Driving assistance control device

Abstract

To provide a driving support control device capable of coping with various driving scenes, and executing proper control according to a risk felt by an individual driver.SOLUTION: A driving support control device includes a storage part for storing a risk level showing the degree of risk felt by a driver, to an obstacle around a vehicle, and learning data of a risk factor which is a factor feeling a risk, and a control part for acquiring data of an obstacle around the vehicle, and outputting, when an obstacle is detected, control information for reducing a risk factor to the obstacle based on learning data, to a system for controlling driving of the vehicle.SELECTED DRAWING: Figure 8

Description

[Technical field]

[0001] The present invention relates to a driving assistance control device. [Background technology]

[0002] In recent years, research and development of autonomous driving technology and driving support technology has been promoted for the purpose of preventing and reducing accidents and reducing the driver's burden. In autonomous driving technology or driving support technology, it is desirable to obtain driving results that the driver can feel safe with. For this reason, various methods have been proposed, such as a method of selecting an optimal driving trajectory taking into account obstacles around the vehicle. In addition, Patent Document 1 proposes a driving support device that sets a driving trajectory taking into account potential risks such as obstacles jumping out from a blind spot.

[0003] Here, the anxiety felt by drivers is subjective, depending on personal experience and sensitivity, as well as how the driver perceives the vehicle's surroundings. For example, when driving on a narrow residential road, some drivers feel a strong risk from utility poles protruding into the road, while others feel a strong risk from objects jumping out from blind spots. For this reason, even if driving conditions are determined based only on physical risks, the driving results obtained may not necessarily be reassuring for all drivers.

[0004] In response to this, Patent Document 2 discloses a driving control device that learns driving operations according to a driving scene for each driver and executes automatic driving control by referring to the learning results. Specifically, Patent Document 2 proposes a driving control device in which a plurality of items for identifying a driving scene of a vehicle are set, and during manual driving control, the driving scene is identified by a combination of environmental item values ​​detected for environmental items, and the driving operation of the driver is learned in association with the driving scene, and during automatic driving control, the environmental item values ​​are detected to identify the driving scene, and automatic driving control is executed by referring to the learning results of the driving operation in the driving scene. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2018-192954 A [Patent Document 2] JP 2015-089801 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, a technology using a method of learning driving scenes as described in Patent Document 2 may not be able to handle unlearned driving scenes. For example, in a scene where a vehicle passes a bicycle approaching from the opposite direction while traveling on a narrow road, if the learned driving scene only includes "the vehicle and the bicycle," if a traffic participant such as a "pedestrian" is added to the vehicle and the bicycle, or if the behavior of the vehicle and the bicycle is different even when there is only the vehicle and the bicycle, or if the surrounding environment is different even when there is only the vehicle and the bicycle, such a scene does not exist in the learned driving scenes, and it is not possible to determine what kind of driving control should be performed.

[0007] The present invention has been made in consideration of the above problems, and an object of the present invention is to provide a driving assistance control device that can respond to various driving situations and execute appropriate control according to the risk perceived by each driver. [Means for solving the problem]

[0008] In order to solve the above problems, according to one aspect of the present invention, a driving assistance control device is provided that includes a memory unit that stores learning data on a risk level indicating the degree of risk perceived by a driver from obstacles around the vehicle and risk factors that cause the driver to perceive the risk, and a control unit that acquires data on obstacles around the vehicle and, when an obstacle is detected, outputs control information to a system that controls the driving of the vehicle for reducing the risk factors from the obstacle based on the learning data.

[0009] In addition, in the above-mentioned driving assistance control device, when multiple obstacles are detected, the control unit may output control information that prioritizes an operation for reducing the risk factor for an obstacle with a high risk level based on learning data.

[0010] In addition, in the above-mentioned driving assistance control device, the vehicle itself is a vehicle capable of executing autonomous driving control, and when multiple obstacles are detected, the control unit may set conditions for autonomous driving control by prioritizing operations for reducing risk factors for obstacles with a high risk level based on learning data, and output information on the control conditions.

[0011] In addition, in the above-mentioned driving assistance control device, the control unit may use a risk map that indicates the risk potential present in obstacles around the vehicle to set a driving trajectory within a drivable route that has a risk that is below an allowable risk that is set depending on the vehicle speed of the vehicle. Effect of the Invention

[0012] As described above, according to the present invention, it is possible to respond to various driving situations and to execute appropriate driving control according to the risk felt by each individual driver. [Brief description of the drawings]

[0013] [Figure 1] 1 is a block diagram showing an example of the configuration of a vehicle control system according to an embodiment of the present invention; [Diagram 2] FIG. 10 is an explanatory diagram showing a risk potential for an object. [Diagram 3] FIG. 11 is an explanatory diagram showing an example of a risk map reflecting potential risks. [Figure 4] FIG. 11 is an explanatory diagram showing an example in which the tolerable risk is increased by reducing the vehicle speed. [Diagram 5] FIG. 11 is an explanatory diagram showing an example in which a reference path can be set as a selected path. [Figure 6]FIG. 11 is an explanatory diagram showing an example of setting a selected path so that a risk value is less than an acceptable risk. [Figure 7] FIG. 13 is an explanatory diagram showing an example of deceleration so that the risk value is less than the allowable risk. [Figure 8] FIG. 11 is an explanatory diagram showing an example in which the gaze ratio of a gaze object and risk factors are reflected in a risk potential. [Figure 9] FIG. 13 is an explanatory diagram showing an example in which a driver's internal characteristics are reflected in the acceptable risk. [Figure 10] 5 is a flowchart showing a learning data storage process performed by the driving assistance control device according to the embodiment; [Figure 11] 2 is an explanatory diagram showing an example of the configuration of data stored by the driving assistance control device according to the embodiment; FIG. [Figure 12] 5 is a flowchart showing a control condition setting process performed by the driving assistance control device according to the embodiment; [Figure 13] 10 is a flowchart showing a process for setting a travel trajectory and a vehicle speed. [Figure 14] FIG. 11 is a block diagram showing a modified example of a driving assistance control device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In this specification and the drawings, components having substantially the same functional configurations are designated by the same reference numerals, and duplicated explanations will be omitted.

[0015] <1. Example of vehicle control system configuration> First, an example of the configuration of a vehicle control system including a driving assistance control device according to an embodiment of the present invention will be described.

[0016] FIG. 1 is a block diagram showing an example of the configuration of a vehicle control system 1 including a driving assistance control device 50 according to this embodiment. The vehicle control system 1 includes an in-vehicle camera 21, an outside-vehicle camera 23, a surrounding environment sensor 25, a vehicle operation / behavior sensor 27, a GPS antenna 29, a map data storage unit 31, an HMI (Human Machine Interface) 43, a vehicle drive control unit 45, and a driving assistance control device 50. The in-vehicle camera 21, the outside-vehicle camera 23, the surrounding environment sensor 25, the vehicle operation / behavior sensor 27, the GPS antenna 29, the map data storage unit 31, the HMI 43, and the vehicle drive control unit 45 are each connected to the driving assistance control device 50 directly or via a communication means such as a CAN (Controller Area Network) or a LIN (Local Inter Net).

[0017] (1-1. In-car camera) The in-vehicle photographing camera 21 generates image data photographing the interior of the vehicle. The in-vehicle photographing camera 21 includes an imaging element such as a CCD (Charge Coupled Device) or a CMOS (Complementary Metal Oxide Semiconductor), and transmits the generated image data to the driving assistance control device 50. The in-vehicle photographing camera 21 is installed so that at least the driver of the vehicle is within the photographing range. Only one in-vehicle photographing camera 21 may be installed, or multiple in-vehicle photographing cameras 21 may be installed.

[0018] (1-2. Exterior camera) The outside-vehicle camera 23 generates image data capturing an image of the surroundings of the vehicle. The outside-vehicle camera 23 may be mounted as a safety feature of the vehicle. The outside-vehicle camera 23 includes an imaging element such as a CCD or a CMOS, similar to the inside-vehicle camera 21, and transmits the generated image data to the driving assistance control device 50. The outside-vehicle camera 23 is made up of one or more cameras provided on the vehicle to capture images in at least one of the directions in front, on the sides, and behind the vehicle.

[0019] (1-3. Ambient environment sensor) The surrounding environment sensor 25 is a sensor that detects people and obstacles around the vehicle. The surrounding environment sensor 25 includes, for example, one or more of a high-frequency radar sensor, an ultrasonic sensor, and a LiDAR. The detected obstacles include other vehicles, bicycles, buildings, utility poles, traffic signs, traffic signals, natural objects, and any other objects that exist around the vehicle. The surrounding environment sensor 25 transmits a sensor signal including the detected data to the driving assistance control device 50.

[0020] (1-4. Vehicle operation / behavior sensor) The vehicle operation / behavior sensor 27 is composed of at least one sensor that detects the operation state and behavior of the vehicle. The vehicle operation / behavior sensor 27 includes at least one of a vehicle speed sensor, an acceleration sensor, and an angular velocity sensor, and detects information on the behavior of the vehicle, such as the vehicle speed, longitudinal acceleration, lateral acceleration, and yaw rate. The vehicle operation / behavior sensor 27 also includes at least one of an accelerator position sensor, a brake stroke sensor, a brake pressure sensor, a steering angle sensor, an engine rotation speed sensor, a brake lamp switch, and a blinker switch, and detects information on the operation state of the vehicle, such as the steering angle of the steering wheel or the steering wheels, the accelerator opening, the amount of brake operation, the on / off of the brake lamp switch, and the on / off of the blinker switch. The vehicle operation / behavior sensor 27 also includes a driving mode changeover switch, and detects setting information of the automatic driving mode or the manual driving mode. The vehicle operation / behavior sensor 27 transmits a sensor signal including the detected information to the driving assistance control device 50.

[0021] (1-5. GPS antenna) The GPS antenna 29 receives satellite signals from GPS (Global Positioning System) satellites. The GPS antenna 29 transmits position information on the map data of the vehicle, which is included in the received satellite signals, to the driving assistance control device 50. Note that instead of the GPS antenna 29, an antenna for receiving satellite signals from another satellite system for identifying the position of the vehicle may be provided.

[0022] (1-6. Map data storage section) The map data storage unit 31 is configured with a storage element such as a RAM (Random Access Memory) or a ROM (Read Only Memory), or a storage medium such as a HDD (Hard Disk Drive), a CD (Compact Disc), a DVD (Digital Versatile Disc), an SSD (Solid State Drive), a USB (Universal Serial Bus) flash, a storage device, etc., and stores map data. The stored map data includes data on a reference path, which is a trajectory that serves as a reference when traveling on each road. A storage medium that stores map data of a navigation system can be used as the map data storage unit 31.

[0023] (1-7.HMI) The HMI 43 is driven by the driving assistance control device 50 and notifies the driver of various information by means of image display, audio output, etc. The HMI 43 includes, for example, a display device and a speaker provided in an instrument panel. The display device may be a display device of a navigation system. The HMI 43 may also include a HUD (head-up display) that displays information on the front window by superimposing it on the scenery around the vehicle.

[0024] (1-8. Vehicle drive control unit) The vehicle drive control unit 45 includes at least one control system that controls the drive of the vehicle. A vehicle to which the driving assistance control device 50 according to this embodiment can be applied is configured to be able to execute at least automatic driving control. The vehicle drive control unit 45 includes an engine control system or a motor control system that controls the drive force of the vehicle, an electric steering system that controls the steering angle of the steering wheel or the steering wheels, and a brake system that controls the braking force of the vehicle. The vehicle drive control unit 45 may also include a transmission system that changes the speed of the output output from the engine or the drive motor and transmits it to the drive wheels.

[0025] (1-9. Driving assistance control device) The driving assistance control device 50 executes control to assist the driver in driving the vehicle. In this embodiment, the driving assistance control device 50 drives the vehicle by automatic driving based on image data transmitted from the exterior camera 23, detection data of the surrounding environment transmitted from the surrounding environment sensor 25, data on the operation state and behavior of the vehicle transmitted from the vehicle operation / behavior sensor 27, and information on the position of the vehicle on the map data transmitted from the GPS antenna 29. Specifically, the driving assistance control device 50 acquires data of the reference path stored in the map data storage unit 31, sets the driving conditions of the vehicle so that the vehicle does not come into contact with pedestrians or obstacles (hereinafter, pedestrians and obstacles are collectively referred to as "obstacles" unless otherwise noted), and transmits a control command to the vehicle drive control unit 45.

[0026] In this embodiment, the driving assistance control device 50 uses a risk level indicating the degree of risk felt by the driver for obstacles around the vehicle and learning data of risk factors that are factors that cause the driver to feel the risk to drive the vehicle in an autonomous driving mode while reducing the risk felt by the driver. In this case, when an obstacle is detected, the driving assistance control device 50 outputs control information for reducing the factors of risk felt by the driver for the obstacle based on the learning data. In other words, the driving assistance control device 50 executes control so that not only the obstacle is avoided but also an operation is performed to reduce the factors that cause the driver to feel risk for the obstacle. This realizes autonomous driving control that the driver can feel safe with, and increases the reliability of the autonomous driving.

[0027] <2. Driving assistance control device> An example of the configuration of the driving assistance control device 50 according to this embodiment will be specifically described below. In this embodiment, an example will be described in which the control unit 51 performs various data analyses while the vehicle is being driven, and the learned data is used for automatic driving control. In this embodiment, the determination of whether the driver senses risk is made based on whether the driver senses "fear."

[0028] The driving assistance control device 50 is configured to include at least an arithmetic processing device such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). A part or all of the driving assistance control device 50 may be configured with updatable firmware or the like, or may be a program module executed by a command from the CPU or the like. As shown in FIG. 1, the driving assistance control device 50 includes a control unit 51, a storage unit 53, and a reference data storage unit 55.

[0029] (2-1. Storage section) The storage unit 53 is configured with a storage element such as a RAM or a ROM, or a storage medium such as a HDD, a CD, a DVD, an SSD, a USB flash, a storage device, etc. The storage unit 53 stores software programs executed by the arithmetic processing device or the image processing device, various parameters used in the arithmetic processing, acquired data, arithmetic results, etc.

[0030] (2-2. Reference Data Storage Unit) The reference data storage unit 55 is configured with an updatable storage medium such as a RAM, HDD, CD, DVD, SSD, USB flash, or storage device, and stores learning data learned by the control unit 51. The learning data is data on the risk level indicating the degree of risk felt by the driver from obstacles around the vehicle, and the learning results of risk factors that cause the driver to feel risk.

[0031] (2-3. Control Unit) The control unit 51 is configured to include an arithmetic processing unit such as a CPU or an MPU, and an image processing unit such as a GPU. The arithmetic processing unit and the image processing unit execute various arithmetic processing by executing programs stored in the storage unit 53. In this embodiment, the control unit 51 includes a surrounding environment detection unit 61, a vehicle data acquisition unit 63, a driver image processing unit 65, a learning unit 67, a driving condition setting unit 69, and a notification control unit 71. Some or all of these units are functions realized by the execution of a program by the arithmetic processing unit or the image processing unit.

[0032] (2-3-1. Surrounding environment detection section) The surrounding environment detection unit 61 detects information about the surrounding environment of the vehicle based on image data transmitted from the exterior camera 23 and data transmitted from the surrounding environment sensor 25. Specifically, the surrounding environment detection unit 61 performs image processing on the image data transmitted from the exterior camera 23 to identify people, other vehicles, bicycles, buildings, natural objects, and other obstacles that exist around the vehicle using object detection technology, and calculates the positions of these objects relative to the vehicle, and the distances and relative speeds between the vehicle and these objects. The surrounding environment detection unit 61 stores data on the detected obstacles around the vehicle in the memory unit 53 as time-series data.

[0033] The surrounding environment detection unit 61 may acquire information transmitted from a device outside the vehicle via a communication means such as vehicle-to-vehicle communication, road-to-vehicle communication, or a mobile communication network, and may identify part of the information about obstacles around the vehicle. The surrounding environment detection unit 61 may also identify the position of the vehicle on map data using vehicle position information acquired by the GPS antenna 29, and identify part of the information about obstacles around the vehicle.

[0034] (2-3-2. Vehicle data acquisition unit) The vehicle data acquisition unit 63 acquires data on the operation state and behavior of the vehicle based on the sensor signal transmitted from the vehicle operation / behavior sensor 27. The data on the operation state and behavior of the vehicle includes, for example, data on the vehicle speed, longitudinal acceleration, lateral acceleration, yaw rate, steering angle of the steering wheel or steering wheels, accelerator opening, brake operation amount, on / off of the brake lamp switch, and on / off of the blinker switch. The data on the operation state and behavior of the vehicle also includes data on the on / off of the autonomous driving mode of the vehicle. The vehicle data acquisition unit 63 stores the acquired data on the operation state and behavior of the vehicle in the memory unit 53 as time-series data.

[0035] (2-3-3. Driver image processing section) The driver image processing unit 65 extracts information about the driver by performing image processing on the image data transmitted from the in-vehicle camera 21. Specifically, the driver image processing unit 65 detects the driver's face using an object detection technique by performing image processing on the image data transmitted from the in-vehicle camera 21. At this time, the driver image processing unit 65 identifies the person by, for example, extracting facial features. The driver image processing unit 65 analyzes the detected face of the driver to detect the direction of the driver's line of sight. The driver image processing unit 65 estimates the driver's gaze target based on the detected direction of the driver's line of sight and data on obstacles around the vehicle detected by the surrounding environment detection unit 61.

[0036] The driver image processing unit 65 also estimates the driver's emotion based on the detected data of the driver's face. For example, the driver image processing unit 65 estimates the driver's emotion by analyzing the data of the driver's facial expression using a facial expression analysis method based on the FACS theory. As described above, in this embodiment, the driver image processing unit 65 estimates at least the driver's emotion of "fear". When the driver's emotion of "fear" is estimated, the driver image processing unit 65 also estimates the degree of "fear". For example, the driver image processing unit 65 analyzes the driver's facial expression and classifies the degree of "fear" into one of a plurality of stages set in advance.

[0037] The driver image processing unit 65 stores the estimated data of the driver's gaze target and emotion as time-series data in the storage unit 53. In this embodiment, the driver image processing unit 65 estimates the driver's gaze target and emotion during manual driving of the vehicle. This makes it possible to obtain information for grasping the individual characteristics that affect the recognition of risks while driving for each driver.

[0038] The method of estimating the emotion is not particularly limited, and other methods may be used. The driver image processing unit 65 may estimate the emotion of the driver by using a biological detection sensor that detects brain waves, pulse, heart rate, etc., other than the in-vehicle photographing camera 21, or by using a combination of these means.

[0039] (2-3-4. Learning Department) The learning unit 67 learns the risk level that the driver feels for each obstacle and the risk factors that cause the driver to feel risk. In this embodiment, the learning unit 67 learns the risk level and risk factors based on data of obstacles around the vehicle detected by the surrounding environment detection unit 61, data of the driver's gaze target and emotion estimated by the driver image processing unit 65, and data of the vehicle's operation state and behavior acquired by the vehicle data acquisition unit 63. As described above, the learning unit 67 learns a state in which the driver feels "fear" as a state in which the driver feels risk.

[0040] The risk level is the degree of "fear" felt by the driver, and corresponds to the data of the degree of "fear" estimated by the driver image processing unit 65. The learning unit 67 analyzes data of obstacles around the vehicle, data of the driver's gaze target and emotions, and data of the vehicle's operating state and behavior, and determines the degree of risk felt by the driver for which object, what factors caused it, and the level of risk.

[0041] For example, the learning unit 67 identifies the time when the emotion of "fear" was estimated from various data stored in the memory unit 53 as time-series data, and extracts data on obstacles around the vehicle and data on the operation state and behavior of the vehicle at that time. The data on obstacles around the vehicle includes, for example, data on the type of obstacle, the size of each obstacle, the distance from the obstacle, and the relative speed from the obstacle. The data on the operation state and behavior of the vehicle includes, for example, data on the vehicle speed, lateral acceleration, longitudinal acceleration, engine speed, steering angle, yaw rate, brake operation amount (brake pressure), accelerator opening, on / off of the brake lamp switch, and on / off of the blinker switch.

[0042] Furthermore, the learning unit 67 identifies the object and the cause of the driver's feeling of "fear" from the extracted data of the obstacles around the vehicle and the data of the vehicle's operation state and behavior. The cause (risk factor) of the driver's feeling of "fear" is identified by, for example, any one or a combination of a plurality of the following: the type of obstacle, the size of each obstacle, the distance from the obstacle, the relative speed from the obstacle, the vehicle speed, the lateral acceleration, the longitudinal acceleration, the engine speed, the steering angle, the amount of brake operation, the accelerator opening, the on / off of the brake lamp switch, and the on / off of the blinker switch.

[0043] For example, by using a known machine learning model, such as a random forest, the learning unit 67 can calculate the importance of each explanatory variable with respect to "fear" when the objective variable is data on the degree of "fear" and the explanatory variables are data on obstacles around the vehicle, data on the driver's gaze target, and data on the vehicle's operation state and behavior. The explanatory variables with high importance obtained by this method are identified as factors that caused the driver to feel "fear." At this time, the explanatory variable with the highest importance may be used as a factor, or multiple explanatory variables with higher importance may be used as factors.

[0044] The learning unit 67 learns the risk level and risk factors for each obstacle for each identified driver. This makes it possible to obtain learning data that reflects the risk factors for each driver's gaze object. The learning unit 67 stores the obtained learning data in the reference data storage unit 55.

[0045] The learning unit 67 may use a known machine learning model to learn the risk level and risk factors for each obstacle using data on obstacles around the vehicle, data on the driver's gaze target and emotions, and data on the vehicle's operation state and behavior as input data for learning. The machine learning model may be, for example, a computational model using a neural network such as a support vector machine, a nearest neighbor method, or a deep learning, or a Bayesian network.

[0046] In addition, in order to reflect factors other than the risk factors for each object gazed upon by the driver, the learning unit 67 may obtain data other than the data on the risk level and risk factors for each obstacle and store the data in the reference data storage unit 55. For example, in this embodiment, the learning unit 67 may obtain the frequency of gaze on an obstacle by the driver and store the data on the frequency of gaze on an obstacle that is perceived as risk in the reference data storage unit 55. It has been suggested that there is a correlation between the frequency of gaze on an obstacle and the way in which the driver recognizes risk, and by using the data on the frequency of gaze on an obstacle, it is possible to set a control condition that better reflects the driving characteristics of the driver. The data on the frequency of gaze on an obstacle may be, for example, the number of times the object is gazed upon within a predetermined reference time, or may be the ratio of the time the object is gazed upon within a predetermined reference time.

[0047] The risk level and risk factors may be learned based on data acquired while the vehicle is in automatic driving mode or based on data acquired while the vehicle is in manual driving mode. Regardless of the data, the risk level and risk factors of the driver with respect to each obstacle can be learned.

[0048] In this embodiment, the learning unit 67 analyzes data on obstacles around the vehicle and data on the vehicle's operation state and behavior, and learns a driving trajectory that takes into account the distance to each obstacle based on data on the driving trajectory and vehicle speed when the driver corrects the driving trajectory to ensure a distance from the obstacle during manual driving. For example, the learning unit 67 identifies a driving scene in which the driver operates the steering wheel, accelerator pedal, or brake pedal to ensure a distance from the obstacle based on data on obstacles around the vehicle detected during manual driving, and data on the vehicle's operation state and behavior. The learning unit 67 generates learning data using data on the distance to the obstacle and the vehicle speed in the identified driving scene as input data. The distance to the obstacle in this case is the distance between the vehicle and the obstacle in a direction perpendicular to the vehicle's traveling direction when the vehicle passes the obstacle. The learning data on the driving trajectory is stored in the data storage unit 55 and is used to set adjustment parameters that reflect the driver's characteristics.

[0049] In addition, although it is not a learning result by the learning unit 67, in this embodiment, data on internal characteristics indicating personal characteristics that affect the risk perception of each driver is stored in advance in the reference data storage unit 55. Examples of such internal characteristics include driving attitude, sensitivity to load during driving, and risk aversion. Specifically, examples of driving attitude include the degree of confidence in driving skills and the degree of reluctance to drive. Examples of sensitivity to load during driving include the degree of tendency to avoid being hindered in driving pace. Examples of risk aversion include the degree of tendency to dare to take unsafe actions if there is a benefit. These internal characteristics are quantified, for example, in five stages, and stored in the reference data storage unit 55 as internal characteristic indexes quantified based on the answers of each driver.

[0050] The items indicating inner characteristics are not limited to the above examples, and any information that can distinguish personal characteristics that affect the risk perception of each individual can be used as appropriate. Furthermore, the data on inner characteristics is not limited to data based on the driver's responses to a questionnaire or the like, and may be evaluations from others or results of a test for evaluating inner characteristics.

[0051] (2-3-5. Driving condition setting section) The driving condition setting unit 69 sets driving conditions when executing the automatic driving control, and transmits information of the set control conditions to the vehicle drive control unit 45. The driving condition setting unit 69 sets at least a driving trajectory and a vehicle speed so that the vehicle does not come into contact with an obstacle when the vehicle is driven in automatic driving along a driving route to a set destination, and transmits a control command to the vehicle drive control unit 45. In this embodiment, the driving condition setting unit 69 sets the vehicle's driving trajectory and vehicle speed using a risk potential that is an index indicating the possibility that the vehicle will collide with an obstacle. The details of the calculation process by the driving condition setting unit 69 will be described in detail later.

[0052] (2-3-6. Notification control section) The notification control unit 71 controls the driving of the HMI 43 to control notifications to the driver. In this embodiment, the notification control unit 71 notifies the driver of the contents of the set control conditions after the vehicle travel control is performed according to the control conditions set based on the learning data of the risk level and risk factors for each obstacle. For example, when the travel trajectory is shifted from the reference path Pr to pass by a detected pedestrian, the notification control unit 71 notifies the driver that "the vehicle passed on the left side of the road to ensure a distance from the pedestrian," and when the vehicle speed is reduced, the notification control unit 71 notifies the driver that "the vehicle slowed down to ensure the safety of the pedestrian." The notification is performed, for example, by at least one of voice and display.

[0053] In addition, when the driving assistance control device 50 sets the control conditions for the automatic driving control, the notification control unit 71 does not need to perform notification.

[0054] <3. Operation example of driving assistance control device> Next, an example of the operation of the driving assistance control device 50 according to this embodiment will be described.

[0055] (3-1. Method for setting automatic driving control conditions) First, a method for setting the automatic driving control conditions by the driving condition setting unit 69 will be described. Below, an example of a basic concept of the method for setting the automatic driving control conditions will be described, and then a method for reflecting the risk level and risk factors that each driver feels for each obstacle in the control conditions will be described.

[0056] (Basic concept) FIG. 2 is an explanatory diagram showing the risk potential for an object. In FIG. 2, a pedestrian is shown as an example of the object. The value of the risk potential (risk value) R i The closer you get to the object (pedestrian), the higher the risk value R i is the distance from the object x i For example, it can be expressed by the following formula (1).

[0057]

number

[0058] R i :Risk value C i : Risk absolute value (gain) x i :Distance from the object τ i :Slope coefficient r i : Radius of object i: Numbering to distinguish objects

[0059] Distance to object x i The absolute risk value C, which is the risk value when i is set for each object as a value that depends on the object. For example, if the object is a "pedestrian" or a "low curb", the absolute risk value C for a "pedestrian" is set as the risk of a collision with a pedestrian being higher than that of a collision with a low curb. i is the absolute risk value C for "low curb" i The slope coefficient τ is set to a value greater than i is a value that is set regardless of the object.

[0060] A risk map is obtained by assigning a risk to each obstacle detected while the vehicle is traveling and adding up the spatial overlap of each risk potential, taking into account the risk of collision with multiple obstacles. In such a risk map, the level of risk can be shown as contour lines on a two-dimensional plane. Here, the risk value in the risk map is R RM However, as mentioned above, the risk value R RM has a two-dimensional distribution. This makes it possible to select a trajectory with a lower risk. At this time, the risk map may be calculated taking into consideration not only the obstacles that are manifested, but also unmanifested risks (latent risks). For example, when the vehicle passes through an area that is a blind spot due to an obstruction after turning, a potential risk may be added on the assumption that a passerby may jump out from the blind spot, and this may be reflected in the risk map.

[0061] Fig. 3 shows an example of a risk map reflecting potential risks. In the risk map shown in Fig. 3, the risk potential of a bicycle Y appearing from a blind spot is reflected in addition to the risk potential given along the road and the risk potential of an obstruction Z detected and made apparent by the surrounding environment detection unit 61. As a result, regardless of whether the bicycle Y is actually present or not, the vehicle X selects a low-risk travel path T away from the blind spot in consideration of the potential risk. This makes it possible to select a safe travel path even in a complex traffic environment such as a residential road.

[0062] (Basic method to reflect driver characteristics) Here, the risk felt by each driver is not necessarily the same as the physical risk. For example, when driving on a narrow residential road, some drivers feel a strong risk from utility poles protruding into the road, while other drivers feel a strong risk from vehicles jumping out from blind spots. For this reason, in this embodiment, as a basic method for setting a driving trajectory, the idea of ​​an index of risk level tolerable to each driver (hereinafter also referred to as "tolerable risk") proposed in the publicly known document "Research on the application of a driver model based on the sense of danger in the forward field of vision to automatic driving control" (Proceedings of the Society of Automotive Engineers of Japan, Vol. 48, No. 3, pp. 763-769) is used.

[0063] To briefly explain the basic method using acceptable risk, the driver's acceptable risk U a is assumed to be related to the vehicle speed V of the host vehicle, and can be expressed by the following equation (2).

[0064]

number

[0065] U a : Acceptable risk w a :gain V:Vehicle speed

[0066] In this embodiment, the gain w a The value of is set for each driver. a The value of may be set to correspond to a legal speed in a situation without the risk of blind spots or other traffic participants being present in the vicinity.

[0067] The allowable risk U shown in the above formula (2) a The higher the vehicle speed V, the smaller the value becomes, and the narrower the allowable range of risk becomes. In other words, the risk value R RM is the allowable risk U a If the vehicle speed V exceeds the allowable risk U a The risk value R RM is the allowable risk U a It is possible to create an area below this.

[0068] Figure 4 shows how the tolerable risk U a 4 is an explanatory diagram showing an example in which the risk potential of the side wall Z1 is enlarged and a drivable route is secured. Fig. 4 is a cross-sectional view perpendicular to the traveling direction of the vehicle, showing a risk map in which the spatial overlap between the risk potential of the side wall Z1 and the risk potential of the pedestrian Y1 is added for a vehicle speed V of 30 km / h and for a vehicle speed V of 20 km / h.

[0069] As shown on the left side of Fig. 4, when the vehicle speed V is 30 km / h, the risk value R RM is the allowable risk U a On the other hand, as shown on the right side of Fig. 4, when the vehicle speed V is 20 km / h, the allowable risk U a The value of increases, and the risk value R RM is the allowable risk U a An area (drivable route) as follows is generated ahead in the traveling direction. In this way, by adjusting the vehicle speed V, it becomes possible to set a driving trajectory within the drivable route.

[0070] 5 to 7 are explanatory diagrams showing examples of a method for setting a travel trajectory and a vehicle speed using a basic method that reflects the driver characteristics. When traveling along a route to a set destination, data on a reference path Pr of the road being traveled is acquired by referring to pre-stored map data. As shown in FIG. 5, when traveling along a road with a side wall Z1 on the left side, the allowable risk U corresponding to the current vehicle speed V (=30 km / h) on the acquired reference path Pr is calculated. a is the risk value R RM As long as the reference path Pr is greater than the selected path Pc, the reference path Pr is set as the selected path Pc.

[0071] Also, as shown in Fig. 6, when the vehicle passes by the utility pole Z2 on the right side of the same road, the allowable risk U corresponding to the current vehicle speed V (= 30 km / h) on the acquired reference path Pr is a is the risk value R RM If the allowable risk U a is the risk value R RM Of the above possible travel routes, the route on the reference path Pr side is set as the selected path Pc.

[0072] Also, as shown in Figure 7, when a vehicle passes by a pedestrian Y1 on the right side on the same road, the risk value R RM is set high, even if the utility pole Z2 shown in Fig. 6 and the pedestrian Y1 shown in Fig. 7 are at the same distance from the vehicle, the allowable risk U corresponding to the current vehicle speed V (= 30 km / h) is a is the risk value R RM In such a case, the vehicle speed V of the vehicle is reduced (V = 20 km / h) to reduce the allowable risk U a increases, and the acceptable risk U a is the risk value R RM The selected path Pc is set by generating the above feasible routes.

[0073] (Method of Reflecting Driver Characteristics in the Present Embodiment) So far, the basic method of setting the control conditions by the driving condition setting unit 69 has been described. In this embodiment, the driving condition setting unit 69 sets the control conditions by further reflecting the learning data obtained by the learning unit 67. As described above, the learning data obtained by the learning unit 67 includes at least data on risk factors for each gaze object of each driver. In this embodiment, the learning data further includes data on the gaze frequency of the driver on the gaze object. In this embodiment, the control conditions are set by reflecting the driver's internal characteristics.

[0074] Specifically, the driving condition setting unit 69 reflects the frequency of gazing at the gazing object and the risk factor for each gazing object in the risk potential of each object. a to reflect the driver's inner characteristics.

[0075] First, a method of reflecting the gaze frequency of a gazed object and the risk factors for each gazed object in the risk potential of each object will be described. In this embodiment, a risk value R that reflects the frequency of gazing at a gaze object and a risk factor for each gaze object is i is expressed by the following formula (3): In the following example, data on the gaze ratio is used as data on the gaze frequency of the gaze object.

[0076]

number

[0077] R i :Risk value C i : Risk absolute value (gain) x i :Distance from the object τ i :Slope coefficient r i : Radius of object GA i : Gaze ratio (0≦GA i ≦1) βi ,γ i : Risk factor parameter (1 + importance of risk factor) i: Numbering to distinguish objects

[0078] Gaze ratio GA i is the ratio of time spent gazing at an object within a given reference time, and is a value between 0 and 1. i The higher the value, the more likely the gaze object is to be perceived as risky, and the risk value R i In addition, the risk factor parameter β i ,γ i is a value that is set according to the importance of the risk factor for each object based on the results of learning by the learning unit 67, and the risk factor parameter β i ,γ i The shape of the risk potential is adjusted by the risk factor parameter β i ,γ i is set to a value (between 1 and 2) according to the importance of the risk factor.

[0079] For example, if a driver feels that there is a risk in terms of the distance to an object, the radius of the object r i The risk factor parameter β i By increasing the value of (approaching 2), the absolute value of the risk potential of the object C i In addition, for drivers who feel risk in terms of the relative speed with respect to the object, the gradient coefficient τ i The risk factor parameter γ i By reducing the value of (approaching 1), the gradient of the object's risk potential can be made gentler and the radius can be enlarged.

[0080] Figure 8 shows the gaze ratio of the gaze object GA i 8 is an explanatory diagram showing the result of reflecting the risk factors and the gaze ratio GA iThe risk map is a cross-sectional view perpendicular to the vehicle travel direction, and is a sum of the spatial overlap between the risk potential of the side wall Z1 and the risk potential of the pedestrian Y1 before and after the reflection of risk factors. The risk value R1 of the side wall Z1 is shown by a solid line, the risk value R2 of the pedestrian Y1 is shown by a dashed line, and the risk value R RM is shown by a dashed line.

[0081] Pedestrian gaze ratio GA i In the case of drivers who feel a risk due to the distance to pedestrians, the gaze ratio GA i The value of and the risk factor parameter β i As a result, as shown in FIG. 8, the radius of the risk potential of pedestrian Y1 is expanded from the original design value, and the risk value R2 is increased. As a result, the risk value R after addition on the side closer to pedestrian Y1 is RM becomes larger, and the selected path Pc is moved away from the pedestrian Y1. In this way, the gaze ratio GA i By reflecting the risk factors in the risk potential, it becomes possible to mainly reflect the driver's characteristics in the setting of the driving trajectory.

[0082] Next, the allowable risk U a This section explains how to reflect the driver's internal characteristics. In this embodiment, the allowable risk U a is expressed by the following formula (4).

[0083]

number

[0084] U a : Acceptable risk w a :gain V:Vehicle speed IV j :Internal characteristic index IV j_max : Maximum value of internal characteristic index M: Number of internal trait indicators α j :Adjustment parameter (0<α j ≦1) j: A number to distinguish the internal characteristics to be considered

[0085] The number M of internal characteristic indexes is the number of items to evaluate internal characteristics. For example, when information on the degree of confidence in driving skills, the degree of reluctance to drive, the degree of tendency to avoid being disturbed in driving pace, and the degree of tendency to take unsafe actions if there is a benefit is used as internal characteristics, the number M of internal characteristic indexes is 4. In addition, when each internal characteristic is evaluated on a 5-point scale, the maximum value IV of the internal characteristic index is j_max is 5, and the internal characteristic index IV j is the evaluation value (value 1 to 5) for each driver. For example, if the index of "confidence in driving skills" is 4, then (1-IV) in the above formula (4) j / IV j_max ) is 0.2. This is calculated for all internal characteristics and averaged. Adjustment parameter α j is a value that is set based on the learning data of the traveling trajectory that is learned by the learning unit 67 and that takes into account the distance to each obstacle.

[0086] Figure 9 shows the internal characteristics of the driver as a function of the acceptable risk U a FIG. 9 is an explanatory diagram showing the results of reflecting the driver's internal characteristics in the permissible risk U a As shown on the left side of Figure 9, the allowable risk U a (=w a / V 2 ) on the reference path Pr, the allowable risk U corresponding to the current vehicle speed V (= 30 km / h) a is the risk value R RM However, a driver who is not confident in his driving skills may feel a risk in the sense of distance between himself and the pedestrian Y1 when passing the reference path Pr.

[0087] In contrast, as shown on the right side of Figure 9, the allowable risk U a By reflecting the internal characteristics of the driver in a In this case, the allowable risk U corresponding to the current vehicle speed V (= 30 km / h) is lowered. a is the risk value R RM Therefore, by reducing the vehicle speed V (V = 20 km / h), the allowable risk U a and increase the acceptable risk U a is the risk value R RM By creating a route that satisfies the above criteria, the selected path Pc is set. This keeps the selected path Pc away from the pedestrian Y1. In this way, the acceptable risk U a By reflecting the internal characteristics of the driver in the vehicle speed and driving trajectory, the internal characteristics of the driver can be reflected in the setting of the vehicle speed and driving trajectory.

[0088] In addition, in formula (4), the adjustment parameter α is set based on the learning data of the driving trajectory taking into account the distance to each obstacle. j Although it is not necessary to use such an adjustment parameter α j By using this, the internal characteristics that appear in the behavior of the vehicle while the driver is manually driving can be expressed as the allowable risk U a Therefore, the vehicle speed and driving trajectory can be set more appropriately.

[0089] In this way, the driving condition setting unit 69 sets the control conditions by reflecting not only the physical risk level of the obstacle that the driver feels is a risk, but also the factors that make the obstacle a risk, in the risk potential. Therefore, in various driving scenes, appropriate driving control can be executed according to the risk that each driver feels for the detected obstacle. In this case, when multiple obstacles are detected, the higher the obstacle's risk level, the lower the absolute value of the risk C iis set to be larger than the basic set value, and the radius of the risk potential of the obstacle is enlarged from the basic set value. Therefore, the control conditions of the vehicle are set so that the operation to reduce the risk factor felt for the obstacle with a higher risk level is prioritized. This realizes safe automated driving for each driver and increases the reliability of the automated driving control.

[0090] (3-2. Flowchart) Next, a flowchart of an operation example of the driving assistance control device 50 will be explained by dividing it into a learning data storage process and a control condition setting process. In the following explanation, detailed explanations of parts that overlap with the contents explained so far will be omitted.

[0091] (3-2-1. Learning data storage processing) 10 is a flowchart showing the learning data storage process. The learning data storage process may be executed at an appropriate timing. For example, the learning data storage process may be executed when the driving of the vehicle is terminated or when the driving of the vehicle in the manual driving mode is terminated, or may be executed at a predetermined time interval or continuously.

[0092] First, the driver image processing unit 65 detects the face of the driver based on the image data transmitted from the in-vehicle camera 21, generates identification data of the driver by analyzing the feature amount of the face, and sets the data in the reference data storage unit 55 (step S11). As a result, data to be learned from this is linked to the driver identification data, and it becomes possible to specify the data as learning data of the driver. The information to identify the driver may be input by the driver.

[0093] Next, the learning unit 67 acquires data on obstacles around the vehicle that are detected by the surrounding environment detection unit 61 and stored as time-series data in the storage unit 53 (step S13). The data on obstacles around the vehicle includes, for example, data on the type of obstacle, the size of each obstacle, the distance from the obstacle, and the relative speed from the obstacle.

[0094] Furthermore, the learning unit 67 acquires data on the vehicle operation state and behavior detected by the vehicle state / behavior sensor 27 and stored as time-series data in the storage unit 53 (step S15). The data on the vehicle operation state and behavior includes, for example, data on the vehicle speed, lateral acceleration, longitudinal acceleration, engine speed, steering angle, brake operation amount, accelerator opening, on / off of the brake lamp switch, on / off of the blinker switch, and the vehicle driving mode.

[0095] The learning unit 67 also acquires data on the driver's gaze target, which is estimated by the driver image processing unit 65 and stored in the storage unit 53 as time-series data (step S17). Furthermore, the learning unit 67 acquires data on the driver's emotions, which is estimated by the driver image processing unit 65 and stored in the storage unit 53 as time-series data (step S19). The data on the driver's emotions includes data on the degree of "fear" felt by the driver (risk level). The risk level can be classified into, for example, five stages.

[0096] Next, the learning unit 67 executes a calculation to learn, based on the acquired data, what factor caused the driver to feel risk for which obstacle (step S21). In this embodiment, the learning unit 67 identifies the time when the emotion of "fear" was estimated, and extracts data on obstacles around the vehicle and data on the vehicle's operation state and behavior at that time. In addition, the learning unit 67 identifies the object for which the driver felt "fear", the degree of "fear" (risk level), and the factor (risk factor) that caused the driver to feel "fear" from the extracted data on obstacles around the vehicle and the vehicle's operation state and behavior data.

[0097] Specifically, the learning unit 67 uses a known machine learning model, for example, a random forest, to calculate the importance of each explanatory variable with respect to "fear" by using data on the degree of "fear" as the objective variable and data on obstacles around the vehicle, data on the driver's gaze target, and data on the vehicle's operation state and behavior as explanatory variables. The explanatory variables with high importance obtained by this method are identified as risk factors that cause the driver to feel "fear" for each obstacle. The risk factors are identified by, for example, any one or a combination of the following: distance to each obstacle, relative speed to the obstacle, vehicle speed, lateral acceleration, longitudinal acceleration, engine speed, steering angle, brake operation amount, accelerator opening, on / off of the brake lamp switch, and on / off of the blinker switch. In this case, the explanatory variable with the highest importance may be the risk factor, or multiple explanatory variables with higher importance may be the factors.

[0098] Furthermore, the learning unit 67 calculates a risk factor parameter β according to the importance of the risk factor for each object based on the risk level and risk factor data for each identified object. i ,γ i Set the risk factor parameter β i ,γ i is set to a value (between 1 and 2) according to the importance of the risk factor, for example.

[0099] Next, the learning unit 67 executes a calculation to learn a driving trajectory (corrected driving trajectory) based on the acquired data, taking into account the distance from each obstacle when the driver corrects the driving trajectory to ensure the distance from the obstacle during manual driving (step S23). For example, the learning unit 67 identifies a driving scene in which the driver operated the steering wheel, accelerator pedal, or brake pedal to ensure the distance from the obstacle, based on data of obstacles around the vehicle detected during manual driving and data of the vehicle's operation state and behavior. The learning unit 67 generates learning data using data of the type of obstacle, the size of the obstacle, the distance from the obstacle, and the vehicle speed in the identified driving scene as input data. The distance from the obstacle in this case is the distance between the vehicle and the obstacle in a direction perpendicular to the vehicle's traveling direction when the vehicle passes the obstacle.

[0100] Furthermore, the learning unit 67 calculates the allowable risk U based on the learned data of the corrected running trajectory. a The adjustment parameter α for adjusting j Set the adjustment parameter α j For example, the greater the distance from the obstacle or the greater the tendency to slow down, the greater the tolerable risk U a is set to a value between 0 and 1 so that is small.

[0101] Next, the learning unit 67 calculates the gaze frequency of the obstacle for which the driver felt a risk (step S25). The data on the gaze frequency of the obstacle may be, for example, the number of times the object was gazed upon within a predetermined reference time, or may be the ratio of the time the object was gazed upon within a predetermined reference time. In this embodiment, the learning unit 67 calculates a gaze ratio (a value between 0 and 1) which is the ratio of the time the driver gazed upon an obstacle for which the driver felt a risk within a predetermined reference time.

[0102] Next, the learning unit 67 stores the learning data obtained as a result of the learning process in steps S21 to S25 in the reference data storage unit 55 in association with the driver's identification data set in step S11 (step S27).

[0103] 11 is an explanatory diagram showing an example of the data configuration stored in the reference data storage unit 55 in this embodiment. The data stored in the reference data storage unit 55 includes object basic data set for each object regardless of the driver, driver data that depends on the driver, and risk data for each object about which the driver feels "fear." Specifically, the object basic data includes a preset absolute risk value C i and the gradient coefficient τ for each object i The driver data includes pre-set driver identification data, gain w a and internal trait index IV for each internal trait i The driver data includes the learned corrected driving trajectory (type of obstacle, distance to obstacle, vehicle speed) and adjustment parameter α j The risk data includes the object that caused the fear, the degree of the fear (risk level), the cause of the fear (risk factor), and the risk factor parameter β i ,γ i and gaze ratio GA i Includes data on.

[0104] In this way, the driving assistance control device 50 learns the risk level and risk factors that each driver perceives for each obstacle, as well as the driving trajectory taking into account the distance to each obstacle and the frequency of looking at the obstacle, based on the data collected while each driver is driving the vehicle, and stores the learned data in the reference data memory unit 55.

[0105] The order of steps S13 to S19 for acquiring various data used in the learning process is not limited to the above example, and may be changed to an appropriate order. Similarly, the order of steps S21 to S25 for calculating each learning data is not limited to the above example, and may be changed to an appropriate order.

[0106] (3-2-2. Control condition setting process) 12 and 13 are flowcharts showing the control condition setting process. First, the driving condition setting unit 69 determines whether the vehicle is in an automatic driving mode (step S31). Whether the vehicle is in an automatic driving mode can be determined based on the setting information of the automatic driving mode. If the vehicle is not in an automatic driving mode (S31 / No), the driving condition setting unit 69 repeats the determination whether the vehicle is in an automatic driving mode until the vehicle is switched to the automatic driving mode.

[0107] On the other hand, if the vehicle is in automatic driving mode (S31 / Yes), the driving condition setting unit 69 acquires identification data of the driver currently operating the vehicle from the driver image processing unit 65, and identifies the driver in reference to the identification data set in the reference data storage unit 55 (step S33). This makes it possible to identify the learning data to be referenced. Note that the information identifying the driver currently operating the vehicle may be input by the driver.

[0108] Next, the driving condition setting unit 69 acquires data on obstacles around the vehicle detected by the surrounding environment detection unit 61 (step S35). The data on obstacles around the vehicle includes at least the type of obstacle, the size of each obstacle, the distance from the obstacle, and the relative speed from the obstacle. Next, the driving condition setting unit 69 acquires data on the operation state and behavior of the vehicle detected by the vehicle state / behavior sensor 27 (step S37). The data on the operation state and behavior of the vehicle includes at least the data on the vehicle speed, lateral acceleration, and longitudinal acceleration. Note that the order of steps S35 and S37 is not limited to this order, and may be interchanged.

[0109] Next, the driving condition setting unit 69 assigns a risk potential to each of the detected obstacles and calculates a risk map by adding up the spatial overlap of each risk potential (step S39). At this time, the driving condition setting unit 69 refers to the learned data of the risk level, risk factor, and gaze ratio stored in the reference data storage unit 55, and calculates the risk potential to be assigned to each obstacle using the above formula (3). Specifically, the driving condition setting unit 69 calculates, from the data stored in the reference data storage unit 55, the risk factor parameter β set for each of the detected obstacles. i ,γ i , the absolute value of risk C i and the gradient coefficient τ i The travel condition setting unit 69 extracts the data of the distance x from the currently detected obstacle. i , the radius of the obstacle r i and the obstacle gaze ratio GAi are input into the above formula (3) to obtain the risk potential (risk value R i ) is calculated. The calculated risk potential reflects the driver's gaze ratio (gazing frequency) of each obstacle and risk factors. In this embodiment, the calculated risk potential also reflects the risk level for each obstacle, such as the sense of distance, relative speed, and acceleration. As a result, the higher the ranking of the risk level felt by the driver for each obstacle, the larger the radius of the risk potential is from the original design value, and the greater the absolute value of the risk C i The driving condition setting unit 69 then calculates a risk map by adding up the spatial overlap of the risk potentials of the respective obstacles.

[0110] Next, the driving condition setting unit 69 determines the driver's tolerable risk U according to the current vehicle speed V. a At this time, the driving condition setting unit 69 refers to the learning data of the risk level and risk factors stored in the reference data storage unit 55 and the data of the driver's internal characteristics, and calculates the allowable risk U aSpecifically, the driving condition setting unit 69 acquires data on the current vehicle speed V, and calculates the driver's internal characteristic index IV from the data stored in the reference data storage unit 55. j , adjustment parameter α j and gain w a Then, the driving condition setting unit 69 inputs the extracted data into the above formula (4) to obtain the driver's tolerable risk U according to the current vehicle speed V. a Calculate the required allowable risk U a reflects the internal characteristics of the driver that may affect his / her risk perception. a The risk level for each obstacle, its distance, relative speed, acceleration, etc. is reflected in the U. a The order of steps S39 and S41 may be interchanged.

[0111] Next, the driving condition setting unit 69 calculates the risk map (risk value R RM ) and allowable risk U a Based on this, the travel path and the vehicle speed are set (step S43). Fig. 13 is a flowchart showing the process of setting the travel path and the vehicle speed.

[0112] First, the driving condition setting unit 69 acquires information on the current position of the vehicle transmitted from the GPS antenna 29 and also acquires data on the reference path Pr, which is a trajectory that serves as a reference when traveling on the road currently being traveled, by referring to the map data storage unit 31 (step S51). Next, the driving condition setting unit 69 acquires data on the risk map and the allowable risk U a Apply the acquired reference path Pr to the risk value R RM is the allowable risk U a It is determined whether or not the following holds (step S53).

[0113] On the reference path Pr, the risk value R ahead in the direction of travel RM is the allowable risk U aIf the vehicle speed V is equal to or lower than the reference path Pr (S53 / Yes), the traveling condition setting unit 69 sets the reference path Pr to the selected path Pc while maintaining the current vehicle speed V (step S55). RM is the allowable risk U a If the risk value R RM is the allowable risk U a It is determined whether or not there exists a possible travel route that satisfies the following (step S57).

[0114] Risk value R ahead RM is the allowable risk U a If there is a possible travel route that satisfies the following criteria (S57 / Yes), the travel condition setting unit 69 sets the route on the reference path Pr side of the possible travel routes as the selected path Pc (step S59). This allows the vehicle to travel on a travel trajectory that poses a low risk to the driver while maintaining the current vehicle speed V.

[0115] On the other hand, the risk value R RM is the allowable risk U a If there is no possible travel route that satisfies the above condition (S57 / No), the travel condition setting unit 69 reduces the allowable risk U obtained by the above formula (4) by decelerating the vehicle. a Increase the risk value R RM is the allowable risk U a The following possible travel route is generated (step S61). Then, the travel condition setting unit 69 sets a selected path Pc within the generated possible travel route (step S63). This realizes travel at a vehicle speed V and on a travel trajectory where the driver feels a small risk.

[0116] 12, the driving condition setting unit 69 outputs information on the driving trajectory and vehicle speed V set in step S43 to the vehicle drive control unit 45. Based on the acquired driving trajectory and vehicle speed V, the vehicle drive control unit 45 sets a control target value for at least one of the engine control system, the motor control system, the electric steering system, the brake system, and the transmission system, and drives the vehicle by automatic driving.

[0117] 4. Effects of the driving assistance control device according to the present embodiment As described above, the driving assistance control device 50 according to this embodiment calculates the risk potential (risk value R ) of each obstacle using the learning data of the risk level and risk factors that the driver feels about the obstacles around the vehicle. i ) and the driver's acceptable risk (U a ), the control conditions for reducing risk factors associated with obstacles during autonomous driving control are set and output to the vehicle drive control unit 45. As a result, the driver's characteristics that affect the perception of risk while the vehicle is traveling are integrated into the risk potential R i This is reflected in the weighting of the allowable risk U a The driving control is therefore executed in accordance with not only the physical risk that does not depend on the type of obstacle, but also the risk level and risk factors that each driver feels for each obstacle. Therefore, it is possible to realize a driving assistance system that can handle any driving situation and that gives each driver a sense of security and trust. As a result, each driver will be able to use the driving assistance system continuously.

[0118] Moreover, according to the driving assistance control device 50 of this embodiment, the higher the risk level that the driver feels for each obstacle, the larger the risk potential is made, and the priority is given to the operation to reduce the risk for that obstacle. i ,γ iis adjusted, and the risk potential for each obstacle is adjusted. As a result, when multiple obstacles are detected, the driving assistance control device 50 prioritizes an operation to reduce the risk factor for an obstacle with a high risk level. Therefore, a driving assistance system that can provide a sense of security and reliability to each driver in accordance with the characteristics of each driver can be realized.

[0119] In addition, according to the driving assistance control device 50 of this embodiment, the internal characteristics that affect the driver's risk perception are calculated based on the acceptable risk U a This allows the vehicle control conditions to be adjusted based not only on the risk the driver feels from the obstacle, but also on the driver's internal characteristics. This makes it possible to realize a driving assistance system that can provide a greater sense of security and trust in response to the characteristics of each driver.

[0120] Moreover, according to the driving assistance control device 50 of this embodiment, the learning unit 67 uses data collected during manual driving of the vehicle to generate learning data of the risk level and risk factors that the driver perceives from an obstacle, and stores the learning data in the reference data storage unit 55. Therefore, the learning data is automatically accumulated while the driver is manually driving the vehicle. Therefore, the learning data of the characteristics that appear during normal driving by the driver and affect the recognition of risk is accurately accumulated, and the accuracy of the set control conditions can be improved.

[0121] <5. Other embodiments> Although the driving assistance control device 50 according to this embodiment has been described so far, various modifications are possible to the driving assistance control device 50. Some modified examples of the driving assistance control device will be described below.

[0122] (5-1. First modified example) The above-described driving assistance control device 50 collects data on obstacles around the vehicle, data on the vehicle's running state and behavior, data on the driver's gaze target and emotions, and the like while the driver is manually driving the vehicle, learns the risk level and risk factors for each obstacle, and stores the learned data in the reference data storage unit 55. In the first modified example, the learned data to be stored in the reference data storage unit 55 is prepared in advance and stored in the reference data storage unit 55.

[0123] For example, data on obstacles around the vehicle, data on the vehicle's running state and behavior, data on the driver's gaze target and emotions, etc. may be collected using another vehicle or a driving simulator, and learning data such as risk levels and risk factors for each obstacle obtained based on these data may be stored in advance in the reference data storage unit 55. In this case, the learning unit 67 provided in the driving assistance control device 50 described above may not be provided. Alternatively, the function of the learning unit 67 may be left as it is, and the previously stored learning data may be updated.

[0124] According to the first modification in which the learning data is stored in advance in the reference data storage unit 55, for example, the results of learning data on risk levels and risk factors that each driver may feel for various objects in various driving scenes can be reflected, and the accuracy of the output control conditions can be improved. Also, when the learning unit 67 is omitted from the driving assistance control device 50, the calculation load of the control unit 51 can be reduced.

[0125] (5-2. Second modified example) In the above-described driving assistance control device 50, the reference data storage unit 55 that stores the learning data is mounted on the vehicle. In the second modified example, the reference data storage unit is provided in an external device that can communicate with the driving assistance control device 50.

[0126] 14 is a block diagram showing an example in which the driving assistance control device 50 is configured to be able to communicate with a reference data storage device 70 that stores learning data. In the second modification, the driving assistance control device 50 is configured to be able to communicate with the reference data storage device 70 via a communication unit 41. The communication unit 41 is an interface compatible with a communication standard for the reference data storage device 70.

[0127] The reference data storage device 70 may be, for example, a server device configured to be able to communicate with the driving assistance control device 50 via a mobile communication network using cloud computing technology. The reference data storage device 70 may also be, for example, a smartphone or a mobile terminal configured to be able to communicate with the driving assistance control device 50 via a communication means such as Bluetooth (registered trademark).

[0128] According to the second variant in which the learning data is stored in a reference data storage device 70 capable of communicating with the driving assistance control device 50, even when the driver changes vehicles or uses multiple vehicles, the accumulated learning data can be used to set control conditions.

[0129] In addition, the configuration example shown in Figure 14 is an example in which pre-prepared learning data is stored in the reference data storage device 70, and the learning unit 67 is omitted, but the driving assistance control device 50 may be configured to include a learning unit 67, and the learning results by the learning unit 67 may be stored in the reference data storage device 70 each time.

[0130] (5-3. Third modified example) The above-described driving assistance control device 50 is configured to set the control conditions for autonomous driving using the learning data stored in the reference data storage unit 55, and output the control conditions to the vehicle drive control unit 45. In the third modified example, the driving assistance control device 50 notifies the driver of driving operation advice even during manual driving.

[0131] For example, when an obstacle is detected during manual driving, the driving condition setting unit 69 sets the vehicle speed V and driving trajectory based on the learning data to reduce the risk factor for the obstacle, and then compares the set vehicle speed V and driving trajectory with the current vehicle speed and driving trajectory to calculate the required steering operation, accelerator operation, or brake operation. In addition, the notification control unit 71 notifies the driver of driving operation advice by driving the HMI 43 according to the calculated steering operation, accelerator operation, or brake operation.

[0132] According to the third variant, even when the vehicle is being manually driven, driving advice is provided based on learning data that affects the driver's perception of risk, thereby providing assistance to the driver to achieve driving operations with less risk.

[0133] Although the preferred embodiment of the present invention has been described in detail above with reference to the accompanying drawings, the present invention is not limited to such an example. It is clear that a person having ordinary knowledge in the technical field to which the present invention pertains can conceive of various modified or altered examples within the scope of the technical ideas described in the claims, and it is understood that these also naturally belong to the technical scope of the present invention.

[0134] For example, in the above embodiment, the driving assistance control device 50 calculates the driver's feeling of "fear" and the factors that caused the driver to feel "fear", but the present invention is not limited to such an example. Either or both of the driver's feeling of "fear" and the factors that caused the driver to feel "fear" may be registered by the driver's own input operation. Specifically, multiple levels of "fear" are set in advance, and when the driver feels "fear", the driver is allowed to select the corresponding level. Alternatively, multiple items are prepared in advance as risk factors, and when the driver feels "fear", the driver is allowed to select the corresponding item. In this case, the input operation may be, for example, an input operation on a touch panel or the like, a switching operation of a changeover switch with items set in advance, or a voice input. [Explanation of symbols]

[0135] 1...vehicle control system, 21...in-vehicle camera, 23...exterior camera, 25...surrounding environment sensor, 27...vehicle state / behavior sensor, 45...vehicle drive control unit, 50...driving assistance control device, 51...control unit, 53...storage unit, 55...reference data storage unit, 61...surrounding environment detection unit, 63...driver image processing unit, 65...learning unit, 67...driving condition setting unit, 69...notification control unit

Claims

1. a storage unit that stores a risk level indicating a degree of risk that a driver feels from an obstacle around the vehicle, learning data of risk factors that are factors that cause the driver to feel the risk, and data of internal characteristics indicating personal characteristics that affect the driver's risk perception; a control unit that acquires data on obstacles around the vehicle, and when an obstacle is detected, sets a risk map showing a distribution of risk values ​​with a maximum value at the position of each obstacle and increasing as the obstacle approaches the obstacle, and an allowable risk value showing a risk value that is tolerable to each driver, and outputs control information for driving the vehicle in an area where the allowable risk value is equal to or greater than the risk value, the control information reducing the risk factor for the obstacle to a system that controls driving of the vehicle, The control unit is setting a maximum value of the risk value, a range of the maximum value, and a gradient along which the risk value changes for each of the obstacles based on data of the traveling state of the host vehicle, data of obstacles around the host vehicle, and the learning data; setting the allowable risk value based on the data of the running state of the host vehicle and the data of the internal characteristics; A driving assistance control device that sets a driving trajectory and a vehicle speed such that the allowable risk value is equal to or greater than the risk value.

2. The smaller the vehicle speed of the host vehicle, the larger the tolerable risk value is set, The control unit is For each of the obstacles, the higher the risk level and the higher the importance of the risk factor, the higher the maximum value of the risk value is increased, the range of the maximum value is expanded, and the gradient of the change in the risk value is made gentler, or at least one of them is changed; The driving assistance control device according to claim 1 , wherein, when it is not possible to set a driving trajectory such that the tolerable risk value is equal to or greater than the risk value, the vehicle speed is reduced to set a driving trajectory such that the tolerable risk value is equal to or greater than the risk value.

3. The control unit is 3. The driving assistance control device according to claim 2, wherein, when the plurality of obstacles are detected, the driving trajectory and the vehicle speed are set by giving priority to an operation for reducing the risk factor for an obstacle having a high risk level based on the learning data.

4. The vehicle is a vehicle capable of executing automatic driving control, The control unit is The driving assistance control device according to claim 3 , further comprising: a vehicle driving control unit configured to execute the automatic driving control, and configured to output information on the set driving trajectory and the set vehicle speed to the vehicle driving control unit.

Images