KR100718941B1 - Apparatus for alarming drowsy and method therefor - Google Patents

Abstract

The present invention provides a drowsiness warning device and a warning method thereof, by using a heart rate variability capable of determining a user's drowsiness most accurately, to determine the user's drowsiness most accurately and quickly and to alert the user.

The drowsiness warning device includes a signal detection / storing unit for detecting HR signals in real time and storing HRV data; A signal analyzer for analyzing the variation value of the HRV data detected and stored in the signal / detection storage unit in real time; A drowsiness detector for analyzing the data analyzed by the signal analyzer in real time to determine whether the driver is drowsy; And a warning unit for warning the driver of the drowsiness state when it is determined that the drowsiness state is detected by the drowsiness detection unit.

Drowsiness driving, heart rate variation, signal analysis, sympathetic nerve, parasympathetic nerve

Description

Drowsiness warning device and drowsiness warning method {APPARATUS FOR ALARMING DROWSY AND METHOD THEREFOR}

1 is an overall configuration diagram for a drowsiness alarm device by real-time autonomic neural analysis.

FIG. 2 is a block diagram illustrating a real-time signal detection / storage unit of FIG. 1. FIG.

3 is a block diagram for explaining a real-time signal analysis unit of FIG.

FIG. 4 is a block diagram illustrating the drowsiness detector of FIG. 1. FIG.

5 and 6 are perspective views showing each embodiment of the drowsiness warning device according to the present invention.

Figure 7 is a flow chart showing a drowsiness warning method using the drowsiness warning device according to the invention.

The present invention relates to a drowsiness warning device, and more particularly, by analyzing the autonomic nerve in real time using a heart rate variance to detect drowsiness accurately and quickly to warn of this or to prevent drowsiness accidents in advance. It relates to a warning device, and also to a drowsiness warning method using such a drowsiness warning device.

In general, when learning or driving in a tired state, or driving for a long time, the learner or driver unconsciously sleeps, which may interfere with the study or cause an accident.

Therefore, a long time ago, various methods for warning or preventing drowsiness of such learners, especially drivers, have been proposed, and various researches and developments are being conducted.

Recently, drowsiness detection or warning devices using various body organs of drivers or functions of the bodies are known. As such known examples, Korean Patent Registration No. 295850, which measures the number of blinks of a driver's eyes using a camera, or determines sleepiness when the heart rate falls below a reference value within a predetermined heart rate using the principle of falling heart rate during sleep , Korean Patent Registration No. 382154 to detect drowsiness by detecting driver's pulse signal, Korean Utility Registration No. 362100 to combat drowsiness with watch type low frequency stimulator, Korean patent registration to reduce drowsiness with watch shape Korean Patent Publication No. 375776, Korean Patent Publication No. 10-2004-101941 which detects drowsiness by comparing the pulse rate in normal operation with the pulse rate in driving. In addition, a method of preventing drowsiness by continuously detecting the brainwave state of the driver and controlling the window and the audio, a method of determining the drowsiness using the biosignal of the driver, and the like have been proposed.

However, such a conventional drowsiness detection method or drowsy driving prevention method has been shown to cause a variety of problems. First, the method of detecting drowsiness by using the blink of the driver has a problem that the performance of the device for tracking the position of the eye determines the accuracy of the measurement and the device is expensive.

In addition, in the case of using the heart rate, there is a case in which the heart rate does not change according to individual differences such as age and gender, and thus there is a problem in that accurate drowsiness cannot be measured.

In addition, the method using the EEG has a problem that the noise is generated during the measurement, the accuracy of accurately measuring or distinguishing the drowsiness or the steady state is significantly reduced.

On the other hand, in recent years, due to the development of medicine and science, research on autonomic nerves is being actively conducted, as well as innovative and accurate research results are being published, and various application technologies using the same have been studied.

Unlike the central nervous system, human autonomic nerves do not move according to our will, but are automatically activated by physical conditions or emotions. The activities of the five intestines related to biological activities are complicated and various organs can be known. It is known to be neither adjustable nor adjustable.

In addition, the organ that produces the energy needed for human activity is the heart, and the heart beats always seem to be constant, but if you look at the interval with actual quantitative figures, it's a little different every beat. This is called heart rate variability (HRV) and is used to evaluate the function of the current autonomic nervous system. Since 1979, attempts have been made to analyze the physiological and pathological clinical relationships caused by HRV. By 1996, the Task-Force team was a member of the European Heart Association and the North American Heart Association. It is organized and has a draft standard for interpreting HRV.

This HRV analysis method is used in the medical community to test the health and psychological state of patients. Recently, attempts have been made to apply related analytical technology to the industry by switching to a real-time processing method by integrating with various information devices.

As such, the autonomic nervous system that controls the activities of the internal organs of the human body is composed of sympathetic and parasympathetic nerves to accurately represent the physical state and arousal state of the human body, and the central nervous system recognizes the state of the autonomic nervous system that humans can recognize. Based on the technical idea that if it can be recognized as a digitized state or if it can be recognized beforehand, various technical situations can be detected in advance, and those skilled in the art will understand such technical ideas as human drowsiness. Active research is being made on the technical problem of applying to detection.

Accordingly, the present invention has been invented to solve the problems and technical problems of the prior art as described above, the object of the present invention is to accurately determine the drowsiness state of the user to warn drowsiness and to prevent accidents due to drowsiness warning It is to provide a device and a warning method thereof.

It is another object of the present invention to provide a drowsiness warning device and a warning method capable of detecting a drowsiness state by detecting the drowsiness state accurately and quickly by analyzing autonomic nerves using heart rate variability in real time.

In order to solve the above object, the drowsiness warning device according to the present invention, the signal detection / storage unit for detecting HR signal in real time to store HRV data; A signal analyzer for analyzing in real time the variation of the HRV data detected and stored in the signal / detection storage unit; A drowsiness detector for analyzing the data analyzed by the signal analyzer in real time to determine whether the driver is drowsy; And a drowsiness warning unit for warning the driver of the drowsiness state when the drowsiness detection unit determines that the drowsiness state.

In addition, to solve the above object, the drowsiness warning method according to the invention, detecting the heart rate from the user's body, and amplifying the detected heart rate and filtering; Converting the filtered effective heart rate signal into a digital heart rate signal, detecting a peak value, and storing the detected HRV data; After preprocessing the set number of data among HRV data, rearranging the preprocessed data according to time, synchronizing HRV data according to time series, connecting resampling data with reverse order data and weighting using cosine function. Applying; After analyzing the weighted data, extract the values of the low frequency band and the high frequency band, and then symbolize the data using the respective frequency bands to calculate the symbolic data for each frequency, and at the same time, the low frequency band data reflecting the sympathetic nerve activity. Calculating normalized data using high frequency band data reflecting the parasympathetic nerve activity; Determining whether the data is in the rising state or the falling state by using the maximum and minimum values received by the data analyzer of the real-time drowsy detector; Calculating activity of the sympathetic nerve of the user or driver using the analyzed and determined data; At the same time, calculating the activity of the parasympathetic nerve of the user or driver using the data analyzed and determined by the symbol data determination unit; Comparing the calculated sympathetic activity value and parasympathetic activity value; If it is determined that the parasympathetic activity value is large as a result of the comparison between the sympathetic activity value and the parasympathetic activity value, determining whether the difference is greater than a prescribed ratio; If the determined ratio is greater than the prescribed ratio as a result of determining the parasympathetic nerve activity ratio, transmitting a warning signal; And alerting the driver when a warning signal is received from the real-time drowsiness detector.

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

1 is a block diagram showing the drowsiness warning device according to the invention as a whole, Figure 2 is a block diagram showing the signal detection / storage of the drowsiness warning device according to the invention in detail, Figure 3 is a drowsiness warning device according to the invention Figure 4 is a block diagram showing in detail the signal analysis unit, Figure 4 is a flow chart showing in detail the drowsiness detection unit of the drowsiness warning device according to the present invention, Figures 5 and 6 are perspective views showing a state in which the drowsiness warning device according to the present invention is applied.

First, referring to FIGS. 1 to 4, the drowsiness warning device according to the present invention includes a signal detection / storage unit 100 for detecting and storing a heartbeat signal from a user such as a student, a driver, or an examiner who needs a drowsiness warning device. do.

The signal detector / storage unit 100 includes a detection sensor 110 for detecting a heartbeat signal of the user. The heartbeat detection sensor 110 may be formed as a two-channel electrode sensor or a photo pulse wave sensor. Here, the photoelectric wave sensor is formed to measure the heart rate from the signal scattered in accordance with the amount of blood flowing through the blood vessel through the light from the fingertip or earlobe using an optical sensor having a light emitting and receiving element.

The heartbeat detection sensor 110 is connected to an amplifier 120 for amplifying the heartbeat signal detected therefrom.

Subsequently, the amplifier 120 is connected with a filter 130 for filtering the heart rate signal amplified therefrom. The filter 130 is preferably formed to remove a 60Hz band signal and high frequency or low frequency noise.

In addition, the filter 130 is connected to the A / D converter 140 for converting the effective heart rate signal filtered therefrom into a digital signal. That is, the A / D converter 140 serves to convert the analog heartbeat signal into a digital heartbeat signal.

The A / D converter 140 is connected to a peak detector 150 for detecting a peak value of the heartbeat signal converted into a digital signal. The peak detector 150 detects the peak value by using an auto-correlation peak detecting method, which is called an autocorrelation peak detection method. In particular, the peak detector 150 calculates heart rate variability data of the user, that is, HRV data using the peak value. In addition, the peak detector 150 may calculate the heart rate per minute from the peak value.

The peak detector 150 is connected with an HRV data store 160 for storing the HRV data detected therefrom. In particular, the HRV data store 160 is stored in order in the HRV data store 160 each time a peak value is calculated at the peak detector 150.

The signal / detection storage unit 100 as described above is connected to a signal analyzer 200 for analyzing in real time the variation value of the detected and stored HRV data. The signal analyzer 200 is preferably configured to analyze in real time whether the current state is drowsy at regular intervals in order to accurately analyze the moment of sleepiness. The analysis cycle of the signal may vary depending on the embodiment or the user's discretion. For example, in the case of the driving state or the high speed operation, the real-time analysis cycle may be as fast as every second. Of course, when the present embodiment is applied to the learner, the signal may be analyzed by setting the analysis period to 5 seconds or more.

In more detail, the signal analysis unit 200 is connected to the HRV signal storage unit 160 to analyze the HRV data or the variance values detected and stored in the signal detection / storage unit 100 in real time at each set analysis period. It is provided with a data preprocessor 210 for pre-processing by receiving or extracting a predetermined number of data stored there. Here, the number of extracted data can be set by a user or an input program. For example, for a relatively accurate measurement, it is preferable to extract 250 to 300 signals and extract about 300 signals. Of course, the preprocessing unit 210 serves to process the lost data by interpolation and to remove artifacts.

The preprocessing unit 210 is connected to a data resampling unit 220 for relocating the data, which has been preprocessed therein, over time. The data resampling unit 220 is, for example, a device for synchronizing preprocessing HRV data prior to the present time according to time series, and HRV data preprocessed in the data preprocessor 210 is generated whenever a peak is extracted without time concept. As the data, this is to relocate the HRV data according to the progress time through resampling. In this case, the resampling is preferably performed using the average heart rate of the HRV as a reference frequency.

The data resampling unit 220 is connected to a weight data generation unit 230 for generating data to which weights are applied to the resampled data therein. The weight data generator 230 may be configured to generate reverse order data that is symmetrical to the resampled data in a manner of converting a signal pattern for the resampled data.

In addition, the weight data generator 230 connects the resampling data and the reverse order data and applies a weight using a value of a cosine function. The reason for combining the original data and the data in the reverse order is to make a complete cycle when applying the weight using the cosine function, to prevent the occurrence of unnecessary noise during fast Fourier transform.

The weight data generator 230 is connected to a frequency analyzer 240 for analyzing the generated weight data. The frequency analyzer 240 analyzes the frequency by extracting values of the low frequency band (LF: 0.04 to 0.15 Hz) and the high frequency band (HF: 0.15 to 0.4 HZ) from the weight data using a fast Fourier transform method. At this time, the value of the low frequency band reflects the sympathetic nerve activity at the present time, while the value of the high frequency band reflects the parasympathetic activity at the current time.

The frequency analysis unit 240 is connected to a data symbol unit 250 for symbolizing data using the frequency band calculated there. The data symbolizing unit 250 compares the low frequency band value calculated by the frequency analyzer 240 with the low frequency band value generated immediately before the value, and is calculated by the frequency analyzer 240 in the same manner. Preferably, the high frequency band is compared with a high frequency band value generated immediately before the value to calculate cumulative symbol data for each frequency.

For example, in the processing of the low frequency band, if the cumulative symbolic data value up to +5 is +5, the low frequency band value at the previous time point is compared with the low frequency value at the present time point, and +1 is added if the current time point is large. +6 is the cumulative symbolic data value, and if the low frequency band value of the previous point is compared with the low frequency value of the present point, and is equal to the current point, +5 plus +0 is the cumulative symbolic data point. If the current time is small, +4 plus -1 is the cumulative symbol data.

Optionally, by collecting cumulative symbol data up to the present time and arranging them on a time series, the flow of change over time of the sympathetic nerve or parasympathetic nerve can be seen.

On the other hand, the frequency analyzer 240 has a data normalization unit 260 for calculating normalized data using low frequency band data reflecting the sympathetic nerve activity value calculated as described above and high frequency band data reflecting the parasympathetic nerve activity value. Connected. The data normalization unit 260 subtracts the ultra low frequency band (VLF) data from the full band (TP) data calculated by the frequency analyzer 240, and then divides the low frequency normalized data and the total band ( TP) subtracts VLF data from the data and divides the data into high frequency band data to calculate high frequency normalized data expressed as a percentage. Low frequency normalized data means sympathetic nerve activity rate in autonomic nerve analysis, and high frequency normalized data means sympathetic nerve activity rate in autonomic nerve analysis.

In addition, the data symbolizing unit 250 and the data normalizing unit 260 are connected to a data storage unit 270 that stores data calculated from each of them, and transmits or provides a signal to a real-time drowsiness detection unit described later.

The signal analyzer 200 configured as described above is connected to a real time drowsiness detector 300 for determining whether the user is drowsy by using and analyzing the data analyzed therefrom.

The real-time drowsy detector 300 is a low frequency band symbolic data flow calculated from the data symbolizing unit 250 and the data normalizing unit 260 and stored in the data storage unit 270, a high frequency band symbolic data flow, By analyzing the flow of sympathetic and parasympathetic nerve activity is configured to detect the user's drowsy state in real time, every set moment.

In more detail, the real-time drowsiness detection unit 300 is provided with low-frequency symbolic data stored in the data storage unit 270 of the signal analysis unit 200, and finds the position of the maximum value and the minimum value therefrom in the rising state or the falling state. A symbolic data determination unit 310 for analyzing the recognition is provided.

The symbolic data determination unit 310 is configured to determine whether the falling state or the rising state, and determines whether the falling time lasts more than the prescribed time or the allowed time, and also determines whether the falling speed increases over a certain range. It is desirable to be.

Symmetric data determination unit 310 is connected to the sympathetic activity calculation unit 320 for calculating the activity of the sympathetic nerve of the user or driver.

In addition, the symbol data determining unit 310 is connected to the parasympathetic activity calculator 330 for calculating the activity of the parasympathetic nerve of the user or driver.

On the other hand, the above-mentioned sympathetic activity calculating unit 320 and parasympathetic activity calculating unit 330 in the activity value comparison unit for comparing the size of the sympathetic activity value and the parasympathetic activity value calculated by each of these calculation unit ( 340 is connected.

In addition, the activity value comparison unit 340 is connected to the parasympathetic activity ratio determination unit 350 that determines whether the difference between the parasympathetic activity value is greater than or equal to the prescribed ratio among the results of the deviation or difference compared therefrom. Here, the reference value of the ratio of the parasympathetic nerve activity is preferably set to 60%, because when the ratio of the parasympathetic nerve activity is 60% or more, it is determined that the driver or the user is in a state of drowsiness or drowsiness.

The parasympathetic nerve activity ratio determining unit 350 is connected to a warning signal transmitting unit 360 for receiving the signal and operating a subsequent warning unit when the ratio determined therefrom becomes a prescribed ratio, for example, 60% or more.

The real-time drowsiness detection unit 300 is connected to the warning unit 400 for warning the driver's drowsy state is confirmed. The warning unit 400 may be configured in an acoustic manner capable of generating an alarm, or may be configured in an acoustic means such as a siren, or may be formed in a manner of stimulating a part of the driver's body.

 On the other hand, the drowsiness alarm device configured as described above can be applied in various ways.

For example, as shown in Figure 5, the drowsiness alarm device according to the first embodiment of the present invention can be applied to the steering wheel when applied to the driver.

In this case, a plurality of heartbeat detection sensors 110 for detecting a heartbeat through the palm of the driver are installed in the vicinity of the steering wheel 500, that is, the part held by the driver. In addition, in the center of the steering wheel to analyze the peak value of the HRV data detected and stored in the signal detection / storage unit 100, the signal / detection storage unit 100 that stores the heart rate data received from the heart rate detection sensor 110 Real-time signal analysis unit 200 for analyzing, the real-time drowsiness detection unit 300 for determining whether the driver is drowsy by analyzing the data analyzed from the signal analysis unit 200 and a warning unit for alerting the driver when the drowsiness determination ( 400 is packaged and installed.

As such, when the drowsiness warning device is installed in the steering wheel, the warning unit 400 may be connected to the warning sound buzzer installed in the steering wheel 500 to warn the drowsiness state of the driver through the warning sound.

On the other hand, as shown in Figure 6, the drowsiness warning device according to the present invention can be applied in the form of hearing aids or earphones to be popular and portable.

For example, the end of the drowsiness warning device 600 formed in the form of earphones or hearing aids is provided with a probe type heart rate detection sensor 110 is inserted into the user's ear, the heart rate received from the heart rate detection sensor 110 in the body. The signal detection / storage unit 100 storing data, the signal analysis unit 200 for peak values of HRV data detected and stored in the signal / detection storage unit 100, and analyzed by the signal analysis unit 200. The real-time drowsiness detection unit 300 for analyzing the data to determine whether the user is drowsy and a warning unit 400 for warning the user when the drowsiness determination is built.

As such, when the drowsiness warning device is implemented in the form of a hearing aid or earphone, the warning unit 400 may be implemented in an alarm or vibration manner to warn.

Hereinafter, the drowsiness warning method using the drowsiness warning device configured as described above will be described in detail.

7 is a flowchart illustrating a drowsiness warning method using the drowsiness warning device according to the present invention.

Referring to FIGS. 7 and 1 to 4, first, a heart rate is measured through a detection sensor 110 connected to or attached to a part of a user's body such as a driver or a learner, and then detected from the heartbeat detection sensor 110. The amplified heart rate is amplified by the amplifier 120, and the filter 130 removes a 60 Hz band signal and high frequency or low frequency noise (S110).

After converting the filtered effective heartbeat signal into the digital heartbeat signal by the A / D converter 140, the peak detector 150 detects the peak value and stores the detected HRV data in the HRV data storage 160. (S120).

Thereafter, the data is transferred to the data preprocessor 210 of the signal analyzer 200 and preprocessed by the set number of HRV data stored in the signal / detection storage unit 100 in the manner described above, and the preprocessed data is completed. The data resampling unit 220 rearranges the data according to time to synchronize the HRV data according to the time series, and then the weight data generator 230 connects the resampling data and the reverse order data to the weighted data using the cosine function. Apply (S210).

When the weight is applied to the resampling data as described above, the frequency analyzer 240 analyzes the weighted data to extract values of the low frequency band (LF: 0.04 to 0.15 Hz) and the high frequency band (HF: 0.15 to 0.4 HZ). In addition, the data symbolizing unit 250 symbolizes data using each frequency band to calculate symbolic data for each frequency, and at the same time, the low frequency band data and parasympathetic data reflecting the sympathetic nerve activity in the data normalizing unit 260. Normalized data is calculated using high-frequency band data reflecting nerve activity values and transferred to the data storage unit 270 to be stored (S220).

As described above, when the analysis of the data is analyzed in real time by the signal analyzer 200 and the processed data value is input to the data storage unit 270 and stored, the maximum value in the data analyzer 310 of the real-time drowsiness detector 300. And find the location of the minimum value and analyze whether it is in the rising state or the falling state (S310-1), and determines whether the analysis result is the falling state or the rising state (S310-2), if the falling state, the falling time is It is determined whether the duration lasts longer than the allowed time (S310-3), and if the falling time lasts for a predetermined time or more, it is determined whether the falling speed is increased beyond the predetermined range (S130-4). In this embodiment, the prescribed falling time is set to 1 minute and the continuous increase range of the falling speed is set to 20%. However, this can be appropriately changed by a user or an input setting value. In addition, if it is determined that the negative (No) in each of the determination step (S310-2, S310-3, S310-4), it is returned to the analysis step (S110) and re-analyzed, or to the determination step before each determination step It will be clear that the analysis and judgment is carried out continuously and repeatedly in a manner that is returned and judged.

Subsequently, the sympathetic nerve activity calculating unit 320 calculates the activity of the sympathetic nerve of the user or the driver as a numerical value by using the above analysis and breaking data in the symbolic data determination unit 310 (S320).

At the same time, the parasympathetic nerve activity calculating unit 330 calculates the activity of the parasympathetic nerve of the user or the driver as a numerical value by using the above analysis and breaking data in the symbol data determination unit 310 (S330).

As described above, when each sympathetic activity and parasympathetic nerve activity are calculated, the activity value comparison unit 340 compares the size of the sympathetic activity value and the parasympathetic activity value (S340).

As a result of comparing the sympathetic nerve activity value and the parasympathetic nerve activity value, it is determined that the parasympathetic nerve activity value is large, the difference of the parasympathetic nerve activity value among the results of the deviation or difference compared in the parasympathetic nerve activity ratio determination unit 350 is a prescribed ratio It judges whether it is abnormal. In this embodiment, the reference value of the ratio of parasympathetic nerve activity is set to 60%, but the judgment reference value may be arbitrarily set and used.

Of course, if it is determined that the negative result (no) in each of the determination step (S340, S350), it will be apparent that the determination and analysis step is carried out repeatedly to return to the analysis step (S110).

After determining the parasympathetic nerve activity ratio, if the determined ratio is determined to be a prescribed ratio, for example, 60% or more, the warning signal transmitter 360 receives the signal and prepares for transmission (S360).

As described above, when the drowsiness state of the driver is confirmed in the real-time drowsiness detection unit 300, the warning unit 400 receives the signal from the warning signal transmitter 360 to warn the driver or user that they are in a drowsy state (S400). ). Here, the warning method for the user may be made of an auditory method using an alarm or a siren, a visual method using a warning lamp, a method of stimulating a part of the body, and the like.

Therefore, by using the driver's or user's biological system, it is possible to accurately and quickly determine whether drowsiness and to warn it.

As described above, the drowsiness warning device according to the present invention uses the heart rate variability that can determine the user's drowsiness state most accurately, it is possible to determine the user's drowsiness state most accurately and quickly to alert the driver There is an effect of improving the reliability and productability.

In addition, it is possible to determine and warn drowsiness by implementing in a variety of ways, such as steering wheel, earphones, there is an advantage that the applicability is improved.

Claims (8)

A signal detector / storage unit for detecting HRV data in real time and storing HRV data, a signal analyzer for analyzing in real time peak values of HRV data detected and stored in the signal / detection storage unit, and the signal analyzer A drowsiness detection unit for determining whether the driver is drowsy by analyzing the data analyzed in real time, and a drowsiness warning unit for warning the user if it is determined that the drowsiness state is the drowsiness detection unit; The real time signal analysis unit A data preprocessor for pre-processing or extracting the HRV data or the peak value detected and stored in the signal detection / storage in real time for each set analysis period; A data resampling unit connected to the preprocessing unit and repositioning the preprocessed data according to time; A weight data generation unit connected to the data resampling unit and configured to generate data to which weight is applied to the resampled data; A frequency analyzer connected to the weight data generator and configured to analyze the generated weight data; A data symbol unit connected to the frequency analyzer and configured to symbolize data using the analyzed frequency band; And a data normalization unit connected to the frequency analyzer and configured to calculate normalization data using low frequency data reflecting the calculated sympathetic activity and high frequency data reflecting the parasympathetic activity. delete delete The method of claim 1, wherein the real-time drowsiness detector A symbolic data determination unit receiving low frequency symbolic data calculated and provided by the real-time signal analysis unit and determining a position of the maximum value and the minimum value therefrom and determining whether the signal is in the rising state or the falling state; A sympathetic activity calculator for calculating a sympathetic activity of a user connected to the determination unit; A parasympathetic neural activity calculating unit coupled to the symbolic data determination unit for calculating the activity of the parasympathetic nerve of the user; An activity value comparison unit connected to the sympathetic activity calculating unit and the parasympathetic activity calculating unit, for comparing the magnitudes of the sympathetic activity values and the parasympathetic activity values calculated by the respective calculating units; And And a parasympathetic nerve activity ratio determining unit connected to the activity value comparing unit and determining whether a difference in parasympathetic activity values is greater than or equal to a prescribed ratio among the results of the deviation or difference compared therefrom. delete delete In the sleepiness warning method using the sleepiness warning device according to claim 1, Detecting a heart rate from the body of the user and amplifying and filtering the detected heart rate; Converting the filtered effective heartbeat signal into a digital heartbeat signal, detecting a peak value, and storing the detected HRV data; After preprocessing the set number of data among the HRV data, rearranging the preprocessed data according to time, synchronizing the HRV data according to time series, connecting the resampling data and the reverse order data, and using the value of the cosine function. Applying a weight; After analyzing the data to which the weight is applied, extract the values of the low frequency band and the high frequency band, and then symbolize the data using the respective frequency bands to calculate symbolic data for each frequency, and at the same time reflect the sympathetic nerve activity. Calculating normalized data using high frequency band data reflecting data and parasympathetic nerve activity; Analyzing and determining whether the data is in the rising state or the falling state using the maximum and minimum values of the analyzed data; Calculating the activity of the sympathetic nerve of the user using the maximum and minimum values of the analyzed and determined data At the same time, calculating the activity of the parasympathetic nerve of the user using the analyzed and determined data; Comparing the calculated sympathetic activity value and parasympathetic activity value; If it is determined that the parasympathetic activity value is large as a result of the comparison between the sympathetic activity value and the parasympathetic activity value, determining whether the difference is greater than a prescribed ratio; If the determined ratio is greater than the prescribed ratio as a result of determining the parasympathetic nerve activity ratio, transmitting a warning signal; Drowsiness warning method comprising the step of alerting the driver when a warning signal is received. The method of claim 7, wherein the analyzing and determining whether the data is in the rising state or the falling state using the maximum and minimum values of the data comprises: Finding positions of the maximum value and the minimum value and analyzing whether each value is in a rising state or a falling state; Determining whether the analysis result is a falling state or a rising state; Judging whether the fall time lasts more than the prescribed time or the allowed time if the falling result is determined; And If the fall time lasts for more than a predetermined time, the drowsiness warning method comprising the step of determining whether the falling speed increases over a certain range.

Images