring system and method thereof, and more particularly, through analysis of meteorological data that is normally collected rather than at the time of a disaster to predict a disaster, a computer detects an abnormality in the operation of a sensor that collects meteorological data. It relates to a disaster observation monitoring system that can be automatically checked and a method therefor.

In urgent disasters such as flooding, heavy rain, gusts, landslides, etc., it is difficult to respond in real time to local disaster situations with conventional weather forecasting methods.

In recent years, in order to ensure the safety of valuable property and life, when a disaster situation occurs, technologies for promptly responding to it are being researched.

Prior literature related to this is Republic of Korea Patent Registration No. 10-2025588 'Internet of Things-Based Intelligent Disaster Response Access Control Method' and Republic of Korea Patent Publication No. 10-2021-0075533 'Video-based Precipitation Information Provision System and Method Using Deep Learning' Research is being conducted on technologies for quickly responding to natural disasters and disasters through artificial intelligence technologies such as deep learning.

However, technologies for promptly responding to such disasters do not have a high frequency of natural disasters and disasters occurring annually, whereas the system for detecting them operates 24 hours a day, 365 days a year, and devices or systems are operated 24 hours a day. It is practically impossible to inspect 365 days a year.

If the reliability of the collected meteorological data, including precipitation, water level, and wind speed, is poor, response to a disaster situation may be delayed, resulting in enormous damage to life and property. For example, if the actual water level in the underground roadway is 2m, which is the dangerous level, but it is measured as 1m due to an abnormality of the sensor, it is displayed as a safe state on the monitor of the central control center, and the timing of prompt disaster response can be practiced.

In particular, physical failure of the sensor or failure of the communication device can be found due to data not being received from the central server, communication failure, etc., but errors in inaccurate numbers due to sensor instability are difficult to find unless a person checks them individually. .

Republic of Korea Patent No. 10-2025588 (2019.09.20.) Republic of Korea Patent Publication No. 10-2021-0075533 (2021.06.23.)

An object of the present invention is to solve the above-described problems, by comparing meteorological data collected from sensors installed in areas where disasters are likely to occur with a risk threshold that is a criterion for disasters, such as flooding, heavy rain, gusts, etc. Not only does it serve as a disaster observation function that enables rapid response by alerting urgent disaster situations, but also automatically checks sensor operation abnormalities such as inaccurate numerical errors due to sensor instability through analysis of weather data collected at all times. It is to provide a disaster observation monitoring system and method equipped with a sensor inspection function.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned will be clearly understood by those skilled in the art from the following description.

In order to achieve the above object, a disaster observation and monitoring system according to the present invention is installed in an area where a disaster may occur and measures meteorological data including precipitation, wind speed, water level, etc., and meteorological data measured from the sensor unit. The data processing unit that collects and transmits the collected weather data to the central processing unit through the communication unit, and the disaster prediction data predicted through the disaster prediction algorithm of the collected weather data received from the data processing unit is the risk that becomes the standard for risk disaster situations It includes a central processing unit that transmits a danger signal to a warning unit in an area exceeding a threshold.

In addition, the central processing unit analyzes the collected meteorological data through the learned artificial intelligence model, classifies the sensor unit in the area where the difference between the inferred value and a certain level or more occurs as a priority inspection target, and provides inspection target information. Includes a monitoring unit characterized by

In addition, the artificial intelligence model stores meteorological data collected from a sensor unit installed in a specific area and meteorological data collected in an area adjacent to the specific area as big data, learns it, infers mutual correlation, and determines the legitimacy of the sensor unit. It is characterized by an artificial intelligence model that judges.

On the other hand, in order to achieve the above object, a disaster observation and monitoring method according to another aspect of the present invention includes a meteorological data measuring step of measuring meteorological data including precipitation, water level, wind speed, etc. in a sensor unit installed in each region, A weather data collection step in which the data processing unit collects the meteorological data measured by the sensor unit and transmits it to the central processing unit through the communication unit, and the central processing unit predicts the occurrence of a disaster through a disaster prediction algorithm using the collected weather data. Occurrence prediction step, when the predicted disaster prediction data exceeds the risk threshold, which is the criterion of a dangerous disaster situation, a disaster risk signal transmission step of transmitting a danger signal to a warning unit in the corresponding area, and a warning unit transmitting a danger signal from the central processing unit. When received, it includes a disaster risk warning step in which an alarm device is activated.

In addition, based on the meteorological data collected by the central processing unit after the meteorological data collection step, the meteorological data of a specific area and the meteorological data of areas adjacent to the particular area are automatically analyzed by the learned artificial intelligence model, and the inferred numerical value and schedule It is characterized in that it further comprises a sensor unit inspection step of providing inspection target information by classifying the sensor unit of the region where the above difference occurs as a priority inspection target.

In addition, the learned artificial intelligence model stores meteorological data collected from a sensor unit installed in a specific area and meteorological data collected in an area adjacent to the specific area as big data, learns it, infers mutual correlation, and It is characterized by an artificial intelligence model that determines legitimacy.

In addition, in the sensor unit inspection step, prior to inspecting the sensor unit through an artificial intelligence model, weather data collected from a sensor unit installed in a specific area and weather data collected in an area adjacent to the specific area are compared and analyzed to determine the legitimacy of the sensor unit. It is characterized by performing the step of determining.

The disaster observation monitoring system and method according to the present invention enable a computer to automatically check abnormalities in the operation of sensors through analysis of weather data collected on a daily basis, so that a person can manually check errors in inaccurate values due to sensor instability. It has the effect of solving the conventional problems that need to be discovered while doing so.

After training the weather data stored as big data for a certain period of time on the artificial intelligence model, the weather data collected in a specific locality and nearby localities is automatically analyzed by the artificial intelligence model, and the sensor in the area where the inferred value and a certain difference occurs is given priority. It is classified as an inspection target to facilitate maintenance such as repair and replacement of sensors, and accordingly, in the event of an actual emergency disaster situation, there is an effect of preventing prompt disaster response timing from being missed.

1 is a configuration diagram showing the configuration of a disaster observation and monitoring system according to the present invention.
Figure 2 is a flow chart showing the overall flow of the disaster observation monitoring method according to the present invention.
3 is a flow chart for explaining the sensor unit inspection step (S250) of FIG.

Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the detailed description of the following embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, only these embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs. It is provided to fully inform the holder of the scope of the invention, and the present invention is only defined by the scope of the claims.

With reference to the accompanying drawings below, specific details for the practice of the present invention will be described in detail. Like reference numbers refer to like elements, regardless of drawing, and "and/or" includes each and every combination of one or more of the recited items.

Terms used in this specification are for describing the embodiments and are not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase. As used herein, "comprises" and/or "comprising" does not exclude the presence or addition of one or more other elements other than the recited elements.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification may be used in a meaning commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless explicitly specifically defined.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a configuration diagram showing the configuration of a disaster observation monitoring system according to the present invention.

Referring to FIG. 1 , the disaster observation and monitoring system according to the present invention largely includes a sensor unit 100, a data processing unit 200, a communication unit 300, a central processing unit 400 and an alarm unit 500.

First, the sensor unit 100 is installed in an area where a disaster may occur and measures meteorological data including precipitation, wind speed, water level, and the like.

The sensor unit 100 may be composed of sensors including a precipitation sensor for measuring precipitation, a water level sensor for measuring water level, a GPS location sensor for acquiring GPS location information, and a wind speed sensor for measuring wind speed. can

The type of the sensor is not limited thereto, and various changes such as an earthquake detection sensor, a displacement sensor, and a fine dust sensor that measure meteorological data in order to predict a disaster in a corresponding region can be applied.

In addition, it goes without saying that the sensor unit 100 may include two or more sensors.

The sensor unit 100 is configured to transmit the measured meteorological data to the location processing unit 200 .

Next, the data processing unit 200 collects weather data measured by the sensor unit 100 and transmits the collected weather data to the central processing unit 400 through the communication unit 300 .

At this time, the data processing unit 200 converts the weather data in analog form collected from the sensor unit 100 into digital weather data and transmits it to the central processing unit 400 .

The communication unit 300 is configured to transmit and receive data with the central processing unit 400 using mobile communication technologies including 5G and LTE, but is not limited thereto and transmits and receives data with the central processing unit 400 through satellite communication. Of course, it can be configured to do so.

Next, the central processing unit 400 predicts a disaster in the corresponding region through a disaster prediction algorithm using the collected meteorological data transmitted from the data processing unit 300 through the communication unit 400 .

The disaster prediction algorithm may refer to a disaster prediction algorithm using big data or artificial intelligence technology, and preferably, based on the collected weather data, usually collected weather data is stored as big data to prevent disasters through this. Artificial intelligence capable of predicting disasters by learning a disaster prediction artificial intelligence model to predict, calculating disaster prediction data through the learned artificial intelligence model, and comparing the calculated disaster prediction data with a risk threshold that is the standard for a disaster situation. I mean the intelligence model. Here, the disaster prediction data means future meteorological data.

In addition, the artificial intelligence technology may refer to deep learning, and the deep learning technology may also refer to an artificial neural network, a deep neural network (DNN), and a convolutional neural network (CNN). , Convolution Neural Network), and at least one artificial neural network technique including RNN (Recurrent Neural Network) may be selected and used, but is not limited thereto.

In addition, the central processing unit 400, if the disaster prediction data predicted through the disaster prediction algorithm of the collected meteorological data transmitted from the data processing unit 200 exceeds the risk threshold that is a criterion for a dangerous disaster situation, exceeds A danger signal is transmitted to the local warning unit 500 through the communication unit 300.

On the other hand, the central processing unit 400 analyzes the collected meteorological data through the learned artificial intelligence model, and classifies the sensor unit in the region where a certain difference or more from the inferred value occurs as a priority inspection target, monitoring to provide inspection target information. includes wealth

The monitoring unit automatically checks the operation abnormality of the sensor unit 100 installed in each region in real time in order to prevent a decrease in reliability in disaster prediction by measuring inaccurate weather data due to a malfunction of the sensor unit 100 perform a function

More specifically, the learned artificial intelligence model stores meteorological data collected from the sensor unit 100 installed in a specific area and meteorological data collected in an area adjacent to the specific area as big data, learns them, and mutually It refers to an artificial intelligence model that determines the legitimacy of the sensor unit by inferring correlation.

In addition, the artificial intelligence model may mean an artificial intelligence model using a linear regression model modeling a relationship between one dependent variable and one or more independent variables, but is not limited thereto, and a logistic regression model using It can mean an artificial intelligence model.

To summarize the monitoring unit, after comparing weather data built with big data collected on a daily basis through artificial intelligence machine learning and building an inference model, weather data from local areas in the field are verified through accuracy verification through artificial intelligence models. It refers to automatically detecting abnormal signs of sensors, automatically classifying them as inspection targets, and displaying them so that administrators can check them.

The central processing unit 400 may be configured to visually display the sensor unit 100 classified as an inspection target or a monitor of the central control center so as to check the area, but is not limited thereto, and the manager terminal It can be configured in any way that can be checked by a person in real time, such as a notification or a separate dedicated application or website.

Next, when the alarm unit 500 receives the danger signal from the central processing unit 400, the alarm unit 500 operates an alarm device.

The warning device may refer to an alarm device that sounds an alarm siren, but is not limited thereto, and may refer to a device for transmitting danger information to an electronic display board installed in a field.

Of course, the alarm unit 300 may consist of two or more alarm devices.

In addition, the alarm device may include an emergency communication device that is operable even in a communication interruption situation when communication is interrupted due to a disaster.

The emergency communication device is configured to transmit a rescue message to government offices including fire departments, police stations, and the like in a communication failure situation.

Hereinafter, a disaster observation and monitoring method according to another aspect of the present invention will be described.

Figure 2 is a flow chart showing the overall flow of the disaster observation monitoring method according to the present invention.

Referring to FIG. 2 , in the disaster observation and monitoring method according to the present invention, first, a meteorological data measuring step (S100) of measuring meteorological data including precipitation, water level, wind speed, etc. in the sensor unit 100 installed in each region is performed. carry out

Next, a weather data collection step (S200) of collecting weather data measured by the sensor unit 100 in the data processing unit 200 and transferring the weather data to the central processing unit 400 through the communication unit 300 is performed.

Next, the central processing unit 400 performs a disaster occurrence prediction step (S300) of predicting the occurrence of a disaster through a disaster prediction algorithm based on the received meteorological data.

The disaster prediction algorithm refers to the above-described disaster prediction algorithm through the disaster prediction monitoring system according to the present invention, and a detailed description thereof will be omitted.

Next, when the predicted disaster prediction data exceeds the risk threshold, which is a criterion for a dangerous disaster situation, a disaster danger signal transmission step (S400) of transmitting a danger signal to the warning unit 500 in the corresponding area is performed.

Next, when the warning unit 500 receives a danger signal from the central processing unit 400, it performs a disaster risk warning step (S500) of operating an alarm device.

On the other hand, based on the meteorological data collected by the central processing unit 400 after the meteorological data collection step (S200), the meteorological data of a specific area and the meteorological data of areas adjacent to the particular area are automatically analyzed in the learned artificial intelligence model, , The sensor unit inspection step (S250) of providing inspection target information by classifying the sensor unit 100 in the area where the inferred numerical value and a predetermined difference or more is first classified as an inspection target is performed.

The sensor unit inspection step (S250) refers to a step continuously and repeatedly performed after the meteorological data collection step (S200).

3 is a flow chart for explaining the sensor unit inspection step (S250) of FIG.

With further reference to FIG. 3 , the sensor unit inspection step (S250) will be described in more detail by taking an example.

Referring to FIG. 3, weather data collected from the sensor unit 100 installed in a specific area before the sensor unit inspection step (S250), that is, before the sensor unit 100 is inspected through the learned artificial intelligence model. A step of determining the legitimacy of the sensor unit 100 through comparative analysis of weather data collected in an area adjacent to the specific area and the specific area (S240) may be performed.

This is to determine whether the sensor unit 100 is operating normally through simple data comparison rather than using artificial intelligence. Since artificial intelligence technology requires a lot of computational processing with high performance, the sensor unit check step (S250) It is inevitable that many calculations will burden the hardware.

Therefore, by performing the first inspection through the step S240 and performing the second inspection on the artificial intelligence model learned after the first inspection, it is possible to have an effect of reducing the burden on this.

In addition, the sensor unit inspection request step of selectively performing the sensor unit inspection step (S250) according to the sensor unit accuracy verification request signal so that the sensor unit inspection step (S250) can be performed according to the sensor unit inspection request. (S230) may be performed. Accordingly, there is an effect of reducing the hardware load for the sensor unit inspection step (S250).

In addition, by rechecking whether or not the sensor unit 100 is normally operating through the learned artificial intelligence model in the second inspection (S250) for the part where the abnormality of the sensor unit 100 was not detected in the first inspection (S240), , it has the effect of increasing reliability.

On the other hand, in another embodiment, for example, the sensor unit 100 in an area with a possibility of inundation measures the amount of precipitation and the water level. First, the amount of precipitation in a nearby area related to the amount of precipitation in the area collected in step S240 is first determined. It is determined whether the amount of precipitation in a local area is appropriate by comparison, and if a large difference between the amount of precipitation in a nearby area and the corresponding area occurs, it may be determined that the amount of precipitation in the area is defective.

Thereafter, in step S250, two or more meteorological data including the amount of precipitation and water level in the local area are input to the learned artificial intelligence model to determine the legitimacy of the water level sensor, and to determine whether the water level sensor is defective, in steps S240 and S250. Of course, some changes may be applied.

In addition, as another example, when the first weather data and the second weather data predict the occurrence of a landslide disaster, the first weather data refers to the amount of precipitation measured through a precipitation measurement sensor, and the second weather data refers to a safe area. It may refer to a position deviation gap between a GPS position sensor installed in the field and a GPS position sensor installed in a landslide risk area.

In addition, when inspecting the sensor unit 100 equipped with a precipitation sensor and a GPS location sensor installed in an area where a landslide may occur, the legitimacy of the precipitation value is verified in step S240, and the precipitation value and GPS location information are used in step S250. It can be changed and performed according to various disaster situations, such as being able to determine whether the sensor unit 100 is normally operating.

In addition, even in the case of the sensor unit 100 including the wind speed sensor, it is possible to check the appropriateness by comparing the related wind speed values of neighboring localities through the learned artificial intelligence model.

In addition, it is not excluded that steps S240 and S250 may be additionally performed according to the case where the sensor unit 100 includes two or more sensors.

As described above, it is possible to check whether or not the normal operation of the sensor unit is abnormal in the same way with respect to various sensors according to various disaster situations.

Although the embodiments of the present invention have been described with reference to the above and accompanying drawings, those skilled in the art to which the present invention pertains can implement the present invention in other specific forms without changing the technical spirit or essential features. You will understand that there is Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

100: sensor unit
200: data processing unit
300: communication department
400: central processing unit
500: alarm unit

Claims (7)

A sensor unit installed in a disaster-probable area to measure meteorological data including precipitation, wind speed, and water level;
a data processing unit that collects meteorological data measured by the sensor unit and transmits the collected meteorological data to a central processing unit through a communication unit; and
A central processing unit that transmits a danger signal to a warning unit in an area where the disaster prediction data predicted through a disaster prediction algorithm from the collected meteorological data transmitted from the data processing unit exceeds a risk threshold that is a criterion for a dangerous disaster situation; observational monitoring system. According to claim 1,
The central processing unit,
Disaster observation characterized in that it includes a monitoring unit that analyzes the collected meteorological data through a learned artificial intelligence model, classifies the sensor unit in the area where the inferred value and a certain difference or more occurs as a priority inspection target, and provides inspection target information monitoring system. According to claim 2,
The artificial intelligence model,
It is an artificial intelligence model that stores the meteorological data collected from the sensor unit installed in a specific area and the meteorological data collected from areas adjacent to the specific area as big data, learns it, and infers the correlation to determine the legitimacy of the sensor unit. Characterized by disaster observation monitoring system. a meteorological data measuring step of measuring meteorological data including precipitation, water level, wind speed, etc., by a sensor unit installed in each region;
a weather data collection step in which a data processing unit collects weather data measured by the sensor unit and transmits the weather data to a central processing unit through a communication unit;
A disaster occurrence prediction step of predicting the occurrence of a disaster through a disaster prediction algorithm based on the collected meteorological data transmitted from the central processing unit;
Disaster danger signal transmission step of transmitting a danger signal to a warning unit in the corresponding region when the predicted disaster prediction data exceeds a risk threshold that is a criterion for a dangerous disaster situation; and
Disaster observation and monitoring method comprising: a disaster risk warning step of operating an alarm device when the alarm unit receives a danger signal from the central processing unit. According to claim 4,
After the meteorological data collection step,
Based on the meteorological data collected by the central processing unit, the meteorological data of a specific area and the meteorological data of areas adjacent to the particular area are automatically analyzed in the learned artificial intelligence model, and the sensor unit in the area where the inferred value and a certain difference occurs is given priority. Disaster observation and monitoring method further comprising a sensor unit inspection step of classifying the inspection target and providing inspection target information. According to claim 5,
The learned artificial intelligence model,
It is an artificial intelligence model that stores the meteorological data collected from the sensor unit installed in a specific area and the meteorological data collected from areas adjacent to the specific area as big data, learns it, and infers the correlation to determine the legitimacy of the sensor unit. Disaster observation monitoring method characterized. According to claim 5,
In the sensor part inspection step,
Prior to checking the sensor unit through the artificial intelligence model, the weather data collected from the sensor unit installed in a specific area and the weather data collected in the area adjacent to the specific area are compared and analyzed to determine the legitimacy of the sensor unit. Disaster observation and monitoring method, characterized in that.

Images