KR20230110856A - HIGHLY RELIABLE IoT SECURITY SENSOR USING BIGDATA - Google Patents

Abstract

The present invention relates to a highly reliable IoT security sensor using big data, by emitting a predetermined sound signal to a sensing area and then analyzing the received sound signal to detect a sound field change pattern, based on a change in the sensing environment of the sensing area, detecting information is obtained from at least one individual sensor, and when there is no change in the status, the value sent from the unit IoT sensor is recognized as a value of the normal state, and when the change in status exists, the value sent from the unit IoT sensor is recognized as the value of a specific sensing environment, thereby recognizing the value sent from the unit IoT sensor as the value of a specific sensing environment. It is about high IoT security sensors. The sound field change pattern is prepared through big data collected from numerous sensing areas and updated in real time.

Description

Highly reliable IoT security sensor using big data {HIGHLY RELIABLE IoT SECURITY SENSOR USING BIGDATA}

The present invention is the result of a low-power and highly reliable security sensor (single integrated product) research project using IOT and big data applied to smart cities in the 2021 non-face-to-face start-up fostering project supported by the Small and Medium Venture Business Agency.

The present invention relates to a highly reliable IoT security sensor using big data, and more particularly, by emitting a predetermined sound signal to a sensing area and then analyzing the received sound signal to detect a sound field change pattern, based on a change in the sensing environment of the sensing area, sensing information is obtained from at least one individual sensor, and when there is no change in the status, the value sent from the unit IoT sensor is recognized as a value in the normal state, and when the change in status exists, the value sent from the unit IoT sensor is recognized as the value of a specific sensing environment. It relates to highly reliable IoT security sensors using data. The sound field change pattern is prepared through big data collected from numerous sensing areas and updated in real time.

Conventional IoT security sensors generally do not have separate countermeasures against malfunctions due to product characteristics or environmental factors. Looking more closely at the cause of the malfunction, it is safe to say that the malfunction is not caused by a defect in the sensor itself, but in most cases, the problem of malfunction is exposed due to the operation standards of each sensor. That is, the actual operation of the sensor is normal, but a problem of malfunction occurs during operation.

Therefore, breaking away from the classical function of existing sensors, it is necessary to measure the change in the sound field state in space in real time to reliably check the pattern of change analyzed more than 10 times for the initial value of the standing wave signal and the change value when an alarm event such as an intruder or fire in the space occurs, or check whether the event has occurred through an artificial intelligence learning model to operate the alarm system alone or to operate in conjunction with a security system.

The highly reliable IoT security sensor using big data according to the present invention is a device that generates sound field signals individually or in multiple ways according to the environment of the site, detects the changing state of the visitor and the site, and based on this, detects the situation of the site through the sensor. In addition, it can be configured to connect devices in parallel mode through multiple ports so that the manager can check the alarm status that operates when multiple devices are connected on the network by connecting one or more devices of single type according to the site conditions.

More specifically, a sound generating device that outputs a sound wave of a specific frequency within a space to be sensed (office, warehouse, etc.), a sound receiving device that receives a sound wave within the space and obtains a sound pressure from the received sound wave, and a frequency analyzer that analyzes the frequency level of the changed sound wave by measuring a change value representing the ratio of the sound pressure obtained by the sound receiving device to the input voltage of the sound generating device in a preparation mode, NC (Normally Close) / NO (No rmally Open) to provide a device that outputs according to the connection signal.

Next, the prior inventions existing in the technical field of the present invention will be briefly described, and then the technical details to be achieved by the present invention to be differentiated from the prior inventions will be described.

First, Korean Registered Patent No. 2312997 B1 (October 7, 2021) relates to a security monitoring method using a multi-tone sound source having different frequencies. A cocktail mode multi-tone sound source output into a security monitoring space is received to measure reference sound field information for each frequency, the current sound field information for each frequency of the received multi-tone sound source is measured, the reference sound field information for each frequency and the current sound field information for each frequency are compared to determine whether a sound field change has occurred, and the reference sound field for each frequency An intrusion or fire situation is determined based on a change in the correlation coefficient between the spectrum and the current sound field spectrum for each frequency, and a security monitoring method using a multi-tone sound source overcomes problems caused by external noise and malfunctions caused by heating and cooling. In addition, a fire and intrusion situation can be accurately distinguished by using a security sensor based on sound field change detection.

Korean Patent Registration No. 2312997 B1 uses a multi-tone sound source to determine an intrusion or fire situation, and uses the multi-tone sound source to overcome problems caused by external noise and malfunction due to heating and cooling. That is, the sound field change detection-based security sensor may malfunction due to external noise or rapid sound field changes caused by heating and cooling.

However, the present invention detects a sound field change pattern to recognize a sudden environmental change, in this case, it is determined that a sensor value (sensing is not a sound field change) is likely to be an error, and an alarm signal is not immediately issued. If the sensor value exceeds a specific range even after the above sound field change is stabilized, a corresponding alarm signal is generated to determine whether the sensor to be sensed is affected by the surrounding environment, and as a result, it relates to an IoT security sensor for preventing the sensor from malfunctioning.

In other words, Korean Patent Registration No. 2312997 B1 seeks to overcome the malfunction problem in a sensor using sound field change, whereas the present invention analyzes the sound field change pattern to determine whether there is a change in the surrounding environment, and determines whether the sound field change has affected the detection signal of a specific sensor (temperature, humidity, motion, etc.), and from this determines whether the corresponding unit IoT sensor does not malfunction.

In addition, Korean Patent Publication No. 2016-0129443 A (2016.11.09) relates to an intelligent sensor device capable of identifying humans and animals and a security system using the same, which is configured to identify the approach of humans and animals, thereby preventing unnecessary malfunction due to the approach of animals such as dogs and cats.

Korean Patent Publication No. 2016-0129443 A is intended to prevent the ultrasonic sensor from malfunctioning due to the approach of animals such as dogs and cats in an ultrasonic sensor. In the present invention, a specific sensor located in a specific place is malfunctioned by the surrounding environment by using a sound field change. The purpose is to finally reduce false alarms caused by malfunctions. Therefore, the sensor device for IoT security with high reliability robust against malfunction of the present invention is clearly different from the prior art in its technical characteristics.

The present invention has been created to solve the above problems, and in providing a highly reliable IoT security sensor that is robust against malfunction, an object of the present invention is to provide a low-power and highly reliable sound field signal in a single (integrated) or multi-type, detect an environmental change in a sensing area according to a change pattern of the sound field signal, and determine whether to transmit an alarm signal to external equipment based thereon.

In addition, another object of the present invention is to provide a highly reliable IoT security sensor robust against malfunctions that can operate regardless of heat or temperature characteristics so that at least one sensor can cope with shielding and intruders based on sound field signals.

In addition, another object of the present invention is to provide a sound source generating device, a sound receiving device, an analysis processor for a sound source and a wavelength, etc. in providing a reliable sensor device for IoT security that is robust against malfunction.

Another object of the present invention is to provide an active sound field change event detection function, to detect the front or back of an obstacle, and to provide a function that can operate regardless of the dark environment of a space.

A highly reliable IoT security sensor that is robust against malfunctions according to an embodiment of the present invention includes a sound source generator that generates a sound source having a specific amplitude and frequency pattern, a sound receiver that receives sound from a detection area, a sound field change analyzer that analyzes changes in the pattern of the generated sound source and the received sound pattern, and a sensor that detects the sensing area from at least one sensor; outputs a result detected by the sensor based on a sound field change, so that the result detected by the sensor is affected by the sound field change It is characterized in that by notifying that the result is the result, a highly reliable detection result robust against malfunction is output. The sound field change pattern is prepared through big data collected from numerous sensing areas and updated in real time.

In addition, the IoT security sensor, characterized in that it further comprises; an output unit for outputting the detection result or an alarm signal according to the detection result to the outside. The output unit judges that there may be a malfunction in the detection result environmentally as a result of analyzing the pattern of the sound field change in the sound field change analyzer for the detection result or the alarm signal, generates a connection signal whether or not to output the alarm signal to an external device according to the determination result, and outputs the alarm signal according to the connection signal.

In addition, the sound field change analysis unit may actively detect a sound field change event, detect a front or rear side of an obstacle, operate regardless of heat or temperature characteristics, or include a combination thereof.

In addition, the IoT security sensor includes a flame detection and smoke detection function, and is characterized in that it operates regardless of the environment of the sensing space.

On the other hand, a sensing method through an IoT security sensor according to another embodiment of the present invention includes a sound source generating step of generating a sound source having a specific amplitude and frequency pattern, a sound receiving step of receiving sound from a sensing area, a sound field change analysis step of analyzing a change in the pattern of the generated sound source and the received sound pattern, and a sensing step of sensing the sensing area from at least one sensor; outputting a result detected by the sensor based on a sound field change, so that the result detected by the sensor is affected by the sound field change By notifying that it is a detection result, it is characterized in that it outputs a detection result that is robust against malfunction and highly reliable.

In addition, the detection method may further include an output step of outputting the detection result or an alarm signal according to the detection result to the outside. In the outputting step, as a result of analyzing the sound field change pattern in the sound field change analysis step for the detection result or the alarm signal, it is determined that there may be a malfunction in the detection result environmentally, and a connection signal is generated whether or not the alarm signal is to be output to an external device according to the determination result, and the alarm signal is output according to the connection signal.

In addition, the sound field change analysis step may actively detect a sound field change event, detect the front or back side of an obstacle, operate regardless of heat or temperature characteristics, or include a combination thereof.

In addition, the detection method includes a flame detection function and a smoke detection function, and is characterized in that it operates regardless of the environment of the detection space.

As described above, the error-resistant and highly reliable IoT security sensor of the present invention emits a predetermined sound signal to the sensing area and then analyzes the received sound signal to detect a sound field change pattern, based on a change in the state of the sensing environment of the sensing area. Obtains detection information from at least one individual sensor, recognizes the value sent from the unit IoT sensor as the value of the normal state, and recognizes the value sent from the unit IoT sensor as the value of the specific sensing environment when the state change exists, thereby preventing malfunction It has robust and reliable effects and advantages.

In addition, since the sound field change pattern is prepared through big data collected from numerous sensing areas and updated in real time, there is an effect of automatically adapting to the latest environment without any extra effort.

1 is a diagram illustrating the configuration of an IoT security sensor network including a highly reliable IoT security sensor using big data according to an embodiment of the present invention.
2 is a block diagram of a highly reliable IoT security sensor using big data according to an embodiment of the present invention.
3 is a diagram illustrating a concept of recognizing an obstacle according to a change in a sound field in a highly reliable IoT security sensor using big data according to an embodiment of the present invention and operating individual sensors accordingly.
4 is a diagram illustrating a concept of recognizing a signal generated according to a change in a sound field at a detection site where various obstacles or events occur in a highly reliable IoT security sensor using big data according to an embodiment of the present invention.
5 is a diagram illustrating a method of extracting an event and an occurrence location of an event by applying input data to an artificial intelligence learning model according to an embodiment of the present invention.
6 is a flowchart illustrating the operation of a highly reliable IoT security sensor using big data according to an embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, a preferred embodiment of a highly reliable IoT security sensor device robust against malfunction of the present invention will be described in detail. Like reference numerals in each figure indicate like members. In addition, specific structural or functional descriptions of the embodiments of the present invention are merely exemplified for the purpose of explaining the embodiments according to the present invention, and unless otherwise defined, all terms including technical or scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and unless explicitly defined in the present specification, it is preferable not to be interpreted in an ideal or excessively formal meaning.

Sound can be defined as a pressure wave radiated from a sound source transmitted in a medium, and a wave caused by this sound is called a sound wave. Here, the pressure wave is created in the process of vibrating the particles of a medium such as air or water, and as the process of increasing and decreasing the density of the surroundings occurs, it creates a waveform. The speed at which sound travels depends on the type and properties of the medium, and also on the temperature, atmospheric pressure, and humidity of the atmosphere. Therefore, if the temperature of the air at the average atmospheric pressure is about 15 degrees Celsius, the speed of sound = 331/45 m/sec + 0.6 m/s ㅧ Celsius temperature is substituted into the formula to know that it is about 340.45 m/sec.

1 is a diagram illustrating the configuration of an IoT security sensor network including a highly reliable IoT security sensor using big data according to an embodiment of the present invention.

As shown in FIG. 1, in the IoT security sensor network, at least one IoT security sensor 100 is connected through an edge (computing) server 200, and each of the plurality of edge computing servers 200 is connected to the IoT cloud server 300 to manage sensing information.

The edge computing server 200 provides a sound field change pattern corresponding to each IoT security sensor or a learning model obtained by learning the sound field change pattern as an artificial intelligence learning model. On the other hand, each IoT security sensor transmits the sensing information detected in the corresponding detection area to the edge computing server or transmits the sound field change of each surveillance area to the edge computing server so that it can be used as an input for the learning model for the sound field change. Here, the sound field change pattern is prepared through big data collected from numerous sensing areas and updated in real time.

The learning model learns various sound field changes through machine learning, and transmits the learning model generated as a result of the learning to the IoT security sensor so that it can be used. That is, the learning model is based on a reinforcement learning model.

The generation of the learning model may be performed in the IoT cloud server 300, and the new learning model generated may be provided to the edge computing server 200 and transmitted to the IoT security sensor 100 as needed. That is, the generation of the learning model can be performed in the IoT cloud server 300 or the edge computing server 200, and the IoT security sensor 100 receives the corresponding learning model from the edge computing server 200 and detects what kind of change in the sound field has occurred in the sensing area. The learning model is created and updated in real time by learning big data related to sound field change patterns collected from numerous sensing areas.

2 is a block diagram of a highly reliable IoT security sensor using big data according to an embodiment of the present invention.

As shown in FIG. 2, the highly reliable IoT security sensor 100 using big data of the present invention includes a sound source generator 110, a sound receiver 120, a detector 130, a sound field change analyzer 140, an output unit 150, a controller 160, and a memory 170.

When outputting a sound source through a speaker, the sound source generator 110 performs a function of amplifying the amplitude of the sound wave or changing its frequency and outputting the sound wave in order to output a sound wave of a desired amplitude and frequency. That is, the sound source generator 110 performs a function of changing the amplitude and frequency of the sound source in order to output sound sources having amplitudes and frequencies of various patterns.

The sound receiver 120 performs a function of amplifying sound input from a microphone for environment analysis. The sound input from the microphone is amplified and converted into a digital signal, and then the sound source generated by the sound source generator 110 is compared with the sound wave converted into a specific pattern using the corresponding digital signal and amplified to a predetermined size to analyze how much it has changed.

The sensing unit 130 receives an analog signal detected by at least one sensor, converts it into a digital signal, and performs signal processing according to the corresponding sensor. If the digitally converted sensor information is determined to be a malfunction according to the analysis result of the sound field change analyzer 140, sensor information or an alarm signal is not output, and if it is determined not to be a malfunction, the corresponding sensor information is output. In addition, by outputting the detection result together with the division result, it is possible to recognize the detection result suitable for the analysis environment.

The sound field change analyzer 140 compares the result of the sound source generated by the sound source generator 110 and output through the speaker and received from the sound receiver 120 through the microphone again. Therefore, since the distorted sensor information is less reliable, it is controlled so that it is not used as an appropriate alarm signal.

In addition, the analysis result of the sound field change analyzer 140 is output together with the detection result detected by the sensor, so that it can be recognized that the detection result is a detection result based on the analysis result.

Here, the sound field change analyzer 140 stores in a memory a pattern in which sound field changes for each of a plurality of obstacles (intruders, etc.) or events (fire, smoke, noise, etc.) are recorded for a plurality of detection areas, and compares sound field changes input through a speaker and received through a sound receiver to detect the corresponding obstacle or event. The change in the sound field is collected from numerous sensing areas, generated in real time, reflected, and updated so that the change in the sound field is always adapted to the latest environment.

On the other hand, the sound field change analyzer 140 receives a learning model obtained by learning sound field changes for each of a plurality of obstacles (intruders, etc.) or events (fire, smoke, noise, etc.) with respect to a plurality of sensing areas, stores the learning model in a memory, and inputs the sound field change input through a speaker and received through the sound receiver to the learning model to detect the corresponding obstacle or event.

The output unit 150 outputs or blocks sensor information (sensing result) detected by a sensor and input through the sensor 130 according to the analysis result output from the sound field change analyzer 140 . The detection result may be output together with an alarm signal or an analysis result. In the case of being output together with an alarm signal, when it is recognized that an emergency situation has occurred in the sensing area, a detection result may be output together with an alarm signal for the corresponding emergency situation. That is, if the detection result is a CCTV image, the result and an alarm signal for an intruder may be simultaneously output. In addition, the detection result may be output together with the analysis result, and the analysis result may have the same or similar form as the alarm signal.

The control unit 160 stores or reads the information in the memory 170 that stores basic information for generating a sound source or programs and data necessary for analyzing sound field changes, and provides the information to the sound source generator 110 and the sound field change analyzer 140. In addition, when sensor information (sensing result) is output from the output unit 150, it is possible to set how sensitively the sound field change analysis result is to be applied. In addition, information or algorithms necessary for analyzing the environment according to the change of the sound field can be provided.

The present invention was created to serve as a sensor for measuring changes in the environment using a sound source. That is, whether a fire has occurred in an area to be detected through a sound source, a specific object has been detected, or no obstacle has appeared can be detected by analyzing the change in the sound field.

The sound field change can be sensed by storing the sound field change pattern and comparing it with the sound received through the microphone. However, in the present invention, a learning model that learns the change of the sound field for each of a plurality of obstacles or events for each sensing area is created and stored through artificial intelligence, and the corresponding obstacle or event can be detected probabilistically despite various nonlinear changes in the sensing area.

This learning model is performed in the edge computing server 200 or the IoT cloud server 300 rather than directly in the IoT sensor device of the present invention, receives only the parameters of the corresponding learning model, and stores them in the memory 170. Then, if necessary, accessed through the control unit 160 so that the sound field change analyzer 140 can detect them.

3 is a diagram illustrating a concept of recognizing an obstacle according to a change in a sound field in a highly reliable IoT security sensor using big data according to an embodiment of the present invention and operating individual sensors accordingly.

As shown in FIG. 3 , when a sound source is emitted from a speaker installed at a monitoring or sensing site, the sound wave received by the microphone will be similar to the change pattern of the sound field for the emitted sound source in an environment without obstacles. In this case, it is a normal environment and the detection result detected by the image sensor also corresponds to a normal result.

If there is a difference in the sound field change between the emitted sound source and the received sound wave, in which an intruder appears in the surveillance site, it can be known that an environmental change has occurred in the surveillance site.

In this case, since it is impossible to determine whether an intruder or a normal obstacle (such as a landlord) has appeared at the surveillance site, it is necessary to check with an image sensor.

In this case, the detection result of the image sensor is very important information together with the analysis result detected from the change in the sound field. In other words, it is impossible to know the exact cause of environmental change only by the analysis result of the environment, so the detection result must be checked at the same time to implement a highly reliable IoT security sensor using big data as a whole.

On the other hand, if a plurality of artificial intelligence learning models in which the type and location of obstacles are labeled in detail are prepared, and then a label with the highest probability is extracted using them, an analysis result that is very precisely adapted to changes in the environment can be obtained. In other words, by finding the label corresponding to the type and exact location of the obstacle or event, and operating the image sensor focused on the location accordingly, it is possible to implement a more precise and reliable IoT security sensor using big data.

4 is a diagram illustrating a concept of recognizing a signal generated according to a change in a sound field at a detection site where various obstacles or events occur in a highly reliable IoT security sensor using big data according to an embodiment of the present invention.

As shown in FIG. 4, the highly reliable IoT security sensor 100 using big data according to the present invention, in a normal environment in which no obstacles exist in the sensing area (FIG. 4 (a)), since it will be detected that the transmitted sound source is received without any change, it is determined that there is no obstacle in the sensing area. In addition, in this case, it is determined that the sensing result of temperature, humidity, image, etc. detected by a separately provided sensor is in a normal state. In this case, the detection results detected by each sensor are collected in a normal environment.

On the other hand, when a specific person exists in the sensing area (FIG. 4(b)), since the transmitted sound source is distorted and received, it is possible to recognize that an obstacle has appeared in the sensing area. However, when a sound source is transmitted and directed to a specific place, when an obstacle appears in the specific place, the modified sound wave is received, so that the position of the obstacle can also be detected. When only a person appears at a specific location in the sensing area, the ambient temperature may change minutely and the image sensor may not be able to detect it in a dark environment. Therefore, by detecting a change in the sound field, an alarm signal can be provided so that the corresponding detection result (temperature, image, etc.) can be further augmented and confirmed.

In this way, if the detection result of at least one sensor is combined with the analysis result of the sound field change, it is possible to implement a highly reliable IoT security sensor using big data for the detection area.

In addition, when a fire occurs in the detection area (FIG. 4(c)), if the sound field change shows a fire pattern with respect to the transmitted sound source, an event called fire can be detected by the sound field change. However, if the noise caused by a fire increases, sound waves with a different pattern from the transmitted sound waves will be received, so complex changes to the environment may be recognized and the environment may not be accurately defined. In this case, if the temperature sensor or image sensor is additionally checked, since the temperature will increase or a flame will be detected in the image, it can be seen that the increase in the corresponding temperature value is caused by a change in the environment, and the cause of the event can be confirmed with the image sensor. If the temperature sensor is used for measuring the temperature, it will be a malfunction if the temperature rise due to fire is recognized as a normal temperature rise. Therefore, in the case of a fire, it can be recognized as a temperature rise, so that a rapid rise in temperature is not recognized as a malfunction of the temperature sensor.

If each situation shown in FIG. 4 is learned with an artificial intelligence learning model, each corresponding learning model is labeled, and the sound input from the microphone is input to each learning model as data, a learning model with the highest probability can be extracted. When the label of the corresponding learning model is checked, the type, location, or combination of these events can be extracted and recognized from the corresponding label.

5 is a diagram illustrating a method of extracting an event and an occurrence location of an event by applying input data to an artificial intelligence learning model according to an embodiment of the present invention.

As shown in FIG. 5, in the highly reliable IoT security sensor using big data according to an embodiment of the present invention, the artificial intelligence learning model used by the sound field change analyzer 140 is generated according to a specific event type (Etype), an event occurrence location (Epos), a microphone location (Mpos), and a speaker location (Spos). Each artificial intelligence learning model created in this way has a unique label. That is, it is labeled as Label: Etype(Fire, invader)_Epos(c, u, b, l, r)_Mpos(c, u, b, l, r)_Spos(c, u, b, l, r). The event type may be a fire or an intruder occurrence, and the event location may be assigned a rough location such as center, top, bottom, left, right, or center, quadrant 1, quadrant 2, quadrant 3, quadrant 4. However, more precisely, each quadrant may be divided into four quadrants and designated.

The microphone position (Mpos) and speaker position (Spos) can be set in the same way.

The label can also be configured by specifying specific coordinates of Epos in the distance and angle of the microphone or speaker, or in a three-dimensional coordinate system. That is, as long as the specific specifications or format of the label is configured to indicate the type of event or obstacle and its occurrence location, freely transforming and displaying it in any format is within the conceived scope of the present invention.

In this setting, a learning model is created for a specific surveillance space, and when acoustic data collected from each microphone is input as input data to all generated artificial intelligence models (e.g., CNN, Convolutional Neural Network), the learning model with the highest probability value for the input data can be extracted, and the event type and event location can be extracted from the label of the extracted learning model to output the analysis result.

From the event type extracted in this way and the occurrence position of the corresponding event, it is possible to detect which event occurred and where within the sensing area. This is performed by the sound field change analyzer 140 .

6 is a flowchart illustrating the operation of a highly reliable IoT security sensor using big data according to an embodiment of the present invention.

As shown in FIG. 6, the reliable IoT security sensor using big data according to the present invention generates a sound source having a specific amplitude and frequency pattern (S110), receives sound from a specific detection area (S120), and analyzes the pattern of the generated sound source and the change in the sound field to analyze the change in the pattern of the received sound (S130). Here, together with step S110, the detection area is sensed from at least one sensor (S140).

Through this process, the sensing method according to the present invention outputs a result detected by the sensor based on a change in the sound field, and informs that the result detected by the sensor is the sensing result affected by the change in the sound field, thereby outputting a highly reliable sensing result that is robust against malfunction.

The detection method further includes outputting the detection result or an alarm signal according to the detection result to the outside (S150). Here, in the process of outputting the detection result or alarm signal, as a result of analyzing the sound field change in the sound field change analysis step, it is determined that there may be a malfunction in the detection result environmentally, and a connection signal is generated whether to output the alarm signal to an external device or not, and the alarm signal is output according to the connection signal.

In addition, the process of analyzing the sound field change is performed by actively detecting a sound field change event, detecting the front or back surface of an obstacle, operating regardless of heat or temperature characteristics, or including a combination thereof.

In addition, the detection method includes flame detection and smoke detection functions, and operates regardless of the spatial environment of the detection area. The flame detection can be detected by pre-extracting a change in a sound field for an event in which a spark occurs, storing the corresponding pattern, and comparing it with the change in the sound field for a sound wave received through a microphone.

In addition, the process of analyzing the sound field change includes extracting a matching probability for the sound field change using at least one learning model, and the learning model is labeled by including the type of obstacle or event, its occurrence location, or a combination thereof.

As described above, the present invention emits a predetermined sound signal to the sensing area and then analyzes the received sound signal to detect a sound field change pattern, thereby obtaining sensing information from at least one or more individual sensors based on a state change of the sensing environment of the sensing area. If there is no state change, the value sent from the unit IoT sensor is recognized as the value of the normal state, and when the state change exists, the value sent from the unit IoT sensor is recognized as the value of the specific sensing environment, so that it is robust against malfunction and has high reliability effects and advantages. there is

As described above, the present invention has been described with reference to the embodiments shown in the drawings, but this is only exemplary, and those skilled in the art to which the technology belongs will understand that various modifications and equivalent other embodiments are possible. Therefore, the technical protection scope of the present invention will be determined by the claims below.

100: IoT security sensor 110: sound source generator
120: sound receiving unit 130: sensing unit
140: sound field change analysis unit 150: output unit
160: control unit 170: memory
200: Edge computing server 300: IoT cloud server
400: DB (database)

Claims (10)

a sound generator generating a sound source having a specific amplitude and frequency pattern;
a sound receiver for receiving sound from the sensing area;
a sound field change analyzer for analyzing a change in the pattern of the generated sound source and the pattern of the received sound; and
A sensing unit configured to sense the sensing area from at least one sensor;
A highly reliable IoT security sensor using big data, characterized in that it outputs a result detected by the sensor based on the sound field change and outputs a highly reliable detection result robust against malfunction by notifying that the result detected by the sensor is the result affected by the sound field change. The method of claim 1,
The IoT security sensor,
Further comprising: an output unit for outputting the detection result or an alarm signal according to the detection result to the outside;
The output unit, as a result of analyzing the sound field change in the sound field change analysis unit for the detection result or the alarm signal, determines that there may be a malfunction in the detection result environmentally, and generates a connection signal whether to output the alarm signal to an external device or not according to the determination result, and outputs the alarm signal according to the connection signal. The method of claim 1,
The sound field change analysis unit,
An IoT security sensor characterized in that it actively detects a sound field change event, detects including the front or back of an obstacle, operates regardless of the characteristics of heat or temperature, or includes a combination thereof. The method of claim 1,
The IoT security sensor,
An IoT security sensor that includes flame detection and smoke detection functions and operates regardless of the spatial environment of the detection area. The method of claim 1,
The sound field change analysis unit,
Extracting a matching probability for the sound field change using at least one learning model, wherein the learning model is labeled including the type of obstacle or event, its occurrence location, or a combination thereof IoT security sensor. In the detection method through a highly reliable IoT security sensor using big data,
A sound source generating step of generating a sound source having a specific amplitude and frequency pattern;
a sound receiving step of receiving sound from the sensing area;
a sound field change analysis step of analyzing a change in the pattern of the generated sound source and the pattern of the received sound; and
and a sensing step of sensing the sensing area from at least one sensor, outputting a result detected by the sensor based on a sound field change, and outputting a highly reliable sensing result that is robust against malfunction by notifying that the result detected by the sensor is the sensing result affected by the sound field change. The method of claim 6,
The detection method is
Further comprising: an output step of outputting the detection result or an alarm signal according to the detection result to the outside;
The output step is
As a result of analyzing the sound field change of the detection result or the alarm signal in the sound field change analysis step, it is determined that there may be a malfunction in the detection result environmentally, a connection signal is generated as to whether or not the alarm signal is to be output to an external device according to the determination result, and the alarm signal is output according to the connection signal. The method of claim 6,
In the sound field change analysis step,
A sensing method characterized in that it actively detects a sound field change event, detects including the front or back of an obstacle, operates regardless of heat or temperature characteristics, or includes a combination thereof. The method of claim 6,
The detection method is
A detection method comprising flame detection and smoke detection functions, characterized in that it operates regardless of the spatial environment of the detection area. The method of claim 6,
In the sound field change analysis step,
And extracting a matching probability for the sound field change using at least one learning model, wherein the learning model is labeled by including the type, occurrence location, or combination of obstacles or events.