Driving assistance device and recording medium

Abstract

To provide a drive support device which can set a drive condition that may reduce anxiety and a sense of discomfort felt by all of occupants when executing drive support control for avoiding an obstacle with automatic driving and provide a recording medium.SOLUTION: A drive support device comprises: a storage unit which stores information on an individual risk potential set to an own vehicle through learning of a risk felt by a driver of the own vehicle with respect to an obstacle in manual driving; an own vehicle risk calculation unit which sets, to the own vehicle an own vehicle, a risk map on which the individual risk potential of an occupant of the own vehicle in automatic driving of the own vehicle is reflected; and a drive condition setting unit which sets a drive condition in the automatic driving of the own vehicle on the basis of the information on an obstacle risk map on which an obstacle risk potential set to each obstacle around the own vehicle is reflected and the information on the own vehicle risk map.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present disclosure relates to a driving assistance device and a recording medium that reflect a driver's driving characteristics in automatic driving control. [Background technology]

[0002] In recent years, vehicles equipped with driving assistance functions and automatic driving functions have been put into practical use, mainly for the purpose of reducing traffic accidents and reducing the driver's burden. For example, a device is known that detects obstacles around the vehicle based on information detected by various sensors such as an exterior camera and LiDAR (Light Detection and Ranging) installed in the vehicle, and assists the driving of the vehicle to avoid collision between the vehicle and the obstacle.

[0003] For example, Patent Document 1 proposes a driving assistance device capable of realizing driving assistance control suited to the actual outside environment of the vehicle and the driver's feelings. Specifically, Patent Document 1 discloses a driving assistance device that generates a distribution of risk potential based on objects recognized in front of the vehicle, and sets a driving area in which the vehicle can travel based on the distribution of risk potential, thereby setting a driving area suited to the actual outside environment of the vehicle and the driver's feelings, thereby realizing suitable driving assistance control. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2009-169535 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, although the driving assistance device disclosed in Patent Document 1 can realize driving assistance control that matches the driver's feelings, the realized driving assistance control does not take into account the feelings of passengers other than the driver, and therefore may cause passengers other than the driver to feel anxious or uncomfortable.

[0006] The present disclosure has been made in consideration of the above problems, and an object of the present disclosure is to provide a driving assistance device and a recording medium that are capable of setting driving conditions that can reduce the anxiety and discomfort felt by all occupants when performing driving assistance control to avoid obstacles through autonomous driving. [Means for solving the problem]

[0007] In order to solve the above problems, according to one aspect of the present disclosure, there is provided a driving assistance device that reflects driving characteristics of a driver learned during manual driving of the host vehicle in automatic driving control of the host vehicle, the driving assistance device including: a memory unit that learns the risks felt by the driver of the host vehicle from obstacles during manual driving and stores information of a personal risk potential that is set for the host vehicle; a host vehicle risk calculation unit that sets a host vehicle risk map that reflects the personal risk potentials of the occupants of the host vehicle during automatic driving of the host vehicle; and a driving condition setting unit that sets driving conditions during automatic driving of the host vehicle based on information on an obstacle risk map that reflects obstacle risk potentials set for each of the obstacles around the host vehicle and information on the host vehicle risk map.

[0008] In addition, in order to solve the above problem, according to another aspect of the present disclosure, a recording medium storing a computer program is provided that causes a processor to execute operations including: reading out information of personal risk potential set for the host vehicle by learning the risks felt by the driver of the host vehicle from obstacles during manual driving; setting for the host vehicle a host vehicle risk map reflecting the personal risk potentials of the occupants of the host vehicle during automatic driving of the host vehicle; and setting driving conditions for the host vehicle during automatic driving based on information of the obstacle risk map reflecting the obstacle risk potentials set for each of the obstacles around the host vehicle and information of the host vehicle risk map. Effect of the Invention

[0009] As described above, according to the present disclosure, when performing driving assistance control to avoid obstacles through autonomous driving, it is possible to set driving conditions that can reduce the anxiety and discomfort felt by all occupants. [Brief description of the drawings]

[0010] [Figure 1] 1 is a schematic diagram illustrating a configuration example of a vehicle equipped with a driving assistance device according to an embodiment of the present disclosure. [Diagram 2] 2 is a block diagram showing a configuration example of a driving assistance device according to the embodiment; FIG. [Diagram 3] FIG. 4 is an explanatory diagram showing an example of an obstacle risk potential. [Figure 4] FIG. 11 is another explanatory diagram showing an example of an obstacle risk potential. [Diagram 5] 5 is a flowchart showing a process of a learning phase performed by the driving assistance device according to the embodiment; [Figure 6] FIG. 2 is an explanatory diagram showing a coordinate system for setting a risk potential and a risk map. [Figure 7] FIG. 11 is an explanatory diagram showing an example of an individual risk potential. [Figure 8] 1 is an explanatory diagram showing the overlap area risk between an obstacle risk map and a vehicle risk map. FIG. [Figure 9]FIG. 11 is an explanatory diagram showing an example of individual risk potentials of drivers with different gradient coefficients. [Figure 10] 5 is a flowchart showing a process of an execution phase performed by the driving assistance device according to the embodiment; [Figure 11] 10 is a flowchart showing an example of a host vehicle risk map setting process performed by the driving assistance device according to the embodiment. [Figure 12] FIG. 10 is an explanatory diagram showing an example of an own vehicle risk map generated by reflecting personal risk potentials of a plurality of occupants; [Figure 13] 10 is a flowchart showing another example of the host vehicle risk map setting process performed by the driving assistance device according to the embodiment. [Figure 14] FIG. 11 is an explanatory diagram showing a method of correcting the personal risk potential of an occupant sitting in the passenger seat. [Figure 15] FIG. 11 is an explanatory diagram showing a method of correcting the personal risk potential of an occupant sitting in the rear seat. [Figure 16] FIG. 13 is an explanatory diagram showing a change in overlap area risk due to differences in travel trajectories. [Figure 17] FIG. 11 is an explanatory diagram showing how the overlap area risk decreases as the vehicle speed decreases. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, preferred embodiments of the present disclosure 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 denoted by the same reference numerals, and redundant description will be omitted.

[0012] <1. Overall vehicle configuration> First, an example of the overall configuration of a vehicle to which a driving assistance device according to an embodiment of the present disclosure can be applied will be described.

[0013] FIG. 1 is a schematic diagram showing an example of the configuration of a host vehicle 1 equipped with a driving support device 50. As shown in FIG. 1 is configured as a four-wheel drive vehicle that transmits drive torque output from a drive force source 9 that generates drive torque for the vehicle to a left front wheel 3LF, a right front wheel 3RF, a left rear wheel 3LR, and a right rear wheel 3RR (hereinafter collectively referred to as "wheels 3" unless a distinction is required). The drive force source 9 may be an internal combustion engine such as a gasoline engine or a diesel engine, a drive motor, or may include both an internal combustion engine and a drive motor.

[0014] The host vehicle 1 may be, for example, an electric vehicle equipped with two drive motors, a front-wheel drive motor and a rear-wheel drive motor, or an electric vehicle equipped with drive motors corresponding to the respective wheels 3. If the host vehicle 1 is an electric vehicle or a hybrid electric vehicle, the host vehicle 1 is equipped with a secondary battery that stores power supplied to the drive motors, and a generator such as a motor or fuel cell that generates power to charge the battery.

[0015] The host vehicle 1 is equipped with a driving force source 9, an electric steering device 15, and brake devices 17LF, 17RF, 17LR, 17RR (hereinafter collectively referred to as "brake devices 17" unless a distinction is required) as devices used to control the driving of the host vehicle 1. The driving force source 9 outputs a driving torque that is transmitted to the front wheel drive shaft 5F and the rear wheel drive shaft 5R via a transmission, a front wheel differential mechanism 7F, and a rear wheel differential mechanism 7R (not shown). The driving of the driving force source 9 and the transmission is controlled by a vehicle control device 41 that includes one or more electronic control devices (ECU: Electronic Control Unit).

[0016] The front-wheel drive shaft 5F is provided with an electric steering device 15. The electric steering device 15 includes an electric motor and a gear mechanism (not shown), and is controlled by a vehicle control device 41 to adjust the steering angles of the left front wheel 3LF and the right front wheel 3RF. During manual driving, the vehicle control device 41 controls the electric steering device 15 based on the steering angle of the steering wheel 13 by the driver. During automatic driving, the vehicle control device 41 controls the electric steering device 15 based on a target steering angle set by the driving assistance device 50.

[0017] The brake devices 17LF, 17RF, 17LR, and 17RR apply braking forces to the front, rear, left, and right drive wheels 3LF, 3RF, 3LR, and 3RR, respectively. The brake devices 17 are configured as hydraulic brake devices, for example, and generate a predetermined braking force by controlling the hydraulic pressure supplied to each brake device 17 by the vehicle control device 41. When the host vehicle 1 is an electric vehicle or a hybrid electric vehicle, the brake devices 17 are used in combination with regenerative braking by the drive motor.

[0018] The vehicle control device 41 includes one or more electronic control devices that control the driving of the driving force source 9 that outputs the driving torque of the host vehicle 1, the electric steering device 15 that controls the steering angle of the steering wheel or the steering wheels, and the brake device 17 that controls the braking force of the host vehicle 1. The vehicle control device 41 may have a function of controlling the driving of a transmission that changes the speed of the output output from the driving force source 9 and transmits it to the wheels 3. The vehicle control device 41 is configured to be able to acquire information transmitted from the driving assistance device 50, and is configured to be able to execute automatic driving control of the host vehicle 1.

[0019] The vehicle 1 also has forward-facing cameras 31LF, 31RF, a rear-facing camera 31R, a LiDAR (Light Detection And Ranging) 31S, an interior camera 33, a vehicle condition sensor 35, a GPS (Global Positioning System) sensor 37 and an HMI (Human Machine Interface) 43.

[0020] The front photographing cameras 31LF, 31RF, the rear photographing camera 31R, and the LiDAR 31S constitute an ambient environment sensor for acquiring information on the ambient environment of the vehicle 1. The front photographing cameras 31LF, 31RF and the rear photographing camera 31R photograph the front or rear of the vehicle 1 and generate image data. The front photographing cameras 31LF, 31RF and the rear photographing camera 31R are equipped with imaging elements such as CCDs (Charged-Coupled Devices) or CMOSs ​​(Complementary Metal-Oxide-Semiconductors), and transmit the generated image data to the driving assistance device 50.

[0021] 1, the front photographing cameras 31LF, 31RF are configured as a stereo camera including a pair of left and right cameras, and the rear photographing camera 31R is configured as a so-called monocular camera, but each may be either a stereo camera or a monocular camera. In addition to the front photographing cameras 31LF, 31RF and the rear photographing camera 31R, the vehicle 1 may also be equipped with cameras provided on the side mirrors 11L, 11R, for example, to photograph the left rear or right rear.

[0022] The LiDAR 31S transmits optical waves and receives reflected waves of the optical waves, and detects an object and the distance to the object based on the time from transmitting the optical waves to receiving the reflected waves. The LiDAR 31S transmits the detection data to the driving assistance device 50. In addition, the host vehicle 1 may be equipped with one or more sensors of a radar sensor such as a millimeter wave radar and an ultrasonic sensor as a surrounding environment sensor for acquiring information about the surrounding environment.

[0023] The interior camera 33 captures images of the interior of the vehicle and generates image data. The interior camera 33 includes an imaging element such as a CCD or a CMOS, and transmits the generated image data to the driving support device 50. In this embodiment, the interior camera 33 is disposed so as to be able to capture images of passengers aboard the vehicle 1. Only one interior camera 33 may be installed, or multiple interior cameras 33 may be installed.

[0024] The vehicle state sensor 35 is composed of at least one sensor that detects the operation state and behavior of the host vehicle 1. The vehicle state sensor 35 includes at least one of a steering angle sensor, an accelerator position sensor, a brake stroke sensor, a brake pressure sensor, and an engine speed sensor, and detects the operation state of the host vehicle 1, such as the steering angle of the steering wheel or steering wheels, the accelerator opening, the brake operation amount, or the engine speed. The vehicle state sensor 35 also includes at least one of a vehicle speed sensor, an acceleration sensor, and an angular velocity sensor, and detects the behavior of the vehicle, such as the vehicle speed, longitudinal acceleration, lateral acceleration, and yaw rate. The vehicle state sensor 35 transmits a sensor signal including the detected information to the driving assistance device 50.

[0025] The GPS sensor 37 receives satellite signals from GPS satellites. The GPS sensor 37 transmits position information on the map data of the vehicle 1 contained in the received satellite signals to the driving assistance device 50. Note that instead of the GPS sensor 37, an antenna for receiving satellite signals from other satellite systems that identify the position of the vehicle 1 may be provided.

[0026] The HMI 43 is driven by the driving support device 50 and presents various information to the driver by means of image display, audio output, etc. The HMI 43 includes, for example, a display device provided in an instrument panel and a speaker provided in the vehicle. 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 1.

[0027] <2. Driving support devices> Next, the driving support device 50 according to this embodiment will be described in detail.

[0028] (2-1. Configuration example) FIG. 2 is a block diagram showing an example of the configuration of the driving support device 50 according to this embodiment. The driving support device 50 is connected to an ambient environment sensor 31, an in-vehicle camera 33, a vehicle state sensor 35, and a GPS sensor 37 directly or via a communication means such as a controller area network (CAN) or a local inter net (LIN). The driving support device 50 is also connected to a vehicle control device 41 and an HMI 43. The driving support device 50 is not limited to an electronic control device mounted on the vehicle 1, and may be a terminal device such as a smartphone or a wearable device.

[0029] The driving support device 50 includes a control unit 51, a storage unit 53, a driving characteristic database 55, and an occupant identification database 57. The control unit 51 includes one or more processors such as a central processing unit (CPU). A part or all of the control unit 51 may be configured with an updatable firmware or the like, or may be a program module or the like executed by a command from the CPU or the like. The storage unit 53 includes a memory such as a random access memory (RAM) or a read only memory (ROM). However, the number and type of the storage units 53 are not particularly limited. The storage unit 53 stores information such as computer programs executed by the control unit 51, various parameters used in the calculation process, detection data, and calculation results.

[0030] The driving characteristics database 55 and the occupant identification database 57 are configured by a memory such as a RAM, or an updatable recording medium such as a hard disk drive (HDD), a compact disk (CD), a digital versatile disk (DVD), a solid state drive (SSD), a USB flash, a storage device, etc. However, the type of the recording medium is not particularly limited.

[0031] The driving characteristics database 55 is a database that stores information on personal risk potentials that reflect the driving characteristics of each driver that are learned when the vehicle 1 is manually driven. The occupant identification database 57 is a database that stores information for identifying occupants of the vehicle 1 (hereinafter also referred to as "occupant identification information"). The occupant identification information may be, for example, an identification number or an identification symbol. However, the occupant identification information is not limited to the above examples. The personal risk potential and occupant identification information will be described in detail later.

[0032] (2-2. Setting of driving conditions based on risk potential) Before describing the specific processing of the driving support device 50, a brief outline of the processing for setting driving conditions based on the risk potential, which is executed by the driving support device 50, will be given.

[0033] 3 and 4 are explanatory diagrams showing the risk potential (obstacle risk potential) set for each obstacle. In FIG. 3 and FIG. 4, examples of risk potentials set for a vehicle are shown. The value of the risk potential (risk value) R i The risk value R is maximum in the area where the obstacle (vehicle) exists, and decreases as the distance from the outer edge of the obstacle (vehicle) increases. i is the distance from the obstacle l i For example, it can be expressed by the following formula (1):

[0034]

number

[0035] R i :Risk value C i : Risk absolute value (gain) l i : Distance from obstacle σ i :Slope coefficient r i : Obstacle radius i: Numbering to distinguish obstacles

[0036] In this embodiment, the risk value R i is specified in the range of "0" to "1", and the distance to the obstacle l i The absolute risk value C, which is the risk value when i is set to "1" and the area is prohibited from travel. However, the absolute risk value C i may be set for each obstacle as a value that depends on the obstacle. For example, when the obstacle is a "vehicle" or a "low curb", the risk absolute value C for the "vehicle" is set as the risk of collision with the vehicle is higher than the risk of collision with the low curb. i is the absolute risk value C for "low curb" i is set to a value greater than

[0037] Gradient coefficient σ i is a value that is set according to the type of obstacle. i may be set according to, for example, a Gaussian function or an exponential function. If the obstacle is a moving object such as a vehicle traveling around the host vehicle 1, the risk in the traveling direction of the vehicle is high, so the depth of the risk in front of the vehicle may be set wider than the risk behind the vehicle, as shown in Fig. 4. In this case, the depth of the risk in front may be variable depending on the vehicle speed of the vehicle or the vehicle speed relative to the host vehicle.

[0038] When the traveling trajectory and acceleration / deceleration of the host vehicle 1 are set using the risk potential, a risk potential is set for each obstacle detected while the host vehicle 1 is traveling, and the spatial overlap of each obstacle risk potential is added to obtain an obstacle risk map (potential field) that indicates the collision risk with multiple obstacles. At that time, the maximum risk value among the risk values ​​of the obstacle risk potentials described above may be used as the risk value of the obstacle risk map at that point. In such an obstacle risk map, the level of risk is shown as contour lines on a two-dimensional plane. As described above, since the risk values ​​have a two-dimensional distribution, it is possible to select a trajectory with a lower risk. The obstacle risk map may be calculated by taking into account not only the obstacles that are manifested, but also the risks that are not manifested (latent risks). For example, when passing through an area that is a blind spot due to an obstruction after a turn, a potential risk may be given on the assumption that a pedestrian or vehicle may jump out of the blind spot area, and this may be reflected in the obstacle risk map.

[0039] The driving assistance device 50 according to this embodiment generates an obstacle risk map based on the obstacle risk potentials set for obstacles around the host vehicle 1, and also sets a host vehicle risk map for the host vehicle 1. The host vehicle risk map reflects the driving characteristics of one or more occupants of the host vehicle 1 during manual driving, making it possible to set driving conditions that can reduce the anxiety and discomfort felt by multiple occupants.

[0040] (2-3. Functional configuration) The driving assistance device 50 is constructed as a device that learns the driving characteristics of each driver when the vehicle 1 is being manually driven, and sets the driving conditions of the vehicle 1 taking into account the driving characteristics of all occupants of the vehicle 1 when the vehicle 1 is being automatically driven.

[0041] 2, the control unit 51 of the driving support device 50 includes a surrounding environment detection unit 61, an occupant detection unit 63, a driving state detection unit 65, an individual risk learning unit 67, a vehicle risk calculation unit 69, and a driving condition setting unit 71. Each of these units is a function realized by the execution of a computer program by a processor such as a CPU.

[0042] (Ambient environment detection section) The surrounding environment detection unit 61 detects the surrounding environment of the vehicle 1 based on the detection data transmitted from the surrounding environment sensor 31. Specifically, the surrounding environment detection unit 61 calculates the type, size (width, height, and depth), and position of an obstacle present around the vehicle 1, the distance from the vehicle 1 to the obstacle, and the relative speed between the vehicle 1 and the obstacle. The detected obstacles include other vehicles in motion, parked vehicles, pedestrians, bicycles, side walls, curbs, buildings, utility poles, traffic signs, traffic signals, natural objects, and any other objects present around the vehicle. The surrounding environment detection unit 61 may also have a lane recognition function, such as detecting boundary lines on a road. The processing by the surrounding environment detection unit 61 will be described in detail later.

[0043] (Occupant detection section) The occupant detection unit 63 executes a process of detecting an occupant of the vehicle 1 based on image data transmitted from the interior photographing camera 33. Specifically, during manual driving of the vehicle 1, the occupant detection unit 63 extracts the occupant identification information of the driver sitting in the driver's seat, who is identified based on the image data transmitted from the interior photographing camera 33, from the occupant identification information stored in the occupant identification database 57, and stores the occupant identification information of the identified driver in the storage unit 53. If the occupant identification information of the corresponding driver is not stored in the occupant identification database 57, the occupant detection unit 63 stores the occupant identification information in the occupant identification database 57 together with the detected driver data.

[0044] In addition, during autonomous driving of the vehicle 1, the occupant detection unit 63 extracts occupant identification information of each occupant identified based on image data transmitted from the interior camera 33 from the occupant identification information stored in the occupant identification database 57, and stores the information on the seating position of each occupant in the storage unit 53. In addition, in this embodiment, during autonomous driving of the vehicle 1, the occupant detection unit 63 detects the direction of the occupant's gaze at each predetermined calculation cycle based on image data transmitted from the interior camera 33. The occupant detection unit 63 may detect the direction of the face instead of the direction of the gaze. The process by the occupant detection unit 63 will be described in detail later.

[0045] (Driving condition detection unit) The driving condition detection unit 65 detects information on the operation state and behavior of the vehicle 1 based on the detection data transmitted from the vehicle condition sensor 35. The driving condition detection unit 65 obtains information on the operation state of the vehicle 1, such as the steering angle of the steering wheel or steering wheels, the accelerator opening, the brake operation amount, or the engine speed, and information on the behavior of the vehicle, such as the vehicle speed, the longitudinal acceleration, the lateral acceleration, and the yaw rate, at predetermined calculation intervals, and stores this information in the memory unit 53.

[0046] (Individual Risk Learning Department) The personal risk learning unit 67 learns the risk the driver feels from an obstacle when manually driving the vehicle 1, and calculates a personal risk potential. The personal risk potential is learned data of the risk (sense of danger of collision) the driver feels from an obstacle depending on the distance from the outer periphery of the vehicle 1, and is learned based on data of the distance between the vehicle 1 and the obstacle acquired during manual driving, and is set for the vehicle 1. In this embodiment, the personal risk learning unit 67 learns parameters (hereinafter also referred to as "personal risk potential parameters") that define the personal risk potential set for the vehicle 1 for each driver, and stores them in the driving characteristics database 55 in association with the driver's occupant identification information.

[0047] In this embodiment, the magnitude of the risk value of the area around the area where the vehicle 1 is located in the personal risk potential or the width of the set range is corrected according to the vehicle speed. The personal risk learning unit 67 calculates an overlap area risk, which is a risk value of an overlap area where an obstacle risk map generated by superimposing obstacle risk potentials set for obstacles around the vehicle 1 and the learned personal risk potential overlap when the vehicle 1 is manually driven. The personal risk learning unit 67 then learns the relationship between the overlap area risk and the vehicle speed for each driver. This allows each driver to learn to what level the overlap area risk is reduced by deceleration. Hereinafter, the level at which each driver reduces the overlap area risk is referred to as the "personal overlap area risk threshold". The personal risk learning unit 67 stores the learned personal overlap area risk threshold in the driving characteristics database 55 in association with the occupant identification information of the driver. The processing by the personal risk learning unit 67 will be described in detail later.

[0048] (Vehicle risk calculation section) During autonomous driving of the host vehicle 1, the host vehicle risk calculation unit 69 sets a host vehicle risk map for the host vehicle 1 that reflects the personal risk potentials of the occupants of the host vehicle 1. Specifically, during autonomous driving of the host vehicle 1, the host vehicle risk calculation unit 69 extracts information on personal risk potentials corresponding to the occupant identification numbers stored in the memory unit 53 from information on personal risk potentials stored in the driving characteristics database 55. In addition, the host vehicle risk calculation unit 69 sets the host vehicle risk map using the extracted information on personal risk potentials. The processing by the host vehicle risk calculation unit 69 will be explained in detail later.

[0049] (Operation condition setting section) During autonomous driving of the host vehicle 1, the driving condition setting unit 71 sets driving conditions for the host vehicle 1 based on information on an obstacle risk map reflecting obstacle risk potentials set for each obstacle around the host vehicle 1 and information on the host vehicle risk map set by the host vehicle risk calculation unit 69. Specifically, during autonomous driving of the host vehicle 1, the driving condition setting unit 71 sets an obstacle risk potential for each obstacle detected by the surrounding environment detection unit 61, and generates an obstacle risk map in which all obstacle risk potentials are superimposed.

[0050] Furthermore, the driving condition setting unit 71 extracts all information on personal risk potentials corresponding to the occupant identification information of the occupants of the vehicle 1 stored in the memory unit 53, and generates an own vehicle risk map. At that time, the driving condition setting unit 71 may correct the personal risk potentials according to the seating positions of the occupants. Furthermore, in this embodiment, the driving condition setting unit 71 extracts all personal overlap area risk thresholds corresponding to the occupant identification information of the occupants of the vehicle 1 stored in the memory unit 53, and sets overlap area risk thresholds to be used in the setting process of the driving conditions.

[0051] The driving condition setting unit 71 sets a planned driving trajectory of the vehicle 1, calculates an overlap area risk for an overlap area between the generated obstacle risk map and the vehicle risk map, and sets the planned driving trajectory with the minimum overlap area risk as the target trajectory of the vehicle 1. In this embodiment, the driving condition setting unit 71 sets a target vehicle speed of the vehicle 1 so that the overlap area risk is equal to or less than an overlap area risk threshold. The driving condition setting unit 71 sets a target steering angle and a target acceleration / deceleration based on the set target trajectory and target vehicle speed information, and transmits this information to the vehicle control device 41. The vehicle control device 41, which has received the information on the driving conditions, controls the drive of each control device based on the set information on the driving conditions. The processing by the driving condition setting unit 71 will be described in detail later.

[0052] <3. Operation of driving support device> Next, an example of the operation of the driving support device 50 according to this embodiment will be specifically described. Below, the example of the operation of the driving support device 50 will be described separately for a learning phase process executed during manual driving and an execution phase process executed during automatic driving.

[0053] (3-1. Learning Phase Processing) FIG. 5 is a flowchart showing an example of a process executed in the learning phase. First, when the in-vehicle system including the driving assistance device 50 is started (step S11), the personal risk learning unit 67 of the control unit 51 determines whether or not the host vehicle 1 is being manually driven (step S13). For example, the personal risk learning unit 67 determines whether or not the driving mode changeover switch is set to the manual driving mode. The driving mode is configured to be switched based on, for example, an operation input by the occupant of the host vehicle 1. If the host vehicle 1 is not being manually driven (S13 / No), the control unit 51 proceeds to processing of the execution phase.

[0054] On the other hand, if the vehicle 1 is being manually driven (S13 / Yes), the occupant detection unit 63 of the control unit 51 executes a process of identifying the driver (step S15). For example, the occupant detection unit 63 executes a face recognition process using image data transmitted from the interior photographing camera 33 to detect an occupant sitting in the driver's seat. The occupant detection unit 63 also extracts facial features of the occupant sitting in the driver's seat and identifies the corresponding occupant in light of the data of the features accumulated in the occupant identification database 57. The occupant detection unit 63 stores the identified occupant identification information in the storage unit 53. If the data of the corresponding occupant is not stored in the occupant identification database 57, the occupant detection unit 63 adds the occupant identification information together with the extracted data of the facial features and stores them in the occupant identification database 57.

[0055] Next, the surrounding environment detection unit 61 of the control unit 51 acquires surrounding environment information of the host vehicle 1 (step S17). Specifically, the surrounding environment detection unit 61 detects obstacles present around the host vehicle 1 based on the detection data transmitted from the surrounding environment sensor 31. The surrounding environment detection unit 61 also calculates the position, type, and size (width, height, and depth) of the detected obstacle, the distance from the host vehicle 1 to the obstacle, and the relative speed between the host vehicle 1 and the obstacle. The detected obstacles include other vehicles in motion, parked vehicles, pedestrians, bicycles, side walls, curbs, buildings, utility poles, traffic signs, traffic signals, natural objects, and any other objects present around the vehicle.

[0056] For example, the surrounding environment detection unit 61 detects an obstacle in front of the vehicle 1 and the type of the obstacle by performing image processing on the image data transmitted from the front photographing cameras 31LF, 31RF, using a pattern matching technique or the like. The surrounding environment detection unit 61 also calculates the position, size, and distance to the obstacle as seen from the vehicle 1, based on the position of the obstacle in the image data, the size of the obstacle in the image data, and the parallax information of the left and right front photographing cameras 31LF, 31RF. Furthermore, the surrounding environment detection unit 61 calculates the relative speed between the vehicle 1 and the obstacle by differentiating the change in distance with respect to time.

[0057] The surrounding environment detection unit 61 may also detect an obstacle based on detection data transmitted from the LiDAR 31S. For example, the surrounding environment detection unit 61 may calculate the position, type, size, and distance from the host vehicle 1 to the obstacle based on information on the time from transmitting an electromagnetic wave from the LiDAR 31S to receiving a reflected wave, the direction in which the reflected wave is received, and the range of the measurement point group of the reflected wave. The surrounding environment detection unit 61 may also calculate the relative speed between the host vehicle 1 and the obstacle by time-differentiating the change in distance.

[0058] The surrounding environment detection unit 61 may also acquire information about an obstacle ahead of the vehicle 1 based on information about the position of the vehicle 1 on map data acquired via the GPS sensor 37 and obstacle position information acquired via a communication means with the outside of the vehicle. The surrounding environment detection unit 61 stores the acquired surrounding environment information in the storage unit 53.

[0059] Next, the personal risk learning unit 67 of the control unit 51 judges whether or not an obstacle to be avoided exists on the traveling trajectory of the vehicle 1 based on the surrounding environment information acquired by the surrounding environment detection unit 61 (step S19). For example, the personal risk learning unit 67 judges whether or not an obstacle exists in an area within a preset distance range from the vehicle 1 when the vehicle 1 travels along the traveling trajectory of the vehicle 1 specified based on the information of the speed, acceleration / deceleration, and steering angle of the vehicle 1 detected by the traveling state detection unit 65. If it is judged that an obstacle to be avoided does not exist on the traveling trajectory of the vehicle 1 (S19 / No), the process returns to step S17, and the process of detecting the surrounding environment information and the process of judging the presence or absence of an obstacle are repeated.

[0060] On the other hand, if it is determined that an obstacle to be avoided is present on the travel path of the vehicle 1 (S19 / Yes), the personal risk learning unit 67 updates the obstacle risk map (step S21). Specifically, the personal risk learning unit 67 sets an obstacle risk potential for each obstacle according to the type, size, relative speed, etc. of the obstacle. In this embodiment, the risk value is defined within a range from "0" to "1", and the risk value of the area in which the obstacle exists is set to "1", making the area untravelable. In addition, the risk value is set so as to gradually decrease with distance from the outer periphery of the area in which the obstacle exists. The gradient at which the risk value decreases may be set according to, for example, a Gaussian function or an exponential function.

[0061] Furthermore, the personal risk learning unit 67 generates an obstacle risk map by overlapping the obstacle risk potentials set for each obstacle. The personal risk learning unit 67 sets an obstacle risk potential for each obstacle, and for areas where the obstacle risk potentials of different obstacles overlap, adds up the risk values ​​at each position to generate an obstacle risk map that shows the distribution of risk values ​​around the vehicle 1. When adding up the overlapping obstacle risk potentials, the maximum risk value among the risk values ​​of multiple obstacle risk potentials at a certain point on a two-dimensional plane may be used as the risk value of the obstacle risk map at that point. When there is one obstacle, the obstacle risk map is generated based on the obstacle risk potential of that obstacle. In such an obstacle risk map, the level of risk is shown as contour lines on a two-dimensional plane.

[0062] Next, the personal risk learning unit 67 stores data on the running trajectory and vehicle speed of the vehicle 1 in the memory unit 53 (step S23). The running trajectory of the vehicle 1 can be represented by a point cloud of the position of the vehicle 1 plotted on the generated obstacle risk map. In addition, the vehicle speed is detected by the running state detection unit 65.

[0063] Next, the personal risk learning unit 67 judges whether or not the host vehicle 1 has completed avoiding the obstacle that was present on the travel path (step S25). For example, the personal risk learning unit 67 judges that the host vehicle 1 has completed avoiding the obstacle when the obstacle recognized in step S19 is no longer detected by the surrounding environment detection unit 61. If it is judged that the host vehicle 1 has not completed avoiding the obstacle that was present on the travel path (S25 / No), the host vehicle risk calculation unit 69 returns to step S21 and repeats the process of updating the obstacle risk map in accordance with the travel of the host vehicle 1, and the process of storing the data of the travel path and vehicle speed of the host vehicle 1.

[0064] On the other hand, when it is determined that the vehicle 1 has completed avoidance of the obstacle that was on the travel path (S25 / Yes), the personal risk learning unit 67 learns the personal risk potential parameters (step S27). In this embodiment, the personal risk learning unit 67 learns the gradient coefficient σ x ,σ y The personal risk learning unit 67 also learns a personal overlap area risk threshold value when the driver reduces the overlap area risk by decelerating when avoiding an obstacle (step S29). The personal risk learning unit 67 calculates a personal risk potential parameter and a personal overlap area risk threshold value based on the obstacle data stored in the memory unit 53 and the travel trajectory and vehicle speed data of the host vehicle 1 until the obstacle avoidance is completed. Hereinafter, the learning process of the personal risk potential parameter and the learning process of the personal overlap area risk threshold value executed by the personal risk learning unit 67 in this embodiment will be described in detail.

[0065] The driver feels the risk of collision depending on the distance from the outer periphery of the vehicle 1 to the obstacle. In this embodiment, the personal risk potential is expressed as a risk value of collision that the driver feels depending on the distance from the outer periphery of the vehicle 1 to the obstacle, within a range from "0" to "1" like the obstacle risk potential shown in Fig. 3. The personal risk potential can be expressed as a function in which the risk value of the area in which the vehicle 1 exists is "1" and the risk value gradually decreases the further away from the outer periphery of the area.

[0066] As shown in FIG. 6, in a two-dimensional coordinate system in which the longitudinal direction of the host vehicle 1 is the x-axis and the width direction of the vehicle is the y-axis, the vehicle length of the host vehicle 1 is L x , the width of vehicle 1 is L y , the gradient coefficient in the x-axis direction is σ x , the gradient coefficient in the y-axis direction is σ y Then, the risk value R of the individual risk potential RP_e e can be expressed by the following formulas (2) to (5).

[0067] (Area where vehicle 1 exists)

number

[0068] (Vehicle width L y Within the range of vehicle length L of vehicle 1 x (Area exceeding

number

[0069] (Vehicle length L x Within the range of vehicle width L y (Area exceeding

number

[0070] (Vehicle length L x and the vehicle width L of the vehicle 1 y (Area exceeding

number

[0071] The gradient coefficient σ of the personal risk potential RP_e expressed by the above formulas (2) to (5) x ,σ y is a parameter that defines the individual risk potential RP_e, and the gradient coefficient σ x ,σ y By adjusting the gradient coefficient σ, the personal risk potential RP_e can be adapted to the driving characteristics of each driver. x ,σ y may be calculated for each of the front (x>0) and rear (x<0) of the host vehicle 1 and the left (y>0) and right (y<0) sides of the host vehicle 1.

[0072] Fig. 7 shows an example of the personal risk potential RP_e set for the host vehicle 1. In the case of a traveling host vehicle 1, since the risk in the traveling direction is high, the depth of the risk potential in front of the host vehicle 1 is set wider than the depth of the risk potential behind the host vehicle 1 to represent the anisotropy of the personal risk potential RP_e. Specifically, the gradient coefficients σ x The depth of the risk potential can be adjusted by setting a coefficient for each of these. In this case, the depth of the risk potential ahead may be variable depending on the speed of the host vehicle 1 or the relative speed between the host vehicle 1 and an obstacle. Alternatively, the center point of the coordinate system shown in FIG. 6 may be moved a distance x c ,y c The anisotropy of the individual risk potential RP_e may be represented by shifting the position by only .times. ...

[0073] Here, as shown in FIG. 8, a case is considered where the host vehicle 1 travels in an environment where there are two parked vehicles, a first parked vehicle 91 and a second parked vehicle 93, in front of the host vehicle 1. The first parked vehicle 91 is parked at the left end of the road, and the second parked vehicle 93 is parked at the right end of the road. The first parked vehicle 91 is parked in front of the second parked vehicle 93. In such an environment, when the host vehicle 1 travels so as not to collide with the first parked vehicle 91 and the second parked vehicle 93, the driver can select various travel trajectories. When the driver travels in a certain environment, the vehicle speed and travel trajectory are determined while comparing the collision risk from an obstacle (obstacle risk potential) with the collision risk according to the distance from the outer periphery of the host vehicle 1 (individual risk potential). In other words, the travel trajectory selected by the driver in a certain environment is considered to be the travel trajectory selected in which the overlapping area risk is the smallest for the driver.

[0074] Specifically, the personal risk learning unit 67 estimates the obstacle risk potential RP 1 ,RP 2and a personal risk potential RP_e is set for the host vehicle 1. The personal risk potential RP_e and each obstacle risk potential RP 1 ,RP 2 The overlap region risk RC in the overlap regions AR1 and AR2 can be expressed by the following formula (6).

[0075]

number

[0076] T p : Time equivalent to the depth in the direction of travel that is taken into account when calculating the trajectory t k : Time for each data sampling period (t=0~Tp) n: number of overlapping regions to consider (2 in this example: AR1 and AR2) m: an integer between 1 and n X i ,Y i :time t k The coordinate points included in the overlapping area ARm

[0077] In other words, the overlapping area risk RC is the coordinate points X i ,Y i Obstacle risk potential RP m The risk value R m and the risk value Re of the individual risk potential RP_e.

[0078] In the environment shown in FIG. 8, the driving trajectory of the vehicle 1 driven by the driver is determined as a trajectory (of the coordinate point group x mr (t k ),y mr (t k )) is set. Therefore, the personal risk learning unit 67 uses the time t k Orbit at (coordinate point X i,Y j ) and calculates a risk value Re that minimizes the overlapping region risk RC calculated based on the above equation (6). Specifically, the personal risk learning unit 67 calculates a gradient coefficient σ of the personal risk potential RP_e that minimizes the overlapping region risk RC. x ,σ y In this manner, the personal risk learning section 67 sets the gradient coefficient σ x ,σ y Learn.

[0079] Figure 9 shows the gradient coefficient σ x ,σ y 8 shows an example of the personal risk potentials RP_e of drivers A and B having different characteristics. In the environment shown in FIG. 8, it is assumed that driver A passes between the first parked vehicle 91 and the second parked vehicle 93 while maintaining equal distances on both the left and right of the host vehicle 1. It is also assumed that driver B passes between the first parked vehicle 91 and the second parked vehicle 93 while maintaining a larger distance to the first parked vehicle 91 on the passenger seat side (left side) than to the second parked vehicle 93 on the driver seat side. In this case, the left and right gradients of the personal risk potential RP_e_A reflecting the driving characteristics of driver A are equal. On the other hand, the left gradient of the personal risk potential RP_e_B reflecting the driving characteristics of driver B is gentler than the right gradient.

[0080] Furthermore, the gradient of the forward / rearward direction of the personal risk potential may also change due to the influence of the driving trajectory. Specifically, when driver A starts steering at a point before driver B when avoiding an obstacle in front of the vehicle 1, driver A is considered to recognize that an overlap area risk RC has occurred at that point, whereas driver B is considered to not recognize that an overlap area risk RC has occurred at that point. In this case, the forward gradient of the personal risk potential RP_e_A reflecting the driving characteristics of driver A will be gentler than the forward gradient of the personal risk potential RP_e_B reflecting the driving characteristics of driver B.

[0081] The gradient coefficient σ of the individual risk potential RP_e for each driver is x ,σ y When setting the driving trajectory, if there is a section where the yaw rate value stored in the storage unit 53 exceeds a preset threshold, the personal risk learning unit 67 may correct the driving trajectory data stored in the storage unit 53 so that the yaw rate value is less than the threshold. This makes it possible to set a driving trajectory within the range of the allowable yaw rate threshold.

[0082] In this embodiment, the personal risk potential RP_e is set to change in accordance with the speed when the host vehicle 1 accelerates or decelerates on the driving path. The change in personal risk potential RP_e due to acceleration or deceleration of the host vehicle 1 during manual driving can be expressed, for example, by the following equation (7).

[0083]

number

[0084] v mr : Vehicle speed after acceleration or deceleration v 0 :Reference speed (arbitrary setting)

[0085] When the host vehicle 1 accelerates, the range of the individual risk potential RP_e expands according to the above formula (7), and the ranges of the overlapping areas AR1, AR2 expand. When the host vehicle 1 decelerates, the range of the individual risk potential RP_e contracts according to the above formula (7), and the ranges of the overlapping areas AR1, AR2 contract. As a result, the number of coordinate points included in the overlapping areas AR1, AR2 changes, and the overlapping area risk RC increases or decreases.

[0086] The personal risk learning unit 67 assumes that the vehicle speed of the vehicle 1 driven by the driver in the environment shown in FIG. 8 is set according to the above formula (7), and kThe relationship between the vehicle speed v and the overlap area risk RC in the vehicle 1 is learned. This allows each driver to learn the personal overlap area risk threshold RC_thr_e when decelerating to reduce the overlap area risk RC. In this case, it is assumed that the deceleration of the host vehicle 1 is set based on the above formulas (6) and (7) according to the magnitude of the difference obtained by subtracting the personal overlap area risk threshold RC_thr_e from the overlap area risk RC.

[0087] In this manner, the personal risk learning unit 67 calculates the gradient coefficient σ x ,σ y In addition, the personal overlap area risk threshold RC_thr_e is learned, and is stored in the driving characteristics database 55 in association with the occupant identification information of the driver (steps S27 to S29).

[0088] Next, the personal risk learning unit 67 judges whether the in-vehicle system has stopped (step S31). If the in-vehicle system has stopped (S31 / Yes), the learning process by the control unit 51 ends. On the other hand, if the in-vehicle system has not stopped (S31 / No), the process returns to step S13 and repeats the processes of the steps described above.

[0089] In this manner, the control unit 51 calculates the gradient coefficient σ 2 that determines the personal risk potential RP_e of each driver based on the data of the driving trajectory and vehicle speed when each driver avoids an obstacle during manual driving of the host vehicle 1. x ,σ y and the individual overlap region risk threshold RC_thr_e. In addition, the control unit 51 learns the learned gradient coefficient σ x ,σ y and the individual overlap area risk threshold RC_thr_e are stored in association with the occupant identification information of the driver in the driving characteristics database 55. Therefore, the collision risk that each driver feels according to the distance from the outer periphery of the vehicle 1 can be learned.

[0090] (3-2. Processing in the Execution Phase) FIG. 10 is a flowchart showing an example of a process executed in the execution phase. First, when the in-vehicle system including the driving assistance device 50 is started (step S41), the driving condition setting unit 71 determines whether the host vehicle 1 is in an autonomous driving state (step S43), similar to the process of step S13. If the host vehicle 1 is not in an autonomous driving state (S43 / No), the control unit 51 proceeds to the learning phase process.

[0091] On the other hand, if the vehicle 1 is in an autonomous driving state (S43 / Yes), the occupant detection unit 63 executes a process of identifying the occupant (step S45). For example, the occupant detection unit 63 executes a face recognition process using image data transmitted from the interior photographing camera 33 to detect the occupants of the vehicle 1 and the seating positions of each occupant. The seating positions can be specified by the positions of the occupants' faces. In addition, the occupant detection unit 63 extracts the facial features of each occupant and specifies the corresponding occupant in light of the data of the features accumulated in the occupant identification database 57. The occupant detection unit 63 associates the specified occupant identification information with the seating position information and stores it in the storage unit 53. If the occupant identification database 57 does not store the data of the corresponding occupant, the occupant detection unit 63 may process the occupant as not existing, or may store information that the occupant identification information is unknown in the storage unit 53 in association with the seating position information.

[0092] Next, the host vehicle risk calculation unit 69 sets the host vehicle risk map RM_e reflecting the personal risk potential RP_e of the occupant of the host vehicle 1 for the host vehicle 1 (step S47). In this embodiment, an overlapping area risk threshold RC_thr is set together with the host vehicle risk map RM_e. Figure 11 is a flowchart showing an example of the host vehicle risk map setting process.

[0093] The host vehicle risk calculation unit 69 determines whether or not there is an occupant other than the driver (step S71). Specifically, the host vehicle risk calculation unit 69 determines whether or not there is occupant identification information of an occupant whose seating position is other than the driver's seat, based on the data of the occupant identification information stored in the storage unit 53.

[0094] If it is determined that there are no occupants other than the driver (S71 / No), the host vehicle risk calculation unit 69 generates a host vehicle risk map that takes into account only the driver (step S79). In this case, the host vehicle risk calculation unit 69 uses the gradient coefficient σ x ,σ y Among the learning data, the gradient coefficient σ x ,σ y Extract the gradient coefficient σ x ,σ y The personal risk potential RP_e defined by the above is set in the host vehicle risk map RM_e.

[0095] Next, the own vehicle risk calculation unit 69 sets the overlap area risk threshold RC_thr considering only the driver (step S81). In this case, the own vehicle risk calculation unit 69 extracts the personal overlap area risk threshold RC_thr_e corresponding to the occupant identification number of the driver from the information of the personal overlap area risk threshold RC_thr_e stored in the driving characteristics database, and sets the personal overlap area risk threshold RC_thr_e to the overlap area risk threshold RC_thr.

[0096] On the other hand, if it is determined in step S71 that there is an occupant other than the driver (S71 / Yes), the host vehicle risk calculation unit 69 determines whether or not the learning data of the personal risk potential parameters and the personal overlap area risk threshold RC_thr_e of the occupant other than the driver are stored in the driving characteristics database 55 (step S73). If it is determined that the above learning data of the occupant other than the driver is not stored in the driving characteristics database 55 (S73 / No), the host vehicle risk calculation unit 69 proceeds to step S79, and performs a process of setting the host vehicle risk map RM_e that takes into account only the driver and a process of setting the overlap area risk threshold RC_thr.

[0097] On the other hand, if it is determined that the above learning data of occupants other than the driver is stored in the driving characteristics database 55 (S73 / Yes), the vehicle risk calculation unit 69 generates a vehicle risk map RM_e taking into account all occupants for which individual risk potential parameters exist (step S75).

[0098] FIG. 12 is an explanatory diagram showing an example of the host vehicle risk map RM_e generated in consideration of multiple occupants. FIG. 12 shows the host vehicle risk map RM_e generated by superimposing the personal risk potential RP_e_A of driver A and the personal risk potential RP_e_B of driver B shown in FIG. 9. In the example shown in FIG. 12, the host vehicle risk calculation unit 69 superimposes the personal risk potential RP_e_A of driver A and the personal risk potential RP_e_B of driver B, and selects the maximum value of either for each coordinate point. Therefore, the gradient of the host vehicle risk map RM_e on the driver's seat side and the front side of the host vehicle 1 is determined by the gradient of the personal risk potential RP_e_A of driver A, and the gradient of the host vehicle risk map RM_e on the passenger's seat side and the rear side of the host vehicle 1 is determined by the gradient of the personal risk potential RP_e_B of driver B. In this way, the host vehicle risk map RM_e is set which covers the collision risks felt by multiple occupants according to the distance from the outer circumferential end of the host vehicle 1.

[0099] Note that Figure 12 shows an example of the host vehicle risk map RM_e reflecting the personal risk potentials RP_e of two occupants, but even when there are three or more occupants, the host vehicle risk calculation unit 69 similarly generates the host vehicle risk map RM_e by superimposing multiple personal risk potentials RP_e.

[0100] After generating the own vehicle risk map RM_e in step S75, the own vehicle risk calculation unit 69 sets the overlap area risk threshold RC_thr in consideration of all occupants who have personal overlap area risk thresholds RC_thr_e (step S77). In this embodiment, the own vehicle risk calculation unit 69 sets the personal overlap area risk threshold RC_thr_e with the smallest value among the personal overlap area risk thresholds RC_thr_e of the multiple occupants to the overlap area risk threshold RC_thr. However, the overlap area risk threshold RC_thr may be set by other methods, such as setting the average value of the personal overlap area risk thresholds RC_thr_e of the multiple occupants to the overlap area risk threshold RC_thr.

[0101] In this way, when the occupant of the host vehicle 1 is the driver only, the host vehicle risk calculation unit 69 sets the driver's personal risk potential RP_e and personal overlap area risk threshold RC_thr_e to the host vehicle risk map RM_e and personal overlap area risk threshold RC_thr_e, respectively. In addition, when the occupants of the host vehicle 1 include occupants other than the driver, the host vehicle risk calculation unit 69 sets the host vehicle risk map RM_e and overlap area risk threshold RC_thr based on the learned data of the occupants' personal risk potential RP_e and personal overlap area risk threshold RC_thr_e present in the driving characteristics database 55. In this way, it is possible to obtain the host vehicle risk map RM_e and overlap area risk threshold RC_thr that reflect the collision risk felt by the occupants of the host vehicle 1 according to the distance from the outer circumferential end of the host vehicle 1.

[0102] Fig. 13 is a flowchart showing another example of the host vehicle risk map setting process. In the example shown in Fig. 13, even if the driving characteristics database 55 contains learning data of the personal risk potential parameters and the personal overlap area risk threshold RC_thr_e, the host vehicle risk map RM_e and the overlap area risk threshold RC_thr are set without taking into consideration the occupant who is not paying attention to the outside of the vehicle.

[0103] Specifically, similarly to the flowchart shown in FIG. 11, when it is determined in step S73 that the learning data of the occupants other than the driver is stored in the driving characteristic database 55 (S73 / Yes), the host vehicle risk calculation unit 69 determines whether or not there is an occupant who is not paying attention to the outside of the vehicle (step S74). Specifically, the host vehicle risk calculation unit 69 acquires information on the line of sight or the direction of the face of the occupant detected by the occupant detection unit 63, and determines whether or not each occupant is paying attention to the outside of the vehicle based on the line of sight or the direction of the face. For example, the host vehicle risk calculation unit 69 may determine that the occupant is not paying attention to the outside of the vehicle when the line of sight or the direction of the face is not directed forward of the host vehicle 1 for a preset time or more. By such a determination process, it is possible to distinguish an occupant who is not paying attention to obstacles, such as an occupant who is sleeping, an occupant who is looking at a mobile terminal or an in-vehicle display device, or an occupant who is looking away for a long time.

[0104] If it is determined that there are no occupants who are not paying attention outside the vehicle (S74 / No), the vehicle risk calculation unit 69 generates a vehicle risk map RM_e taking into account all occupants for whom there is learning data for individual risk potential parameters and individual overlap area risk threshold RC_thr_e, and sets the overlap area risk threshold RC_thr, in a procedure similar to steps S75 and S77 of the flowchart shown in Figure 11.

[0105] On the other hand, if it is determined that there is an occupant who is not paying attention to the outside of the vehicle (S74 / Yes), the host vehicle risk calculation unit 69 generates a host vehicle risk map RM_e by taking into consideration all occupants for whom learning data of personal risk potential parameters exists in the driving characteristics database 55 and who are paying attention to the outside of the vehicle (step S83). Next, the host vehicle risk calculation unit 69 sets an overlap area risk threshold RC_thr by taking into consideration all occupants for whom learning data of personal overlap area risk threshold RC_thr_e exists in the driving characteristics database 55 and who are paying attention to the outside of the vehicle (step S85). The process of generating the host vehicle risk map RM_e and the process of setting the overlap area risk threshold RC_thr are performed in the same manner as steps S75 and S77, or steps S79 and S81.

[0106] In this way, when there is an occupant who is not paying attention to the outside of the vehicle, the host vehicle risk calculation unit 69 sets the host vehicle risk map RM_e and the overlap area risk threshold RC_thr without taking the occupant into consideration. This makes it possible to obtain the host vehicle risk map RM_e and the overlap area risk threshold RC_thr that reflect the collision risk felt by an occupant who is paying attention to the outside of the vehicle and who may sense the risk of collision with an obstacle according to the distance from the outer periphery of the host vehicle 1.

[0107] 11 and 13, when there is no learning data of the individual risk potential parameter and the individual overlap area risk threshold RC_thr_e in the driving characteristics database 55 for an occupant detected by the occupant detection unit 63, the information of the occupant is not taken into consideration, and instead, a substitute value of the individual risk potential parameter and the individual overlap area risk threshold RC_thr_e set in advance may be applied to the occupant. The substitute value may be, for example, the average value of each of the individual risk potential parameter and the individual overlap area risk threshold RC_thr_e stored in the driving characteristics database 55. Alternatively, the substitute value may be the value of the individual risk potential parameter and the individual overlap area risk threshold RC_thr_e calculated in advance assuming an exemplary skilled driver or a virtual reference driver.

[0108] Furthermore, the host vehicle risk calculation unit 69 may generate the host vehicle risk map RM after correcting the personal risk potential RP_e of each occupant based on the detected seating position of each occupant. Specifically, the personal risk potential parameters are data learned when each occupant is seated in the driver's seat. The range of the blind spot area as seen by the occupant may vary depending on the seating position of the occupant. For this reason, the change in the range of the blind spot area due to differences in the seating position of the occupant may be reflected in the personal risk potential RP_e, and the left / right or front / rear distribution of the personal risk potential RP_e may be changed.

[0109] FIG. 14 is an explanatory diagram showing a method of correcting the personal risk potential RP_e of an occupant sitting in the passenger seat. For example, assume that the personal risk potential RP_e of the occupant (driver) sitting in the driver's seat, which is learned when the vehicle 1 is being manually driven, is distributed evenly between the left and right (see the left diagram in FIG. 14). In this case, the widths from the outer periphery of the personal risk potential RP_e set on both the left and right sides of the vehicle 1 will be equal (W RR =W RL On the other hand, the width L of the blind spot on the right side (driver's seat side) of the vehicle 1 as seen from the driver's seat RR is the width L of the blind spot area on the left side (passenger seat side) of the vehicle 1. RL will be smaller than

[0110] Taking into account this change in the range of the blind spot, when the occupant is seated in the passenger seat, the widths of the personal risk potential RP_e on both the left and right sides are corrected based on the ratio of the widths of the blind spots (see the right diagram in FIG. 14). Specifically, when the occupant is seated in the passenger seat, the width W RR The ratio of the width of the blind spot to the width of the blind spot (L LR / L RR ), the width W RR In addition, the width W of the personal risk potential RP_e on the left side (passenger seat side) of the vehicle 1 is increased. RL The ratio of the width of the blind spot to the width of the blind spot (LLL / L RL ), the width W RL In this case, the width of the blind spot L RR ,L RL ,L LR ,L LL The value of may be a preset value, or may be set according to information on the height of the occupant input in advance or the position of the occupant's face or eyes detected by the occupant detection unit 63.

[0111] 15 is a diagram showing a method of correcting the personal risk potential RP_e of a passenger seated in the rear seat. FF is the width L of the blind spot area in front of the vehicle 1 as seen from the rear seat. RF Taking this change in the range of the blind spot into consideration, when the occupant is seated in the rear seat, the width of the forward personal risk potential RP_e is corrected according to the ratio of the width of the blind spot (see the lower diagram in Figure 15). Specifically, when the occupant is seated in the rear seat, the forward width W FF The ratio of the width of the blind spot to the width of the blind spot (L RF / L FF ) to obtain the width W FF In this case, the width of the blind spot L RF ,L FF The value of may be a preset value, or may be set according to information on the height of the occupant input in advance or the position of the occupant's face or eyes detected by the occupant detection unit 63.

[0112] In this way, by reflecting changes in the range of the blind spot area due to differences in the seating position of the occupant in the individual risk potential RP_e and changing the left / right or front / rear distribution of the individual risk potential RP_e, it is possible to expand or contract the width of the individual risk potential RP_e in accordance with the expansion or contraction of the range of the blind spot area of ​​the individual risk potential RP_e depending on the seating position. In this way, the host vehicle risk calculation unit 69 can generate the host vehicle risk map RM_e using the individual risk potential RP_e that takes into account changes due to differences in the seating position in the collision risk felt by each occupant depending on the distance from the outer periphery of the host vehicle 1.

[0113] 10, after the host vehicle risk map RM_e is set in step S47, the surrounding environment detection unit 61 acquires the surrounding environment information of the host vehicle 1 in the same procedure as step S17 in the flowchart shown in Fig. 5 (step S49). The surrounding environment detection unit 61 stores the acquired surrounding environment information in the memory unit 53.

[0114] Next, the driving condition setting unit 71 determines whether or not there is an obstacle to be avoided on the traveling path of the vehicle 1 (step S51) in the same procedure as step S19 of the flowchart shown in Fig. 5. If it is determined that there is no obstacle to be avoided on the traveling path of the vehicle 1 (S51 / No), the process returns to step S49, and the process of detecting the surrounding environment information and the process of determining the presence or absence of an obstacle are repeated.

[0115] On the other hand, if it is determined that an obstacle to be avoided is present on the driving trajectory of the vehicle 1 (S51 / Yes), the driving condition setting unit 71 updates the obstacle risk map in a procedure similar to step S21 of the flowchart shown in Figure 5 (step S53).

[0116] Next, the driving condition setting unit 71 determines a target trajectory (x mr (t k ),y mr (t k)) (step S55). Specifically, the driving condition setting unit 71 calculates the overlap area risk RC when the host vehicle 1 is caused to travel along a planned travel trajectory that is set on an obstacle risk map that has been superimposed with the obstacle risk potentials RPi set for each obstacle, and calculates the planned travel trajectory that minimizes the overlap area risk RC (see FIG. 8). More specifically, the driving condition setting unit 71 calculates the overlap area risk RC in the overlap area between the host vehicle risk map RM_e that reflects the driving characteristics of the occupant of the host vehicle 1 and the obstacle risk map for a plurality of settable planned travel trajectories, using the above formula (6), and calculates the planned travel trajectory (x mr (t k ),y mr (t k The driving condition setting unit 71 determines a planned driving trajectory (x mr (t k ),y mr (t k )) to the target trajectory (x mr (t k ),y mr (t k )).

[0117] FIG. 16 is an explanatory diagram showing changes in overlap area risk RC due to differences in driving trajectories. The example shown in FIG. 16 shows a driving scene in which the host vehicle 1 passes beside a parked vehicle 95 parked on the left side of the road. A curb 97 is present on the right edge of the road. Driving trajectory T2 is set so as to pass on the curb 97 side compared to driving trajectory T1. Note that the obstacle risk potential RP 3 and the obstacle risk potential RP of curb 97 4 The maximum value for both is set to "1".

[0118] In this driving scene, the vehicle risk map RM_e and the obstacle risk potential RP 3 The overlapping area risk RC1 of the overlapping area between the vehicle risk map RM_e and the obstacle risk potential RP 4The overlap area risk RC is the sum of the overlap area risk RC1 and the overlap area risk RC2 of the overlap area with the curb 97. When the driving trajectory T2 that is relatively closer to the curb 97 is set, the overlap area risk RC1 value is smaller and the overlap area risk RC2 value is larger than when the driving trajectory T1 is set. By using the above formula (6), the driving trajectory that minimizes the overlap area risk RC as a sum is calculated.

[0119] Furthermore, the driving condition setting unit 71 sets a target vehicle speed at which the overlap area risk RC is equal to or less than the overlap area risk threshold RC_thr (step S57). Specifically, when the minimum overlap area risk RC calculated in step S55 exceeds the overlap area risk threshold RC_thr, the driving condition setting unit 71 calculates the target vehicle speed after deceleration using the above formulas (6) and (7) based on the difference value obtained by subtracting the overlap area risk threshold RC_thr from the overlap area risk RC. More specifically, the driving condition setting unit 71 calculates the target vehicle speed after deceleration v, which can be set, by subtracting the overlap area risk threshold RC_thr from the vehicle speed v in the above formula (7). mr The risk value R entered as e (x, y, v) is the risk value R in the above formula (6). e (X i (t k ),Y j (t k )) to calculate a vehicle speed v at which the overlap area risk RC becomes the overlap area risk threshold RC_thr. The driving condition setting unit 71 sets the calculated vehicle speed v to a target vehicle speed v mr Set to.

[0120] 17 is an explanatory diagram showing how the range of the host vehicle risk map RM changes and the overlap area risk RC decreases in response to a decrease in vehicle speed. e When setting the area where the vehicle 1 exists (R e = 1) surrounding risk value R e, decreases as the vehicle speed decreases. Therefore, even if the same travel trajectory T1 as the travel trajectory T1 shown in FIG. 16 is set, the overlap area risk RC decreases. In this way, the driving condition setting unit 71 reduces the vehicle speed to set the target vehicle speed v at which the overlap area risk RC is equal to or less than the overlap area risk threshold RC_thr. mr can be set.

[0121] Next, the driving condition setting unit 71 of the control unit 51 sets driving conditions for the automatic driving control of the vehicle 1 based on the information of the set target trajectory and target vehicle speed (step S59). Specifically, the driving condition setting unit 71 sets a target steering angle based on the target trajectory, and sets a target acceleration / deceleration based on the target vehicle speed. The driving condition setting unit 71 transmits information on the set target steering angle and target acceleration / deceleration to the vehicle control device 41. The vehicle control device 41, which has received the information on the target steering angle and target acceleration / deceleration, sets control target amounts for each device, such as the electric steering device, the brake device, and the driving force source, based on the information on the target steering angle and the target acceleration / deceleration, and controls the traveling of the host vehicle 1.

[0122] Next, the driving condition setting unit 71 judges whether or not the host vehicle 1 has completed the avoidance of the obstacle that was on the travel path (step S61) in accordance with the same procedure as step S25 of the flowchart shown in Fig. 5. If it is judged that the host vehicle 1 has not completed the avoidance of the obstacle that was on the travel path (S61 / No), the driving condition setting unit 71 returns to step S53 and repeats the process of updating the obstacle risk map in accordance with the travel of the host vehicle 1, the setting of the target path and target vehicle speed of the host vehicle 1, and the setting and transmission process of the driving conditions.

[0123] On the other hand, if it is determined that the vehicle 1 has completed avoiding the obstacle that was on the travel path (S61 / Yes), the driving condition setting unit 71 determines whether the in-vehicle system has stopped (step S63). If the in-vehicle system has stopped (S63 / Yes), the driving condition setting process by the control unit 51 ends. On the other hand, if the in-vehicle system has not stopped (S63 / No), the process returns to step S43 and repeats the process of each step described above.

[0124] In this way, during automatic driving of the host vehicle 1, the control unit 51 generates a host vehicle risk map that reflects the individual risk potential parameters of each occupant learned during manual driving. In addition, the control unit 51 sets, as the target trajectory, a planned driving trajectory that minimizes the overlap area risk, which is a risk value of an overlap area between the generated host vehicle risk map and an obstacle risk map that reflects the obstacle risk potential set for obstacles around the host vehicle 1. Therefore, it is possible to reduce the anxiety and discomfort felt by not only the driver of the host vehicle 1 but also the occupants of the host vehicle 1 regarding obstacles.

[0125] <4. Summary> As described above, the driving support device 50 according to this embodiment learns the personal risk potential RP_e of each driver during manual driving of the host vehicle 1, and stores it in the driving characteristics database 55. Furthermore, during automatic driving of the host vehicle 1, the driving support device 50 sets the host vehicle risk map RM_e reflecting the learned personal risk potential RP_e of the occupant of the host vehicle 1 to the host vehicle 1, and sets the driving conditions of the host vehicle 1 based on information on the obstacle risk map reflecting the obstacle risk potential RPi set for each obstacle around the host vehicle 1 and information on the host vehicle risk map RM_e. For this reason, the target trajectory and target vehicle speed of the host vehicle 1 are set so as to reduce not only the risk of collision from the obstacle, but also the risk of collision felt by the occupant depending on the distance from the outer circumferential end of the host vehicle 1, thereby reducing the anxiety and discomfort felt by the occupant.

[0126] Furthermore, when the host vehicle 1 is being driven autonomously and there are multiple occupants in the host vehicle 1, the driving assistance device 50 according to this embodiment generates a host vehicle risk map RM_e that reflects information on the personal risk potentials RP_e of the multiple occupants. This makes it possible to set driving conditions that can reduce the anxiety and discomfort felt by occupants other than the driver.

[0127] Furthermore, when there are multiple occupants of the host vehicle 1 during autonomous driving of the host vehicle 1, the driving assistance device 50 according to this embodiment sets the host vehicle risk map RM_e as the risk value Re at each coordinate position in a two-dimensional coordinate system including a plane consisting of the front-rear and left-right directions of the host vehicle 1, using the maximum value of the personal risk potentials RP_e of the multiple occupants at each coordinate position as the risk value Re at each coordinate position. Therefore, the host vehicle risk map RM_e can be generated for each direction as seen from the host vehicle 1, in accordance with the occupant who feels the most risk.

[0128] In addition, the driving support device 50 according to the present embodiment calculates an overlapping area risk RC, which is a risk value of an overlapping area between the obstacle risk map and the vehicle risk map RM_e when the vehicle 1 is driven along the planned driving path, and sets the planned driving path that minimizes the overlapping area risk RC as the target path. Therefore, it is possible to set driving conditions that can reduce the sum of the risks felt by each occupant of the vehicle 1.

[0129] In the driving support device 50 according to this embodiment, the personal risk potential RP_e is set so that the risk value Re around the area in which the host vehicle 1 exists becomes smaller as the vehicle speed of the host vehicle 1 becomes relatively slower, and the driving characteristics database 55 further stores information on the personal overlap area risk threshold RC_thr_e when each driver decelerates the host vehicle 1 when the host vehicle 1 passes beside an obstacle during manual driving of the host vehicle 1. In addition, the driving support device 50 sets the target vehicle speed to a vehicle speed at which the overlap area risk RC is equal to or smaller than the smallest personal overlap area risk threshold RC_thr_e among the personal overlap area risk thresholds RC_thr_e of the multiple occupants aboard the host vehicle 1. For this reason, anxiety and discomfort can be reduced based on the occupant who is most susceptible to risk among the occupants of the host vehicle 1.

[0130] Furthermore, the driving assistance device 50 according to this embodiment corrects the personal risk potential RP_e of each occupant based on the difference in the blind spot area between the seating position of each occupant and the driver's seat. Therefore, the host vehicle risk map RM_e can be generated using the personal risk potential RP_e that takes into account the change in collision risk felt by each occupant depending on the distance from the outer periphery of the host vehicle 1 due to the difference in the seating position of each occupant. Therefore, the anxiety and discomfort felt by each occupant can be reduced depending on the seating position of each occupant.

[0131] Furthermore, the driving assistance device 50 according to this embodiment sets the host vehicle risk map RM_e and the overlap area risk threshold RC_thr excluding information on the personal risk potential RP_e of occupants who are not paying attention to the outside of the vehicle, among multiple occupants aboard the host vehicle 1. This makes it possible to reduce anxiety and discomfort of occupants who are paying attention to the outside of the vehicle and who may sense the risk of a collision with an obstacle.

[0132] Furthermore, the driving assistance device 50 according to this embodiment includes an individual risk learning unit 67 that learns the individual risk potential RP_e based on data on the distance between the host vehicle 1 and obstacles around the host vehicle 1, obtained during manual driving. This makes it possible to learn the driving characteristics of the driver who drove the host vehicle 1, without having to previously store a database that accumulates the driving characteristics of drivers.

[0133] Although the preferred embodiment of the present disclosure has been described in detail above with reference to the accompanying drawings, the present disclosure is not limited to such examples. It is clear that a person having ordinary knowledge in the technical field to which the present disclosure belongs can conceive of various modified or amended 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 disclosure.

[0134] For example, the technical scope of the present disclosure also includes a computer program that causes a processor constituting the control unit 51 to execute operations including reading out information on a personal risk potential that is set for the host vehicle based on data on the distance between the host vehicle 1 and obstacles around the host vehicle 1 obtained during manual driving, which is learned as the risk that the driver of the host vehicle perceives from obstacles; setting a host vehicle risk map for the host vehicle that reflects the personal risk potentials of the occupants of the host vehicle when the host vehicle is being driven autonomously; and setting driving conditions for the host vehicle when it is being driven autonomously based on information on the obstacle risk map that reflects the obstacle risk potentials set for each of the obstacles around the host vehicle and information on the host vehicle risk map, and a recording medium having such a computer program recorded thereon.

[0135] In the above embodiment, all of the functions of the driving support device 50 are installed in the vehicle 1, but the present disclosure is not limited to such an example. For example, some of the functions of the driving support device 50 may be provided in a server device that can communicate via a mobile communication means, and the driving support device 50 may be configured to transmit and receive data to and from the server device.

[0136] In addition, the following exemplary embodiments also fall within the technical scope of the present disclosure. In the above embodiment, when there are multiple occupants in the host vehicle during autonomous driving of the host vehicle, the host vehicle risk calculation unit sets a host vehicle risk map by taking the maximum value of the personal risk potentials of the multiple occupants at each coordinate position in a two-dimensional coordinate system including a plane consisting of the forward / backward and left / right directions of the host vehicle as the risk value at each coordinate position. In the above embodiment, the personal risk potential is set such that the risk value around the area in which the host vehicle is located becomes smaller as the vehicle speed of the host vehicle becomes relatively slower, the memory unit further stores information on the personal overlap area risk threshold obtained by calculating the overlap area risk when the driver decelerates the host vehicle as the host vehicle passes beside an obstacle during manual driving, and the driving condition setting unit sets the target vehicle speed to a vehicle speed at which the overlap area risk is equal to or less than the smallest personal overlap area risk threshold among the personal overlap area risk thresholds of a plurality of occupants riding in the host vehicle. In the above embodiment, the driving assistance device includes an individual risk learning unit that learns an individual risk potential that reflects the risk that the driver perceives from an obstacle, based on the distance between the host vehicle and an obstacle when the driver of the host vehicle passes beside the obstacle during manual driving. A computer program that causes a processor to execute operations including: reading out information of a personal risk potential that is set for the host vehicle by learning the risk that the driver of the host vehicle perceives from obstacles during manual driving; setting for the host vehicle a host vehicle risk map that reflects the personal risk potentials of the occupants of the host vehicle during automatic driving of the host vehicle; and setting driving conditions for the host vehicle during automatic driving based on information of an obstacle risk map that reflects the obstacle risk potentials set for each of the obstacles around the host vehicle and information of the host vehicle risk map. [Explanation of symbols]

[0137] 1...vehicle, 31...surrounding environment sensor, 33...in-vehicle camera, 35...vehicle state sensor, 50...driving assistance device, 51...control unit, 55...driving characteristic database, 57...occupant identification database, 61...surrounding environment detection unit, 63...occupant detection unit, 65...driving state detection unit, 67...personal risk learning unit, 69...own vehicle risk calculation unit, 71 driving condition setting unit

Claims

1. A driving assistance device that reflects driving characteristics of a driver learned during manual driving of a vehicle in automatic driving control of the vehicle, a storage unit that learns a risk felt by a driver of the host vehicle from an obstacle during manual driving and stores information on a personal risk potential that is set for the host vehicle; a host vehicle risk calculation unit that sets a host vehicle risk map for the host vehicle, the host vehicle risk map reflecting the personal risk potential of an occupant of the host vehicle during autonomous driving of the host vehicle; a driving condition setting unit that sets driving conditions during autonomous driving of the host vehicle based on information on an obstacle risk map that reflects obstacle risk potentials set for each of the obstacles around the host vehicle and information on the host vehicle risk map; A driving assistance device comprising:

2. The driving assistance device according to claim 1 , wherein when there are multiple occupants of the host vehicle during autonomous driving of the host vehicle, the host vehicle risk calculation unit sets the host vehicle risk map based on information of the personal risk potentials of the multiple occupants.

3. The driving assistance device according to claim 1 or 2, wherein the driving condition setting unit calculates an overlap area risk, which is a risk value of an overlap area between the obstacle risk map and the vehicle risk map when the vehicle is driven along a planned driving trajectory, and sets the planned driving trajectory that minimizes the overlap area risk as a target trajectory.

4. 4. The driving assistance device according to claim 1, wherein the host vehicle risk calculation unit corrects the personal risk potential of each of the occupants based on a difference in a blind spot area between a seating position of each of the occupants and a driver's seat.

5. The vehicle risk calculation unit sets the vehicle risk map excluding information on the personal risk potential of occupants who are not paying attention to the outside of the vehicle, among multiple occupants riding in the vehicle.

6. The processor: learning a risk felt by a driver of the vehicle in relation to an obstacle during manual driving and reading out information of a personal risk potential set for the vehicle; setting a host vehicle risk map for the host vehicle, the host vehicle risk map reflecting the personal risk potential of an occupant of the host vehicle during automatic driving of the host vehicle; Setting driving conditions during autonomous driving of the host vehicle based on information on an obstacle risk map reflecting obstacle risk potentials set for each of the obstacles around the host vehicle and information on the host vehicle risk map; A recording medium storing a computer program for causing the computer to execute operations including the steps of:

Images