JP2015064584A - Driver feeling estimation device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow estimation of a driver state indicating a feeling of a driver.SOLUTION: In a region specified by use of environment difficulty required for driving operation and driving skill of a driver, a "pleasant region PA" in which the driving skill is improved on the basis of pleasure and security sense is set as a driver state, a "boring region PC" in which a feeling of boredom is felt on a side of lower environment difficulty than the "pleasant region PA" is set, and an "uneasy region PB" in which uneasiness is felt on a side of higher environment difficulty than the "pleasant region PA" is set. The driver state is estimated according to a position in which the environment difficulty required for the driving operation of the driver calculated according to external environment of a vehicle and the driving skill evaluated on the basis of the driving operation of the driver are positioned.

Description

The present invention relates to a driver emotion estimation apparatus and method.

Vehicles tend to be equipped with driving assistance devices that assist the driving of the driver. For example, Patent Document 1 proposes assistance for driving operation according to the driver's driving skill, and Patent Document 2 determines information to be presented to the driver based on the driver's recognition state. What to do has been proposed. As a result, the driver can perform driving accurately with driving assistance.

JP 2005-319872 A Japanese Patent Laid-Open No. 2005-4414

However, in any of the above patent documents, the driver's feelings, intentions, thoughts, etc. when receiving driving assistance are not taken into account, and driving assistance according to the setting is executed by ignoring them. Is done. For this reason, even if a driver with low driving skill wants to improve driving skill or is not satisfied with driving assistance, the driving assistance device will provide driving assistance without permission. Take away the opportunity to improve.

On the other hand, this inventor pays attention to the fixed relationship between a specific emotion and the improvement of driving skill. Among specific emotions, especially enjoyment (self-fulfillment) is between the improvement of driving skill and, as shown in FIG. 1, "It is fun because driving skill improves and driving skill improves" This is because if the driver creates a loop-like relationship and makes the driver enjoyable, the driving skill can be improved effectively. It is also possible to contribute to the reduction of CO2 and the reduction of traffic accidents through the effective improvement of driving skills.

The present invention has been made in view of such circumstances, and a technical problem thereof is a driver emotion estimation apparatus and method capable of estimating a driver state, that is, a driver's emotion state. Is to provide.

In order to achieve the above technical problem, the following solution is adopted in the device of the present invention. That is, as described in claim 1,
An environmental difficulty level calculation means for calculating the environmental difficulty level required for the driving operation of the driver by the external environment of the vehicle;
Driving skill evaluation means for evaluating the driving skill based on the driving operation of the driver;
A driver state estimating unit that estimates a driver state with respect to an external environment from the environmental difficulty level calculated by the environmental difficulty level calculating unit and the driving skill level evaluated by the driving skill evaluation unit;
It is supposed to be equipped with.

Preferred embodiments based on the above solution are as described in claims 2 to 6 in the claims. That is,
The driver state estimating means estimates whether or not the driver state is a state where the driver has a feeling of fun (corresponding to claim 2).

In the coordinate system set by the environmental difficulty level and the driving skill, further has a storage means that divides the driver state into regions,
The driver state estimation means estimates the driver state by comparing the storage difficulty calculated by the environmental difficulty calculation means and the driving skill evaluated by the driving skill evaluation means with the storage means. ,
(Corresponding to claim 3).

  As the area of the driver state stored in the storage means, a “fun area” having a feeling that the driver is fun and a driver having a feeling that the driver is bored in a direction having a lower environmental difficulty than the fun area. The “boring area” and the “anxiety area” in which the driver has a feeling of anxiety are set to those who have a higher environmental difficulty than the “fun area” (corresponding to claim 4).

The environmental difficulty level is calculated in consideration of vehicle driving information in addition to vehicle external information,
The driving skill is one of a skill related to an image of a driver's vehicle trajectory, a skill related to a vehicle width interval image, and a skill related to a driving operation.
(Corresponding to claim 5).

In order to achieve the technical problem, the following solution is adopted in the method of the present invention. That is, as described in claim 6,
A first step of calculating an environmental difficulty level required for a driver's driving operation according to an external environment of the vehicle;
A second step of evaluating the driving skill based on the driving operation of the driver;
A third step of estimating the driver state with respect to the external environment from the environmental difficulty calculated in the first step and the driving skill evaluated in the second step;
It is supposed to be equipped with.

ADVANTAGE OF THE INVENTION According to this invention, the driver | operator state which a driver | operator feels with respect to an external environment can be estimated using the environmental difficulty of an external environment and a driver | operator's driving skill.

The conceptual diagram which shows the idea of this invention. Explanatory drawing which shows the vehicle carrying the driving skill training apparatus which concerns on embodiment. The block diagram which shows the input-output relationship of the driving skill training apparatus which concerns on embodiment. Explanatory drawing which illustrates notionally the control content of the vehicle-mounted control apparatus which concerns on embodiment. Explanatory drawing explaining a driver emotion map and driving assistance. Explanatory drawing explaining correction of driving assistance. Explanatory drawing explaining the relationship between assist amount, driving skill, and "fun area". Explanatory drawing explaining the correction method of a driver emotion map (anxiety area | region, pleasant area | region, and boundary line AL1). Explanatory drawing explaining the method of correction | amendment of a driver emotion map (a boring area | region, a pleasant area | region, and boundary line AL2). The block diagram which shows the driving skill training apparatus (vehicle-mounted control apparatus) which concerns on embodiment. The figure which shows an example of the driver emotion map (Map A, Map B, Map C) about each driving skill. Explanatory drawing which shows each state example of the difficulty of a parking lot. The flowchart which shows the example of control of the vehicle-mounted control apparatus which concerns on embodiment. 14 is a flowchart showing a continuation of FIG. Explanatory drawing explaining the method of calculating an assist amount in the driver emotion map about the driving skill regarding the image of a vehicle locus. Explanatory drawing explaining the method of calculating an assist amount in the driver emotion map about the driving skill regarding the image of a vehicle width interval. Explanatory drawing explaining the method of calculating an assist amount in the driver emotion map about the driving skill regarding a driving operation. The figure which shows the parking result display example in a display apparatus.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings, taking as an example a case where the embodiment is used for driving support.
First, a conceptual embodiment of the overall driving assistance of the present invention will be described.
In FIG. 2, reference numeral 1 denotes a vehicle (own vehicle), and the vehicle 1 is equipped with a driving support device 2 as a driving skill training device that trains driving skills. As shown in FIG. 3, the driving support device 2 includes an external information acquisition device 3, a vehicle information acquisition device 4, a driver action acquisition device 5, a driver emotion acquisition device 6, an information communication device 7, an in-vehicle device. Thus, the driving support device 2 can perform parking assistance, lane keeping assistance, inter-vehicle distance assistance, and the like as driving assistance.

The external information acquisition device 3 acquires external information such as parking spaces and obstacles. Specifically, a camera or a radar is used as the external information acquisition device 3. The vehicle information acquisition device 4 acquires vehicle information such as vehicle speed, steering angle, accelerator opening, and brake switch signal. Specifically, sensors such as a steering angle sensor are used as the vehicle information acquisition device 4. It is done. The driver action acquisition device 5 acquires the driver's line of sight as driver action information. Specifically, a vehicle interior camera or the like is used as the driver action acquisition device 5. The driver emotion acquisition device 6 acquires a driver's emotion based on the driver's facial expression, voice, sweating, brain waves, intracerebral blood flow, and the like. A vehicle interior camera is used to acquire the driver's facial expression. . In the present embodiment, one in-vehicle camera 14 serves as the driver action acquisition device 5 and the driver emotion acquisition device 6 (see FIG. 2). The information communication device 7 transmits / receives information to / from an external service center (infrastructure) 9. The in-vehicle control device 8 includes an information processing device 10 and an information storage device 11. The information processing apparatus 10 receives various information as input information from the external information acquisition apparatus 3, the vehicle information acquisition apparatus 4, the driver action acquisition apparatus 5, the driver emotion acquisition apparatus 6, and the information communication apparatus 7 and performs driving support. The control or display signal is output to the vehicle drive device 12 such as a steering assist actuator or a brake assist actuator or the information display device 13 such as a display. In addition, the information storage device 11 includes a driver action data unit 11a, a driver skill data unit 11b, and a driver emotion map unit 11c (see FIG. 10 described later), each of which stores specific data that they handle. To do.

The vehicle-mounted control device 8 is supposed to realize driving support in the “fun area PA” that is felt enjoyable for the driver based on the processing there. This is because driving assistance that feels fun brings about an improvement in driving skill with respect to the assisted driving (see FIG. 1).

First, the conceptual contents of the control of the in-vehicle control device 8 will be described with reference to FIG.
The in-vehicle control device 8 performs processing of a short period period and processing of a long period period longer than the short period period.

The purpose of the short-term period processing is to determine whether or not driving assistance is necessary or what kind of driving assistance is appropriate when driving assistance is performed. The processing of the period cycle is performed by interrupt processing every 100 msec, for example.

In the process of this short-term period, first, the environmental difficulty level calculation is performed in S (S indicates a step) 51. That is, in S51, detection data detected by the external information acquisition device 3 and the vehicle information acquisition device 4 is read, and external information such as obstacles and the environmental difficulty level associated with the external data are shared from the shared data DB1. The corresponding relationship data is read. Then, by using the detection data as the correspondence data, the environmental difficulty level corresponding to the current situation outside the vehicle is calculated, and the environmental difficulty level is stored in the shared data DB1.
As another aspect, when calculating the environmental difficulty level, the in-vehicle control device 8 directly acquires the difficulty level of the external environment from the shared data DB 1 or the external service center, or, for example, in the case of parking assistance, As information, parking space (vacant) space information, vehicle trajectory for parking in the parking space, the number of turnovers, the presence or absence of obstacles in the parking space, etc. are obtained from an external service center or directly from them, and externally based on it The environmental difficulty level may be calculated. In this case, as a result of calculating the external environment difficulty level, those that should be shared with the calculated data are transmitted to the external service center for use by other vehicles.

Next, in S52, the driving skill stored in the shared data DB2 is read. This driving skill is calculated based on the long-term period processing described later after the data detected by the driver action acquisition device 5 is stored in the shared data DB 2, and is again shared. It is stored in the data DB2. In the present embodiment, for example, the most recent stored driving skill is used (read).
The shared data DB1, DB2 may be provided in the information storage device 11 of the in-vehicle control device 8, or may be provided in the external service center 9 on condition that transmission / reception is performed via the information communication device 7, and environmental difficulty is also high. The calculation of the degree and the calculation of the driving skill may be performed at an external service center.

When the calculation of the environmental difficulty level and the reading of the driving skill are finished, selection and execution of driving assistance are performed in S53. In selecting the driving assistance, first, a driver emotion map (in which a “fun area PA” is set in the coordinate system of the environmental difficulty level and the driving skill) is read from the individual data DB 4. As shown in FIG. 5, the driver emotion map is for calculating an assist (driving support) amount aa, and is created according to the individual driver in the long-term period processing described later. . Specifically, as shown in FIG. 5, the driver emotion map is a map of driver emotions with respect to driving skill and environmental difficulty, and in the driver emotion map, the driver receives driving assistance and is fun. The “fun area PA” that feels like, the “anxiety area PB” that feels anxiety, and the “bored area PC” that feels bored are divided by boundary lines AL1 and AL2.

Next, when selecting driving assistance, an assist amount aa is obtained in the driver emotion map. In obtaining the assist amount aa, the initial coordinate point P1 is obtained from the current environmental difficulty level DR1 of the vehicle obtained in S51 and the current driving skill DT1 of the driver read in S52, and the driving skill DT1 is calculated. On the broken line, the difference to the reference value P5 provided on the center line AL3 of the boundary lines AL1 and AL2 of the “fun area PA” is obtained. This difference is the assist amount aa.

Next, the driving assistance content corresponding to the assist amount aa is selected and executed. That is, according to the environmental difficulty level DR1 and the driving skill DT1, the environmental difficulty level can be substantially reduced by the assist amount aa corresponding to the driving skill, so that the assist amount aa (selection according to the driving skill of the driver is selected) The driver is positioned in the “fun area PA” by executing the driving assistance content), so that the driver can feel fun. As shown in FIG. 5, since there is a relationship of environmental difficulty = assist amount + driving skill, the driving skill of the driver can be obtained by subtracting the assist amount from the environmental difficulty level.

Regarding the driving assistance contents corresponding to the assist amount aa, when the driving assistance is, for example, parking assistance, the driving skill can be classified into a skill related to the image of the vehicle trajectory, a skill related to the image of the vehicle width interval, and a skill related to the driving operation. Accordingly, the assist amount aa also corresponds to them. In this case, the assist amount aa corresponds to the fineness of the display of the vehicle trajectory until parking, the display timing, and the like for the skill relating to the image of the vehicle trajectory, and the assist amount increases as the display fineness increases and the display timing becomes earlier. aa increases. That is, as the assist amount increases, the target locus and advice are displayed from an early stage when the deviation from the target locus is small. Regarding the skill relating to the image of the vehicle width interval, the timing of displaying the distance display between the vehicle 1 and the obstacle corresponds to the assist amount aa, and the assist amount aa increases as the display timing becomes earlier. That is, as the assist amount increases, the inter-vehicle distance is displayed from the time when the inter-vehicle distance is large. Regarding the skill related to the driving operation, the operation timing of each automatic operation system of the vehicle corresponds to the assist amount aa, and the assist amount aa increases as the operation timing becomes earlier. That is, as the assist amount increases, the operation timing of automatic steering and automatic braking is advanced.

Further, regarding the assist amount aa, when the driver's emotion was measured for each driving skill by changing the assist amount aa, it was as shown in FIG. 7 under a certain environmental difficulty level. According to FIG. 7, in any case of each driving skill, the lower the driving skill, the larger the assist amount aa (see the example of the arrow in FIG. 7) required to belong to the “fun area PA”. Showed a tendency to become. In this case, the “fun area PA” is an area in which the driver's emotions are detected for “fun” and “reassurance”, and the uppermost part of the two overlaps.

When driving support is executed in S53, the emotion / psychology of the driver at that time is estimated in S54. This is because it is determined whether or not driving support is appropriate through the driver's feelings, and whether or not driving support is corrected. In this case, the shared data DB 3 stores correspondence data between the driver's emotion / psychology and the driver's biometric information in advance, and the in-vehicle control device 8 captures the correspondence data and adds the driver's emotion acquisition device. 6 is used to estimate the driver's emotion / psychology at that time.
At this time, the acquired driver's emotion and the assist amount aa at that time are stored in the shared data DB 3 and the individual data DB 4.
In this case, the in-vehicle control device 8 receives information on the correspondence between emotion and biological information from the external service center, and estimates the driver's emotion based on the driver's ecological information and information from the external service center. It is good.

When the driver's emotion / psychology is estimated in S54, it is determined in S55 whether or not the driver feels pleasure. This is because effective driving skills cannot be improved unless the driver feels fun. In S55, when it is estimated that the driver feels fun, the processing of the assist amount aa is terminated without being corrected, assuming that the driver emotion map (see FIG. 5) is suitable for the driver. . On the other hand, if the driver does not feel fun in S55, the process proceeds to S56 and the driving assistance is corrected.

In the correction of the driving support in S56, for example, as shown in FIG. 6, when it is detected that the driver emotion map is not suitable for the driver and the driver's emotion is in an “anxiety” state at the reference point P5. (The black point in FIG. 6 indicates that the reference point P5 is in an “uneasy” state), and this and the assist amount aa at that time are both stored in the individual data DB4. Then, a second assist amount aa + Δaa is obtained based on a predetermined correction amount Δaa, the process proceeds to S53, and driving assistance based on the second assist amount aa + Δaa is executed again. Is estimated in S54. In this way, S54 to S56 and S53 are repeatedly executed until the driver emotion is estimated to be “fun”, and the driver emotion and the assist amount at that time are stored in the individual data DB4. (In FIG. 6, the reference point P5 (black point) at the first assist amount aa is in the “anxiety” state, the reference point P5 ′ (black point) at the second assist amount aa + Δaa is in the “anxiety” state, and the third assist amount aa + 2Δaa. The reference point P5 ″ (white point) in “represents a“ fun ”state).
As a result, the driver can receive driving assistance and drive in a fun state, and the driving skill of the driver gradually increases from driving skill DT1 to DT2, as shown in FIG.

Next, the process of the long period period by the vehicle-mounted control apparatus 8 is demonstrated.
The processing of the long period cycle is aimed at creating and correcting the emotion map of the driver, and the creation and correction of the driver emotion map is performed, for example, every week by using the data ( (Accumulated data). For this reason, the vehicle-mounted control device 8 stores data for each process of the short-term period every time driving assistance is performed.

This will be specifically described. First, as shown in FIG. 4, in S61, the environmental difficulty level for each driving support implemented during the past week is read from the shared data DB1. This environmental difficulty level is stored in the shared data DB 1 as one of accumulated data at the time of each driving support (processing in a short-term period).

Next, the process proceeds to S62, and the driving skill evaluation at the time of each driving support performed during the past one week is performed. The in-vehicle control device 8 stores model data at that time, vehicle speed according to the operation of the driver, and the like for each driving support performed during the past week, and these are read from the shared data DB 2. The proximity of the operation of the driver to the model data is determined by comparison with the model data. The higher (closer) this proximity is, the higher the driving skill of the driver is, and it is stored in the shared data DB 2. For example, if the driving assistance is parking assistance, the amount of deviation from the model data when parked in the parking space is calculated, and the smaller the deviation, the higher the driving skill of the driver is evaluated, and this is saved. Is done.
Instead of the above method, the driver emotion map stored in the individual data DB 4 may be obtained by subtracting the assist amount from the environmental difficulty level according to “environment difficulty level = assist amount + driving skill”.

Next, the process proceeds to S63 where emotion / psychological estimation is performed. In the in-vehicle control device 8, the emotion data of each driver at that time is stored in the shared data DB 3 for each driving support performed during the past week, and these are read out, and the in-vehicle control device 8 The emotion data is divided into “fun”, “bored”, and “anxiety”. For this reason, the driver's facial expression may be detected by a vehicle interior camera or the like during driving assistance (when processing in a short period cycle), and “fun” and “boring” may be detected by the vehicle interior camera. Detection and “anxiety” may be detected by a sweat sensor, brain waves, voice or the like.

Next, the process proceeds to S64, and the driver emotion map is corrected so that the driver's individual characteristic learning, that is, the driver emotion map matches each driver. The driver emotion map is corrected by calculating the environmental difficulty (S61), driving skill evaluation (S62), and data based on emotion / psychological estimation (S63) (such as driving support data stored during one week). Is used, and the driver emotion map is revised every week.

Specifically, it is as follows. For example, as shown in FIG. 8, when driving assistance for a certain environmental difficulty level, the driving skill at the time of driving assistance is DT1 (evaluated in S62), and the coordinates are P5 due to the initial assistance amount aa. However, even if the driver's emotion is in the “anxiety” state (represented by ● in FIG. 8) and the coordinates are P5 ′ due to the second assist amount aa + Δaa, the driver's emotion is in the “anxiety” state (FIG. 8). When the driver's emotion is in a “fun” state (represented by a circle in FIG. 8) for the first time when the coordinate is P5 ″ due to the third assist amount aa + 2Δaa (FIG. 5). In the driving skill DT1, it is determined that the boundary between the “fun area PA” and the “anxiety area PB” exists between P5 ′ and P5 ″.

Such processing is similarly performed for other driving assistance data (such as driving assistance data stored for one week). When driving assistance for another environmental difficulty level, when the driving skill at the time of the driving assistance is DT2 (evaluated in S62), the driver's feelings may be felt even if the coordinates are P6 due to the initial assist amount aa. Even if the driver is in an “anxiety” state (represented by ● in FIG. 8) and the coordinates become P6 ′ due to the second assist amount aa + Δaa, the driver's emotion is in an “anxiety” state (represented by ● in FIG. 8). Yes, only when the coordinates become P6 ″ due to the third assist amount aa + 2Δaa, when the driver's emotion is in a “fun” state (indicated by a circle in FIG. 8), in the driving skill DT2, the “fun area PA” And “anxiety region PB” are determined to exist between P6 ′ and P6 ″.

When such processing is finished, the processing result is used, and the boundary line AL1 between the “fun area PA” and the “anxiety area PB” of the driver's emotion map is P5 ′ and P5 ″ in the above example. And the boundary line AL1 ′ connecting the intermediate point between P6 ′ and P6 ″ is stored in the individual data DB4. In the subsequent short-term cycle processing, the driving assistance assist amount is calculated from a driver emotion map in which each region is divided by the corrected boundary line AL1 'and the original boundary line AL2.

Contrary to the above, when the boundary line AL2 between the “fun area PA” and the “bored area PC” does not match the driver, the following processing is performed. For example, as shown in FIG. 9, when driving assistance for a certain environmental difficulty level, the driving skill at the time of driving assistance is DT1 (evaluated in S62), and the coordinates become P5 by the initial assistance amount aa. However, even if the driver's emotion is in the “bored” state (represented by ▲ in FIG. 9) and the coordinates are P5 ′ due to the second assist amount aa−Δaa, the driver's emotion is in the “bored” state ( The driver's emotion became “fun” (represented by a circle in FIG. 9) only after the coordinates became P5 ″ due to the third assist amount aa−2Δaa. Sometimes, in the driving skill DT1, it is determined that the boundary between the “fun area PA” and the “bored area PC” is between P5 ′ and P5 ″.

Such processing is similarly performed for other driving assistance data (such as driving assistance data stored for one week). When driving assistance for another environmental difficulty level, the driving skill at the time of driving assistance is DT2 (evaluated in S62), and the driver's emotions even if the coordinate is set to coordinate P7 by the first assist amount aa Is in a “bored” state (represented by ▲ in FIG. 9), and the driver's emotion is in a “bored” state (in FIG. 9, ▲ in FIG. 9) even if the coordinates become coordinates P7 ′ by the second assist amount aa−Δaa. When the driver's emotion is in a “fun” state (represented by a circle in FIG. 9) for the first time when the coordinates become coordinate P7 ″ by the third assist amount aa−2Δaa. In DT2, it is determined that the boundary between the “fun area PA” and the “bored area PC” exists between P7 ′ and P7 ″.

When such processing is finished, the processing result is used, and the boundary line AL2 between the “fun area PA” and the “boring area PC” of the driver's emotion map is P5 ′ and P5 ″ in the above example. And the boundary line AL2 ′ connecting the intermediate points of P7 ′ and P7 ″ are stored in the individual data DB4. In the subsequent short-term cycle processing, the driving assistance assist amount is calculated by a driver emotion map in which each region is divided by the corrected boundary line AL2 'and the original boundary line AL1.

(Embodiment in parking support)
Next, an embodiment in which parking assistance is used as the driving assistance will be described.
In the present embodiment, the driving support device 2 is set to perform parking support as driving support, and the driving support device 2 includes an external information acquisition device 3 as specifically shown in FIG. A camera outside the vehicle (hereinafter, the same symbol 3 is used), various sensors for acquiring vehicle information (hereinafter, the same symbol 4 is used) as the vehicle information acquisition device 4, a vehicle interior camera 14 as the driver action acquisition device 5, and a driver emotion acquisition device 6 includes a vehicle interior camera 14 and a sweat sensor 6a, and an information communication device 7 is provided with a transmitter 7a and a receiver 7b that perform transmission and reception with the transceiver 16 of the parking lot service center 15. Furthermore, a GPS (Global Positioning System) sensor 17 is also provided to detect the position of the vehicle.
Further, the in-vehicle control device 8 has the specific configuration shown in FIG. 10, and based on the processing there, parking support in the “fun area PA” that is felt to be fun for the driver is to be realized. . This is because the parking support that feels fun brings about an improvement in driving skill regarding parking (see FIG. 1).

The in-vehicle control device 8 having the above contents includes a storage unit 18, a parking space / obstacle detection unit 19, and a learning process unit 20 of the vehicle 1.
The storage unit 18 stores the vehicle size of the host vehicle and the host vehicle position information received from the GPS sensor 17, and the host vehicle information stored in the storage unit 18 is transmitted to the parking lot service center by the transmitter 7a. 15 transmitters / receivers 16 are transmitted.
The parking space / obstacle detection unit 19 receives parking (vacant) space position data of the parking lot, obstacle information on the route, etc. by the transceiver 15 and the receiver 7b of the parking lot service center 15, and also outside the vehicle Obstacle information is also input from the camera 3, and based on this information, the parking space / obstacle detection unit 19 detects a parking space and an obstacle.

The learning process unit 20 of the vehicle 1 includes a driver action recording / analysis unit 21, a comparison / analysis unit 22 with model data, a driver skill evaluation unit (driving skill evaluation means) 23, and a driver emotion map creation / correction unit ( Setting means) 24, and in the present embodiment, these are operated at intervals of one week.

The driver action recording / analysis unit 21 includes various information such as detection result information from the parking space / obstacle detection unit 19, vehicle position information from the GPS sensor 17, vehicle speed from various sensors 4, steering angle, and accelerator opening. The vehicle information and the driver's line-of-sight position information from the in-vehicle camera 14 (driver action acquisition device 5) are input, and the driver action record / analyzer 21 calculates driver action characteristics from these driver action information and vehicle information. Is required. The driver behavior characteristics and the like are recorded (saved) in the driver behavior data section 11a in the information storage device 11, and the driver behavior characteristics and the like are taken out as necessary.

The comparison / analysis unit 22 with the model data is input with data from the driver behavior data unit 11a via the driver behavior recording / analysis unit 21 and model data from the parking lot service center 15 (models up to parking). Data) is input, and the comparison / analysis unit 22 with the model data compares and analyzes both data.

The driver skill evaluation section (driving skill evaluation means) 23 receives analysis result information from the comparison / analysis section 22 with model data, and the driver skill evaluation section 23 evaluates the skill of the driver. In the present embodiment, three types of skills are evaluated as a driver's skill: a skill that imagines a vehicle trajectory when parking, a skill that imagines a vehicle width interval, and a skill of driving operation. The skill to image the vehicle trajectory is determined based on the amount of deviation from the model data (trajectory), and the skill to image the vehicle width interval is determined based on the deviation from the parking space partition line during parking. The The driving skill is determined based on the amount of deviation from the timing of the model data at the time of the driving operation. These skills are stored in the driver skill data section 11b, and are extracted from the driver skill data section 11b as necessary.

The driver emotion map creation / correction unit (setting means) 24 inputs the driver skill evaluation information in the driver skill evaluation unit 23, the parking difficulty level (environmental difficulty level) in the parking difficulty level calculation unit 25 described later, and the like. At the same time, driver emotion information from the vehicle interior camera 14 and the sweat sensor 6a (driver emotion acquisition device 6, emotion detection means) is input via the driver emotion detection unit 28, and based on these information, driver emotion information is input. As shown in FIG. 11, as a map, an emotion map (hereinafter referred to as a map A) regarding a skill that images a vehicle trajectory during parking, and an emotion map (hereinafter referred to as a map B) regarding a skill that images a vehicle width interval. Then, an emotion map (hereinafter referred to as map C) regarding the skill of the driving operation is created (corrected). Such maps A, B, and C are stored in the driver emotion map unit 11c in the information storage device 11, and are extracted as necessary.

The in-vehicle control device 8 includes a parking difficulty level calculation unit (parking difficulty level calculation unit) 25, an assist amount calculation unit (driver state identification unit, driving support selection unit) 27, a driver emotion detection unit 28, and an assist amount correction unit. (Driver state specifying means, driving support restoration means) 29, and a feedback model teaching unit 30 (display instruction means).

Detection result information of the parking space / obstacle detection unit 19 is input to the parking difficulty calculation unit 25, and the parking difficulty calculation unit 25 is based on the detection result information of the parking space / obstacle detection unit 19. Then, the trajectory from the position of the parking space to the host vehicle is estimated, and the parking difficulty level is calculated from the number of turnovers and the presence or absence of obstacles around the parking space. This parking difficulty level calculation may be performed by the parking lot service center 15 and received by the in-vehicle control device 8. The output information from the parking difficulty level calculation unit 25 is output to the information display device 13, and the display device 13 is based on the output information from the parking difficulty level calculation unit 25 as shown in FIG. In this case, the parking difficulty level in the parking space may also be displayed.

Data (driving skill) of the driver skill data unit 11b is input to the assist amount calculation unit 27 via the driver skill evaluation unit 23, and each map data of the driver emotion map unit 11c is a driver emotion map creation / correction unit. The parking difficulty level obtained by the parking difficulty level calculation unit 25 is also input. In the assist amount calculation unit 27, as shown in FIG. 11, on the maps A, B, and C based on the environmental difficulty level calculated by the parking lot difficulty level calculation unit 25 and the driving skills in the driver skill data unit 11b. The initial coordinate point P1 is plotted. When the initial coordinate point P1 is located in the “fun area PA” in each map, the driver is in a fun state and the driver's driving skill can be improved. Although no control or display signal is output to the device 12 or the display device 13 and driving assistance is not performed, as shown in FIG. 11, when the initial coordinate point P1 is not located within the “fun area PA” in each map, driving is performed. As it is necessary to provide support, the difference between the initial coordinate point P1 and the reference value of the “fun area PA” under the current driving skill of the driver is obtained as the assist amount aa (arrow in FIG. 11). In order to perform driving support according to the assist amount aa, a control or display signal is output from the assist amount calculation unit 27 to the vehicle drive device 12 or the display device 13.Thereby, as driving assistance, the vehicle drive device 12 performs assist driving such as automatic braking, automatic steering, and automatic accelerator, and the display device 13 performs assist display.

In addition to signals from the vehicle interior camera 14 and the sweat sensor 6a, an output signal from the assist amount calculation unit 27 is input to the driver emotion detection unit 28, and the driver emotion detection unit 28 receives the assist amount calculation unit 27. Detects the driver's emotions when driving assistance based on the assist amount aa calculated by.

The assist amount correction unit 29 receives the output information from the driver emotion detection unit 28 (unpleasant, driver feelings such as anxiety). Based on the output information, the assist amount calculation unit 27 described above assists. When the driver emotion is not determined to be in the “fun” state even after the execution of the amount aa (see FIG. 11), as shown in FIG. 6, the assist amount is corrected until the driver emotion is in the “fun” state (aa + Δaa, aa + 2Δaa...), the assist amount correcting unit 29 outputs a control or display signal to the vehicle drive device 12 or the display device 13 in order to perform driving support corresponding to the corrected assist amount. Of course, at this time, detection by the driver emotion detection unit 28 is performed to determine the driver emotion, and the detection result is stored.

Output information from the driver emotion detection unit 28 is input to the feedback model teaching unit (display instruction unit) 30, and the feedback model teaching unit 30 causes the display device 13 to display a parking result. Then, the analysis contents of the driving result of the driver are displayed.

A control example of the vehicle driving skill training device (on-vehicle control device) will be specifically described based on the flowcharts shown in FIGS. 13 and 14. In addition, the code | symbol S in FIG. 13, FIG. 14 shows a step.

In the driving support device 2 (vehicle-mounted control device 8), two controls with different execution timings are performed. One is performed with a long-period period of every week, and the other is performed with a short-term period of every 100 msec.

First, control under a short period period every 100 msec will be described. The vehicle information is transmitted to the transceiver 16 of the parking lot service center 15 (management server) as the infrastructure every 100 msec from the start of control (S1, S2). The own vehicle information includes the own vehicle position, the vehicle size, and the like. Based on this, the parking lot service center 15 transmits parking space information and model data (vehicle trajectory for parking in the parking space, information on the number of turn-backs, etc.), and the vehicle receives this (S3). Thereafter, the vehicle 1 detects position information of obstacles (vehicles, etc.) around the parking space by the outside camera 3 (S4), and based on the information of S4 and the information from the parking lot service center 15, The difficulty level of the parking lot is calculated (S5). 12A to 12D show an example of calculation of the difficulty level of the parking lot. The difficulty increases as it goes from FIG. 12A to FIG. 12D, and the determination of the difficulty is based on the parking space information from the parking lot, the vehicle trajectory for parking in the parking space, the number of turnovers, and obstacles around the parking space. This is done based on information such as the presence or absence of objects. 12A to 12D, the hatched portion is a space where the host vehicle 1 is parked.

When the difficulty level of the parking lot is calculated, the driving skill of the driver is read from the driver skill data calculated and stored for each long-term period (S6). In the present embodiment, as the driving skill of the driver, an image of a vehicle trajectory, an image of a vehicle width interval, and a driving operation are read. And it is discriminate | determined whether this driver | operator's driving skill is below a predetermined value (S7). This is because when the driving skill is insufficient, it is necessary to assist for safety regardless of the driver's feelings. The predetermined value of S7 as a determination criterion is set from this viewpoint. For this reason, when S7 is YES, in S8, driving assistance of automatic driving is forcibly performed, and the process ends. On the other hand, when S7 is NO, the process proceeds to S10, and the assist amount is calculated from the driver emotion maps A, B, and C thereafter.

In S10, it is determined whether or not the coordinate point P1 composed of each driving skill of the driver in S6 and the parking difficulty level for the parking space in the parking lot belongs to the “fun area PA” in the map A of FIG. . This is to determine whether or not it is necessary to provide driving assistance regarding the skill of the image of the vehicle trajectory.

When S10 is determined to be YES and the initial coordinate point P1 is in the “fun area PA” of the map A, the driver is in a fun state (self-fulfilled state), so no assistance is provided for the vehicle trajectory, and the process proceeds to S11. Transition. In S11, it is determined whether or not the initial coordinate point P1 belongs to the “fun area PA” in the map B in order to determine whether or not driving assistance regarding the image of the vehicle width interval should be performed. When this S11 is YES, it is a time when the initial coordinate point belongs to the “fun area PA” in the map B. At this time, the assist to the vehicle interval is not performed, and the process proceeds to S12. In S12, it is determined whether or not the initial coordinate point P1 belongs to the “fun area PA” in the map C in order to determine whether or not driving assistance related to the driving operation should be performed. When this S12 is YES, it is a time when the initial coordinate point P1 belongs to the “fun area PA” in the map C. At this time, it is determined that there is no necessity for assisting the driving operation, and the parking completion determining step in S13. Proceed to

On the other hand, when S10 is NO, it is necessary to provide driving assistance with respect to the image of the vehicle trajectory, and the assist amount aa related to the image of the vehicle trajectory corresponds to the environmental difficulty level of the initial coordinate point P1 and the “fun area in the map A”. Based on the difference from the reference value of “PA” (see FIG. 15), assistance regarding the image of the vehicle trajectory according to the magnitude of the assist amount aa is calculated (S14). Specifically, the assist amount aa relating to the image of the vehicle trajectory has a corresponding relationship with, for example, the fineness of display of the vehicle trajectory until parking (vehicle trajectory, steering angle), and the display timing thereof. As the value of aa increases, the display of the vehicle trajectory becomes finer and the display timing thereof becomes earlier. FIG. 12 shows an example of driving assistance related to the image of the vehicle trajectory. The vehicle trajectory (arrow) and the steering angle are displayed according to the degree of driving assistance, and the display timing is adjusted.

After S14, the process proceeds to S11, where it is determined whether or not the initial coordinate point P1 is within the “fun area” of the map B. When this S11 is NO, it is necessary to provide driving assistance regarding the image of the vehicle width interval, and in S16, a map B shown in FIG. As shown in FIG. 4, the assist amount aa is calculated, and the driving assistance regarding the image of the vehicle width interval according to the assist amount aa is calculated (S16). Specifically, the assist amount aa related to the vehicle width interval is in correspondence with the display timing of the distance between the vehicle and the obstacle displayed by the corner sensor and the front, side, and rear cameras. As the assist amount aa is larger, the display timing is earlier.

After S16, the process proceeds to S12, in which it is determined whether or not the initial coordinate point P1 is within the “fun area” of the map C. When this S12 is NO, it is a time when the initial coordinate point does not belong to the “fun area PA” in the map C and driving assistance should be provided for the driving operation. At this time, the map C in FIG. As shown, the assist amount aa is calculated. (S17). Specifically, the assist amount aa related to the vehicle width interval has a corresponding relationship with the operation timing of the automatic operation system of the vehicle. The larger the value of the assist amount aa, the earlier the operation timing of the automatic operation system. .

Then, the driving assistance obtained in S14, S16, S17 is executed in S15. At that time, in order to confirm the effect, the driver's emotion / psychology is estimated, and the parking lot difficulty, the assist amount, The driving skill and driver feeling are stored (S18). The estimation of the driver emotion / psychology is performed by detecting emotions such as enjoyment and security from the vehicle interior camera 14 and the sweat sensor 6a (driver emotion acquisition device 6), and the detection result and the emotion / psychology estimation in the information storage device 11 are detected. The degree of emotion of the driver is specified by comparing and collating with the data.
The execution timing of “assist for vehicle trajectory” by map A executed in S15, “assist for vehicle width interval” by map B, and “assist for driving operation” by map C are different, respectively. Since the emotion / psychological estimation in S18 is performed immediately after each assist is performed, it is estimated whether each assist is “fun”, “anxiety”, or “bored”.

After S18, in S27, it is determined whether the driver emotion is estimated to be “fun” for all the assists. When S27 is YES, the process proceeds to S13. When S27 is NO, the process proceeds to S19, and it is determined whether or not NO in S27 is immediately after the vehicle track assist is performed. When S19 is YES, it is determined that the assist amount based on the map A is inappropriate, and the process proceeds to S22, where the assist amount is determined based on the predetermined correction amount as described with reference to FIG. Will be corrected. When the process of S22 is completed, the process proceeds to S20, and in S20 as well, it is determined whether or not NO in S27 is immediately after the vehicle width interval assist is performed. When S20 is YES, it is determined that the assist amount based on the map B is inappropriate, and the process proceeds to S23, where the assist amount is determined based on the correction amount determined in advance according to the map B as shown in FIG. Will be corrected. When the process of S23 is completed, the process proceeds to S21. In S21 as well, it is determined whether or not NO in S27 is immediately after the execution of the assist operation. When S21 is YES, it is determined that the assist amount based on the map C is inappropriate, and the process proceeds to S24, where the assist amount is determined based on the correction amount determined in advance according to the map C as described in FIG. Will be corrected. After S24, the process proceeds to S15 again, and the processes after S15 are repeatedly executed as described above.
When S19, S20, and S21 are NO, the assist amount is not corrected assuming that the assistance by the map A, the map B, and the map C is appropriate.
Thus, the assist amount is corrected until YES is determined in S27, and if YES is determined in S27, the process proceeds to S13.

When the process proceeds to the parking completion determination step of S13, it is determined whether or not parking is completed. When S13 is NO, the process returns to S18. When S13 is YES, in S25, the driver's action data from the start of parking (when transmission / reception to / from the parking lot transmitter / receiver starts) until the completion of parking is performed. It is memorize | stored in the action data part 11a. In this case, as the driver action data, specifically, a difficulty level of the parking lot, a signal such as a vehicle speed according to the driver's operation, model data at that time, driver feeling, and the like are stored.

When the action data of the driver is stored (S25), the parking result is displayed in the next S26. For example, FIG. 18 shows an example of the parking result in the display device 13. On the screen, problems and weak points are analyzed from the driver's driving result (shown by solid line: f1) and the ideal vehicle trajectory (dotted line: f0), and driving points are advised. Specifically, in the left half of the screen of the display device 13, for example, the steering operation is slow in B1, the speed is too high when backing in B2, the steering operation amount is large in B3, and the right rear in B4 In the right half of the screen of the display device 13, the evaluation of the driving skill related to the parking result is indicated by the magnitude of the value, and the reason for the evaluation and the parking result are displayed. Based on the guidance (response) content, further alerting content is displayed.
Further, by improving the driving skill by displaying a display that visually indicates how much the driving skill is now as the character CH in the screen grows.

Next, control under a long-term period of every week will be described with reference to FIGS. In the first week from the start of control or in the first week from the previous execution of control, the action data of the driver of the vehicle is read from the driver action data section 11a (S1 ', S2'). Specifically, the driver behavior data section 11a reads the difficulty level of the parking lot, a signal such as the vehicle speed according to the driver's operation, model data at that time, the driver's emotion at that time, and the like. When the driver action data and the like are read, the driving skill of the driver is calculated based on the data, and the driving skill of the driver is sent to the next S4 'and stored in the driver skill data section 11b (S3'). In S4 ', a driver's emotion map is created (corrected) in the same manner as described with reference to FIGS. As described above, the driver emotion map is created (corrected) for each driver based on the calculated driver driving skill, driver emotion, and the like, and is stored in the driver emotion map unit 11c. Such a series of processes is repeated again after one week.

As described above, in this embodiment, the driver can drive in a fun state with respect to parking, and can improve the driving skill based on self-realization at that time. In addition, even when the driving skill (initial coordinate point P1) with respect to the environmental difficulty level is located in the bored region, by making the driving operation difficult, that is, by gradually reducing the assist amount, Driving skill can be improved.

1 Vehicle 2 Driving support device (vehicle driving skill development device)
3 External camera (external information acquisition device)
6 Driver emotion acquisition device (Emotion detection means)
6a Sweating sensor (Emotion detection means)
8 On-vehicle controller 14 Car interior camera (Emotion detection means)
23 Driver skill evaluation section (driver skill evaluation means)
24 Driver emotion map creation / correction part (setting means)
25 Parking difficulty level calculator (environmental difficulty level calculation means)
27 Assist amount calculation unit (driver state specifying means, driving support selecting means)
28 Driver Emotion Detection Unit (Emotion detection means)
29 Assist amount correction unit (driver state specifying means, driving support selecting means)
30 Feedback model teaching unit (display instruction means)
PA Fun Area (Skill Improvement Area)
P1 Initial coordinate point (driver state)
P2 Transition coordinate point (driver state)
aa Assist amount (deviation amount)

Claims (6)

An environmental difficulty level calculation means for calculating the environmental difficulty level required for the driving operation of the driver by the external environment of the vehicle;
Driving skill evaluation means for evaluating the driving skill based on the driving operation of the driver;
A driver state estimating unit that estimates a driver state with respect to an external environment from the environmental difficulty level calculated by the environmental difficulty level calculating unit and the driving skill level evaluated by the driving skill evaluation unit;
A driver emotion estimation device comprising: In claim 1,
The driver state estimation device estimates whether or not the driver state is a state having a feeling that the driver is happy as the driver state. In claim 1,
In the coordinate system set by the environmental difficulty level and the driving skill, further has a storage means that divides the driver state into regions,
The driver state estimation means estimates the driver state by comparing the storage difficulty calculated by the environmental difficulty calculation means and the driving skill evaluated by the driving skill evaluation means with the storage means. ,
A driver emotion estimation device characterized by that. In claim 3,
As the area of the driver state stored in the storage means, a “fun area” having a feeling that the driver is fun and a driver having a feeling that the driver is bored in a direction having a lower environmental difficulty than the fun area. A driver emotion estimation apparatus characterized in that a “boring region” and an “anxiety region” in which the driver has a feeling of anxiety are set in a higher environmental difficulty than the “fun region”. In any one of Claims 1 thru | or 4,
The environmental difficulty level is calculated in consideration of vehicle driving information in addition to vehicle external information,
The driver emotion estimation apparatus, wherein the driving skill is one of a skill related to an image of a driver's vehicle trajectory, a skill related to a vehicle width interval image, and a skill related to a driving operation. A first step of calculating an environmental difficulty level required for a driver's driving operation according to an external environment of the vehicle;
A second step of evaluating the driving skill based on the driving operation of the driver;
A third step of estimating the driver state with respect to the external environment from the environmental difficulty calculated in the first step and the driving skill evaluated in the second step;
A driver emotion estimation method characterized by comprising:

Images