JP6015754B2 - Operating state estimation device - Google Patents

Description

  The present invention relates to an apparatus for estimating a driving state of a driver.

  As an apparatus for estimating the driving state of the driver, there is an apparatus described in Patent Document 1, for example. In the device described in Patent Document 1, when the vehicle speed is within a preset vehicle speed range, for example, a steering angle is acquired as the driving operation amount of the driver, and the distribution characteristics of the steering angle are based on the acquired history of the steering angle. The amount (difference amount) is obtained. And the apparatus of patent document 1 determines whether the driving | operation operation of a driver | operator is unstable from the feature-value.

JP 2009-9495 A

In the above prior art, if the vehicle speed is within a preset vehicle speed range, the feature amount is calculated using information on the steering angle acquired during that time. However, depending on the traveling state of the vehicle, for example, the steering angle may be changed regardless of the steering input by the driver, for example, by traveling on an uneven road surface. In this case, in the above prior art, there is a possibility that the estimation accuracy of the driving state is temporarily lowered.
The present invention has been made paying attention to the above points, and aims to improve the estimation accuracy of the driving state.

To solve the above object, one aspect of the present invention comprises a first driving state estimation unit for estimating a driving state from the operation information of the driver-operable driving operator. Further, according to one aspect of the present invention, the driving state by the first driving state estimation unit using at least one of the operation information of the driving operator that can be operated by the driver, the information on the vehicle state, and the information around the vehicle. A traveling state determination unit that determines a specific traveling state that is a traveling state of the vehicle that is estimated to affect the estimation is provided. According to one aspect of the present invention, a traveling state influence determination unit that determines the influence of the specific traveling state determined by the traveling state determination unit on the driving state estimation performed by the first driving state estimation unit; The second driving state estimation unit that estimates a driving state that is not influenced by the specific driving state than the first driving state estimation unit, and the degree of influence determined by the driving state influence degree determination unit is greater than or equal to a preset reference A state selection unit that selects the estimation by the second driving state estimation unit instead of the estimation by the first driving state estimation unit, in the case of the setting influence degree state that is an extremely high state, The influence degree determination unit determines the influence degree based on the driving state history estimated by the first driving state estimation unit.
Second driving state estimation unit above SL is an operating state without using the control information of the driving operator may estimate when the impact is determined that the setting impact state.

  According to one aspect of the present invention, the driving state is estimated in consideration of the degree of influence of the determined driving state. As a result, according to one aspect of the present invention, the estimation accuracy of the driving state is improved.

It is a figure which shows the structural example of this embodiment. It is a figure which shows the structure of a driving | running state estimation process part. It is a figure which shows the example of an event which concerns on embodiment based on this invention. It is a figure explaining a steering pattern. It is a figure explaining a noise steering pattern. It is a figure explaining the structure of a driving | running state distribution calculation part. It is a figure explaining the process of the driving | running state estimation process part which concerns on 1st Embodiment based on this invention. It is a figure which shows the example of several event generation | occurrence | production. It is a figure explaining the process of the driving | running state estimation process part which concerns on 2nd Embodiment based on this invention. It is a figure explaining the determination of the influence degree which concerns on 2nd Embodiment based on this invention.

“First Embodiment”
First, a first embodiment according to the present invention will be described with reference to the drawings.
(Constitution)
FIG. 1 is a diagram illustrating a configuration of a vehicle on which the driving support device according to the present embodiment is mounted.
As shown in FIG. 1, the vehicle according to the present embodiment includes a steering wheel 1 as a steering indicator, an accelerator pedal opening sensor 2, a brake pedal operation amount sensor 3, a steering angle sensor 4, a wheel speed sensor 5, and a winker detection sensor. 6, meter display 7, navigation device 8, shift sensor 9, G sensor 10, yaw rate sensor 11, light operation detection unit 12, audio operation detection unit 13, operation detection flag 14 of VDC (vehicle dynamics control), LDP (lane departure) (Prevention: Lane departure prevention support system) operation detection flag 14 and controller 100.

The accelerator pedal opening sensor 2 detects the opening amount (acceleration instruction amount) of the accelerator pedal. The detected opening amount is output to the controller 100. The brake pedal operation amount sensor 3 detects an operation amount (braking instruction amount) of the brake pedal as a braking instruction amount. The detected operation amount is output to the controller 100.
The steering angle sensor 4 is an angle sensor attached to the steering column, the steering wheel 1 or the like, for example, and detects the steering angle by the driver's steering from the rotation of the steering shaft. The detected steering angle is output to the controller 100.

  The wheel speed sensor 5 detects the vehicle speed by detecting the number of rotations of the wheel, for example. The detected vehicle speed is output to the controller 100. The wheel speed sensor 5 may detect the vehicle speed based on a signal to the meter display 7. The turn signal detection sensor 6 detects the turn signal state of the turn signal lever. The detected blinker state is output to the controller 100. The shift sensor 9 is provided in a shift lever or a transmission and detects shift position information (shift information). The detected shift position information is output to the controller 100.

The information presenting device outputs an alarm or other presentation by sound or image in accordance with a control signal from the controller 100. The information presentation device includes, for example, a speaker that provides information to the driver by a buzzer sound or voice, and a display unit that provides information by displaying an image or text. For example, the display unit of the navigation device 8 may be used as the display unit.
The navigation device 8 includes a GPS receiver, a map database, a display monitor, and the like, and is a system that performs route search, route guidance, and the like. The navigation device 8 can acquire information such as the type of road on which the host vehicle is traveling and the road width based on the current position of the host vehicle obtained from the GPS receiver and the road information stored in the map database.

  The G sensor 10 detects longitudinal acceleration and lateral acceleration generated in the vehicle. The detected acceleration is output to the controller 100. The yaw rate sensor 11 detects the yaw rate generated by the vehicle and outputs the detected yaw rate to the controller 100. The light operation detection unit 12 detects the operation operation of lighting the light and outputs the information to the controller 100. The audio operation detection unit 13 detects an audio operation and outputs the information to the controller 100. The detection flag 14 of the VDC operation is a flag that is turned on by the operation of the VDC. The LDP operation detection flag 15 is a flag that is turned ON by the operation of the LDP.

The controller 100 is an electronic control unit composed of a CPU, a storage unit such as a ROM and a RAM, and other CPU peripheral components, and executes a preset process by operating a program stored in the storage unit. To do.
The controller 100 includes an operating state estimation processing unit 100A. The driving state estimation processing unit 100A estimates the driving state of the driver from the steering angle information detected by the steering angle sensor 4. Of the processing of the controller 100, the driving state estimation processing unit 100A analyzes the driving characteristics of the driver based on signals detected by the accelerator pedal opening sensor 2, the brake pedal operation amount sensor 3, the steering angle sensor 4, and the like. The degree of driving instability such as the messiness of the driving operation of the driver is determined. Then, a warning or other information is presented to the driver in accordance with the degree of driving instability, and processing for alerting the driver is performed.

  As shown in FIG. 1, the driving support device including the driving state estimation processing unit 100A according to the present embodiment estimates the driving state based on the information of the steering angle sensor 4 and the like, and the estimation results are presented as information presentation devices 7, 8, and 16. Present via. In FIG. 1, visual information presentation and auditory information presentation are illustrated as the presentation by the information presentation device. For visual information presentation, for example, the display unit of the meter display 7 or the navigation device 8 is used. For example, a speaker 16 is used for presenting auditory information.

The driving state estimation processing unit 100A includes a driving state acquisition unit 20, a driving state determination unit 21, a driving state influence degree determination unit 22, a driving state distribution calculation unit 23, a driving state determination unit 24, and an information presentation unit 25.
Here, the driving state estimation processing unit 100A of the present embodiment estimates the driving state from the operation information of the driving operator that can be operated by the driver. In the present embodiment, the operable driver is the steering wheel 1 as a steering operator, and the steering angle information corresponds to the operation information.

The driving state acquisition unit 20 acquires driving state data as operation information of a driving operator that can be operated by the driver. The driving state data is information for determining driving characteristics. The driving state acquisition unit 20 of the present embodiment uses the steering angle information as driving state data.
Here, as information that can be the driving state data in the present invention, in addition to steering angle information, acceleration / deceleration information based on the operation of an accelerator pedal or a brake pedal can be exemplified. The calculation of the operating state distribution (operating state distribution) and the amount of difference between the distributions using these operating state data is described in, for example, the gazette of International Publication No. WO2009 / 013815 (Japanese Patent Application No. 2009-524342). What is necessary is just to calculate by a well-known method.

  The traveling state determination unit 21 determines the traveling state of the vehicle using at least one of operation information of a driving operator that can be operated by the driver, vehicle state information, and vehicle surrounding information. The traveling state determination unit 21 of the present embodiment determines a specific traveling state that is a traveling state that is estimated to affect the driving state estimation obtained from the steering angle information. An event that causes such a specific running state is called an event.

  Here, the “running state estimated to affect the driving state estimation” is estimated to affect the driving operation by an input unrelated to the driving operation of the driver, that is, the driving operation of the driver. This is a traveling state in which operation information is estimated to be disturbed. Since the present embodiment is intended for a steering operation, it refers to a traveling state that occurs in an event that affects the steering input regardless of the driver's steering operation.

The traveling state determination unit 21 according to the present embodiment determines the specific traveling state when detecting the occurrence of any of the events shown in FIG.
Here, as shown in FIG. 3, the types of the above event examples include a first event that requests some operation from the driver, a second event that generates a sprung behavior of the vehicle, and an unsprung behavior of the vehicle. It is classified into a third event to be requested and a fourth event in which a specific vehicle behavior occurs.

Examples of the first event include running on a curved road, changing lanes (eg, turning left and right), accelerating / decelerating operation (eg, braking operation), shifting operation, operation of switch / lever, and tunnel entrance / exit.
Examples of the second event are road swell and acceleration / deceleration.
An example of the third event is a joint passage on the road surface.
Examples of the fourth event are occurrence of side slip, VDC control operation, and LDP control operation.

The traveling on the curved road can be detected by, for example, the steering angle or the lateral G. In the present embodiment, a curved road having a predetermined curvature or more is detected as the event. Here, since the driving state is estimated based on the steering angle information, it is preferable to detect the event using information other than the steering angle.
The lane change is detected by detecting a turn signal operation, for example.
The acceleration / deceleration operation can be detected by detecting an operation of an accelerator pedal or a brake pedal, for example. In the present embodiment, for example, a case where the pedal operation speed is equal to or higher than a preset speed is detected as an event.

The shift operation is detected, for example, by detecting a shift lever operation or shift information.
The operation of the operation switch / lever is detected by detecting the operation of switches that can be operated by the driver, such as a wiper operation signal, a light operation signal, and an audio state change operation.
The entrance and exit of the tunnel are detected by detecting a light operation (ON to OFF or OFF to ON), for example.

The undulation of the road surface is detected by the size of the lateral G detected by the G sensor 10. For example, it is determined that an event has occurred when the preset lateral G is equal to or greater than that.
The occurrence of the acceleration / deceleration is detected by the size of the front and rear G detected by the G sensor 10. For example, it is determined that an event has occurred when the value is greater than or equal to the preset front and rear G.
The joint passage on the road surface is detected based on a change in wheel speed based on the wheel speed sensor 5. An event occurrence is detected when the change in wheel speed per unit time is equal to or greater than a preset threshold value.

The occurrence of the side slip is detected as an event when a yaw rate equal to or higher than a preset threshold value is generated based on the detection value of the yaw rate sensor 11.
The operation of the VDC control is detected as an event occurrence when a VDC operation flag indicating that the VDC control is operating is ON. The operation of the LDP control is detected as an event occurrence when an LPD operation flag indicating that the LPD control is operating is ON.

  In addition, although the case where the detection of an event was detected from the information generated inside the vehicle was illustrated in the above description, it may be detected based on information outside the vehicle. Each event may be detected by acquiring travel environment information based on a combination of map information and GPS information, or an image obtained by capturing the surroundings of the vehicle with a front camera (not shown). For example, a white line may be detected by image processing to detect the curvature of the road, or the position and traveling direction of the vehicle may be detected by a combination of map information and GPS information.

Further, the traveling state influence degree determination unit 22 determines the influence degree of the traveling state determined by the traveling state determination unit 21 to the driving state estimation obtained from the steering angle information.
The traveling state influence degree determination unit 22 according to the present embodiment determines the influence degree when the traveling state determination unit 21 determines the specific traveling state associated with the occurrence of the event. Note that the initial value of the influence degree is “0”. Specifically, as shown in FIG. 4, the traveling state influence determination unit 22 of the present embodiment is set in advance after the traveling state determination unit 21 detects the event (detects a specific traveling state). When a steering pattern within a temporal range (for example, 5 seconds) is acquired and it is determined that the acquired steering pattern is a preset noise steering pattern, the degree of influence is set to “1”. When it is determined that the acquired steering pattern is not a preset noise steering pattern, the influence degree is set to “0”.

  For example, the traveling state influence degree determination unit 22 detects a change in the steering angle for one cycle or a half cycle of several deg from the event detection time as a steering pattern, and a steering pattern defined by its amplitude, frequency, etc., For example, when the steering pattern is equal to or larger than a preset amplitude, it is determined as a noise steering pattern. An example of the noise steering pattern is shown in FIG. For example, in the case of a waveform for one period less than a preset amplitude (eg, several deg) and a preset frequency (eg, 1 Hz) within a preset time (eg, 0.5 seconds) after the occurrence of an event The noise steering pattern is determined. In addition, for example, in the case of a waveform corresponding to a half cycle equal to or lower than a preset frequency within a preset time (for example, 0.5 seconds) after the occurrence of an event, the noise pattern is determined.

  The driving state distribution calculation unit 23 calculates a plurality of driving state distributions having different time ranges while referring to the driving state data acquired by the driving state acquisition unit 20 and the degree of influence determined by the traveling state influence degree determination unit 22. . The driving state distribution calculating unit 23 according to the present embodiment has a first driving state distribution obtained from the steering angle information acquired in a first relatively long time range set in advance, and a time period that is longer than the first driving state distribution. The second operating state distribution obtained in the second temporal range with a short range is calculated.

  At this time, the driving state distribution calculation unit 23 according to the present embodiment is a period during which the influence degree is determined as the set influence degree state when the influence degree determined by the traveling state influence degree determination unit 22 is “1” ( For example, the steering angle information (driving state data) in a predetermined time range (for example, 5 seconds) after the event is detected is excluded, that is, the driving state estimation method is changed to change the second driving state distribution. Perform the calculation.

  Even when the driving state data in the period when the influence level is determined to be the set influence level state is excluded, the length of the time range for obtaining the second driving state distribution is the same as the second time range. It is preferable. However, when excluding the operation state data during the period when the influence degree is determined to be the set influence state, the length of the time range for obtaining the second operation state distribution is different from the second time range. For example, a time range longer than the second time range may be set.

  Here, the first relatively long time range set in advance is a time range in which normal driving characteristics of the target driver can be acquired, and is set to a value of, for example, 30 minutes or more. In addition, the second time range of the second operating state distribution is a time range in which the current driving characteristic (most recent driving characteristic) can be determined, for example, a time range from about 3 minutes before the current time. To do. Each time range described above is merely an example, and may be set based on experiments, theory, or the like based on a steering angle information acquisition cycle. In this embodiment, steering is performed at a preset sampling interval (100 msec). Get corner information.

  Further, the driving state distribution calculation unit 23 stores data stored for calculating each driving state distribution (frequency distribution, etc.) every time a steering angle as driving state data is acquired for each driving state distribution. In addition to updating, each operating state distribution is updated (calculated). However, as described above, at least the second driving state distribution of the first and second driving state distributions is calculated by excluding the driving state data during the period when the influence degree is determined to be the set influence degree state. The process of excluding the above operating state data during the period when the influence degree is determined to be the set influence degree state is, as will be described later, when the set influence degree state and the data before the determination are determined as the set influence state. This is realized by overwriting the data or stopping the update of the data when the setting influence state is determined.

As shown in FIG. 6, the operation state distribution calculation unit 23 includes a first operation state distribution calculation unit 23A, a second operation state distribution calculation unit 23B, a distribution storage unit 23C, a distribution selection unit 23D, and a distribution setting. Part 23E.
Based on the steering angle information (driving state data) that is sequentially updated as described above, the first driving state distribution calculation unit 23A calculates the first driving state distribution in the relatively long first time range as described above. calculate.

Based on the steering angle information (driving state data) sequentially updated as described above, the second driving state distribution calculation unit 23B calculates the second driving state distribution in the relatively short second time range as described above. calculate.
The distribution storage unit 23C repeatedly calculates the third operating state distribution in the temporal range at a preset third time interval (for example, every 5 seconds), and stores the calculated third operating state distribution in the storage unit. To do.

  When the travel state influence degree determination unit 22 determines that the influence degree = “1”, the distribution selection unit 23D stores the third driving state distribution stored before the travel state determination unit 21 detects the event, or the above One of the first operating state distributions is selected. As a result, the driving state estimation is selected. In the present embodiment, since it is determined whether the influence degree is “1” or “0”, for example, the third operating state distribution is selected. Here, when the influence degree is determined in multiple stages, when the influence degree is not less than a preset threshold value and less than the first upper limit threshold value, the third operating state distribution is selected, and the influence degree is the above-described first degree. If it is equal to or greater than the upper threshold value, the first operating state distribution may be set to be selected.

  When the distribution selection unit 23D selects the third operation state distribution that is the operation state distribution, the distribution setting unit 23E changes the second operation state distribution and changes the state estimation method. Specifically, the distribution setting unit 23E replaces the second operating state distribution with the third operating state distribution when the influence degree is “1”. This process is the same as the period set in advance after determining the steering angle information for a preset period (for example, 5 seconds) after determining the specific traveling state, immediately before determining the specific traveling state. This is synonymous with replacing the steering angle information obtained in the time range.

  Here, as described above, the process of replacing the second driving state distribution with the third driving state distribution is performed without using steering angle information for a preset period (for example, 5 seconds) after determining the specific driving state. The second driving state distribution may be obtained using the front and rear steering angle information. In other words, processing for prohibiting updating of stored data for calculating the second driving state distribution (frequency distribution, etc.) is performed until a predetermined period elapses after it is determined as the specific traveling state. May be.

The driving state determination unit 24 determines the driving state based on a difference amount (relative entropy) between the first driving state distribution and the second driving state distribution calculated by the driving state distribution calculation unit 23.
The information presentation unit 25 performs a process of presenting information to the driver based on the driving state determined by the driving state determination unit 24.

Next, the processing of the driving state estimation processing unit 100A will be described with reference to FIG. The process of the driving state estimation processing unit 100A is performed at a preset control cycle (for example, every 100 msec).
In step S110, the driving state estimation processing unit 100A acquires traffic environment information. That is, in step S110, information on the traffic environment in which the host vehicle travels, such as road surface information, road surface input, tunnels, branching / merging, curves, toll gates, road surface inclination, and the like is acquired. In the present embodiment, traffic environment information required to detect the event is acquired.

In step S120, the driving state estimation processing unit 100A acquires vehicle information. In step S120, operation information of the driver's driving operator and vehicle behavior information are acquired. In the present embodiment, the steering angle information for estimating the driving state and the vehicle information required to detect the event are acquired.
Here, in this embodiment, as shown in FIG. 3, the information for detecting an event can be detected from information that the vehicle has. That is, in steps S110 and S120, as described above, information on the operation of the accelerator pedal / brake pedal, the blinker operation, the shift operation, the navigation / audio operation, and the like are acquired as the operation information of the driver's driving operator. Moreover, information, such as a vehicle speed, front-rear G, lateral G, and wheel speed, is acquired as vehicle data system information indicating the vehicle state.

  The driving state estimation processing unit 100A obtains, for example, information on toll gates, tunnels, merging / merging, curves, road surface inclination, and the like from the navigation device 8 as traffic environment information or other road environment information that is information around the vehicle, for example. May be used. Such information can be acquired by using the map database information of the navigation device 8.

Next, in step S130, driver operation information is acquired. In the present embodiment, the driving state acquisition unit 20 acquires steering angle information.
Next, in step S140, the traveling state determination unit 21 determines the traveling state. In the present embodiment, when the event is detected, it is determined that a specific traveling state has occurred.
Next, in step S150, the traveling state influence degree determination unit 22 sets the influence degree = “0” when it is determined in step S140 that the traveling state influence degree is not the specific traveling state. In addition, when it is determined in step S140 that the travel state influence degree determination unit 22 is in the specific travel state, the travel state influence degree determination unit 22 detects the event based on the steering angle information acquired in advance within a set time range (for example, 5 seconds) from the event detection. The obtained steering pattern is acquired, and it is determined whether the acquired steering pattern is a noise steering pattern. When the acquired steering pattern is determined to be a noise steering pattern, “1” is set as the influence degree. Otherwise, “0” is set as the degree of influence even if it is determined as the specific running state.

Next, in step S160, the first driving state distribution obtained from the steering angle information acquired in the first relatively long time range set in advance and the second time range shorter than the first driving state distribution. The second driving state distribution obtained from the steering angle information acquired in the time range is calculated.
It should be noted that every third time interval set in advance (for example, 5 seconds), a third operating state distribution is calculated with the third time interval as a time range, and the calculated third operating state distribution is stored in the storage unit. Perform the process.

  Furthermore, in step S160, the driving state distribution to be employed is selected based on the degree of influence determined by the traveling state influence degree determination unit 22. Specifically, when the influence level = “0”, the process proceeds to step S170 without replacing the second operating state distribution. On the other hand, when it is determined that the influence degree = “1”, the second operating state distribution is replaced with the stored third operating state distribution. Thereafter, the process proceeds to step S170. As described above, when the influence level is “1”, the second driving state distribution may be calculated without using the steering angle information acquired in a preset period from the occurrence of the specific traveling state. .

Here, as described above, the first driving state distribution and the second driving state distribution are updated every time the steering angle information that is the driving state data is acquired.
In step S170, the amount of difference (relative entropy) between the distributions of the first driving state distribution and the second driving state distribution (or the second driving state distribution after replacement if replaced) is calculated by the steering entropy method. . Thereafter, the process proceeds to step S180.

  Specifically, in step S170, based on the steering angle signal when the driver performs the steering operation, how the driver's current driving operation is different from the normal driving operation, that is, the normal driving operation and A difference amount for determining whether or not the state is unstable is calculated. That is, in step S170, relative entropy (feature amount, instability) is calculated as a value representing the unsmoothness of the driving operation. In general, in a state where the driver's attention is not concentrated on driving, the time during which steering is not performed becomes longer than in normal driving where the driving is concentrated, and a large steering angle error is accumulated. Therefore, the corrected steering amount when the driver's attention returns to driving increases. In this embodiment, relative entropy is calculated using this characteristic. Specifically, the steering error distribution (driving state distribution) accumulated for a long time before the past or the present and the current driver's steering error distribution (driving state distribution) acquired for a short time, that is, the temporal range. A plurality of operating state distributions having different values are calculated. Then, the relative entropy is calculated from the long-time steering error distribution and the current short-time operation error distribution, using the long-term steering error distribution as much as the normal driving characteristic as a reference.

Here, the relative entropy is a physical quantity representing a difference amount (distance) between two steering error distributions (driving state distributions), and the degree of difference between the two steering error distributions, that is, how far the two steering error distributions are separated. Represents. Based on the calculated relative entropy value, it is possible to evaluate the stability of the current latest driving state with respect to the past long-time driving state (normal driving characteristics).
The first driving state distribution (steering error distribution) accumulated for a relatively long time, the second driving state distribution of the current driver (steering error distribution) acquired for a relatively short time, and the distribution using the first driving state distribution (steering error distribution). As for the calculation of the difference amount (relative entropy), a known method for obtaining relative entropy, such as a processing method described in Japanese Patent Application Laid-Open No. 2009-9495, may be employed.

In step S170, the driving state is estimated. Specifically, the unstable operation state is determined based on the difference amount (relative entropy). In step S170 of the present embodiment, the calculated difference amount is compared with a preset determination threshold. And when a difference amount is larger than a determination threshold value, it determines with an unstable driving | running state. Thereafter, the process proceeds to step S180.
In step S180, when the state determined to be the unstable operation state in step S170 continues for a predetermined time or more for an unstable determination threshold (for example, 5 seconds), information presentation processing is performed. In addition, you may interrupt information presentation during the period determined as a specific driving state.
As an example of information presentation, a warning is displayed and a warning is displayed with a voice such as “Driving is disturbed.

(Operation other)
When it can be considered that the normal driving characteristics of the driver can be acquired based on the travel time from the start of data collection, the driving state is estimated using the steering entropy method.
At this time, an amount of difference between the first driving state distribution representing the driving characteristics of the driver at the normal time and the second driving state distribution representing the latest driving characteristics is calculated, and the driving state is estimated from the magnitude of the difference amount. Then, it is determined whether or not an unstable operation state is present. This makes it possible to accurately detect an unstable traveling state regardless of the traffic environment. That is, it is possible to detect an unstable state with high accuracy by adapting to an individual's usual characteristics regardless of differences in traffic environment.

At this time, it is desired to detect only the steering angle information (driving state data) for evaluating the instability of the driver due to the driver's operation, but it is caused by the driving state such as the intentional steering of the driver or the road environment. The steering operation may be disturbed due to the influence of the steering input. If the driving state distribution using the steering angle information including this disturbance is used, the detection accuracy of driving instability may be deteriorated. In particular, since the second operating state distribution has a short time range, it is easily affected by the above.
From such an idea, in the present embodiment, it is determined whether or not the vehicle is in a specific traveling state generated by an event that is estimated to disturb the steering operation by the driver.

  Furthermore, even in a specific traveling state due to such an event, there may be a case where the influence on the steering operation is actually small. For this reason, in this embodiment, when the specific running state is determined, the degree of influence on the steering operation by the specific running state is determined, and if the degree of influence is greater than a preset value, the second The driving state distribution is changed, that is, the driving state estimation method is changed.

As described above, in the present embodiment, when the driving state is estimated to be disturbed and the influence of the driving state on the steering operation is large, it is determined as the setting influence state. The operation state instability is determined by replacing the second operation state distribution with the third operation state distribution that does not include the operation state data when the vehicle is operating, that is, by changing the estimation method.
As a result, it is possible to prevent erroneous detection of an unstable driving state by a steering input unrelated to the driver's steering operation, that is, the driver state.

Further, at this time, in this embodiment, when it is determined that the driving state is estimated to be disturbed by the steering operation, the degree of influence due to the driving state is further determined without changing the estimation method unconditionally. The steering angle information at the time of the traveling state in which the steering operation is estimated to be disturbed is not used only when the influence degree is in the set specific influence state.
As a result, it is possible to reduce scenes in which the state estimation method is changed as false detection scenes.

(Modification)
(1) Here, in the above-described embodiment, as the state estimation method to be changed when the determined influence degree is the set specific influence degree state (impact degree = “1”), the above-described specific influence degree and The case where the second operating state distribution is replaced (reconstructed) with the third operating state distribution not including the operating state data at the time of determination is illustrated. The change of the state estimation method is not limited to this. You may employ | adopt the other state estimation method which does not use the driving | running state data at the time of determining with the said specific influence degree. For example, the second operating state distribution is overwritten with the first operating state distribution. That is, a state estimation method of resetting the difference amount may be employed.

(2) At this time, for example, when the influence degree is a specific setting influence degree state (impact degree = “1”), the state estimation to be changed according to the detected event The method may be selected individually.
For example, according to the type of event, the third driving state distribution is adopted for the first type of event, the first driving state distribution is adopted for the second type of event, and the second driving is performed with the adopted driving state distribution. The state estimation method is changed by replacing the state distribution.

  The first type event is assumed to be an event in which the driving state of the driver is assumed to be continued before and after the specific driving state period. The first type event is, for example, an event that occurs when traveling in a specific road environment set in advance. The specific road environment includes the uneven shape of the road surface, the inside of the tunnel, the merging / branching road, the curved road, the vicinity of the toll booth, the interruption of another vehicle to the destination of the host vehicle, the traffic congestion state, and the like. In this case, an unstable state can be determined in a state before a specific influence degree occurs. In this case, the driving state data that causes the erroneous detection is removed by replacing the driving state data when the first influence degree is generated with the previous driving state data. Specifically, data replacement is performed in units of the third time range.

  The second type of event is an event that is assumed that the driving state of the driver changes before and after the specific driving state period. The second type of event is an event that occurs by operating a specific driving operation preset by the driver, for example. Note that the second type event is determined with priority over the first type event. The specific driving operation includes, for example, a lane change operation, an accelerator pedal operation, a brake pedal operation, a winker operation, an operation of the navigation device 8, an audio operation, a light operation, and the like. In this case, since the second operating state distribution acquired before the specific degree of influence occurs cannot be used, the second operating state distribution is replaced with the first operating state distribution having a relatively long time range. The second operating state distribution is reset. In this case, since the first operating state distribution and the second operating state distribution match, the difference amount (relative entropy) is “0”. In this case, the data collection for the second operating state distribution may be newly started from the start of the next third time range after the occurrence of the specific influence degree is finished. .

(3) In the above embodiment, as the determination of the influence degree, the case of the stepwise determination of the influence degree = “0”, “1” is exemplified. The influence degree may be determined as a digital value of three or more levels or a continuous value.
(4) In step S180 of the above embodiment, when the state determined as the unstable operation state continues as the information presentation for a time longer than a preset unstable determination threshold (for example, 5 seconds), the information presentation is performed. Processing is in progress. The content of the information presentation may be changed based on the content, frequency, and degree of influence of the event that generated the specific running state.
For example, according to the content of the detected event, “the vehicle is staggering”, “the steering operation is disturbed”, or the like may be selected.

  Further, as shown in FIG. 8, when a plurality of events occur within a preset period, the information presentation content may be changed according to the number of events. For example, when the number of events that occur per unit time is small, “How about a break” is presented according to the driver's operation status, and when there are many events, the external Present that the vehicle is staggering, taking into account the influence of the factors. At this time, the content of information presentation may be determined according to the event that occurs most frequently among the plurality of events that have occurred.

However, if the number of events that occurred within a preset period is equal to or greater than the preset number, it is estimated that the accuracy of the driving state estimation will be low, so information presentation on the driving state may be prohibited. .
Here, the driving | running state distribution calculation part 23 comprises a 1st driving | running state estimation part and a 2nd driving | running state estimation part. State estimation method when cormorants make changes in the second operating state distribution corresponds to a first driving state estimation unit, a second operational state distribution, or when the first operational state to restore replaced by a third operating state distribution A state estimation method in the case of resetting estimation information by replacing with a distribution corresponds to the second operating state estimation unit . This is also applies to other embodiments.
The degree of influence = “1” corresponds to the influence degree being set in a preset influence degree state.

(Effect of this embodiment)
Next, the effect of this embodiment will be described.
(1) A 1st driving | running state estimation part estimates a driving | running state from the operation information of the driving operator which a driver | operator can operate. The traveling state determination unit 21 determines the traveling state of the vehicle using at least one of the operation information of the driving operator that can be operated by the driver, the vehicle state information, and the information around the vehicle. The traveling state influence degree determination unit 22 determines the influence degree of the traveling state determined by the traveling state determination unit 21 to the driving state estimation performed by the first driving state estimation unit. The second driving state estimation unit estimates the driving state by estimating the driving state different from the first driving state estimation unit. When the influence degree determined by the traveling state influence degree determination part 22 is a preset influence degree state, the distribution selection part 23D replaces the estimation by the first driving state estimation part with the second driving state. Select estimation by the estimation unit.
According to this configuration, even if it is determined that the driving state may be disturbed, the driving state is estimated in consideration of the influence level of the driving state. As a result, the estimation accuracy of the driving state is improved.

(2) The traveling state determination unit 21 determines whether or not the specific traveling state is a traveling state that is estimated to affect the driving state estimation by the first driving state estimation unit. The traveling state influence degree determination unit 22 determines the influence degree when the traveling state determination unit 21 determines the specific traveling state.
According to this configuration, the influence degree is determined in the case of a specific traveling state in which the driving state may be disturbed. As a result, it is possible to more reliably determine whether the traveling state affects the driving state.

(3) The traveling state influence determination unit 22 determines the influence based on vehicle information in a preset time range or a driving state estimated by the first driving state estimation unit.
The behavior of the vehicle according to the running state can be determined by the vehicle information or the driving state estimated by the first driving state estimation unit. And it becomes possible to determine an influence degree reliably by determining an influence degree by the driving state estimated by the vehicle information or the 1st driving state estimation part.

(4) The travel state influence degree determination unit 22 determines the influence degree from the operation information of the driving operator used by the first driving state estimation unit.
According to this configuration, it is possible to determine the degree of influence more reliably by using the information of the driving operator itself that estimates the driving state.
(5) The driving operator used by the first driving state estimation unit is a steering operator that instructs steering. The traveling state influence degree determination unit 22 determines the influence degree based on the steering pattern of the steering operator.
When the driving state is disturbed by an event that generates a specific traveling state, it is estimated that a specific steering pattern occurs. In this way, it is possible to determine the degree of influence based on the steering pattern.

(6) The vehicle information used in the traveling state determination unit 21 is information related to the sprung behavior change and the unsprung behavior change.
According to this configuration, it is possible to detect an event that disturbs the steering input such as the road surface shape and the acceleration / deceleration of the vehicle.
(7) The driving operator used in the travel state determination unit 21 is at least one of the steering wheel 1, the accelerator pedal, the brake pedal, the winker operation instruction unit, and the wiper operation instruction unit.
According to this configuration, it is possible to detect an event caused by a driver's operation.

(8) A 2nd driving | running state estimation part estimates a driving | running state, without using the operation information of the said driving operator when the influence is determined to be a setting influence degree state.
This makes it possible to estimate the driving state without using information that affects the estimation of the driving state.
(9) The information presentation unit 25 provides driving information to the driver based on the estimated driving state. The information presenting unit 25 determines driving information to provide information according to the degree of influence determined by the traveling state influence degree determining unit 22.
According to this configuration, it is possible to provide appropriate information according to the degree of influence.
(10) The information presenting unit 25 responds to the frequency at which the influence degree determined by the traveling state influence degree determination unit 22 is determined as the preset influence degree state or the traveling state determined by the traveling state determination part 21. Determine driving information to provide information, or decide to stop providing information.
According to this configuration, it is possible to provide appropriate information according to the degree of influence.

“Second Embodiment”
Next, a second embodiment will be described with reference to the drawings. In addition, about the structure similar to the said embodiment, the same code | symbol is attached | subjected and demonstrated.
The basic configuration of this embodiment is the same as that of the first embodiment. However, in the present embodiment, the processes of the traveling state influence determination unit 22 and the distribution selection unit 23D are different.
When the traveling state determining unit 21 determines that the traveling state influence unit 21 is in the specific traveling state, the traveling state influence determining unit 22 of the present embodiment sets a preset elapsed time (for example, 5 seconds) preset from the start of the specific traveling state, that is, the detection of the event. Referring to the amount of difference (feature amount) between the first driving state distribution and the second driving state distribution until the time when the specific running state is determined (immediately after the event occurs) The amount of increase or the maximum value is taken as the impact level.

The distribution selection unit 23D selects the first state estimation method when the influence degree is less than the first influence degree threshold value set in advance, and the influence degree is equal to or greater than the first influence degree threshold value. When the influence level is less than the second influence degree threshold value greater than the first influence degree threshold value, the second state estimation method is selected, and when the influence degree is equal to or greater than the second influence degree threshold value, the third state estimation is performed. Select a method.
The distribution setting unit 23E performs the second operation state distribution changing process based on the state estimation method selected by the distribution selection unit 23D. Specifically, when the distribution selection unit 23D selects the first estimated state, the second operation state distribution changing process is not performed. When the distribution selection unit 23D selects the second estimated state, the second operating state distribution is replaced with the third operating state distribution. When the distribution selection unit 23D selects the third estimated state, the second operation state distribution is replaced with the first operation state distribution.

Next, the process of the driving | running state estimation process part 100A of this embodiment is demonstrated, referring FIG. The process of the driving state estimation processing unit 100A is performed at a preset control cycle (for example, every 100 msec).
Here, steps S110 to S140 and steps S170 to S180 are the same as those in the first embodiment, and a description thereof will be omitted.
Further, the present embodiment is different from the first embodiment in that the process of step S200 is added.
In step S200, a process of storing the difference amount (relative entropy, feature amount) between the distributions obtained in step S170 for at least the latest preset period is performed. The preset period is, for example, 10 seconds.

  In step S150 of the present embodiment, when the event is detected in step S140 and the specific running state is determined, the past difference amount stored in step S200 is referred to determine the specific running state, and then set in advance. The maximum difference amount within the set period (for example, 5 seconds) is acquired as the detected difference amount. Then, the difference between the detected difference amount and the difference amount immediately before the specific traveling state occurs is obtained as the influence degree. The degree of influence made up of the difference in the amount of difference indicates that the greater the degree of influence, the greater the influence on steering by the specific traveling state.

Here, as shown in FIG. 10, when steering is disturbed due to the occurrence of an event, the amount of difference represented by relative entropy tends to increase. Therefore, it is possible to determine the degree of influence from the increase in the amount of difference. In the above description, the case is described in which the degree of influence is obtained from the increment of the difference amount due to the occurrence of the event.
In step S160, the first driving state distribution obtained from the steering angle information acquired in the first relatively long time range set in advance and the second temporal state having a shorter time range than the first driving state distribution. The second driving state distribution obtained from the steering angle information acquired in the range is calculated.
It should be noted that every third time interval set in advance (for example, 5 seconds), a third operating state distribution is calculated with the third time interval as a time range, and the calculated third operating state distribution is stored in the storage unit. Perform the process.

Furthermore, in step S160, the driving state distribution to be employed is selected based on the degree of influence determined by the traveling state influence degree determination unit 22. Specifically, when the influence degree is equal to or greater than a preset influence degree threshold, the second driving state distribution is replaced with the third driving state distribution stored before determining the specific running state. Alternatively, when the second driving state distribution is calculated, the steering angle information is acquired in the second time range except for the steering angle information within a preset period (for example, 5 seconds) after determining the specific traveling state. The second driving state distribution is obtained from the steering angle information.
When it is determined that the influence level is equal to or greater than the threshold threshold value that is larger than the influence threshold value, the second operating state distribution is replaced with the first operating state distribution, and the relative value is reset (initialized).
Other processes are the same as those in the first embodiment.

(Operation other)
When it can be considered that the normal driving characteristics of the driver can be acquired based on the travel time from the start of data collection, the driving state is estimated using the steering entropy method.
At this time, an amount of difference between the first driving state distribution representing the driving characteristics of the driver at the normal time and the second driving state distribution representing the latest driving characteristics is calculated, and the driving state is estimated from the magnitude of the difference amount. Then, it is determined whether or not an unstable operation state is present. This makes it possible to accurately detect an unstable traveling state regardless of the traffic environment. That is, it is possible to detect an unstable state with high accuracy by adapting to an individual's usual characteristics regardless of differences in traffic environment.

At this time, it is desired to detect only the steering angle information (driving state data) for evaluating the instability of the driver due to the driver's operation, but it is caused by the driving state such as the intentional steering of the driver or the road environment. The steering operation may be disturbed due to the influence of the steering input. If the driving state distribution using the steering angle information including this disturbance is used, the detection accuracy of driving instability may be deteriorated. In particular, since the second operating state distribution has a short time range, it is easily affected by the above.
From such an idea, in this embodiment, it is determined whether the vehicle is in a specific traveling state due to an event that is estimated to disturb the steering operation by the driver.

  Furthermore, even in a specific traveling state due to such an event, there may be a case where the influence on the steering operation is actually small. For this reason, in this embodiment, when the specific running state is determined, the degree of influence on the steering operation by the specific running state is determined, and if the degree of influence is greater than a preset value, the second The driving state distribution is changed, that is, the driving state estimation method is changed.

As described above, in the present embodiment, when the driving state is estimated to be disturbed and the influence of the driving state on the steering operation is large, it is determined as the setting influence state. The operation state instability is determined by replacing the second operation state distribution with the third operation state distribution that does not include the operation state data when the vehicle is operating, that is, by changing the estimation method.
As a result, it is possible to prevent erroneous detection of an unstable driving state by a steering input unrelated to the driver's steering operation, that is, the driver state.

Further, at this time, in this embodiment, when it is determined that the driving state is estimated to be disturbed by the steering operation, the degree of influence due to the driving state is further determined without changing the estimation method unconditionally. The steering angle information at the time of the traveling state in which the steering operation is estimated to be disturbed is not used only when the influence degree is in the set specific influence state.
As a result, it is possible to reduce scenes in which the state estimation method is changed as false detection scenes.
Here, the case where the influence degree is equal to or larger than the influence degree threshold corresponds to the influence degree being in a preset influence degree state.

(Effect of this embodiment)
In addition to the effects of the first embodiment, the following effects are achieved.
(1) The traveling state influence degree determination unit 22 determines the influence degree based on the driving state history estimated by the first driving state estimation unit.
It is estimated that the more disturbed the driving state is, the higher the influence degree due to the driving state is. As described above, the influence degree can be determined based on the history of the driving state.
(2) The traveling state influence degree determination unit 22 is characterized in that the influence degree is determined based on the temporal change of the driving state.
When the driving state is disturbed by the occurrence of an event, it is estimated that a temporal change occurs in the driving state. As described above, the influence degree can be determined based on the temporal change of the driving state.

As described above, the entire contents of the Japanese Patent Application 2012-138809 (filed on June 20, 2012) to which the present application claims priority form part of the present disclosure by reference.
Although the present invention has been described with reference to a limited number of embodiments, the scope of rights is not limited thereto, and modifications of each embodiment based on the above disclosure are obvious to those skilled in the art.

1 Steering wheel (driver)
20 driving state acquisition unit 21 traveling state determination unit 22 traveling state influence determination unit 23 driving state distribution calculation unit ( first driving state estimation unit, second driving state estimation unit)
23A 1st driving | running state distribution calculation part 23B 2nd driving | running state distribution calculation part 23C Distribution memory | storage part 23D Distribution selection part (state selection part)
23E Distribution setting unit 24 Driving state determination unit 25 Information presentation unit 100 Controller 100A Driving state estimation processing unit

Claims (10)

A first driving state estimating unit that estimates a driving state from operation information of a driving operator that can be operated by the driver;
It is estimated that the driving state estimation by the first driving state estimation unit is affected by using at least one of the operation information of the driving operator that can be operated by the driver, the vehicle state information, and the vehicle surrounding information. A traveling state determination unit that determines a specific traveling state that is a traveling state of the vehicle to be operated;
A driving state influence determination unit that determines the influence of the specific driving state determined by the driving state determination unit on the driving state estimation performed by the first driving state estimation unit;
A second driving state estimation unit that estimates a driving state that is not affected by the specific driving state than the first driving state estimation unit;
In the case of a setting influence degree state in which the influence degree determined by the traveling state influence degree determination unit is higher than a preset reference, the second driving state is used instead of the estimation by the first driving state estimation unit. A state selection unit for selecting estimation by the estimation unit;
With
The running state influence judging unit, the first driving state estimation unit driving state estimation apparatus, wherein a determining influence on the basis of the history of the operating state estimating. A first driving state estimating unit that estimates a driving state from operation information of a driving operator that can be operated by the driver;
It is estimated that the driving state estimation by the first driving state estimation unit is affected by using at least one of the operation information of the driving operator that can be operated by the driver, the vehicle state information, and the vehicle surrounding information. A traveling state determination unit that determines a specific traveling state that is a traveling state of the vehicle to be operated;
A driving state influence determination unit that determines the influence of the specific driving state determined by the driving state determination unit on the driving state estimation performed by the first driving state estimation unit;
A second driving state estimation unit that estimates a driving state that is not affected by the specific driving state than the first driving state estimation unit;
In the case of a setting influence degree state in which the influence degree determined by the traveling state influence degree determination unit is higher than a preset reference, the second driving state is used instead of the estimation by the first driving state estimation unit. A state selection unit for selecting estimation by the estimation unit;
With
The second driving state estimation unit, impact settings influence state determination has been to estimate the operating state without using the operation information of the OPERATION operator driving state estimation device according to claim when that . The running state influence judging unit, the first driving state estimation device described to determine the degree of influence based on the operating state of estimating the 請 Motomeko 2 you characterized in driving state estimation unit. The running state influence judging unit driving state estimation apparatus according to claim 2, characterized in that to determine the influence from the operation information of the driving operating element the first driving state estimation unit is used. The driving operator used by the first driving state estimating unit is a steering operator that instructs steering,
The driving state estimation device according to claim 4, wherein the traveling state influence degree determination unit determines the influence degree based on a steering pattern of the steering operator.   The driving state estimation device according to claim 1, wherein the traveling state influence degree determination unit determines the influence degree based on a temporal change of the driving state.   The driving state according to any one of claims 1 to 6, wherein the vehicle state information used in the traveling state determination unit is information on a sprung behavior change and an unsprung behavior change. Estimating device.   8. The driving operator used in the traveling state determination unit is at least one of a steering wheel, an accelerator pedal, a brake pedal, a winker operation instruction unit, and a wiper operation instruction unit. The driving | running state estimation apparatus described in any 1 item | term. An information presentation unit that provides driving information to the driver based on the driving state estimated by the first driving state estimation unit or the second driving state estimation unit;
The said information presentation part determines the driving information to provide information according to the influence degree which the said driving state influence degree determination part determined, It described in any one of Claims 1-8 characterized by the above-mentioned. Driving state estimation device.   The information presenting unit provides driving according to the frequency at which the degree of influence determined by the traveling state influence degree determining unit is determined as the preset influence degree state or the driving state determined by the traveling state determining unit. 10. The driving state estimation apparatus according to claim 9, wherein information is determined or information provision is stopped.

Images