KR100517119B1 - System for preventing drowsiness while driving using unconscious respiratory detection - Google Patents

Abstract

The present invention relates to a drowsy driving prevention system using an insensitive breathing detection system that detects a driver's breathing in a situation where the driver is not aware of the driver's breathing, and detects a driver's breathing signal; Respiratory analysis means (12) for analyzing the respiratory signal and outputting an alarm signal when it is drowsy; And an alarm means (13) for receiving an alarm signal and generating an alarm; in particular, the respiratory detection means (11) includes a polymer compound plate for detecting tensile force and restoring force applied by a driver's breath, and a polymer. It is equipped with a strain gauge sensor that changes resistance when it is attached to the compound plate and receives tensile and restoring force, and a bridge circuit that detects a resistance change and converts it into a digital breathing signal, and is attached to a seat belt, and the breath analysis means 12 periodically calculates the ratio α of the power of a specific frequency band to the total power in the FFT power spectrum of the breathing signal, and outputs an alarm signal when the α value continuously exceeds a certain value for a predetermined number of times in succession. Characterized in that.

Description

Drowsy driving prevention system using insensitive breath detection {SYSTEM FOR PREVENTING DROWSINESS WHILE DRIVING USING UNCONSCIOUS RESPIRATORY DETECTION}

The present invention relates to a driver's drowsiness prevention system, and more particularly, to a drowsiness driving prevention system using an involuntary breath detection method for detecting a driver's drowsiness by detecting a driver's breathing in a situation where the driver is not aware.

Drowsy driving is an important cause of traffic accidents, with 36% of fatal traffic accidents and 43-54% of all traffic accidents resulting from drowsy driving. Therefore, a system that detects a driver's drowsiness and generates a warning sound to the driver is a very important technology to reduce traffic accidents.

Until now, research on the technology of detecting drowsiness while driving has been mostly conducted by electroencephalogram (EEG), Blinking Duration, or bowing, which show relatively large changes when entering drowsiness. The progress has been made around precision.

Korean Patent No. 209610 discloses a drowsiness preventing device using an EEG signal, but there is a problem in that the mounting of sensors for detecting EEG and heart rate may be sensitive to the driver's control. Therefore, there is a need for a way to detect drowsiness in situations where the driver is not aware.

Korean Patent No. 291378 detects whether the driver's eyes are closed and detects drowsiness using the duration of a closed eye state. In general, the principle of operation of the system using the eyelid opening and closing time is to attach a small camera 5-6 cm size with an ultrasonic sensor and an infrared sensor on the steering wheel portion of the vehicle to identify the driver's eyes. When the computer analyzes the driver's eyes blinking and the time when the eyes are closed, the alarm device sounds. This has some disadvantages, such as the detection sensor interferes with the natural light from the outside or reflected light from the glasses. In addition, obstacles such as blinding may occur due to other obstacles, and there is a disadvantage that the initial installation cost is expensive.

Korean Patent No. 180385 determines the drowsiness state by measuring the angle change amount of the isoline and the headline. Another method is a system that detects the vehicle's meandering amount and steering wheel operation by lane-recognizing cameras and sensors, and allows the computer to determine the driver's level of attention and voice warning with indication on the speed display. have. This is a system that determines whether the driver is drowsy by detecting the meandering rate according to the steering amount and the shaking of the steering wheel. However, it is difficult to accurately determine whether the driver is drowsy only by detecting the driving situation, since the steering amount of the steering wheel is small on the highway where the straight road continues. Therefore, in order to accurately determine whether the state is close to drowsiness, it is essential to detect the biosignal of the driver.

In order to solve the above problems, the present invention is focused on the fact that there is a clear difference between the driver's breathing patterns when waking up and sleepy, and can detect and analyze breathing in a situation where the driver is not aware of preventing drowsy driving. It is an object to provide a system.

The present invention, to achieve the above object, the breath detection means for detecting a breathing signal of the driver (11); Respiratory analysis means (12) for analyzing the respiratory signal and outputting an alarm signal when the user is drowsy; And alarm means (13) for receiving the alarm signal and generating an alarm. The drowsiness driving prevention system using the insensitive inhalation detection is provided.

According to another feature of the invention, the respiratory detection means 11, the polymer compound plate 111 for detecting the tensile force and the restoring force applied by the driver's breathing, and is attached on the polymer compound plate from the polymer compound plate Strain gauge sensor 112, the resistance is changed when the tensile force and the restoring force is transmitted, and the bridge circuit 113 is connected to the strain gauge sensor and detects the resistance change and converts it into a digital breathing signal, It is characterized by being attached.

According to another feature of the invention, the respiratory analysis means 12 periodically calculates the ratio α of the power of a particular frequency band to the total power in the Fast Fourier Transform (FFT) power spectrum of the respiratory signal and When the α value continuously exceeds a predetermined value for a predetermined number of times, an alarm signal is output.

First of all, the breathing state during sleep can be divided into non rapid eye movement (NREM) sleep and rapid eye movement (REM) sleep.

The drowsy sleep state is NREM sleep, and the NREM sleep is developed into shallow sleep stages 1 and 2 and slow wave sleep stages 3 and 4. The first stage and the second stage during a short period of time hindered by the awake state are called unstable NREM sleep, and drowsiness corresponds to unstable NREM sleep.

The main feature of unstable NREM sleep is periodic breathing. Periodic breathing is a phenomenon in which the size of the breathing increases and decreases regularly at the beginning of sleep or during sleepiness. That is, when sleep enters an unstable NREM sleep state, the breathing gradually increases, then decreases again, then stops for a few seconds and starts the next cycle. Apart from this breathing, less movement can also be a sign of drowsiness. When you are drowsy or sleepy, your movements naturally decrease. In special situations when driving, immobility can also be used to detect drowsiness because of the large change in movement.

The present invention analyzes the breathing characteristics of the driver as described above to determine the drowsiness state, hereinafter will be described an embodiment of the present invention with reference to the accompanying drawings.

1 is a block diagram of the drowsiness driving prevention system according to the present invention, the present invention is largely composed of a breath detection means 11, a breath analysis means 12 and the alarm means (13).

First, the respiratory detection means 11 is composed of a strain gauge sensor 112, a polymer compound plate 111 and a bridge circuit 113. Strain gauge sensor 112 is attached to the polymer compound plate 111 is fixed to the seat belt across the shoulder and belly of the driver. The tension and restoring force are alternately applied to the seat belt by the driver's breathing, and this force is again transmitted to the sensor 112 through the polymer plate 111. The resistance change of the sensor 112 due to this force is detected by the bridge circuit 113 and converted into a digital signal.

The respiratory analysis means 12 analyzes the digital respiratory signal through the Fast Fourier Transform (FFT) power spectrum to analyze the difference between the respiratory signal when awake and sleepy.

As shown in Fig. 2A, in awake breathing, the frequency component is spread over a relatively large area. In other words, there is a signal group that appears as the main lobe near the highest frequency, and there are two or three side lobes with similar sizes around it.

On the contrary, in the sleepy state, as shown in FIG. 2B, the signal is concentrated on one main lobe. Sidelobes appear around, but very small compared to the main lobe. That is, it can be seen that breathing in the drowsy state is more regular.

To distinguish between the awake state and the drowsiness state, the ratio of the power of a specific frequency band (0.25 to 0.35) to the total power in the FFT spectrum is calculated as an alpha value as shown in Equation 1 below. Where P (f) is the power of the signal spectrum

delete

3A and 3B are graphs showing alpha values over time in waking and drowsy states, respectively.

Looking at the change in alpha value over time, it can be seen that the overall value is larger in the drowsy state. As shown in FIG. 3A, the alpha value in the waking state is smaller than 1, while in the drowsy state as in FIG. 3B, a value smaller than 1 exists, but most have an alpha value greater than 1.

In this embodiment, the FFT analysis is performed in real time every second, so that an alarm signal is output when the alpha value is greater than 1 three times, and other appropriate values may be set according to the embodiment.

In the case of a predetermined drowsy condition, the alarm signal is output from the respiratory analysis means 12 as described above, and the alarm means 13 receives this and sounds an alarm. As for the alarm method, various applications such as buzzer and vibration are possible.

Since the present invention uses a biosignal, it is possible to detect drowsiness more accurately than other drowsiness detection systems, and it is possible to detect a biosignal in a situation where the driver is not aware, and thus there is no objection to the driver's care. It can be easily distributed due to the low cost compared to the drowsiness prevention system, which can contribute to the prevention of traffic accidents caused by drowsy driving.

1 is a block diagram of a drowsiness driving prevention system according to the present invention,

2A and 2B are graphs showing the respiratory fast Fourier transform (FFT) power spectra when awake and drowsy, respectively;

3A and 3B are graphs showing alpha values over time in waking and drowsy states, respectively.

Claims (3)

A device for detecting a breathing signal of a driver and analyzing the detected signal to generate an alarm when the driver is drowsy, The means for detecting the breathing is Polymer compound plate for sensing the tension and restoring force applied by the driver's breathing; A strain gauge sensor attached to the polymer compound plate to change resistance when tensile force and restoring force are transmitted through the polymer compound plate; And A bridge circuit connected to the strain gauge sensor and detecting a change in the resistance and converting the change into a digital signal; Wherein the means for detecting breathing calculates a ratio (α) of the power of a specific frequency band to the power of the entire frequency band in the FFT (Fast Fourier Transform) power spectrum of the breathing signal, wherein the value of α The drowsiness driving prevention device using insensitive breath detection characterized in that an alarm is generated when the number of consecutive times exceeds a predetermined number of times. delete delete