JP5084701B2 - Unconscious driving detection device, driving support device, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately determine the consciousness deteriorating state of a driver. <P>SOLUTION: A frequency analyzing part 18 calculates a power value f<SB>i</SB>of each frequency band i from the time series data s of a steering angle signal input from a steering angle sensor 12, and a total power value calculation part 22 reads weighting factor w<SB>i</SB>showing a large value from a weighting factor DB 20 according as the level of difference between a power value ld<SB>i</SB>in a frequency band i when driver's consciousness deteriorates and a power value in the other frequency band and the level of difference between a power value when the driver's consciousness deteriorates and a power value when the driver is awakened in the same frequency band is large, and multiplies the power value f<SB>i</SB>of each frequency band i by the weighting factor w<SB>i</SB>, and calculates a total power value p by adding the value of w<SB>i</SB>&times;f<SB>i</SB>of each frequency band for prescribed frequency bands, and a determination part 24 calculates a cumulative value (a) of a total power value p, and determines whether or not the cumulative value (a) is larger than a prescribed value to determine the consciousness deteriorating state of the driver. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to a low consciousness driving detection device, a driving support device, and a program, and in particular, a low consciousness driving detection device and a driving support device that determine a driver's low consciousness state based on a vehicle running state and a driver steering state. And programs.

Conventionally, data based on the driving state of the vehicle and the operation state of the driver is detected, based on the detected data, it is determined whether or not the driver is in a state of reduced consciousness, and an alarm is issued to the driver. Assisting the driver's driving has been done. For example, in the driving concentration determination apparatus of Patent Document 1, a value corresponding to a power value near 1 Hz is calculated as a middle value by combining two moving average processes for a steering angle signal detected by a steering angle sensor, and the middle value is calculated. It has been proposed to determine the degree of driving concentration based on the magnitude of a determination value obtained by adding a value for a predetermined time.
Japanese Patent Laid-Open No. 2007-026271

  However, the technique described in Patent Document 1 determines a value corresponding to data in a frequency band centered on 1 Hz on the assumption that a measurement signal in a frequency band centered on 1 Hz is easily detected immediately before going to sleep. However, as shown in FIG. 12, the frequency band that is easily detected when the consciousness is reduced and the frequency band where the difference in power value is large between the consciousness lowered time and the normal time depends on the environment such as vehicle characteristics and driving conditions. Different. The technique described in Patent Document 1 does not take into consideration how much such a decrease in concentration (decrease in consciousness) affects which frequency band, and accurately determines the driver's consciousness decrease state. There was a problem that I could not do it.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a low-consciousness driving detection device, a driving support device, and a program that can accurately determine a driver's low-consciousness state. .

  In order to achieve the above object, the consciousness reduction driving detection apparatus according to the present invention continuously detects a physical quantity indicating a running state of the own vehicle or a steering state when the driver steers the own vehicle, and at the time of the detected physical quantity. Detection means for outputting series data, frequency analysis means for performing frequency analysis of the time series data of the physical quantity output by the detection means, and calculating a frequency distribution indicated by a distribution of power values for each frequency band, and consciousness A value obtained by multiplying the corresponding power value for each frequency band by the weight coefficient for each frequency band calculated so as to increase as the degree of difference between the frequency distribution at the time of decrease and the frequency distribution at the time of awakening increases. Total power value calculating means for calculating the total power value over a predetermined frequency band as a total power value, and the total power value calculated by the total power value calculating means within a predetermined time There has been configured to include by determining more than a predetermined value, a determination unit drowsiness of the driver, the.

  In addition, the low consciousness driving detection program according to the present invention continuously detects a physical quantity indicating a running state of the host vehicle or a steering state when the driver steers the host vehicle, and time-series data of the detected physical quantity. Frequency analysis means for calculating the frequency distribution indicated by the power value distribution for each frequency band by performing frequency analysis of the time-series data of the physical quantity output by the detection means for outputting the frequency distribution when the consciousness is reduced The weight coefficient for each frequency band calculated so as to become heavier as the degree of difference from the frequency distribution at the time of awakening increases, and a value obtained by multiplying the corresponding power value for each frequency band is added over a predetermined frequency band A total power value calculating means for calculating the sum as a total power value, and a sum of the total power values calculated by the total power value calculating means within a predetermined time is less than a predetermined value; By determining, it is intended to function as a determination unit drowsiness of the driver.

  According to the low consciousness driving detection apparatus and the low consciousness driving detection program according to the present invention, the detection means continuously detects a physical quantity indicating a running state of the own vehicle or a steering state when the driver steers the own vehicle, The time-series data of the detected physical quantity is output, and the frequency analysis means performs a frequency analysis of the output time-series data of the physical quantity, and calculates a frequency distribution indicated by a power value distribution for each frequency band. Then, the total power value calculation means calculates the weight coefficient for each frequency band calculated so as to increase as the difference between the frequency distribution at the time of consciousness reduction and the frequency distribution at the time of awakening increases. The sum of the value multiplied by the power value added over a predetermined frequency band is calculated as the total power value, and the judgment means determines whether the sum of the total power values within a predetermined time is greater than or equal to the predetermined value, thereby reducing driver awareness. Determine the state.

  As described above, the consciousness decline is based on the sum of the total power values within a predetermined time obtained as the sum of the values multiplied by the weighting factors based on the difference in frequency distribution between the consciousness decline and the wakefulness of the corresponding frequency band. By determining whether or not the frequency band affected by the lowering of consciousness and the degree of the influence are taken into consideration, the driver's consciousness lowered state can be accurately determined.

  The consciousness reduction driving detection apparatus according to the present invention continuously detects a physical quantity indicating a running state of the own vehicle or a steering state when the driver steers the own vehicle, and outputs time-series data of the detected physical quantity. Detecting means, frequency analyzing means for performing frequency analysis of the time-series data of the physical quantity output by the detecting means to calculate a frequency distribution indicated by a power value distribution for each frequency band, and the frequency at the time of consciousness decline Each frequency band calculated so as to increase as the difference between the distribution and the frequency distribution at the time of awakening obtained from the initial time-series data of the physical quantity output by the detection means within a predetermined time from the start of traveling increases. The total power value calculation means for calculating the total power value by adding a value obtained by multiplying the corresponding power value for each frequency band by the weighting factor of the frequency band over a predetermined frequency band. And determining means for determining whether the driver's consciousness declined state by determining whether the sum of the total power values calculated by the total power value calculating means within a predetermined time is equal to or greater than a predetermined value. ing.

  In addition, the low consciousness driving detection program of the present invention continuously detects a physical quantity indicating a running state of the own vehicle or a steering state when the driver steers the own vehicle, and time-series data of the detected physical quantity is obtained. Perform frequency analysis of the time-series data of the physical quantity output by the output detection means, calculate the frequency distribution indicated by the power value distribution for each frequency band, the frequency distribution at the time of consciousness decline, Weighting coefficient for each frequency band calculated so as to become heavier as the degree of difference from the awakening frequency distribution obtained from the initial time-series data of the physical quantity output by the detection means within a predetermined time from the start of traveling increases The total power value calculation is calculated as a total power value by adding a value obtained by multiplying the corresponding power value for each frequency band over a predetermined frequency band. And determining whether the sum of the total power values calculated by the total power value calculation means within a predetermined time is equal to or greater than a predetermined value, thereby functioning as a determination means for determining the driver's consciousness reduction state It is.

  According to the low consciousness driving detection apparatus and the low consciousness driving detection program according to the present invention, the total power value calculation means calculates the frequency distribution at the time of consciousness drop and the physical quantity output by the detection means within a predetermined time from the start of running. The frequency obtained by multiplying the power value for each corresponding frequency band by the weighting factor for each frequency band calculated so that it becomes heavier as the degree of difference from the awakening frequency distribution obtained from the initial time series data increases. The total added over the band is calculated as the total power value.

  In this way, since the total power value is calculated based on the weighting coefficient calculated using the time series data of the actually detected physical quantity, the influence of the environment such as the vehicle characteristics and the driving state is taken into consideration, and the driver's It is possible to accurately determine the state of reduced consciousness.

  Further, the weighting factor of the present invention can be calculated based on the information amount of the band. In order to calculate the weighting factor based on the information amount of the band, the weighting factor can be calculated according to the following equation.

Here, pd _i is a band probability obtained by dividing the power value of the frequency band i in the frequency distribution at the time of consciousness reduction by the sum of the power values of the predetermined frequency band in the frequency distribution at the time of consciousness reduction, and pa _i is the wake-up time. This is a band probability obtained by dividing the power value of the frequency band i in the frequency distribution by the sum of the power values of the predetermined frequency band in the frequency distribution at awakening.

  Further, the weighting factor of the present invention can be calculated according to the following formula.

However, μ _di is an average value of power values in the frequency band i in the frequency distribution at the time of consciousness decrease, σ _di is a dispersion value of power values in the frequency band i in the frequency distribution at the time of consciousness decrease, and μ _ai is at the time of awakening An average value of power values in the frequency band i in the frequency distribution, σ _ai is a variance value of the power values in the frequency band i in the frequency distribution at the time of awakening.

  Further, the predetermined value is set to an average value of the sum within a predetermined period from the start of traveling, or a cumulative period when the driving state is determined to be stable based on the vehicle speed and yaw rate of the vehicle is determined in advance. It can be an average value of the sums within the cumulative period when the period is exceeded. Further, it may be started when the continuous period when it is determined that the driving state is stable based on the vehicle speed and the yaw rate of the vehicle becomes equal to or longer than a predetermined period. In addition, when the vehicle speed of the vehicle is smaller than a predetermined value and the yaw rate of the vehicle is larger than a predetermined value, it can be determined that the driving state is not stable. Thus, by determining the predetermined value for determining whether or not the state of consciousness is lowered based on the total power value calculated from the actually detected data, the magnitude depends on the vehicle characteristics and the individuality of the driver. It is possible to determine a state of reduced consciousness with higher accuracy for different power values.

  Furthermore, the predetermined value can be corrected based on the vehicle speed and the curve size within a predetermined period from the start of traveling.

  Further, the predetermined value can be calculated based on at least one value among a vehicle speed, a yaw rate, a yaw angle, and a lateral acceleration. The predetermined value can also be calculated according to the following formula.

However, thr ₀ is the predetermined value initially set, v is the current vehicle speed, v ₀ is the average value of the vehicle speed within a predetermined period from the start of travel, y is the yaw rate of the current vehicle, and y ₀ is The average value of the yaw rate within a predetermined period from the start of travel, θ is the yaw angle obtained by time integration of the yaw rate, and θ ₀ is the average value of the yaw angle. In particular, when the vehicle speed is high or the curve of the road is large, it has been experimentally observed that the power value is large, and a predetermined value for determining whether or not the state of consciousness is lowered. By increasing the values when the vehicle speed and the size of the curve are larger than the average value, it is possible to determine the state of reduced consciousness with higher accuracy. Note that y cos θ is an index that indicates the magnitude of the curve of the traveling road represented by the yaw rate and the yaw angle.

  Further, a driving support device according to the present invention includes the above-described consciousness reduction driving detection device, and support means for performing support for returning to a wakeful state when a consciousness reduction state is detected by the determination means. It is configured.

  In the driving support device of the present invention, when the driver operates a device other than the steering provided in the vehicle within the predetermined time, or when the driver performs an abrupt accelerator operation, the vehicle speed is predetermined. The apparatus may further include support stop means for stopping the support provided by the support means when the speed is lower than the speed or when another device for supporting driving is activated. Under such circumstances, there is a high possibility that an erroneous determination will occur in the determination of the consciousness lowering state. Therefore, the support can be stopped by the support stop means, and erroneous support can be prevented. Note that other devices for assisting driving include, for example, a lane keeping system and auto cruise control.

  As described above, according to the low-consciousness driving detection device, the driving support device, and the program of the present invention, the frequency band affected by the low-consciousness and the degree of the influence are taken into consideration, so the driver's low-consciousness state can be accurately determined. The effect that it can be determined is obtained.

  Hereinafter, with reference to the drawings, an embodiment in which the present invention is applied to a consciousness reduction driving detection apparatus mounted on a vehicle will be described in detail.

  As shown in FIG. 1, the consciousness reduction driving detection apparatus 10 according to the first embodiment determines a consciousness reduction state based on a steering angle sensor 12 that detects a steering angle and an output from the steering angle sensor 12. A computer 14 is provided that executes a processing routine and causes the alarm device 16 to output an alarm to the driver according to the determination result.

  The steering angle sensor 12 detects the steering angle of the steering, generates a signal corresponding to the steering angle, and outputs it to the computer 14 as time series data.

  The computer 14 is a CPU that controls the entire consciousness lowering operation detection apparatus 10, a ROM as a storage medium that stores a program for processing a consciousness lowering state detection that will be described later, a RAM that temporarily stores data as a work area, and these It is comprised including the bus which connects.

  If the computer 14 is described by function blocks divided for each function realizing means determined based on hardware and software, as shown in FIG. 1, time-series data corresponding to the steering angle input from the steering angle sensor 12 is used. A frequency analysis unit 18 that performs frequency analysis of the frequency, a total power value calculation unit 22 that calculates a total power value based on the analysis result of the frequency analysis unit 18 and the weighting factor for each frequency band stored in the weighting factor DB 20, and A configuration including a determination unit 24 that determines whether the driver is in a state of reduced consciousness based on the calculation result of the total power value calculation unit 22 and controls the alarm device 16 so that an alarm is output based on the determination result. Can be expressed as

The frequency analysis unit 18 performs high-pass filtering processing on the time-series data of the steering angle within a predetermined time input from the steering angle sensor 12 in order to remove the influence on the steering angle frequency component due to the curve, A discrete Fourier transform is performed to calculate a power value f _i (frequency distribution) for each frequency band i of the steering frequency.

The total power value calculation unit 22, by the power value f _i for each frequency band i calculated (1) in the frequency analysis unit 18, calculates the total power value p.

  Here, n is a predetermined constant and represents that the total power value p is obtained in a predetermined frequency band from frequency bands 1 to n. Since the predetermined frequency band needs to be a frequency band corresponding to a frequency band in which a weighting coefficient described later is calculated, a minimum value to a maximum value of the steering frequency at the time of consciousness lowering and awakening are obtained in advance by experiments or the like. This range is set as a predetermined frequency band. It is desirable that the set predetermined frequency band corresponds to substantially the entire frequency band of the steering frequency that is actually detected.

Further, w _i is a weighting factor for each frequency band, and a value calculated in advance by equation (2) is stored in the weighting factor DB 20.

Here, pd _i is obtained by dividing the power value ld _i of the frequency band i in the frequency distribution when the consciousness is lowered by the sum of the power values of the predetermined frequency band from the frequency band 1 to n, and pa _i is the wake-up time. The power value la _i of the frequency band i in the frequency distribution is divided by the sum of the power values of the predetermined frequency band from the frequency band 1 to n. The power value ld _i and the power value la _i are values calculated as the average of the power values in the same frequency band with respect to the frequency distribution for the predetermined number of times obtained by detecting the time-series data of the predetermined number of physical quantities. .

Equation (2) is a KL (Kullbuck-Leibler) that symmetrizes the difference between probability distributions when the consciousness is lowered and when aroused when the shape of the frequency distribution (average spectrum) of the average power value is regarded as a discrete probability distribution. ) Equivalent to the amount of information calculated. As shown in FIG. 2, this is because the difference between the power value ld _i in the frequency band i and the power value in the other frequency band when the consciousness is lowered, and the power value ld _i when the consciousness is lowered in the same frequency band i. And the power value la _i at the time of awakening, the larger the difference, the larger the amount of information.

  Therefore, the value is large in a frequency band having a large power value compared to other frequency bands, or in a frequency band in which the difference between the power value when the consciousness is lowered and the power value when awakening is large. That is, the frequency band weighting coefficient expressed by the equation (2) becomes heavy.

  The weighting coefficient DB 20 is a built-in or external storage unit such as a hard disk drive (HDD) or a CD-ROM, and is composed of a medium that can store a weighting coefficient for each frequency band.

  The determination unit 24 stores the total power value p calculated by the total power value calculation unit 22 in a FIFO buffer (not shown) having a predetermined size, and calculates a cumulative value a of the stored total power value p. The FIFO type buffer has a size capable of storing a predetermined number (for example, 6) of total power values p, and updates the oldest data when storing the predetermined number + 1 of the total power values p. Further, when the storable size is converted into time, the detection time (for example, 10 seconds) of the time series data s of the steering angle used to calculate the predetermined number × total power value p is, for example, 6 × 10 seconds = 60 seconds. It can be memorized. And the determination part 24 determines whether the accumulated value a of the total power value p is more than predetermined value, and controls the alarm device 16 so that an alarm is output by the alarm device 16 based on the determination result.

  Next, with reference to FIG. 3, a processing routine for detection of a reduced consciousness state according to the first embodiment will be described.

  In step 100, the time-series data s of the steering angle signal is input from the steering angle sensor 12 to the frequency analysis unit 18.

Next, at step 102, the time-series data s of the steering angle is subjected to a high-pass filtering process to obtain s ′. Next, at step 104, a discrete Fourier transform is applied to s ′ to obtain a power value f for each frequency band i. _i is calculated.

Next, in step 106, the calculated power value f _i for each frequency band i is read from the weighting coefficient DB 20 and multiplied by the weighting coefficient for each corresponding frequency band i, and w _i × f _i for each frequency band is multiplied. The total power value p is calculated by adding the values over a predetermined frequency band (frequency bands 1 to n) (equation (1)).

  Next, in step 108, the calculated total power value p is stored in the FIFO type buffer. Next, in step 110, the cumulative value a of the total power value p stored in the FIFO type buffer is calculated, and it is determined whether or not the cumulative value a is larger than a predetermined value. If it is equal to or less than the predetermined value, it is determined that the consciousness is not yet lowered, and the process returns to step 100 to repeat the process. If it is greater than the predetermined value, it is determined that the state of consciousness has been lowered, and the routine proceeds to step 112 where an alarm is issued from the alarm device 16 to the driver.

  As described above, according to the consciousness reduction driving detection apparatus 10 according to the first embodiment, in order to make a determination based on the total power value in consideration of the frequency band affected by the consciousness reduction and the degree of the influence, The driver's consciousness deterioration state can be determined with high accuracy, and since the determination is always made based on the cumulative value of the total power value for the latest predetermined time, the driver can be alerted with good timing and driving assistance with high accuracy. It can be performed.

  Next, a second embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted. As shown in FIG. 1, the configuration of the second embodiment is the same as the configuration of the first embodiment. However, the calculation method of the weighting factor for each frequency band stored in the weighting factor DB 20 is the first. This is different from the first embodiment.

In the consciousness reduction driving detection apparatus 10 in the second embodiment, the weighting coefficient w _{i for} each frequency band calculated in advance by the equation (3) is stored in the weighting coefficient DB 20.

Here, as shown in FIG. 4, μ _di is the average value of the power values in the frequency band i in the frequency distribution at the time of consciousness decrease, and σ _di is the variance of the power values in the frequency band i in the frequency distribution at the time of consciousness decrease. The value, μ _ai, is an average value of the power values in the frequency band i in the frequency distribution at awakening, and σ _ai is a variance value of the power values in the frequency band i in the frequency distribution at the time of awakening.

In the frequency band i, the probability distribution of the steering frequency at the time of consciousness decline is pd _i (x), and the probability distribution of the steering frequency at the time of waking is pa _i (x), and the probability distribution pd _i (x) and the probability distribution pa _i ( When the degree of difference from x) is expressed by the amount of KL information, equation (4) is obtained.

Assuming that the probability distribution pd _i (x) and the probability distribution pa _i (x) are normal distributions, the expression (4) is expanded and symmetrized to obtain the expression (3). As shown in FIG. 4, the weighting coefficient represented by the expression (3) includes the degree of variation in the power value of the frequency band i in the frequency distribution at the time of consciousness decrease and the variation in power value at the time of consciousness decrease in the same frequency band i. The higher the degree of difference between the degree and the degree of variation in power value at awakening, the heavier.

  As described above, according to the reduced consciousness operation detection apparatus in the second embodiment, the determination is made based on the total power value in consideration of the variation of the power value for each frequency band. Can be determined.

  Next, a third embodiment will be described. In addition, about the part used as the structure and process similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted. The third embodiment is different from the first embodiment in that the weighting factor is calculated using the frequency distribution of the steering frequency detected within a predetermined time from the start of traveling.

  As shown in FIG. 5, the consciousness reduction driving detection apparatus 210 in the third embodiment is functionally stored in the frequency analysis unit 18, the analysis result of the frequency analysis unit 18, and the frequency distribution (ld) DB 220. A weighting factor calculation unit 226 that calculates a weighting factor for each frequency band based on the frequency distribution of the steering frequency when the consciousness is lowered, an analysis result of the frequency analysis unit 18, and a weight for each frequency band calculated by the weighting factor calculation unit 226 The computer 214 includes a total power value calculation unit 22 that calculates a total power value based on the coefficients, and a determination unit 24.

  The frequency analysis unit 18 calculates a frequency distribution for the initial steering frequency based on the time-series data of the steering angle detected within a predetermined time from the start of traveling. Assuming that the driver is not in a state of reduced consciousness at the beginning of the predetermined time from the start of driving, the frequency distribution of the steering frequency calculated based on the time-series data of the steering angle acquired during this period is calculated at the time of awakening. This is a frequency distribution.

Based on the frequency distribution at the time of awakening calculated by the frequency analysis unit 18 and the frequency distribution at the time of lowering of consciousness stored in the frequency distribution (ld) DB 220, the weight coefficient calculation unit 226 calculates the above equation (2) or ( 3) The weighting coefficient w _i for each frequency band _i is calculated from the equation.

  Note that a combination of the weighting factor calculation unit 226 and the total power value calculation unit 22 of the third embodiment corresponds to the total power value calculation unit of claim 2.

  Next, with reference to FIG. 6, a processing routine for detecting a reduced consciousness state according to the third embodiment will be described.

In step 300, the counter value t is set to 1. Next, in step 100, the time series data s of the steering angle signal is obtained from the steering angle sensor 12, and then in step 102, the time series data s of the steering angle is obtained. Is subjected to high-pass filtering processing to obtain s ′, and then in step 104, s ′ is subjected to discrete Fourier transform to calculate a power value f _i (frequency distribution) for each frequency band i.

  Next, in step 302, it is determined whether or not the counter value t is larger than a predetermined value T. When it is larger than the predetermined value T, the routine proceeds to step 106, and when it is equal to or smaller than the predetermined value T, the routine proceeds to step 304. The predetermined value T is set based on a predetermined time from the start of running necessary to obtain a frequency distribution at awakening. That is, the detection time × T of the time-series data s of the steering angle signal is set to be a predetermined time from the start of traveling.

In step 304, it is determined whether or not the counter value t is T. When t is smaller than T, the predetermined time from the start of running necessary to obtain the frequency distribution at the time of awakening has not elapsed, so the determination is negative and the routine proceeds to step 306 where the frequency band i calculated at step 104 is determined. Each power value f _i (f (t = 1)) is stored in a frequency buffer (not shown) which is a predetermined storage area.

  Next, in step 312, the counter value t is incremented, the process returns to step 100, and the processes of steps 100 to 104, steps 302 to 306, and step 312 are repeated until t = T.

  If it is determined in step 304 that t = T, the process proceeds to step 308, where f (t = 1) to f (t = T) stored in the frequency buffer is used as the initial frequency distribution at awakening. The frequency distribution of the steering frequency is calculated. Specifically, when calculating the weighting coefficient based on the equation (2), as shown in FIG. 2, an average spectrum of f (t = 1) to f (t = T) is calculated, and (3 ), When calculating the weighting coefficient, f (t = 1) to f (t = T) are calculated as a distribution as shown in FIG.

Next, in step 310, based on the frequency distribution of the initial steering frequency calculated in step 308 and the frequency distribution at the time of consciousness decrease stored in the frequency distribution (ld) DB, the expression (2) or ( 3) The weighting coefficient w _i for each frequency band _i is calculated from the equation.

  Next, the processing of step 100 to step 104 is repeated through step 312, and when the process proceeds to step 302, since t becomes larger than the predetermined value T, the determination is affirmed and the process proceeds to step 106. In step 106 to step 112, the determination of the consciousness reduction state and the warning processing to the driver are performed by the same processing as in the first embodiment. At this time, the weighting factor calculated in step 310 is used to calculate the total power value p used for determination of consciousness decrease.

  As described above, according to the consciousness reduction driving detection apparatus in the third embodiment, in order to calculate the weighting factor for determining the consciousness reduction state based on the steering angle data detected by actual traveling. Individual differences are also considered in the determination, and the driver's consciousness reduction state can be determined with higher accuracy.

  Next, a fourth embodiment will be described. In addition, about the part used as the structure and process similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted. In the fourth embodiment, the first embodiment is that a value of a predetermined value for determining whether or not the consciousness is lowered is set using a power value calculated within a predetermined time from the start of traveling. Is different.

  As shown in FIG. 7, the consciousness reduction driving detection apparatus 410 in the fourth embodiment detects a steering angle sensor 12, a computer 414, an alarm device 16, a vehicle speed sensor 418 for detecting the vehicle speed, and a vehicle yaw rate. The yaw rate sensor 420 is configured to be included.

  Functionally, the computer 414 includes a frequency analysis unit 18, a weight coefficient DB 20, a total power value calculation unit 22, and a determination unit 424. The determination unit 424 determines whether or not the driver is in a state of reduced consciousness based on whether or not the total power value calculated by the total power value calculation unit 22 is equal to or greater than a predetermined value. The predetermined value for determination is determined based on the total power value calculated by the total power value calculation unit 22, and is corrected based on the vehicle speed detected by the vehicle speed sensor 418 and the yaw rate detected by the yaw rate sensor 420. .

  Next, with reference to FIG. 8, a processing routine for detection of reduced consciousness according to the fourth embodiment will be described.

  In step 300, 1 is set to the counter value t. Next, the total power value p is calculated by the processing in steps 100 to 106, and then in step 500, the total power value p is stored in the FIFO type buffer.

  Next, in step 502, it is determined whether or not the counter value t is greater than a predetermined value T. If it is greater than the predetermined value T, the process proceeds to step 518, and if it is less than the predetermined value T, the process proceeds to step 504.

  In step 504, it is determined whether or not the counter value t is T. If t is smaller than T, the predetermined time from the start of running necessary to determine the predetermined value for determining the state of consciousness has not elapsed, so the determination is negative and the routine proceeds to step 506, where the FIFO type buffer is stored. It is determined whether or not a predetermined number (for example, 6) of total power values p are stored. If stored, in step 508, the cumulative value a of the total power value p in the FIFO buffer is calculated and stored in a cumulative value buffer (not shown) which is a predetermined storage area. Proceed to 516. If the predetermined number of total power values p are not stored, the process proceeds to step 516 as it is, and the counter value t is incremented and the process returns to step 100. By repeating the processing of step 506 and step 508 until t = T, the cumulative value a of a predetermined number of total power values p is stored in the cumulative value buffer.

If it is determined in step 504 that t = T, the process proceeds to step 510 to acquire the vehicle speed v (t) and yaw rate y (t) detected from the start of travel to time T. Next, in step 512, an average value of the accumulated value a of the total power value p stored in the accumulated value buffer is calculated, and a coefficient k is added to this to obtain an initial value thr ₀ of a predetermined value for determination. calculate. The coefficient k is corrected so that the initial value thr ₀ of the predetermined value does not become an unreasonably low value in consideration of the possibility of calculating the total power value p lower than usual by driving carefully at the start of traveling. Is to do. The coefficient k can be a value that is twice the standard deviation of the cumulative value a of the total power value p stored in the cumulative value buffer, for example.

Next, in step 514, the value of the predetermined initial value thr ₀ is substituted into the predetermined value thr. Next, in step 516, the counter value t is incremented and the processing returns to step 100. The processing in steps 100 to 106 and step 500 is repeated, and when the processing proceeds to step 502, t becomes larger than the predetermined value T. Then, the process proceeds to step 518.

  In step 518, the current vehicle speed v and yaw rate y are acquired. Next, in step 520, the value of the predetermined value thr is corrected by the following equation (5).

However, v ₀ is the average value of the vehicle speed v (t) from the travel start acquired at step 510 to time T, and y ₀ is the yaw rate y (t) from the travel start acquired at step 510 to time T. , Θ is a yaw angle obtained by time integration of the yaw rate y, and θ ₀ is an average value of the yaw angle θ. Note that y cos θ is an index (an index indicating the magnitude of the curve of the travel path) indicating the influence of the curve size of the travel path on the steering angle. Note that lateral acceleration may be used as an index indicating the magnitude of the curve of the travel path.

  Next, in step 110, the predetermined value thr corrected by the equation (5) is used as a threshold value to determine whether or not the consciousness is lowered.

  As described above, according to the consciousness reduction driving detection apparatus in the fourth embodiment, the consciousness reduction state is determined based on the total power value calculated from the steering angle data detected by actual traveling. In order to calculate the threshold value (predetermined value thr), for example, in the case of a driver having a characteristic of steering a large steering, the predetermined value thr becomes a large value, and in the case of a driver having a characteristic of steering a small steering, Since the criterion for determination is set according to the normal steering amount, for example, the predetermined value thr is a small value, individual differences are also considered in the determination, and the driver's state of reduced awareness can be determined more accurately. Can do. Furthermore, since the predetermined value thr is corrected based on the vehicle speed and the yaw rate in consideration of the vehicle speed that changes from moment to moment and the steering amount that changes depending on the curve of the travel path, the determination accuracy can be further improved.

In the fourth embodiment, the case where the predetermined value thr is corrected by using an index indicating the magnitude of the curve of the traveling road represented by the yaw rate and the yaw angle has been described. etc. map information, and calculates the magnitude r of the curve of the road, using the ratio between the average value r ₀ of the magnitude of the curve of the running path within a predetermined time and magnitude r of the curve of the current traveling road The predetermined value thr may be corrected.

  Next, a fifth embodiment will be described. In addition, about the part used as the structure and process similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted. The fifth embodiment is different from the first embodiment in that driving assistance is not performed under predetermined conditions.

  As shown in FIG. 9, the consciousness reduction driving detection device 610 according to the fifth embodiment includes a steering angle sensor 12, a computer 614, an alarm device 16, an accelerator sensor 618 for detecting an operation amount of an accelerator, and an operation switch such as a winker. 620, a vehicle speed sensor 622 for detecting the vehicle speed of the vehicle, and a GPS 624 for detecting the curvature of the travel path.

  The computer 614 functionally includes a frequency analysis unit 18, a weight coefficient DB 20, a total power value calculation unit 22, and a determination unit 624. The determination unit 624 controls the alarm device 16 to output an alarm when the total power value calculated by the total power value calculation unit 22 is equal to or greater than a predetermined value and a predetermined condition is satisfied.

  Next, with reference to FIG. 10, the processing routine of the consciousness fall state detection of 5th Embodiment is demonstrated.

  In step 100 to step 108, the total power value p is calculated and stored in the FIFO buffer. Next, in step 110, the accumulated value a of the total power value p stored in the FIFO buffer is a predetermined value. It is determined whether or not the state of consciousness is lowered by determining whether or not it has become larger than thr. If it is greater than the predetermined value thr, the process proceeds to step 700. If it is equal to or smaller than the predetermined value thr, the process returns to step 100 and the process is repeated.

  Next, in step 700, the speed v (t) detected within a predetermined time is acquired, and a value obtained by multiplying the speed v (t) by a weighting factor wa that differs depending on the time t is added for a predetermined time, and the vehicle speed weight is added. The attached linear sum aopr is calculated. For example, as shown in FIG. 11, the weighting coefficient is a value such that Σ | wa | = 1.0. Thus, the weighted linear sum aopr of the vehicle speed takes a large value when a sudden change occurs in the vehicle speed, and takes a small value when no sudden change occurs in the vehicle speed.

  Next, in step 702, it is determined whether or not the weighted linear sum aopr is greater than a predetermined value athr. If it is equal to or less than the predetermined value athr, it is determined that there has been no sudden change in the vehicle speed within a predetermined time. If no, the process proceeds to step 704.

  In step 704, data detected by the accelerator sensor 618 and the operation switch 620 is acquired. Next, in step 706, it is determined whether or not an accelerator operation and an operation switch such as a blinker have been operated within a predetermined time. As for the accelerator operation, in order to detect only an abrupt accelerator operation, the accelerator operation may be detected when the accelerator is depressed more than a predetermined amount. If there is no operation, it is determined that there is no operation that affects the steering within a predetermined time, and the determination is negative and the routine proceeds to step 708.

  In step 708, the current vehicle speed v is acquired. Next, in step 710, it is determined whether the vehicle speed v is smaller than a predetermined value vthr. The predetermined value vthr is a value regulated in an urban area such as 70 km / h, for example. Accordingly, when the current speed v is smaller than the predetermined value vthr, it can be determined that the vehicle is traveling in an urban area where the change in the steering amount is large. If the speed v is greater than or equal to the predetermined value vthr, it is determined that the change in the steering amount is small, such as when traveling on a highway, for example, and the result is negative and the routine proceeds to step 712.

  In step 712, the curvature r of the traveling road is calculated from the current position and map information acquired by the GPS 624. Next, in step 714, it is determined whether the curvature r of the traveling road is larger than a predetermined value rthr. If the curvature r is equal to or less than the predetermined value rthr, it is determined that the influence on the steering angle is small due to the curvature of the traveling road, and the result is negative and the routine proceeds to step 716. Note that the value of y cos θ represented by the yaw rate y and the yaw angle θ described in the fourth embodiment may be used as an index of curvature of the travel path.

  In step 716, it is determined whether or not a signal indicating that another driving support device such as a lane keeping system or auto cruise control is operating within a predetermined time is received. If not received, it is determined that there is no influence on the steering angle due to the operation of another driving support device, the result is negative, the process proceeds to step 112, an alarm is output to the driver, and the process is terminated.

  If affirmative in step 702, step 706, step 710, step 714, and step 716, it is determined that the determination of the state of reduced consciousness is likely to cause an erroneous determination due to changes in the driving environment, and an alarm is output. Without returning to step 100, the process is repeated.

  As described above, according to the consciousness reduction driving detection apparatus in the fifth embodiment, when the weighted linear sum apo of the vehicle speed is larger than the predetermined value athr, the accelerator and the operation switch are operated within the predetermined time. If the vehicle speed v is smaller than the predetermined value vthr, the curvature r of the travel path is larger than the predetermined value rthr, or if any other driving assistance device is operating, the consciousness is lowered. Since there is a high possibility that an erroneous determination occurs in the state determination, it is possible to prevent an erroneous alarm from being performed by stopping the output of the alarm.

  Next, a sixth embodiment will be described. In addition, about the part used as the structure and process similar to 1st-5th embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted. In the sixth embodiment, the other point is that a predetermined value for determining whether or not the state of consciousness is lowered is set using the power value based on the steering angle detected when the driving state is stable. This is different from the embodiment.

  As shown in FIG. 12, a consciousness reduction driving detection device 710 in the sixth embodiment detects a steering angle sensor 12, a computer 714, an alarm device 16, a vehicle speed sensor 720 that detects the vehicle speed of the vehicle, and a yaw rate of the vehicle. A yaw rate sensor 722 and an operation switch 724 such as a winker are included.

  The computer 714 functionally includes a frequency analysis unit 18, a weighting coefficient DB 20, a total power value calculation unit 22, and a determination unit 724. The determination unit 724 determines whether or not the driver is in a state of reduced consciousness based on whether or not the total power value calculated by the total power value calculation unit 22 is equal to or greater than a predetermined value. The predetermined value thr for the determination is a total power value calculation unit based on a steering angle signal when it is determined that the driving state is stable based on detection signals of the vehicle speed sensor 720, the yaw rate sensor 722, and the operation switch 724. It is determined based on the total power value calculated in 22.

  Next, with reference to FIG. 13, a processing routine for detection of a reduced consciousness state according to the sixth embodiment will be described.

  In step 800, 0 is set to the counter value t1, and then in step 802, the current vehicle speed v and yaw rate y are acquired.

  Next, in step 804, it is determined whether or not the vehicle speed v is equal to or higher than the predetermined speed vx, the absolute value of the yaw rate y is equal to or lower than the predetermined yaw rate yx, and the operation switch 724 is not operated. This is to determine whether or not the operating state is stable. For example, a state where the vehicle is traveling on a straight road at a certain speed or more is regarded as a state in which the driving state is stable, and a state in which a blinker that is one of the operation switches 724 is operated in order to change lanes or the like A state where the vehicle is traveling on a road having a large curvature can be set to a state where the driving state is not stable.

  Note that, instead of the yaw rate, it may be determined whether the driving state is stable by using a lateral acceleration or a steering angle signal that is a source of the vehicle signal of the yaw rate and the lateral acceleration. If the operating state is stable, the process proceeds to step 806, and if not, the process returns to step 800.

  Next, in step 806, the counter value t1 is incremented. Next, in step 808, it is determined whether or not the counter value t1 is equal to or greater than a predetermined value T1. This is to determine whether or not the continuous period of the period in which the operation state determined in step 804 is stable is equal to or longer than a predetermined period. When the continuous period of the stable state is equal to or longer than the predetermined period determined by the predetermined value T1, the process proceeds to step 300 in order to acquire a steering angle signal for calculating the predetermined value thr, and when it is less than T1, step 802 is performed. Returning to step 802, the processes in steps 802 to 806 are repeated until the continuous period of the stable state becomes a predetermined period or more.

  In step 300, 1 is set to the counter value t, and the total power value p is calculated by the processing in steps 100 to 106. Next, at step 810, it is determined whether or not the counter value t is smaller than a predetermined value T. The predetermined value T is a value based on an accumulation period necessary for calculating the predetermined value thr for determining the consciousness lowered state. If t is smaller than T, the process proceeds to step 812. If t is equal to or greater than T, it is determined that the predetermined value thr for determining the consciousness reduction state has already been calculated, and the process proceeds to step 500.

  Next, the current vehicle speed v and yaw rate y are acquired in step 812 by the same processing as in steps 802 and 804, and in step 814, it is determined whether or not the driving state is stable. If the operating state is stable, the process proceeds to step 500. If the operating state is not stable, the process returns to step 100.

  Next, in step 500, the total power value p is stored in the FIFO type buffer. Next, in step 502, it is determined whether or not the counter value t is greater than a predetermined value T. If it is greater than the predetermined value T, the process proceeds to step 110. If it is less than the predetermined value T, the process proceeds to step 504.

  In step 504, it is determined whether or not the counter value t is T. If t is smaller than T, the accumulated time necessary for calculating the predetermined value thr for determining the consciousness lowering state has not elapsed, so a negative determination is made and the routine proceeds to step 506 where a predetermined number (in the FIFO type buffer ( It is determined whether or not (for example, 6) total power values p are stored. If stored, in step 508, the cumulative value a of the total power value p in the FIFO buffer is calculated and stored in a cumulative value buffer (not shown) which is a predetermined storage area. Proceed to 516. If the predetermined number of total power values p are not stored, the process proceeds to step 516 as it is, and the counter value t is incremented and the process returns to step 100. By repeating the processing of step 506 and step 508 until t = T, the cumulative value a of a predetermined number of total power values p is stored in the cumulative value buffer.

  If it is determined in step 504 that t = T, the process proceeds to step 816, where the average value of the accumulated value a of the total power value p stored in the accumulated value buffer is calculated, and the coefficient k is added to this. Thus, a predetermined value thr for determination is calculated. The coefficient k is used to correct the predetermined value thr so that it does not become an unreasonably low value in consideration of the possibility that a total power value p lower than usual is calculated by careful driving at the start of driving. It is. The coefficient k can be a value that is twice the standard deviation of the cumulative value a of the total power value p stored in the cumulative value buffer, for example.

  Next, in step 516, the counter value t is incremented and the process returns to step 100. After the processing of step 102 to step 106, the process proceeds to step 810. Since t> T, the result is negative and the process proceeds to step 500. . After step 500, the process proceeds to step 502. Affirmative determination is made at step 502 and processing proceeds to step 110.

  In step 110, when the predetermined value thr calculated in step 816 is set as a threshold value and the accumulated value a stored in the accumulated value buffer is larger than the predetermined value thr, it is determined that the state of consciousness is lowered, and step 112 is performed. To go to step 100, output an alarm, and if the accumulated value a is less than or equal to thr, return to step 100.

  As described above, according to the consciousness reduction driving detection apparatus in the sixth embodiment, the consciousness reduction state is determined based on the total power value calculated from the steering angle data detected by actual traveling. When the threshold value (predetermined value thr) is calculated, since the predetermined value is calculated based on the steering angle data when the driving state is stable, the difference between the consciousness drop time and the awakening time can be determined more clearly. A predetermined value can be calculated.

  In the sixth embodiment, as in the fourth embodiment, the calculated predetermined value is used as the initial value of the predetermined value, and the predetermined value is corrected based on the vehicle speed and the yaw rate. Good.

  In the above embodiment, the case where the detected physical quantity is the steering angle has been described. However, any signal may be used as long as it changes when the consciousness decreases, such as the yaw rate, the lateral position, the inter-vehicle distance, and the steering angular velocity.

  Further, in the above embodiment, the case where the predetermined frequency band for calculating the total power value is set to substantially the entire frequency band of the detected steering frequency has been described. By setting the weight coefficient to “0”, the total power value can be calculated only from the power value in the desired frequency band.

  In the first embodiment and the second embodiment, the case where the weight coefficient calculated in advance and stored in the weight coefficient DB is used to calculate the total power value has been described. The frequency distribution and the frequency distribution at the time of awakening are stored in advance, and when calculating the total power value, the weight coefficient is calculated from the stored frequency distribution at the time of lowering consciousness and the frequency distribution at the time of awakening. Also good.

  In the above-described embodiment, the case where a warning is output when it is determined that the driver is in a state of reduced consciousness has been described. However, the driving of the driver is supported by other methods such as control to apply a brake. You may do it.

It is the schematic which shows the structure of the consciousness reduction driving | running detection apparatus which concerns on 1st Embodiment. It is a figure for demonstrating the calculation method of the weighting coefficient in 1st Embodiment. It is a flowchart which shows the processing routine of a consciousness fall state detection of 1st Embodiment. It is a figure for demonstrating the calculation method of the weighting coefficient in 2nd Embodiment. It is the schematic which shows the structure of the consciousness reduction driving | running detection apparatus which concerns on 3rd Embodiment. It is a flowchart which shows the processing routine of a consciousness fall state detection of 3rd Embodiment. It is the schematic which shows the structure of the consciousness fall driving | running detection apparatus which concerns on 4th Embodiment. It is a flowchart which shows the processing routine of a consciousness fall state detection of 4th Embodiment. It is the schematic which shows the structure of the consciousness reduction driving | running detection apparatus which concerns on 5th Embodiment. It is a flowchart which shows the processing routine of the consciousness fall state detection of 5th Embodiment. It is a figure which shows an example of the weighting coefficient for the linear sum calculation with a weighting coefficient of the vehicle speed in 5th Embodiment. It is the schematic which shows the structure of the consciousness reduction driving | running detection apparatus which concerns on 6th Embodiment. It is a flowchart which shows the processing routine of a consciousness fall state detection of 6th Embodiment. It is a figure which shows the influence of the consciousness fall by the difference in an environment.

Explanation of symbols

10, 210, 410, 610, 710 Decreased driving detection device 12 Steering angle sensor 14, 214, 414, 614, 714 Computer 16 Alarm device 18 Frequency analysis unit 20 Weight coefficient DB
22 total power value calculation unit 24, 424, 624, 724 determination unit 220 frequency distribution (ld) DB
226 Weight coefficient calculation unit

Claims (16)

Detection means for continuously detecting a physical quantity indicating a running state of the host vehicle or a steering state when the driver steers the host vehicle, and outputting time-series data of the detected physical quantity;
Frequency analysis means for performing frequency analysis of the time-series data of the physical quantity output by the detection means, and calculating a frequency distribution indicated by a distribution of power values for each frequency band;
The weighting factor for each frequency band calculated so as to become heavier as the difference between the frequency distribution at the time of consciousness reduction and the frequency distribution at awakening is larger is multiplied by the power value for each corresponding frequency band. A total power value calculating means for calculating a total power value by adding the values over a predetermined frequency band;
Determining means for determining whether the driver's consciousness declined state by determining whether the sum of the total power values calculated by the total power value calculating means within a predetermined time is greater than or equal to a predetermined value;
Consciousness descent driving detection device. Detection means for continuously detecting a physical quantity indicating a running state of the host vehicle or a steering state when the driver steers the host vehicle, and outputting time-series data of the detected physical quantity;
Frequency analysis means for performing frequency analysis of the time-series data of the physical quantity output by the detection means, and calculating a frequency distribution indicated by a distribution of power values for each frequency band;
Calculated so that the greater the degree of difference between the frequency distribution at the time of consciousness reduction and the frequency distribution at the time of awakening obtained from the initial time-series data of the physical quantity output by the detection means within a predetermined time from the start of running, the greater the difference A total power value calculating means for calculating a total power value by adding a value obtained by multiplying the corresponding weight value for each frequency band by a power value for each corresponding frequency band over a predetermined frequency band;
Determining means for determining whether the driver's consciousness declined state by determining whether the sum of the total power values calculated by the total power value calculating means within a predetermined time is greater than or equal to a predetermined value;
Consciousness descent driving detection device.   The consciousness reduction driving | running detection apparatus of Claim 1 or Claim 2 which calculates the said weighting coefficient based on the information content of a zone | band. The consciousness reduction driving | running detection apparatus of any one of Claims 1-3 which calculates the said weighting coefficient according to the following formula | equation.

Here, pd _i is a band probability obtained by dividing the power value of the frequency band i in the frequency distribution at the time of consciousness reduction by the sum of the power values of the predetermined frequency band in the frequency distribution at the time of consciousness reduction, and pa _i is the wake-up time. This is a band probability obtained by dividing the power value of the frequency band i in the frequency distribution by the sum of the power values of the predetermined frequency band in the frequency distribution at awakening. The consciousness reduction driving | running detection apparatus of any one of Claims 1-3 which calculates the said weighting coefficient according to the following formula | equation.

However, μ _di is an average value of power values in the frequency band i in the frequency distribution at the time of consciousness decrease, σ _di is a dispersion value of power values in the frequency band i in the frequency distribution at the time of consciousness decrease, and μ _ai is at the time of awakening An average value of power values in the frequency band i in the frequency distribution, σ _ai is a variance value of the power values in the frequency band i in the frequency distribution at the time of awakening.   The consciousness reduction driving | running detection apparatus of any one of Claims 1-5 which made the said predetermined value the average value of the said sum in a predetermined period from driving | running | working start.   The predetermined value is the average value of the sums within the cumulative period when the cumulative period when the driving state is determined to be stable based on the vehicle speed and yaw rate of the vehicle is equal to or greater than a predetermined period. The consciousness reduction driving | running detection apparatus of any one of Claims 1-5 which were said.   The unconscious driving according to claim 7, wherein the cumulative period is started when a continuous period when it is determined that the driving state is stable based on a vehicle speed and a yaw rate of the vehicle is equal to or longer than a predetermined period. Detection device.   The consciousness reduction driving detection according to claim 7 or 8, wherein when the vehicle speed of the vehicle is smaller than a predetermined value and the yaw rate of the vehicle is larger than a predetermined value, it is determined that the driving state is not stable. apparatus.   The consciousness reduction driving | running detection apparatus of any one of Claims 1-9 which correct | amends the said predetermined value based on the vehicle speed and the magnitude | size of a curve in the predetermined period from a driving | running | working start.   The consciousness reduction driving | running detection apparatus of any one of Claims 1-10 which calculate the said predetermined value based on at least 1 or more values among a vehicle speed, a yaw rate, a yaw angle, and a horizontal direction acceleration. The consciousness reduction driving | running detection apparatus of any one of Claims 1-11 which calculate the said predetermined value according to the following formula | equation.

However, thr ₀ is the predetermined value initially set, v is the current vehicle speed, v ₀ is the average value of the vehicle speed within a predetermined period from the start of travel, y is the yaw rate of the current vehicle, and y ₀ is The average value of the yaw rate within a predetermined period from the start of travel, θ is the yaw angle obtained by time integration of the yaw rate, and θ ₀ is the average value of the yaw angle. The consciousness reduction driving | running detection apparatus of any one of Claims 1-12,
Support means for performing support for returning to a wakeful state when a state of reduced consciousness is detected by the determination means;
A driving support device including   When the driver operates a device other than the steering wheel provided in the vehicle within the predetermined time, when the driver performs an abrupt accelerator operation, when the vehicle speed is equal to or lower than the predetermined speed, and assists driving The driving support device according to claim 13, further comprising support stop means for stopping the support provided by the support means in any case when another device for operating is activated. Computer
The frequency of the physical quantity time-series data output by the detecting means for continuously detecting the physical quantity indicating the running state of the own vehicle or the steering state when the driver steers the own vehicle and outputting the time-series data of the detected physical quantity. Frequency analysis means for performing an analysis and calculating a frequency distribution indicated by a power value distribution for each frequency band;
The weighting factor for each frequency band calculated so as to become heavier as the difference between the frequency distribution at the time of consciousness reduction and the frequency distribution at awakening is larger is multiplied by the power value for each corresponding frequency band. A total power value calculating means for calculating a total power value by adding the values over a predetermined frequency band;
Determining means for determining whether the driver's consciousness declined state by determining whether the sum of the total power values calculated by the total power value calculating means within a predetermined time is greater than or equal to a predetermined value;
Consciousness descent driving detection program to make it function. Computer
The frequency of the physical quantity time-series data output by the detecting means for continuously detecting the physical quantity indicating the running state of the own vehicle or the steering state when the driver steers the own vehicle and outputting the time-series data of the detected physical quantity. Frequency analysis means for performing an analysis and calculating a frequency distribution indicated by a power value distribution for each frequency band;
Calculated so that the greater the degree of difference between the frequency distribution at the time of consciousness reduction and the frequency distribution at the time of awakening obtained from the initial time-series data of the physical quantity output by the detection means within a predetermined time from the start of running, the greater the difference A total power value calculating means for calculating a total power value by adding a value obtained by multiplying the corresponding weight value for each frequency band by a power value for each corresponding frequency band over a predetermined frequency band;
Determining means for determining whether the driver's consciousness declined state by determining whether the sum of the total power values calculated by the total power value calculating means within a predetermined time is greater than or equal to a predetermined value;
Consciousness descent driving detection program to make it function.

Images