JP4814779B2 - Vehicle attention monitoring device - Google Patents

Description

  The present invention relates to a vehicle attention monitoring apparatus, and more particularly to a vehicle attention monitoring apparatus useful for maintaining a safe driving of a vehicle by monitoring a decrease in the attention of a vehicle driver.

Estimate the level of attention the vehicle driver has, display the estimated level of attention on the LCD panel of the instrument panel, etc., and warn the driver when it is determined that the level of attention has decreased. An apparatus that performs the above has been developed and proposed in Patent Document 1.
In the device disclosed in Patent Document 1, the driver's arousal level is estimated based on the driving state quantity corresponding to the meandering rate when the vehicle is traveling, the driving operation status with respect to the vehicle, and the like. It is used as a degree.

More specifically, the driver's arousal level is repeatedly estimated every minute based on the meandering rate of the vehicle and the driving operation status with respect to the vehicle, and stored in the memory. Each time the estimated awakening level is updated, it is displayed in time series. Further, when the arousal level obtained in this way is lower than the determination level, the display color and the display form are changed to warn that the driver's attention is reduced.
JP-A-9-267660

In such a device, the update interval of the arousal level is 1 minute, but according to a study conducted by the applicant, the fluctuation period of the arousal level is about 25 seconds to 251 seconds, and the arousal level is 1 minute. In the case of a decrease in a short period of less than, there is a problem that it is impossible to promptly warn of a decrease in arousal level.
In order to solve such a problem, it is conceivable to shorten the estimation interval and display interval of the arousal level, but in this case, the information amount of the estimated arousal level increases and the display becomes too fine. Problems arise. In this case, when the driver's arousal level is temporarily recovered by the warning, the actual awakening level is gradually recovered, but the awakening level is immediately displayed by the recovered arousal level. Since it is updated, there is a possibility that the driver's arousal level is lowered again with peace of mind.

Therefore, in order to solve such a problem, the driving state amount of the vehicle necessary for estimating the driver's arousal level is repeatedly detected at a predetermined detection timing, and a predetermined time interval is included so as to include the latest detection timing. It is conceivable to estimate the driver's arousal level based on the average value of the driving state quantities detected over the predetermined detection period while updating and setting the predetermined detection period.
In this case, it is possible to cope with a sudden decrease in the arousal level by setting the predetermined detection timing interval to be shorter than 25 seconds that is the minimum fluctuation period of the arousal level. In addition, by estimating the arousal level using the average value of the driving state quantity during the predetermined detection period, even if there is a temporary recovery of the arousal level, the history of the decrease in the arousal level is reflected in the average value. Therefore, the display of the arousal level can be changed in accordance with the subsequent recovery of the actual arousal level.

Furthermore, the time interval for updating the predetermined detection period is changed from a time interval shorter than the minimum fluctuation period of the arousal level according to the fluctuation of the arousal level to a time interval similar to that of the device of Patent Document 1, It is conceivable to provide an appropriate display that is easy for the driver to recognize.
However, when the average value of the operation state quantities repeatedly detected in the predetermined detection period is used as described above, the older operation state quantity in the predetermined detection period is far from the current operation state quantity. Probability is high. For this reason, when the driving state quantity detected in the predetermined detection period is simply integrated to obtain an average value, it is excessively influenced by the driving state quantity far from the current driving state quantity, and the arousal level May not be accurately estimated.

  The present invention has been made in view of such problems, and an object of the present invention is for a vehicle capable of accurately monitoring an attention level while appropriately responding to a driver's attention level fluctuation. It is to provide an attention monitoring device.

In order to achieve the above object, the vehicle attention monitoring apparatus according to the present invention includes a driving state amount detecting means for repeatedly detecting a driving state amount corresponding to a driving state of the vehicle at a predetermined detection timing, and the latest detection timing. The first detection period having a predetermined length including the first detection period is updated and set at a predetermined time interval, and the first detection period is detected among the plurality of operation state quantities detected by the operation state detection unit over the first detection period. With respect to the operating state quantity from the first detection timing that is the earliest in the period to the second detection timing after a predetermined period from the first detection timing, the older operating state quantity is weighted smaller , and as weighting than the operation state quantity in the period of the inner first detection period excluding the period from the first detection timing to the second detection timing is reduced And then integrating the operating state quantity for each of the first detection periods to repeatedly obtain the operating state quantity integrated value, and based on the operating state quantity integrated value integrated by the integrating means. An average value calculating means for obtaining an average value of the driving state quantity for each of the first detection periods, and a note of the driver of the vehicle based on the average value of the driving state quantity calculated by the average value calculating means Characterized by comprising: an attention calculation means for calculating an attention index indicating the degree of force; and a monitoring means for monitoring a decrease in the attention based on the attention index calculated by the attention calculation means. (Claim 1).

According to the vehicle attention monitoring apparatus configured as described above, the driving state amount detection unit repeatedly detects the driving state amount representing the driving state of the vehicle at a predetermined detection timing.
The integrating means updates and sets a first detection period having a predetermined length including the latest detection timing at predetermined time intervals, and the first detection period among a plurality of operating state quantities detected over the first detection period. Of the first operation timing amount, the first operation timing amount from the first detection timing to the second detection timing after the first detection timing to the second detection timing is set so that the older operation state amount is weighted and the first The weighting is performed so that the weighting is smaller than the driving state quantity in the period within the first detection period excluding the period from the detection timing to the second detection timing, and then the driving state quantity is set for each first detection period. Accumulate and repeatedly calculate the operating state quantity integrated value.

The average value calculating means obtains an average value of the operation state quantity for each first detection period based on the operation state quantity integrated value obtained repeatedly in this way.
Then, the attention calculation means calculates an attention index indicating the degree of driver's attention based on the average value of the driving state amount calculated by the average value calculation means, and based on the calculated attention index, A monitoring means monitors the decline in attention.

In the vehicle attention monitoring apparatus, the integrating means can change the second detection timing according to a driver of the vehicle, and changes the weighting according to the change of the second detection timing. (Claim 2).
According to the vehicle attention monitoring apparatus configured as described above, by making the second detection timing at which the latest driving state quantity to be weighted is detected when integrating the driving state quantity variable, The period during which the operating state quantity to be weighted is detected is variable. Accordingly, the number of operating state quantities to be weighted becomes variable, and accordingly, the weighting for these operating state quantities is changed.

  More specifically, in the vehicle attention monitoring apparatus, the integrating means performs the weighting by multiplying the driving state quantity by a weighting coefficient having a value less than 1, and the average value calculating means includes An average value of the operating state quantities is obtained by dividing the operating state quantity integrated value by the number of the detection timings in a second detection period shorter than the first detection period, and the first detection period The sum of the weighting coefficient multiplied by each operation state quantity from the first detection timing to the second detection timing and the number of operation state quantities not weighted is the second detection period. It is equal to the number of the detection timings in (claim 3).

According to the vehicle attention monitoring apparatus configured as described above, the weighting of the driving state quantity by the integrating means is performed by multiplying the driving state quantity by a weighting coefficient having a value of less than 1.
The average value calculation means obtains the average value of the driving state quantity by dividing the driving state quantity integrated value by the number of detection timings in the second detection period shorter than the first detection period. At this time, the sum of the weighting coefficient multiplied by each operation state quantity from the first detection timing to the second detection timing and the number of operation state quantities that are not weighted is the detection timing in the second detection period. Is equal to the number of

Therefore, the average value of the driving state quantity calculated by the average value calculating means corresponds to the driving state quantity detected at one detection timing.
More specifically, in the vehicle attention monitoring apparatus, the monitoring means includes a predetermined number of attention indices selected from the plurality of attention indices repeatedly calculated by the attention calculation means. A display means for displaying in time series is provided (claim 4).

According to the vehicle attention monitoring apparatus configured as described above, a predetermined number of attention indices selected from the plurality of attention indices calculated by the attention calculation means are displayed by the display means of the monitoring means. Displayed in time series.
More specifically, in the vehicle attention monitoring apparatus, the monitoring unit includes an alarm unit that issues an alarm when the attention index calculated by the attention calculation unit is lower than a predetermined determination value. It is characterized (claim 5).

  According to the vehicle attention monitoring apparatus configured as described above, when the attention index calculated by the attention calculation means becomes lower than a predetermined determination value, an alarm is given by the alarm means of the monitoring means.

According to the vehicle attention monitoring apparatus of the present invention, when the driving state quantity integrated value necessary for obtaining the attention index is obtained, the first detection timing at least in the first detection period is the first detection timing. The operation state quantity from the first detection timing to the second detection timing after a predetermined period is weighted so that the older operation state quantity is weighted , and the period from the first detection timing to the second detection timing is excluded. Since the weighting is performed so that the weighting is smaller than the driving state quantity in the period within the first detection period, the old driving state quantity that is likely to be far from the current driving state quantity in the first detection period. It is possible to accurately estimate the attention index.

  According to the vehicle attention monitoring apparatus of claim 2, the number of the driving state quantities to be weighted is variable by changing the period during which the driving state quantity to be weighted is detected. . Accordingly, it is possible to change the degree of influence of the relatively old driving state amount on the attention index within the first detection period, and appropriately calculate the attention index according to individual differences in the fluctuation cycle of the attention. It becomes possible to monitor the attention with high accuracy.

In addition, since the weighting for the driving state quantity is also changed at this time, it is possible to appropriately weight each driving state quantity even when the period during which the driving state quantity to be weighted is detected is changed. It becomes possible to calculate the attention index appropriately.
Further, according to the vehicle attention monitoring apparatus of claim 3, the average value of the driving state quantity calculated by the average value calculating means corresponds to the driving state quantity detected at one detection timing. .

  Therefore, even when the detection period of the driving state quantity to be weighted is changed or the weighting for the driving state quantity is changed, it is based on the average value of the driving state quantity accurately regardless of these changes. It is possible to calculate the attention index. Further, the calculation of the attention index based on the average value of the driving state quantities can be performed in the same manner regardless of these changes, so that the calculation process can be simplified.

According to the vehicle attention monitoring apparatus of claim 4, a predetermined number of attention indices selected from the plurality of attention indices calculated by the attention calculation means are displayed in time series by the display means. Is displayed. Therefore, the driver can recognize the change in the attention level in time series, and can take an appropriate measure according to the change in the attention level.
Further, according to the vehicle attention monitoring apparatus of the fifth aspect, when the attention index calculated by the attention calculation means becomes lower than a predetermined determination value, an alarm is issued by the warning means. Accordingly, it is possible to make the driver strongly recognize a decrease in focus, and appropriate measures can be taken according to the decrease in attention.

A vehicle attention monitoring apparatus according to an embodiment of the present invention will be described below with reference to the drawings.
FIG. 1 is a system configuration diagram of a vehicle attention monitoring apparatus. As shown in FIG. 1, the vehicle attention monitoring apparatus includes an electronic control unit (hereinafter referred to as ECU) 1, a display device 2, and an alarm device 3.

  Information necessary for control performed by the ECU 1 is input to the ECU 1 from various sensors and switches provided in the vehicle, and the ECU 1 operates as an attention index indicating the degree of driver's attention based on the input information. The degree of arousal of a person is calculated. Then, the ECU 1 controls the display device 2 and the alarm device 3 based on the calculated arousal level, whereby the degree of driver's attention is displayed on the display device 2, and an alarm is issued from the alarm device 3 as necessary. Done.

  Information input to the ECU 1 includes, as driving operation information for the vehicle, a winker signal (Sw) indicating the operation state of the vehicle winker switch, an exhaust brake signal (Se) indicating the operation state of the exhaust brake, and operation of the clutch pedal. A clutch signal (Sc) indicating a state is included, and a steering pulse (Ps) corresponding to a steering angle detected by a steering angle sensor incorporated in a steering wheel as vehicle steering information, and a steering neutral pulse indicating a steering neutral position (Pn) is included. Furthermore, as information indicating the running state of the vehicle, for example, a vehicle front image (Img) imaged by a camera disposed in the upper center of the windshield, and a vehicle speed signal corresponding to the vehicle running speed detected by the vehicle speed sensor ( V) and the like are input.

  The ECU 1 includes a CPU 4 that executes a control program corresponding to the control performed by the ECU 1, a memory such as a ROM 5 and a RAM 6 that stores program instructions and data, an input interface 7 for taking in the various information described above into the ECU 1, and a control by the ECU 1 An output interface 8 for outputting a signal to the display device 2 and the alarm device 3 and a timer 9 for counting time required in the arithmetic processing executed by the CPU 4 are configured.

FIG. 2 shows a functional configuration diagram corresponding to the control executed by the vehicle attention monitoring apparatus configured as described above.
As shown in FIG. 2, the ECU 1 is based on the input information, and repeatedly calculates a driving state quantity corresponding to the driving state of the vehicle at a predetermined detection timing to detect a driving state quantity detection unit (driving state quantity detection). 10), and the monotonicity of the driving operation with respect to the vehicle, the corrected steering amount with respect to the traveling direction of the vehicle, and the meandering rate of the vehicle are used as the driving state quantities.

In order to obtain these driving state quantities, the driving state quantity detection unit 10 includes a monotonicity calculation unit 10a, a corrected steering amount calculation unit 10b, and a meandering rate calculation unit 10c.
The monotonicity calculating unit 10a calculates the monotonicity of the driving operation by the driver from the frequency of the driving operation on the vehicle based on the winker signal Sw, the exhaust brake signal Se, the clutch signal Sc, and the like. The corrected steering amount calculation unit 10b calculates a corrected steering amount with respect to the traveling direction of the vehicle based on the steering pulse Ps and the steering neutral pulse Pn. Further, the meandering rate calculation unit 10c obtains the meandering rate of the vehicle from the change in the horizontal position of the white line on the road recognized using the vehicle front image Img and the vehicle speed signal V.

The detection of these operation state quantities is repeatedly performed at a predetermined detection timing based on the time count of the timer 9 of the ECU 1, and for example, 15 detection timings per second are set at equal intervals. . Accordingly, the monotonicity, the corrected steering amount, and the meandering rate are each calculated at 15 pieces per second.
The monotonicity, the corrected steering amount, and the meandering rate repeatedly detected by the driving state quantity detection unit 10 in this way are sent to the integration unit (integration means) 11 and are detected over a predetermined first detection period. The corrected steering amount and the meandering rate are integrated. Since the driver's arousal level fluctuates in a cycle of 25 seconds to 251 seconds as described above, in the present embodiment, the first detection period is set to 6 minutes, which is longer than the longest fluctuation cycle.

  In addition, the first detection period is updated at predetermined time intervals, and in the present embodiment, the first detection period is updated every 15 seconds. Therefore, the monotonicity, the corrected steering amount, and the meandering rate are obtained by integrating the data for the past six minutes including the latest data, and the respective integrated values are obtained. Each unit of the meander rate (hereinafter referred to as an integration unit) is updated and set every 15 seconds.

The configuration of the integration unit thus obtained is as shown in FIG. 3, and FIG. 3 representatively shows the integration unit for one kind of driving state quantity among the monotonicity, the corrected steering amount, and the meandering rate. ing.
As shown in FIG. 3, the integration unit U1 corresponding to the integration value obtained at time t0 is set so as to integrate the operation state quantities detected at the respective detection timings in the first detection period T1 (6 minutes). Even after time t0, similar integration units are sequentially set at predetermined time intervals Δt (15 seconds) as U2 to U5 in the figure.

Here, each integrating unit is divided into segments every one minute, and as shown in FIG. 3, the segments are S1, S2, S3, S4, S5 and S6 from the latest data side.
As described above, the monotonic degree, the corrected steering amount, and the meandering rate are each detected 15 times per second, and corresponds to any one of the driving state quantities of the monotonic degree, the corrected steering amount, and the meandering rate. One segment includes 900 pieces of operation state quantity data, and one integration unit contains 5400 pieces of operation state quantity data, which is six times as much.

  In the present embodiment, instead of simply integrating the operation state quantities contained in one integration unit, the integration unit U1 shown in FIG. 3 will be described as an example. The first detection first in the first detection period T1. Each operation state quantity is weighted by multiplying the operation state quantity included in the segment S6 including the timing t1 and the operation state quantity included in the next segment 5 by a weighting coefficient smaller than 1.

Therefore, the latest detection timing t2 in the segment S5 corresponds to the second detection timing of the present invention, and the second detection timing t2 is after 1799 detection timings (a predetermined period in the present invention) after the first detection timing t1. Become.
More specifically, all operating state quantities in segment S6 are multiplied by a weighting factor K6 set to a value of 0.33, and all operating state quantities in segment S5 have a value of 0.67. Is multiplied by the weighting coefficient K5 set. On the other hand, all the operation state quantities in the segments S1 to S4 are multiplied and multiplied by weighting coefficients K1 to K4 having a value of 1, respectively. Therefore, all the operation state quantities in the segments S1 to S4 are integrated as they are without weighting.

The integrating unit 11 integrates other integrating units such as the integrating units U2 to U5 in the same manner as the integrating unit U1, and integrates the monotonicity, the corrected steering amount, and the meandering rate for each integrating unit. The integrated value, the corrected steering amount integrated value, and the meandering rate integrated value are output to the average value calculation unit (average value calculation means) 12.
In the average value calculation unit 12, the monotonic degree is obtained by dividing the monotonic degree integrated value, the corrected steering amount integrated value, and the meandering rate integrated value by the number of detection timings in the second detection period T2 shorter than the first detection period T1. The average values of the corrected steering amount and the meandering rate are obtained for each integration unit.

As shown in FIG. 3, the second detection period T2 has a length corresponding to five segments from the latest segment S1 to the segment S5. Such a length of the second detection period T2 is set in association with the values of the weighting coefficients K5 and K6 used for weighting in the integrating unit 11.
That is, the values of the weighting coefficients K5 and K6 used for weighting the operation state quantity in the integrating unit 11 are 0.33 and 0.67, respectively. When both are added, the total value of the weighting coefficients multiplied by one type of operating state quantity in segments S5 and S6 matches the number of one type of operating state quantity in one segment. ing. In other words, a value obtained by weighting and integrating the operation state quantities of the segments S5 and S6 corresponds to a value obtained by integrating one kind of operation state quantity in one segment without performing weighting.

That is, the sum of the total value of such weighting coefficients and the number of one type of operation state quantity in the segments S1 to S4 that are not weighted is equal to the number of one type of operation state quantity in the five segments.
Therefore, by dividing each of the monotonic degree integrated value, the corrected steering amount integrated value, and the meandering rate integrated value obtained by the integrating unit 11 by the number of detection timings in the second detection period T2 corresponding to five segments. The respective average values corresponding to the monotonicity, the corrected steering amount, and the meandering rate at one detection timing are obtained.

  When determining the average values of the monotonicity, the corrected steering amount and the meandering rate obtained in this way, the monotonicity, the corrected steering amount and the oldest segment and the second oldest segment in one integrating unit are determined. Weighting is performed for each meandering rate. Since the oldest segment is weighted less than the next oldest segment, the older information that is likely to be far from the current monotonicity, corrected steering amount, and meander rate is less affected. In addition, each average value of the monotonicity, the corrected steering amount, and the meandering rate can be obtained with high accuracy.

As described above, since the integration unit is updated every 15 seconds, the average value calculation unit 12 obtains each average value of monotonicity, corrected steering amount, and meandering rate every 15 seconds. (Attention calculation means) 13
The attention calculation unit 13 uses the average values of the monotonicity, the corrected steering amount, and the meandering rate obtained in this way, and the degree of driver's attention by fuzzy inference based on a predetermined membership function. The arousal level is obtained as an attention level index indicating, and the obtained arousal level is output to the display control unit 14 and the alarm control unit 15 corresponding to the monitoring means of the present invention. Accordingly, the awakening level is output to the display control unit 14 and the alarm control unit 15 every 15 seconds.

Based on the arousal level output every 15 seconds from the attention calculation unit 13, the display control unit 14 displays the temporal change in the arousal level up to the present time as a 10-element (10 time points) bar graph as shown in FIG. In addition, the display device 2 is controlled so as to switch the display in accordance with the state of decrease in the arousal level.
Further, the alarm control unit 15 warns the driver of a decrease in attention by voice according to the display by the display device 2 in accordance with the degree of decrease in the arousal level output every 15 seconds from the attention calculation unit 13. .

In the display device 2, the magnitude of the arousal level is displayed at the height of the bar graph by the control performed by the display control unit 14, the driver's arousal level is sufficiently high, and the driver is alert. In a normal case where it is not necessary to recognize the decrease, each bar graph is displayed in green as indicated by A in FIG.
When the arousal level falls within a certain range, each bar graph is displayed in orange as indicated by B in FIG.

Further, when the arousal level is significantly reduced, the display of the arousal level by the bar graph is stopped as shown by C in FIG. 4, the attention character is enlarged and displayed in red, and the display color is displayed to the driver. It is made to recognize the decline of attention power strongly.
When the arousal level is displayed in time series by the bar graph, the display control unit 14 displays the latest arousal level at the right end and the 10 new arousal levels are displayed by 10 bar graphs. 2 is controlled.

  The arousal level is output from the attention calculation unit 13 every 15 seconds, but when the arousal level is sufficiently high and the display according to A in FIG. 4 is performed, the display update by the bar graph is performed every minute. It is like that. Accordingly, the driver can visually recognize the transition of the arousal level for a maximum of 10 minutes from the displayed bar graph, and can recognize the change of his / her attention based on the transition of the arousal level based on the bar graph and take necessary measures. It becomes possible.

Each time 1 minute elapses, the height of the rightmost bar graph is changed to the height corresponding to the latest awakening level, and the oldest awakening level is removed from the 10 awakening levels displayed so far. The display is updated so that the nine arousal levels are shifted to the left from the previous display positions one by one and displayed by the remaining nine bar graphs.
On the other hand, when the arousal level output from the attention calculator 13 falls below the first determination value, the display is immediately updated as described above, and the display color is switched to orange as indicated by B in FIG. Since the arousal level is output from the attention calculation unit 13 every 15 seconds, when the awakening level is reduced, the display is updated at a time interval shorter than the normal display update time interval of 1 minute. It becomes possible to make a person recognize a decrease in attention promptly.

While the awakening level is lower than the first determination value, the display is updated every minute and the display color remains orange while the level is substantially the same.
When the awakening level further decreases and falls below the second determination value lower than the first determination value, the display control unit 14 immediately updates the display described above, and the display color remains orange. At this time, the alarm control unit 15 controls the alarm device 3 in response to such a decrease in the arousal level. As the first alarm from the alarm device 3, for example, "Are you tired?" Sounds such as “Is it falling?” By such display of the display device 2 and the first alarm of the alarm device 3, the driver can immediately recognize a further decrease in attention.

Next, while the awakening level is lower than the second determination value, while the level is substantially the same, the above-described display update is performed every minute and the display color remains orange. Further, the first alarm by the alarm device 3 is not performed.
Further, when the arousal level decreases and falls below a third determination value lower than the second determination value, the display control unit 14 immediately switches the display on the display device 2 to C in FIG. The display color is red. In addition, the alarm control unit 15 controls the alarm device 3 in response to such a decrease in the arousal level. As the second alarm from the alarm device 3, for example, “attention is decreasing. "Etc." is output.

Then, after the arousal level falls below the third determination value, the display control unit 14 causes the display device 2 to update the display of C in FIG. 4 by the display device 2 every minute while the level is substantially the same. 2, the alarm control unit 14 controls the alarm device so that a second alarm is issued from the alarm device 3 every minute.
The operation of the vehicle attention monitoring apparatus according to the present embodiment when the driver's arousal level gradually decreases is as described above, but when the driver's arousal level gradually recovers, the following is performed. The vehicle attention monitoring device operates as described above.

When the driver's arousal level is in a recovery trend, the display on the display device 2 is constantly updated every minute, and the warning by the warning device 3 is not performed.
That is, even if the arousal level recovers to a level that is greater than or equal to the third determination value and less than the second determination value, the display device 2 continues to display C in FIG. 4 until one minute has elapsed since the previous display update. In this way, the display control unit 14 controls the display device 2. If one minute has elapsed since the last update of the display and the awakening level is still at a level equal to or higher than the third determination value and lower than the second determination value, the display control unit 14 displays the display of the display device 2 in FIG. Switching to B, the latest arousal level is set to the right end, and 10 new arousal levels are displayed as a bar graph, and the display color is set to orange.

  While the awakening level is maintained at a level equal to or higher than the third determination value and lower than the second determination value, the display color is maintained orange and the height of the rightmost bar graph is updated every time one minute elapses. Change to a height corresponding to the awakening level, and of the 10 awakening levels displayed so far, the 9 awakening levels except the oldest awakening level are left one by one from the previous display position. After shifting, the display control unit 14 updates the display so that the remaining nine bar graphs are displayed.

  Even when the arousal level further recovers and becomes a level that is equal to or higher than the second determination value and lower than the first determination value, such display update is continued, and the awakening level is recovered to a level equal to or higher than the first determination value. The display control unit 14 controls the display device 2 so that the display is switched to A in FIG. 4 when one minute has passed since the last display update. That is, the display color of the display device 2 is switched to green, and the awakening level display is updated by the bar graph every minute.

As described above, by monitoring the decrease in the driver's attention according to the fluctuation of the arousal level, when the arousal level greatly decreases beyond a predetermined determination value, it is shorter than the fluctuation period of the arousal level. The display can be immediately updated at time intervals, and a warning can be promptly recognized by the driver to recognize a reduction in attention.
In addition, the driver may temporarily recover the arousal level due to an alarm or the like, but if the display is updated immediately in response to such a recovery of the arousal level, the driver will feel secure and the arousal level will be restored again. May decrease. However, as described above, when the arousal level is in a recovery trend, the display is updated at intervals of one minute longer than 15 seconds, which is the calculation period of the arousal level, thereby avoiding such a problem. Is possible.

  Furthermore, when calculating the average values of the monotonicity, the corrected steering amount, and the meandering rate required for calculating the arousal level, as described above, they are included in the oldest segment and the oldest segment in one integrating unit. Weighting is performed for each of the monotonicity, the corrected steering amount, and the meandering rate, and the oldest operating state amount is weighted so that the oldest segment is less weighted than the next oldest segment. Weighting is performed as follows.

As a result, it is possible to reduce the influence of old information that is likely to be far from the current monotonicity, the corrected steering amount, and the meandering rate, and to accurately calculate the average values of monotonicity, corrected steering amount, and meandering rate. It becomes. Accordingly, it is possible to accurately estimate the driver's arousal level based on the average values of the monotonicity, the corrected steering amount, and the meandering rate.
For example, the driver's arousal level recovers relatively slowly, but when the average value of the driving state quantity is calculated without weighting the driving state quantity over a predetermined period at all. Therefore, there is a possibility that proper attention monitoring corresponding to such recovery of the arousal level cannot be performed.

  In other words, until the driving state quantity when the predetermined period has been updated several times and the arousal level has dropped from the predetermined period, this driving state quantity continues to contribute to the average value of the driving state quantity. Become. When this driving state quantity deviates from the predetermined period, the driving state quantity immediately does not contribute to the average value of the driving state quantity, so that the arousal level is estimated as if it has recovered rapidly.

However, by weighting the old driving state quantity as in the present embodiment, the degree that the driving state quantity when the arousal level is reduced contributes to the average value of the driving state quantity gradually decreases. It becomes possible to monitor appropriate attention corresponding to the actual recovery of arousal level.
In the above embodiment, one integrating unit is composed of six segments S1 to S6 per minute, and the operation state quantity in the oldest segment S6 and the next oldest segment S5 is weighted. However, the number of segments included in one integrating unit and the segments to be weighted are not limited to this.

In the above embodiment, the four segments S1 to S4 are not targeted for weighting. This is because the fluctuation period of the driver's arousal level is 25 to 251 seconds, so that the arousal level can be properly estimated even in the longest fluctuation period, and is divided into four segments, that is, 4 minutes (240 seconds). Thus, no weighting is performed on the operation amount of the minute.
However, instead of weighting in such an embodiment, the number of segments included in one integration unit is changed according to the driver in consideration of the individual difference of the driver regarding the fluctuation period of the arousal level, and the weighting The number of target segments, that is, the second detection timing may be changed. Therefore, such a modification will be described below.

The vehicle attention monitoring apparatus according to this modification is different from the above-described embodiment only in the accumulation of the driving state quantity performed by the integrating unit 11, and the other configurations are the same as those in the above-described embodiment. Therefore, below, it demonstrates centering on integration | accumulation of the driving | running state amount in the integrating | accumulating part 11, and abbreviate | omits description about another part.
FIG. 5 shows the values of the weighting coefficients applied to the segments included in one integration unit for each weighting pattern used in the integration unit 11, and the smaller the number, the newer the segment S1. The latest one.

As shown in FIG. 5, in the integrating unit 11, the weighting pattern is changed to any one of P1 to P3 according to the driver, and the number of segments included in one integrating unit is changed according to the adopted pattern. It is supposed to be.
In the pattern P1, six segments S1 to S6 are included in one integrating unit as in the above embodiment, and the same weighting as in the above embodiment is performed.

That is, weighting is not performed on the operating state quantities from the newest segment S1 to segment S4 by using the weighting coefficients K1 to K4 having a value of 1. Further, the operation state quantity of the oldest segment S6 is multiplied by a weighting coefficient K6 having a value of 0.33, and the operation state quantity of the next oldest segment S5 is multiplied by a weighting coefficient K5 having a value of 0.67. The
After weighting in this way, by integrating the operating state quantities in one integrating unit, an integrated value corresponding to the integrated value of the operating state quantities substantially included in the five segments as in the above embodiment. Is obtained.

Therefore, the average value calculation unit 12 divides the integrated value of the operation state quantity obtained by the integration unit 11 by the number of detection timings in the second detection period T2 corresponding to five segments as in the above embodiment. Thus, an average value corresponding to the operation state amount at one detection timing is obtained.
Further, in the pattern P2, seven segments S1 to S7 are included in one integrating unit, and the weighting is performed so that the older the operation state quantity of the segments S4 to S7 is, the smaller the weight is.

  That is, weighting is not performed on the operating state quantities from the newest segment S1 to segment S3 by using the weighting coefficients K1 to K3 having a value of 1. In addition, the operation state quantity of the oldest segment S7 is multiplied by a weighting coefficient K7 having a value of 0.2. Hereinafter, the operation state quantity of the segment S6 is weighted by a weighting coefficient K6 having a value of 0.4 and the operation of the segment S5. The state quantity is multiplied by a weighting coefficient K5 having a value of 0.6, and the operation state quantity of the segment S4 is multiplied by a weighting coefficient K4 having a value of 0.8.

  Since the total value of the weighting coefficients K4 to K7 applied to the operation state quantities of the segments S4 to S7 is 2, the total value of the weighting coefficients multiplied by one operation state quantity in the segments S4 to S7 is two segments. This corresponds to the number of one type of operation state quantity in. In other words, the value obtained by weighting and integrating the operation state quantities of the segments S4 to S7 corresponds to the value obtained by integrating one kind of operation state quantity in the two segments without weighting.

That is, the sum of the total values of the weighting coefficients K4 to K7 and the number of one kind of operation state quantity in the segments S1 to S3 that are not weighted is 1 in the five segments as in the case of the pattern P1. It is equal to the number of types of operating state quantities.
Therefore, the average value calculation unit 12 performs the same calculation process as in the case of the pattern P1, and calculates the integrated value of the operation state quantity obtained by the integration unit 11 within the second detection period T2 corresponding to five segments. By dividing by the number of detection timings, an average value corresponding to the operation state quantity at one detection timing is obtained.

Further, in the pattern P3, eight segments S1 to S8 are included in one integrating unit, and the weighting is performed so that the older the operation state quantity of the segments S3 to S8 is, the smaller the weight is.
That is, weighting is not performed on the operating state quantities of the newest segment S1 and segment S2 because the weighting coefficients K1 and K2 having a value of 1 are used. In addition, the operation state quantity of the oldest segment S8 is multiplied by a weighting coefficient K8 having a value of 0.14. Hereinafter, the operation state quantity of the segment S7 is weighted by a weighting coefficient K7 having a value of 0.29 and the operation of the segment S6. The state quantity has a weighting coefficient K6 having a value of 0.43, the operating state quantity in segment S5 has a value of 0.57, the weighting coefficient K5 has a value, and the operating state quantity in segment S4 has a weighting of 0.71. The coefficient K4 and the operating state quantity of the segment S3 are respectively multiplied by a weighting coefficient K3 having a value of 0.86.

  Since the total value of the weighting coefficients K3 to K8 applied to the operating state quantities of the segments S3 to S8 is 3, the total value of the weighting coefficients multiplied by one operating state quantity in the segments S3 to S8 is 3 segments. This corresponds to the number of one type of operation state quantity in. In other words, the value obtained by weighting and integrating the operation state quantities of the segments S3 to S8 corresponds to the value obtained by integrating one kind of operation state quantity in the three segments without performing weighting.

That is, the sum of the total values of the weighting coefficients K3 to K8 and the number of one kind of operation state quantity in the segments S1 and S2 that are not weighted is the same as in the case of the patterns P1 and P2. It becomes equal to the number of one kind of operation state quantity in.
Therefore, the average value calculation unit 12 performs the same calculation process as in the patterns P1 and P2, and calculates the integrated value of the operation state quantity obtained by the integration unit 11 in the second detection period T2 corresponding to five segments. By dividing by the number of detection timings, an average value corresponding to the operation state quantity at one detection timing is obtained.

  By changing the weighting pattern according to the driver in this way, the number of segments to be weighted is increased in the pattern P2 compared to the pattern P1, and the old driving with respect to the average value of the driving state quantity is increased. The degree to which the state quantity contributes decreases. Therefore, it is possible to appropriately estimate the arousal level for a driver whose fluctuation period of the arousal level is shorter than that of the driver to which the pattern P1 is applied.

Further, since the number of segments to be weighted is increased in the pattern P3 compared to the pattern P2, the degree of contribution of the old driving state quantity to the average value of the driving state quantity is further reduced. Therefore, it is possible to appropriately estimate the arousal level for a driver whose fluctuation period of the arousal level is shorter than that of the driver to which the pattern P2 is applied.
In this way, by changing the weighting pattern for the driving state quantity according to the driver, the awakening level can be appropriately adjusted in a wide range from a driver having a relatively long fluctuation period of awakening level to a driver having a relatively long period of time. It is possible to estimate and monitor the driver's attention well.

Further, when the weighting pattern for the driving state quantity is changed in this way, the integrated value of the driving state quantity obtained by integrating by the integrating unit 11 is as described above by appropriately distributing the weighting coefficient. In any pattern, it corresponds to an integrated value obtained by integrating the operation state quantities included in the five segments without weighting.
Therefore, the average value calculation unit 12 can calculate the average value of the driving state quantities by performing the same calculation for any pattern. For this reason, it is not necessary to change the calculation process of the average value calculation unit 12 every time the weighting pattern is switched, and the calculation process of the average value calculation unit 12 can be simplified.

Furthermore, the average value of the driving state quantity calculated by the average value calculation unit 12 is always an average value corresponding to the driving state quantity at one detection timing regardless of the change of the weighting pattern for the driving state quantity. Therefore, regardless of the change of the weighting pattern, the arousal level can always be estimated with high accuracy, and the driver's attention can be well monitored.
Note that the switching of the weighting pattern according to the driver may be performed manually by the driver himself by a switch operation or the like, or the driver's age is estimated from the driving state quantity, and according to the estimation result. You may make it perform automatically.

This is the end of the description of the vehicle attention monitoring apparatus according to the embodiment of the present invention and the modification thereof, but the present invention is not limited to the embodiment and the modification thereof.
For example, the number of segments included in one integrating unit and the number of segments to be weighted are not limited to those shown in the above-described embodiment and modifications, but are limited to the specifications of the vehicle attention monitoring device. Changes can be made accordingly.

Moreover, in the said embodiment, although the time interval which updates an integrating | accumulating unit was 15 seconds, this time interval is not limited to this, It can change suitably.
Further, in the above embodiment and the above modification, the weighting coefficient having a common size is applied to one segment for weighting. However, the weighting coefficient applied to one segment is changed. You may make it make it. In this case, the variation range of the weighting factor applied in the segment is larger than the variation range of the weighting factor applied in the segment immediately before the segment, and the variation range of the weighting factor applied in the segment after the segment. It is desirable to make it smaller.

Moreover, in the said embodiment and the said modification, although it was made to use three, a monotonicity, a correction steering amount, and a meandering rate, as a driving | running state amount, a driving | running state amount is not restricted to this, a driver | operator's amount is not limited to this. It may be anything that fluctuates in relation to changes in attention.
Furthermore, the information used for detecting the monotonicity, the corrected steering amount, and the meandering rate is not limited to that used in the above embodiment.

  Further, the form of monitoring the driver's attention based on the obtained arousal level is not limited to the display and alarm as in the above embodiment, and the vehicle drive device is adapted to the reduction in the attention. Various forms are possible, such as changing the operating state of the braking device.

1 is a system configuration diagram of a vehicle attention monitoring apparatus according to an embodiment of the present invention. It is a functional block diagram corresponding to the control performed by the vehicle attention level monitoring apparatus of FIG. It is a conceptual diagram of the integrating | accumulating unit used with the vehicle attention level monitoring apparatus of FIG. It is a figure which shows the display state by a display apparatus. It is a figure which shows the relationship between the weighting pattern and the weighting coefficient in the modification of the vehicle attention monitoring apparatus of FIG.

Explanation of symbols

10 Operating state detector (Operating state detection means)
11 Accumulation unit (accumulation means)
12 Average value calculation unit (average value calculation means)
13 Attention calculation unit (attention calculation means)
14 Display control unit (monitoring means)
15 Alarm control unit (monitoring means)

Claims (5)

Driving state quantity detecting means for repeatedly detecting a driving state quantity corresponding to the driving state of the vehicle at a predetermined detection timing;
A first detection period having a predetermined length including the latest detection timing is updated and set at predetermined time intervals, and the plurality of operation state quantities detected by the operation state detection unit over the first detection period. The older operating state amount is weighted smaller than the operating state amount from the first detection timing that is the earliest in the first detection period to the second detection timing that is a predetermined period after the first detection timing. And weighting so that the weight is smaller than the operating state quantity in the period within the first detection period excluding the period from the first detection timing to the second detection timing . Integration means for accumulating the operation state quantity for each detection period and repeatedly obtaining the operation state quantity integration value;
An average value calculating means for obtaining an average value of the operating state quantity for each of the first detection periods based on the operating state quantity integrated value integrated by the integrating means;
Attention power calculation means for calculating an attention index indicating the degree of attention the driver of the vehicle has based on the average value of the driving state quantities calculated by the average value calculation means;
A vehicle attention level monitoring device comprising: monitoring means for monitoring a decrease in the attention level based on the attention level index calculated by the attention level calculation unit.   The said integrating | accumulating means can change the said 2nd detection timing according to the driver | operator of the said vehicle, The said weighting is changed with the change of the said 2nd detection timing, The said weighting is characterized by the above-mentioned. Vehicle attention monitoring device. The integrating means performs the weighting by multiplying the operating state quantity by a weighting coefficient having a value less than 1.
The average value calculation means obtains an average value of the driving state quantities by dividing the driving state quantity integrated value by the number of the detection timings in a second detection period shorter than the first detection period. ,
The sum of the total value of the weighting coefficients multiplied by the respective operation state quantities from the first detection timing to the second detection timing in the first detection period and the number of operation state quantities not subjected to the weighting are: The vehicle attention monitoring apparatus according to claim 1 or 2, wherein the number of detection timings in the second detection period is equal to the number of detection timings.   The monitoring means includes a display means for displaying a predetermined number of attention indices selected from the plurality of attention indices repeatedly calculated by the attention calculation means in time series. The vehicle attention monitoring apparatus according to any one of claims 1 to 3.   The vehicle according to any one of claims 1 to 4, wherein the monitoring unit includes an alarm unit that issues an alarm when the attention index calculated by the attention calculation unit is lower than a predetermined determination value. Attention monitoring device.

Images