JP2015219771A - Drowsy driving prevention device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a drowsy driving prevention device which outputs an alarm at a proper timing thereby preventing drowsy driving, neither requiring a face condition or biological information of a driver nor increasing operation load applied to the driver.SOLUTION: A drowsy driving prevention device (8) prevents a driver from falling asleep at the wheel, and comprises: a drowsiness level detection part (16) for detecting a drowsiness level of the driver based on drive operation of the driver and/or a behavior of a vehicle (1); an alarm output part (18) for outputting an alarm when the drowsiness level detected by the drowsiness level detection part is higher than a predetermined alarm threshold; an alarm threshold setting part (20) for setting the alarm threshold, based on drive operation and/or the behavior of the vehicle since travel start of the vehicle until passing of a predetermined learning time; and an alarm threshold correction part (22) for correcting the set alarm threshold, based on the set alarm threshold and detected drowsiness level, after setting the alarm threshold.

Description

  The present invention relates to a snooze driving prevention apparatus, and more particularly, to a snoozing driving prevention apparatus that prevents a snoozing driving based on a driver's drowsiness level detected based on a driving operation by a driver and / or a behavior of a vehicle.

  2. Description of the Related Art Conventionally, a snooze driving prevention device that prevents a driver from snoozing is known.

  For example, Patent Document 1 discloses a dozing determination apparatus that includes a photographing unit that photographs a driver's face and determines the driver's dozing based on the state of the face photographed by the photographing unit. There is also known an apparatus that detects biological information such as the opening degree of the driver's eyelids, brain waves, heartbeats, and the like, and determines doze based on the biological information.

  Further, in Patent Document 2, a driver's drowsiness is detected based on vehicle information such as the steering angle of the vehicle, throttle opening, vehicle speed, number of brake operations, etc., and if a drowsiness is detected, a warning is given to the driver. A device for preventing snoozing is disclosed.

JP 2007-233475 A JP 2009-251973 A

  However, in the device that determines doze based on the driver's face state or biometric information like the device of Patent Document 1 described above, a device for detecting the driver state must be provided in the vehicle. Costs will increase.

In addition, like the device of Patent Document 2 described above, a device that detects a driver's doze based on vehicle information increases costs compared to a device that determines a driver's doze based on the driver's face state or biological information. Even if the driver does not feel so sleepy or tired, the driver may determine that the driver is asleep and output a warning. In this case, the driver feels that the warning is bothersome. End up.
Therefore, the device of Patent Document 2 includes a reset switch that is operated by a driver and outputs a warning stop signal. When the reset switch is operated, the warning output is stopped. However, in order to stop the output of the warning, the driver must operate the reset switch, which increases the operation burden on the driver.

  The present invention has been made to solve the above-described problems of the prior art, and does not require the driver's face state or biometric information, and does not increase the driver's operation burden. An object of the present invention is to provide a drowsiness prevention device capable of outputting a warning and preventing a driver from drowsing.

In order to achieve the above object, a snooze driving prevention device of the present invention is a snooze driving prevention device that prevents a driver from falling asleep, and the driver's drowsiness level is determined based on the driving operation by the driver and / or the behavior of the vehicle. Drowsiness level detection means for detecting, alarm output means for outputting an alarm when the drowsiness level detected by the drowsiness level detection means is higher than a predetermined alarm threshold, and a predetermined learning time elapses after the vehicle starts running An alarm threshold value setting means for setting an alarm threshold value based on driving operation and / or vehicle behavior in the period up to and after setting the alarm threshold value, setting based on the set alarm threshold value and the detected sleepiness level And an alarm threshold value correcting means for correcting the alarm threshold value.
In the present invention configured as described above, the warning threshold value setting means sets the warning threshold value based on the driving operation and / or the behavior of the vehicle after the vehicle starts running until a predetermined learning time elapses. The alarm threshold value correcting means corrects the set alarm threshold value based on the set alarm threshold value and the detected drowsiness level after the alarm threshold value is set, so it is considered that the driver does not feel sleepy. An alarm threshold can be set based on the driving operation and / or vehicle behavior immediately after the start of traveling, and an alarm is output at an appropriate frequency according to the relationship between the actually detected drowsiness level and the alarm threshold. The alarm threshold value can be corrected so that the driver's face state and biometric information are not required, and the driver's operation burden is not increased. In it is possible to prevent drowsy driving of the driver and outputs an alarm.

In the present invention, it is preferable that the alarm threshold value correction means increases the alarm threshold value when a drowsiness level higher than the set alarm threshold value is continuously detected over a first determination time longer than the learning time. To correct.
In the present invention configured as described above, the driver does not feel so drowsy and continues driving for the first determination time even after the alarm is output due to the drowsiness level exceeding the alarm threshold. In this case, the alarm threshold value can be corrected upward, and the frequency of alarm output can be reduced, thereby outputting the alarm at an appropriate timing and making the driver doze without causing the driver to feel annoyance. Can be prevented.

In the present invention, it is preferable that the alarm threshold value correction means has a number of times that the detected drowsiness level exceeds a set alarm threshold value during a second determination time longer than the learning time, a predetermined determination number or more. If so, the alarm threshold value is corrected upward.
In the present invention configured as described above, when the driver does not feel so sleepy and the driving is continued despite the alarm being output more than the determination times during the second determination time, The alarm threshold can be corrected upward to reduce the frequency at which the alarm is output, thereby preventing the driver from falling asleep by outputting an alarm at an appropriate timing without making the driver feel bothersome be able to.

In the present invention, it is preferable that the warning threshold value correcting means is a correction that is proportional to the average value of the drowsiness level detected during the third determination time longer than the learning time after the vehicle starts running. A determination reference value is calculated, and a total time in which the calculated drowsiness determination reference value is smaller than the set alarm threshold and a drowsiness level higher than the correction determination reference value is detected is longer than the learning time. If it is longer than the determination time, the alarm threshold is corrected downward.
In the present invention configured as described above, the drowsiness level of the driver greatly increases compared to the drowsiness level until the third determination time elapses after the vehicle starts running, and the state is the sum total. When it is maintained for the fourth determination time or more and there is a high possibility that the driver feels drowsy, the alarm threshold value can be corrected downward to increase the frequency at which the alarm is output. An alarm can be output at a proper timing to prevent the driver from falling asleep.

  According to the snooze driving prevention apparatus according to the present invention, the driver's face state and biometric information are not required, and the driver's snooze driving is performed by outputting an alarm at an appropriate timing without increasing the driver's operation burden. Can be prevented.

It is a block diagram which shows the electric constitution of the snooze driving | running prevention apparatus by embodiment of this invention. It is a flowchart of the dozing driving prevention process which the dozing driving prevention apparatus by embodiment of this invention performs. It is a flowchart of the alarm threshold value correction process which the dozing driving prevention apparatus by embodiment of this invention performs. It is a diagram which shows the time change of the drowsiness level referred when the alarm threshold value correction | amendment part of the snooze driving | running prevention apparatus by embodiment of this invention determines whether an alarm threshold value is upwardly corrected. It is a diagram which shows the time change of the drowsiness level referred when the alarm threshold value correction | amendment part of the snooze driving | running prevention apparatus by embodiment of this invention determines whether an alarm threshold value is corrected below.

Hereinafter, a drowsiness prevention device according to an embodiment of the present invention will be described with reference to the accompanying drawings.
First, with reference to FIG. 1, the configuration of a snooze driving prevention apparatus according to an embodiment of the present invention will be described.
FIG. 1 is a block diagram showing an electrical configuration of a snooze driving prevention apparatus according to an embodiment of the present invention.

  First, in FIG. 1, the code | symbol 1 shows the vehicle carrying the doze driving prevention apparatus by embodiment of this invention. The vehicle 1 includes a steering angle sensor 2 that detects a rotation angle of a steering wheel, a vehicle speed sensor 4 that detects a vehicle speed, and a yaw rate that detects a rotation angular velocity (yaw rate) of the vehicle 1 around a vertical axis (yaw axis). It has a sensor 6. The steering angle sensor 2, the vehicle speed sensor 4, and the yaw rate sensor 6 output the detected values to the doze driving prevention device 8.

  Further, the vehicle 1 is provided with an output device 10 that outputs a doze prevention alarm in accordance with the control of the doze driving prevention device 8. The output device 10 includes a display 12 that outputs character information and image information, and a speaker 14 that outputs audio information.

As shown in FIG. 1, the snooze driving prevention device 8 includes a drowsiness level detection unit 16 that detects the drowsiness level of the driver based on the driving operation by the driver and the behavior of the vehicle 1, and the sleepiness level detected by the drowsiness level detection unit 16. Has an alarm output unit 18 for outputting an alarm from the output device 10 when the alarm threshold is exceeded. Here, the “sleepiness level” refers to a scale representing the degree of sleepiness and fatigue felt by the driver.
In addition, the snooze driving prevention device 8 includes an alarm threshold value setting unit 20 that sets an alarm threshold value that serves as a reference for determining whether the alarm output unit 18 outputs an alarm, an alarm threshold value correction unit 22 that corrects the alarm threshold value, And it has the memory | storage part 24 which memorize | stores the data used for the correction of the alarm threshold value by the alarm threshold value setting by the alarm threshold value setting part 20, and the alarm threshold value correction | amendment part 22. FIG.
These elements included in the snooze driving prevention device 8 include a CPU, various programs that are interpreted and executed on the CPU (including basic control programs such as an OS and application programs that are activated on the OS to realize specific functions), And a computer having an internal memory such as a ROM or RAM for storing programs and various data.

Next, referring to FIG. 2 and FIG. 3, a description will be given of the dozing driving prevention process performed by the dozing driving prevention device 8.
FIG. 2 is a flowchart of the dozing operation prevention process performed by the dozing operation prevention apparatus 8 according to the embodiment of the present invention, and FIG. 3 is a flowchart of the alarm threshold value correction process performed by the dozing operation prevention apparatus 8 according to the embodiment of the present invention. is there.

  First, the main routine of the dozing driving prevention process will be described with reference to FIG. The dozing driving prevention process shown in FIG. 2 is performed when the ignition of the vehicle 1 is turned on and the vehicle 1 starts to travel (for example, when the vehicle speed detected by the vehicle speed sensor 4 becomes 5 km / h or more). It is started and executed repeatedly.

  As shown in FIG. 2, when the dozing driving prevention process is started, in step S <b> 1, the alarm threshold setting unit 20 determines the steering angle detected by the steering angle sensor 2 and the yaw rate detected by the yaw rate sensor 6. Obtained as a set of sleepiness data. In other words, as the drowsiness and fatigue felt by the driver increases, the finer steering operations decrease and the rapid correction operations occur more frequently, and the vehicle behavior changes, so the steering angle and vehicle behavior corresponding to this steering operation are changed. The corresponding yaw rate can be used as sleepiness data for detecting the sleepiness level. The sleepiness data acquired here is temporarily stored in the memory.

  Next, in step S <b> 2, the alarm threshold setting unit 20 determines whether or not a predetermined learning time (for example, 10 minutes) has elapsed after the vehicle 1 starts traveling. As a result, when the learning time has not elapsed, the alarm threshold value setting unit 20 returns to step S1 and repeats acquisition of sleepiness data.

  On the other hand, when the learning time has elapsed, the warning threshold value setting unit 20 proceeds to step S3, and stores the set of sleepiness data acquired during the learning time in the storage unit 24 as learning data.

Next, it progresses to step S4 and the warning threshold value setting part 20 becomes the reference | standard which determines whether the warning output part 18 outputs a warning based on the learning data memorize | stored in the memory | storage part 24 in step S3. Set the initial threshold value.
Specifically, the alarm threshold value setting unit 20 calculates the Mahalanobis distance of each sleepiness data for the learning data that is a set of sleepiness data acquired during the learning time, and calculates the median value of the calculated Mahalanobis distances. A value obtained by multiplying by a predetermined coefficient is calculated as an initial value of the alarm threshold.

  In step S5, the drowsiness level detection unit 16 acquires the steering angle detected by the steering angle sensor 2 and the yaw rate detected by the yaw rate sensor 6 as a set of drowsiness data.

Next, in step S6, the sleepiness level detection unit 16 detects the sleepiness level based on the sleepiness data acquired in step S5 and the learning data stored in the storage unit 24 in step S3.
Specifically, the sleepiness level detection unit 16 calculates the Mahalanobis distance of the sleepiness data acquired in step S5 for the learning data that is a set of sleepiness data acquired during the learning time, and calculates the Mahalanobis distance. The sleepiness level.

  Next, in step S7, the alarm output unit 18 determines whether or not the drowsiness level detected in step S6 has exceeded the set alarm threshold value (that is, the detected drowsiness level is not less than the alarm threshold value). It is determined whether or not the value has increased to a value larger than the threshold value. As a result, when the drowsiness level exceeds the alarm threshold value, the process proceeds to step S <b> 8, and the alarm output unit 18 outputs an alarm from the output device 10.

  When the drowsiness level does not exceed the alarm threshold value in step S7, or after outputting an alarm from the output device 10 in step S8, the process proceeds to step S9, and the alarm threshold value correcting unit 22 corrects the set alarm threshold value. The alarm threshold value correction process is executed.

After executing the warning threshold value correction process, the process proceeds to step S10, where the warning output unit 18 determines whether or not the ignition of the vehicle 1 is turned off. As a result, if the ignition is not turned off, the process returns to step S5. Thereafter, the snooze driving prevention device 8 repeats the processes of steps S5 to S9 until the ignition of the vehicle 1 is turned off.
On the other hand, if the ignition of the vehicle 1 is turned off in step S10, the dozing operation prevention device 8 ends the dozing operation prevention process.

  Next, the alarm threshold value correction process will be described with reference to FIG. As shown in FIG. 3, when the warning threshold value correction process is started, in step S21, the warning threshold value correction unit 22 is in a state where a drowsiness level higher than the set warning threshold value is detected in step S6. It is determined whether it continues over 1st determination time T1 (for example, 20 minutes) longer than learning time.

As a result, when the state in which the drowsiness level higher than the set alarm threshold is detected in step S6 continues for the first determination time T1, the process proceeds to step S22, and the alarm threshold correction unit 22 The threshold is corrected upward. For example, the alarm threshold value correction unit 22 multiplies the set alarm threshold value by a correction coefficient larger than 1, or adds a predetermined correction value to the set alarm threshold value.
After correcting the alarm threshold value upward in step S22, the alarm threshold value correction unit 22 ends the alarm threshold value correction process and returns to the main routine.

  In step S21, when the state in which the drowsiness level higher than the set alarm threshold is detected in step S6 does not continue for the first determination time T1, the process proceeds to step S23 and the alarm threshold correction unit 22 is performed. Is the number of times the alarm is output from the output device 10 during the second determination time T2 (for example, 20 minutes) longer than the learning time (that is, the detected drowsiness level exceeds the set alarm threshold) It is determined whether or not the number of times is equal to or greater than a predetermined number of times of determination (for example, 3 times).

  As a result, when the number of times the alarm is output from the output device 10 during the second determination time T2 is equal to or greater than the predetermined determination number, the process proceeds to step S22, and the alarm threshold correction unit 22 corrects the alarm threshold upward. To do. Thereafter, the alarm threshold value correction unit 22 ends the alarm threshold value correction process and returns to the main routine.

  On the other hand, if the number of times that the alarm is output from the output device 10 during the second determination time T2 is not greater than or equal to the predetermined number of determinations (less than the number of determinations) in step S23, the process proceeds to step S24 to correct the alarm threshold value. The unit 22 determines whether or not the total number of times that the alarm is output from the output device 10 after the vehicle 1 starts traveling is less than three.

  As a result, when the total number of times the alarm is output from the output device 10 after the vehicle 1 starts traveling is less than 3, the process proceeds to step S25, where the alarm threshold value correction unit 22 learns the learning time after the vehicle 1 starts traveling. An average value S of drowsiness levels detected in step S6 until a longer third determination time T3 (for example, 1 hour) elapses is calculated.

  Next, proceeding to step S26, the alarm threshold value correction unit 22 has a drowsiness level detected at step S6 smaller than the alarm threshold value, and adds a predetermined proportionality coefficient (for example, 2) to the average value S calculated at step S25. It is determined whether or not it is larger than a correction determination reference value that is a multiplied value.

  As a result, when the drowsiness level detected in step S6 is smaller than the alarm threshold value and larger than the correction determination reference value, the process proceeds to step S27, where the alarm threshold value correction unit 22 has a drowsiness level higher than the correction determination reference value. It is determined whether or not the total time detected in step S6 is equal to or longer than a fourth determination time T4 (for example, 20 minutes) longer than the learning time.

As a result, when the total time in which the drowsiness level higher than the correction determination reference value is detected in step S6 is equal to or longer than the fourth determination time T4, the process proceeds to step S28, and the alarm threshold value correction unit 22 lowers the alarm threshold value. To correct. For example, the alarm threshold value correction unit 22 multiplies the set alarm threshold value by a correction coefficient smaller than 1, or subtracts a predetermined correction value from the set alarm threshold value.
After correcting the alarm threshold value downward in step S28, the alarm threshold value correction unit 22 ends the alarm threshold value correction process and returns to the main routine.

  In step S24, when the total number of times the alarm is output from the output device 10 after the start of traveling of the vehicle 1 is not less than three (three or more), the drowsiness level is smaller than the alarm threshold in step S26. And when it is not larger than the correction determination reference value (the drowsiness level is not less than the alarm threshold value or not more than the correction determination reference value), or in step S27, the total time when the drowsiness level higher than the correction determination reference value is detected, If it is not equal to or longer than the fourth determination time T4 (less than the fourth determination time T4), the alarm threshold correction unit 22 ends the alarm threshold correction processing without correcting the alarm threshold, and returns to the main routine.

Here, a situation where the alarm threshold value is corrected by the alarm threshold value correction unit 22 will be described with reference to FIGS. 4 and 5.
FIG. 4 is a diagram showing a temporal change in drowsiness level that is referred to when the alarm threshold value correction unit 22 determines whether to correct the alarm threshold value upward in the alarm threshold value correction process, and FIG. It is a diagram which shows the time change of the drowsiness level referred when the warning threshold value correction | amendment part 22 determines whether a warning threshold value is correct | amended below in a correction | amendment process. 4 and 5, the horizontal axis represents time, and the vertical axis represents the detected sleepiness level. 4 and 5, the solid line indicates the drowsiness level, the broken line indicates the alarm threshold value before correction, and the alternate long and short dash line indicates the alarm threshold value after correction.

  For example, in the example of FIG. 4A, after the drowsiness level exceeds the alarm threshold, the drowsiness level remains higher than the alarm threshold for the first determination time T1 (the determination in step S21 is “ YES ”). In this situation, since the driver continues driving for the first determination time T1 even after the alarm is output because the sleepiness level exceeds the alarm threshold, the driver does not feel so much sleepiness or fatigue. there is a possibility. In that case, if the alarm output is repeated, the driver feels bothersome. Therefore, the alarm threshold value correction unit 22 corrects the alarm threshold value upward to reduce the frequency with which the alarm is output (step S22).

  In the example of FIG. 4B, the number of times the alarm is output from the output device 10 during the second determination time T2 is three (the determination in step S23 is “YES”). In this situation, the driver continues driving despite the alarm being output three times during the second determination time T2, so the driver may not feel so much sleepiness or fatigue. . In that case, if the alarm output is repeated, the driver feels bothersome. Therefore, the alarm threshold value correction unit 22 corrects the alarm threshold value upward to reduce the frequency with which the alarm is output (step S22).

  On the other hand, in the example of FIG. 4C, the drowsiness level is not maintained higher than the alarm threshold over the first determination time T1 (the determination in step S21 is “NO”), and the second determination time. The number of times the alarm is output from the output device 10 during T2 is two (the determination in step S23 is “NO”). In this situation, the driver is unlikely to feel alarming, and the alarm threshold correction unit 22 does not correct the alarm threshold upward.

  In each example of FIGS. 5A, 5B, and 5C, the drowsiness level is smaller than the alarm threshold value, and the correction determination reference value (the third determination time T3 after the vehicle 1 starts traveling). The total time of the state in which the sleepiness level is higher than the average value S of the sleepiness level detected until the time elapses is equal to or longer than the fourth determination time T4 (the determination in step S27 is “YES”) ). In this situation, the drowsiness level of the driver is greatly increased compared to the drowsiness level until the third determination time T3 elapses after the vehicle 1 starts to travel, and it is possible to feel drowsiness and fatigue. Despite this, no alarm is output. Therefore, the alarm threshold value correcting unit 22 corrects the alarm threshold value downward to increase the frequency at which the alarm is output (step S28).

  Next, the operational effects of the above-described dozing operation prevention device 8 of the present embodiment will be described.

  First, the warning threshold value setting unit 20 sets a warning threshold value based on the steering angle and the yaw rate until a predetermined learning time elapses after the vehicle 1 starts running, and the warning threshold value correction unit 22 sets the warning threshold value. After setting, since the set alarm threshold value is corrected based on the set alarm threshold value and the detected drowsiness level, the steering angle and the yaw rate immediately after the start of driving that the driver does not feel drowsiness or fatigue are considered. Can be set as a reference, and the alarm threshold can be corrected so that an alarm is output at an appropriate frequency according to the relationship between the actually detected drowsiness level and the alarm threshold, As a result, the driver's face state and biometric information are not required, and an alarm is output at an appropriate timing without increasing the driver's operation burden. It is possible to prevent.

  In particular, the alarm threshold correction unit 22 corrects the alarm threshold upward when a drowsiness level higher than the set alarm threshold is continuously detected over the first determination time T1 longer than the learning time. If the driver does not feel so much sleepiness or fatigue and the driver continues driving for the first determination time T1 even after the alarm is output because the sleepiness level exceeds the alarm threshold, the alarm threshold is increased upward. Correction can be made to reduce the frequency at which the alarm is output. This makes it possible to prevent the driver from falling asleep by outputting an alarm at an appropriate timing without making the driver feel bothersome.

  In addition, the alarm threshold correction unit 22 generates an alarm when the number of times that the detected drowsiness level exceeds the set alarm threshold during the second determination time T2 longer than the learning time is equal to or greater than the predetermined determination number. Since the threshold value is corrected upward, the driver does not feel so much sleepiness or fatigue, and the driving is continued even though the alarm is output more than the determination times during the second determination time T2. The alarm threshold can be corrected upward to reduce the frequency at which the alarm is output, thereby preventing the driver from falling asleep by outputting an alarm at an appropriate timing without making the driver feel bothersome be able to.

  Further, the warning threshold value correction unit 22 is a correction determination that is twice the average value S of the sleepiness level detected until the third determination time T3 longer than the learning time elapses after the vehicle 1 starts traveling. A reference value is calculated, and a fourth determination in which the total time in which the calculated drowsiness level is smaller than the set alarm threshold and higher than the correction determination reference value is longer than the learning time is calculated. When the time T4 is equal to or greater than the time T4, the alarm threshold value is corrected downward, so that the driver's drowsiness level increases significantly compared to the drowsiness level until the third determination time T3 elapses after the vehicle 1 starts traveling. If the state is maintained for the fourth determination time T4 or more in total and there is a high possibility that the driver feels drowsiness or fatigue, the alarm threshold is corrected downward and the frequency of the alarm is output. Can be raised Accordingly, it is possible to prevent drowsy driving the driver by outputting an alarm at the right time.

Finally, a modification of the snooze driving prevention device 8 will be described.
In the above-described embodiment, the steering angle and the yaw rate are used as the sleepiness data for detecting the sleepiness level. However, information related to other driving operations and / or the behavior of the vehicle 1 (for example, brake pedal force, vehicle speed, acceleration / deceleration) , Yaw acceleration, etc.) may be used as sleepiness data.

DESCRIPTION OF SYMBOLS 1 Vehicle 2 Steering angle sensor 4 Vehicle speed sensor 6 Yaw rate sensor 8 Drowsiness prevention device 10 Output device 12 Display 14 Speaker 16 Drowsiness level detection part 18 Alarm output part 20 Alarm threshold value setting part 22 Alarm threshold value correction part 24 Storage part

Claims (4)

A drowsiness prevention device that prevents a driver from falling asleep,
Drowsiness level detection means for detecting the drowsiness level of the driver based on the driving operation by the driver and / or the behavior of the vehicle;
An alarm output means for outputting an alarm when the sleepiness level detected by the sleepiness level detection means is higher than a predetermined alarm threshold;
An alarm threshold value setting means for setting the alarm threshold value based on the driving operation and / or the behavior of the vehicle after a predetermined learning time elapses after the vehicle starts to travel;
After setting the alarm threshold value, the drowsiness driving prevention apparatus, comprising: an alarm threshold value correcting unit that corrects the set alarm threshold value based on the set alarm threshold value and the detected drowsiness level. .   The alarm threshold value correcting means corrects the alarm threshold value upward when a drowsiness level higher than the set alarm threshold value is continuously detected over a first determination time longer than the learning time. The snooze driving prevention device described in 1.   When the number of times that the detected drowsiness level exceeds the set alarm threshold during a second determination time longer than the learning time is greater than or equal to a predetermined determination number, The snooze driving prevention device according to claim 1 or 2, wherein the threshold value is corrected upward.   The alarm threshold value correction means calculates a correction determination reference value that is proportional to the average value of the sleepiness level detected during the third determination time that is longer than the learning time after the vehicle starts running, The total time during which the calculated correction determination reference value is smaller than the set alarm threshold and a drowsiness level higher than the correction determination reference value is detected is equal to or longer than a fourth determination time longer than the learning time. 4. If there is, the drowsiness prevention device according to any one of claims 1 to 3, wherein the alarm threshold value is corrected downward.

Images