Driving support device

Abstract

To provide a driving support device capable of considering a driver's brain model (driver internal model) to perform appropriate driving support optimized for each driver.SOLUTION: A driving support device includes: traveling scene recognition means 31; driving target calculation means 32 for calculating a target driving operation and target vehicle behavior; driving operation detection means 33 for detecting a driver's driving operation; driving detection means 34 for detecting behaviors of a vehicle; driver determination means 36 for determining the level of the driver's skill in a traveling scene subject to determination; and support mode setting means for setting the mode of notification by HMI 26 depending on the determination of the level of the driver's skill by the driver determination means 36.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a driving assistance device that assists the driver of a vehicle such as an automobile, and in particular to an improvement that enables the provision of assistance optimized for each individual driver by performing driving assistance that takes into account the driver's internal brain model (driver's internal model). [Background technology]

[0002] In recent years, vehicles such as automobiles have begun to be equipped with driving assistance devices (e.g., advanced driver assistance systems (ADAS)) that assist drivers in driving, and various technologies have been proposed for such driving assistance devices.

[0003] For example, Patent Document 1 (JP 2022-178816 A) proposes a technology that can improve a driver's driving skills by observing the driver's driving (driving performance) in a vehicle control device, dividing it into cognition, judgment, and operation, and providing driving assistance (coaching) at the optimal timing for driving performance that is determined to be insufficient. Patent Document 2 (JP Patent No. 6428748) also proposes a technology in a driving assistance system that guides the vehicle driver toward a target emotional state (e.g., a comfortable state) by providing appropriate sensory stimuli (such as providing visual and auditory stimuli, changing accelerator sensitivity or steering wheel sensitivity, etc.). Patent Document 3 (JP Patent No. 6221776) also proposes a technology in which a driving evaluation device evaluates driving operations and notifies the driver of the evaluation results, allowing the driver to recognize whether the driving operations they performed were appropriate. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2022-178816 [Patent Document 2] Patent No. 6428748 [Patent Document 3] Patent No. 6221776 Summary of the Invention [Problem to be solved by the invention]

[0005] As described above, technologies for improving the driving skills and improving the emotions of drivers have been proposed for conventional driving assistance devices. However, the conventional technology did not include the idea of ​​optimizing driving assistance for each driver by considering (modeling) the processing in the brain of each driver. In other words, even when conventional driving assistance devices notify the driver (such as by issuing an alarm) or intervene in driving operations, they are uniform and do not take into account the brain state of each individual driver (internal driver model), and therefore cannot be said to be optimized for improving the driving skills and improving the emotions of the driver who is the target of driving assistance.

[0006] The present invention has been made with a focus on these problems, and aims to provide a driving assistance device that can provide appropriate driving assistance that is optimized for each individual driver by taking into account the driver's internal brain model (driver's internal model). [Means for solving the problem]

[0007] In order to achieve the above object, the present invention adopts the following solution: That is, as described in claim 1, a driving assistance device that assists a driver of a vehicle in driving includes driving scene recognition means that recognizes driving scenes of the vehicle, driving detection means that detects a driving operation of the driver and / or a behavior of the vehicle, driving target calculation means that calculates a target driving operation and / or a target vehicle behavior that matches the driving scene, interface means that notifies the driver of driving information, driving history storage means that stores the driving operation of the driver and / or the behavior of the vehicle for each driving scene, driver determination means that determines the driver's proficiency in the driving scene to be determined based on a comparison between the driver's past driving operation and / or the vehicle behavior in the driving scene to be determined and the target driving operation and / or the target vehicle behavior, and assistance mode setting means that sets a mode of notification to the driver by the interface means based on the determination of the driver's proficiency by the driver determination means.

[0008] According to the above solution, the manner in which the interface means (e.g., HMI 26) notifies the driver is set according to the driver's level of familiarity with the target driving scene, so that notifications are made that are precisely suited to the driver's level of familiarity, and the driver's prediction error (motor system prediction error 2A) can be effectively reduced, thereby improving the driver's driving skills and improving their emotional well-being.

[0009] A preferred embodiment based on the above-described solution is as set forth in claim 2 and subsequent claims. That is, the assistance mode setting means sets the notification mode by the interface means for a driver who has been determined to have a high level of proficiency with respect to a driving scene during driving to a mode in which notification regarding vehicle behavior is performed without notification regarding the driving operation itself (corresponding to claim 2). In this case, for a highly proficient driver (experienced driver), unnecessary notification regarding the driving operation itself, which an experienced driver could perform unconsciously, is not performed, while information regarding vehicle behavior that is useful even for an experienced driver is not provided. Therefore, it is possible to promote appropriate independent driving by experienced drivers and effectively achieve improvement in the driver's mood.

[0010] The notification regarding the vehicle behavior includes a notification of the difference between the vehicle behavior and the target vehicle behavior when the difference between the vehicle behavior and the target vehicle behavior becomes equal to or greater than a predetermined value (corresponding to claim 3). In this case, an experienced driver can accurately grasp problems regarding the vehicle behavior itself, and therefore, the motion system prediction error can be smoothly reduced.

[0011] The assistance mode setting means sets the mode of notification by the interface means to a mode of providing notification regarding the driving operation itself for a driver who is determined to have low proficiency with respect to a driving scene during driving (corresponding to claim 4). In this case, a driver with low proficiency (a novice driver) can receive notification regarding the driving operation itself (provision of necessary information and coaching), so that prediction errors in the motor system can be effectively reduced for driving operations that the driver is unable to perform accurately, thereby promoting improvement in the driver's driving technique and improving the driver's emotional state accordingly.

[0012] The driver determination means determines that the driver has a high level of proficiency for a driving scene to be determined if the driver's driving operation for a predetermined period of time is within a predetermined driving operation target range from the target driving operation and the vehicle behavior for the predetermined period of time is within a predetermined vehicle behavior target range from the target vehicle behavior (corresponding to claim 5). In this case, the driver's proficiency is determined using both the driver's driving operation and the vehicle behavior, so that an accurate determination can be made.

[0013] The vehicle is provided with a driver monitoring means capable of detecting physiological values ​​and / or behavior of the driver, and the assistance mode setting means reduces the level of notification by the interface means when the physiological values ​​and / or behavior indicating tension of the driver detected by the driver monitoring means in the same driving scene are decreased in the current driving scene compared to the previous driving scene (corresponding to claim 6). In this case, a decrease in the physiological values ​​and / or behavior indicating tension of the driver suggests an improvement in the driver's driving technique or improved mood in that driving scene, so the level of unnecessary notification (for example, the amount of information) can be reduced to encourage the driver to drive more autonomously.

[0014] The vehicle is provided with a driving operation intervention means for intervening in the driving operation of the driver, and the assistance mode setting means changes the mode of driving assistance provided by the driving operation intervention means based on the determination by the driver determination means (corresponding to claim 7). In this case, in addition to notification by the interface means, the driving operation intervention means (e.g., driving operation intervention means 27) also intervenes in the driving operation, so that more effective driving assistance can be provided according to the driver's level of proficiency. [Effects of the Invention]

[0015] According to the present invention, the manner of notification from the interface means is appropriately changed according to the driver's proficiency, so that highly skilled drivers can drive more autonomously and comfortably, while less skilled drivers can reduce prediction errors in difficult driving situations at an early stage. This effectively improves the driver's driving skills and improves their emotional well-being. [Brief explanation of the drawings]

[0016] [Figure 1] FIG. 2 is a diagram showing a driver internal model in the present invention. [Figure 2] FIG. 2 is a block diagram showing an example of a control system of the driving assistance device of the present invention. [Figure 3] FIG. 10 is a diagram showing an example of a display for an experienced driver in an HMI. [Figure 4] FIG. 10 is a diagram showing an example of a display for a novice driver in an HMI. [Figure 5] 10 is a flowchart illustrating an example of control according to the present invention. [Figure 6] 4 is a flowchart showing an example of control for driver determination according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0017] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. FIG. 1 shows an internal driver model 1 that models the internal brain state of a driver while driving a vehicle. This internal driver model 1 was constructed by the inventor of the present invention by applying the idea of ​​the free energy principle related to human brain function to the internal brain processing of a driver, particularly while driving a vehicle. The present invention configures a driving assistance device based on this internal driver model 1. First, the contents of the internal driver model 1 will be described in detail.

[0018] The free energy principle is a theory proposed by Friston that provides a unified explanation of the various functions of the human brain as a system that attempts to minimize free energy. According to the free energy principle, the internal model of the human brain is updated (modified) so as to minimize prediction error.

[0019] As shown in Fig. 1, the driver internal model 1 includes a motor system internal model 1A and a body system internal model 1B. Here, the motor system internal model 1A models the function in the brain of the driver while driving, which makes inference about the motor system (i.e., predicts the behavior of the vehicle in response to driving operations) and controls the body movement (driving operations) based on this inference. On the other hand, the body system internal model 1B models the function in the brain of the driver while driving, which makes inference about the body system (i.e., predicts the body states of the internal organs, blood vessels, etc. while driving a vehicle), and controls the body states (states of the internal organs, blood vessels, etc.) based on this inference.

[0020] In the brain of a driver while driving a vehicle, two processing loops are carried out: a motor system processing loop 3A (in the figure, a processing loop including motor system internal model 1A → result prediction 11 → driving operation 12 → vehicle behavior 13 → perception and cognition 14 → motor system prediction error 2A) for reducing motor system prediction error 2A centered on motor system internal model 1A, and an internal system processing loop 3B (in the figure, a processing loop including internal system model 1B → internal state prediction 15 → internal state 16 → internal system prediction error 2B) for reducing internal system prediction error 2B centered on internal system internal model 1B.

[0021] Here, the motor system prediction error 2A is recognized by the motor system internal model 1A as the difference between the vehicle behavior predicted by the driver (predicted result 11) and the actual vehicle behavior 13 (realized result). On the other hand, the internal body system prediction error 2B is recognized by the internal body system model 1B as the difference between the prediction 15 of the driver's internal body state and the actual internal body state 18. In the driver internal model 1, the motor system internal model 1A and the internal body system model 1B are updated in the process of reducing the motor system prediction error 2A and the internal body system prediction error 2B in the motor system processing loop 3A and the internal body system processing loop 3B.

[0022] To explain in more detail below, while driving a vehicle, information about the vehicle's driving situation (for example, the driving scene in which the vehicle is driving and various accompanying information) is perceived by the driver (perceptual perception 10). Based on this perceptual perception 10, the motor system internal model 1A infers the vehicle's behavior that should be executed in that situation (driving scene), and this is used as a result prediction 11 (predicted vehicle behavior).

[0023] Furthermore, the motor system internal model 1A infers a driving operation 12 required to obtain a result prediction 11 (predicted vehicle behavior). That is, the motor system internal model 1A has a hypothesis (belief) for inferring (predicting) what driving operation should be performed to obtain the predicted vehicle behavior, and based on this hypothesis, it infers the driving operation 12 to be performed. The driver will perform the driving operation 12 in accordance with this inference (judgment).

[0024] For example, in a driving scene where a lane change is required, the vehicle behavior required for an appropriate lane change (e.g., a driving path for an appropriate lane change) is predicted (result prediction 11) based on perceptual recognition of the presence of vehicles between the destination vehicles and the presence of vehicles behind, and specific driving operations such as steering to change lanes and depressing the pedal to adjust the vehicle speed are inferred and executed as driving operations 12 required to achieve the predicted result.

[0025] The driving operation 12 thus executed results in an actual vehicle behavior 13, and this actual vehicle behavior 13 (realized result) is perceived by the driver (perceptual recognition 14). Specifically, the vehicle behavior 13 is perceived by confirming the vehicle behavior visually (optical flow) or aurally, or by bodily experiencing the sense of acceleration (G value).

[0026] The vehicle behavior 13 thus perceived is compared with the result prediction 11, and a motion system prediction error 2A is recognized as the difference between the result prediction 11 (predicted vehicle behavior) and the actual vehicle behavior 13 (realized result). For example, if the timing (e.g., steering timing) of a driving operation 12 determined to obtain the result prediction 11 (predicted vehicle behavior) is actually too late to obtain the result prediction 11, the delay in the vehicle behavior 13 relative to the result prediction 11 is recognized as a motion system prediction error 2A.

[0027] When driving a vehicle, driving operation 12 is adjusted to reduce motor system prediction error 2A. Then, in motor system internal model 1A, the inference content (hypothesis for inference) regarding driving operation is changed so as to reduce motor system prediction error 2A, thereby updating motor system internal model 1A. Such positive updating of motor system internal model 1A (updating that reduces motor system prediction error 2A) enables the driver to perform appropriate driving operations, and the driver's driving skills improve.

[0028] For example, if the timing of steering when changing lanes was delayed, preventing smooth lane changes, then if the internal model of the motor system 1A is updated so that the prediction error of the motor system 2A is reduced (a new hypothesis is adopted that steering must be performed at an earlier timing in order to change lanes), the timing of steering when changing lanes will be appropriately advanced, allowing steering to be performed at the appropriate time.

[0029] Meanwhile, in the internal body system processing loop 3B, the internal body system model 1B predicts (internal body state prediction 15) the internal body state (state of internal organs, blood vessels, etc.) based on the perceptual cognition 10. In other words, the internal body system model 1B has a hypothesis (belief) for inferring (predicting) what the internal body state will be in relation to the state recognized by the perceptual cognition 10, and the internal body state is predicted based on this hypothesis.

[0030] This internal body state prediction 15 is compared with the actual internal body state 16, and the difference between the internal body state prediction 15 and the actual internal body state 16 is the internal body system prediction error 2B. The internal body system model 1B is updated to reduce the internal body system prediction error 2B, and as a result, the internal body system (autonomic nerves, etc.) is adjusted so that the internal body state prediction 15 and the actual internal body state 16 match.

[0031] In the brain processing described above, the internal motor system model 1A and the internal body system model 1B interact with each other, and if the internal motor system model 1A changes, this change will affect the internal body system model. Furthermore, the inference in the internal body system model 1B and the internal body system prediction error 2B are the cause of the driver's emotions 17.

[0032] For example, if the vehicle is in a difficult driving situation (such as attempting to merge onto a difficult main road), the internal model 1B of the body will predict that driving will not go well in that driving situation and that the heart rate will rise, and the actual heart rate will also rise. Furthermore, such negative inference can also cause negative emotions such as anxiety (fear) about driving and a lack of enjoyment from driving.

[0033] In contrast, if the internal motor model 1A is improved (positively updated) and driving skills are improved, resulting in smoother driving operations, this improvement in the internal motor model 1A will have a positive effect on the internal circumstantial model 1B. In other words, the internal circumstantial model 1B will be positively updated to not predict an increase in heart rate, reflecting confidence in driving in that driving situation, and the actual increase in heart rate will also disappear. Furthermore, positive inferences about that driving situation (predictions that driving operations will be performed well) will also induce positive emotions such as confidence and a sense of security (calmness) and joy in driving a vehicle (happiness) when faced with that driving situation.

[0034] In this way, if the internal motor system model 1A is updated appropriately, the driving skill in that driving situation will naturally improve, and the driver's emotions 17 will also improve. However, if the internal motor system model 1A is not updated appropriately in a positive manner, the driving skill will not improve and the driver will continue to perform inappropriate driving operations (driving operations with large motor system prediction errors 2A). In addition, negative emotions such as anxiety about driving and a lack of enjoyment from driving will continue.

[0035] The present invention aims to improve the driver's driving skills and emotions by providing appropriate driving assistance for such negative situations. Specifically, when appropriate driving assistance is provided for driving operations (e.g., guidance for appropriate driving operations and evaluation of driving operations using an HMI) while driving a vehicle, the driver can drive appropriately in that situation, and the driver's internal motor system model 1A is rewritten positively (i.e., the motor system prediction error 2A is reduced by the driving assistance-guided actions and attention, resulting in the driver's belief that they can perform driving operations appropriately). When the internal motor system model 1A is rewritten in this way, the interaction 1B between the internal motor system model 1A and the internal physiology model 1B also rewrites the internal physiology model 1B positively, and the driver's emotions (and furthermore, sensibility, including value judgments) improve positively (e.g., anxiety and fear about driving operations due to prediction of a negative outcome are reduced).

[0036] In relation to this, the inventor of the present invention has found that "people cultivate a sense of self-efficacy by accumulating small successful experiences, and eventually this self-efficacy generalizes, making it easier for people to feel happy on a regular basis (happiness level increases)." The positive updating of motor system internal model 1A and physiology internal model 1B according to the present invention is a small successful experience, and is thought to lead to an increase in the driver's happiness level.

[0037] Therefore, by repeatedly performing the appropriate driving assistance of the present invention while driving a vehicle, positive experiences (feelings of happiness) through the driving assistance are accumulated as small successful experiences, and ultimately, a sense of self-efficacy (for example, confidence that one can handle the vehicle) is cultivated, and as this self-efficacy generalizes, a state in which the driver's happiness level (constant sense of happiness) is enhanced can be achieved.

[0038] In the above-described internal driver model 1, the specific processing contents in the motor system processing loop vary depending on the characteristics of the driver. For example, the processing contents for reducing the prediction error vary greatly depending on whether the driver is proficient in driving in the target driving scene (in the case of an experienced driver) or not (in the case of a novice driver) in driving in the driving scene.

[0039] To explain in more detail, in the case of an experienced driver who is proficient in driving in the target driving scene, the driver can accurately predict the outcome of the driving operation in that driving scene, so the motor system prediction error 2A is small (to the extent that fine adjustments are required for specific driving), and it is thought that there is no need to make major corrections to the driving operation 12 inferred to obtain the outcome prediction 11.

[0040] That is, in the case of an experienced driver, the driving operations required in that driving scene are performed unconsciously (that is, the driving operations such as steering and pedal operation that are fixed in the brain are performed like unconscious muscle movements in response to the driving scene that is seen, without much conscious judgment), and it is thought that the process of adjusting specific driving operations 12 so as to reduce motor system prediction error 2A based on recognized vehicle behavior 13 does not have a large weight in brain processing (to the extent of making fine adjustments when driving).

[0041] Here, if we call the part of the motor system processing loop consisting of motor system internal model 1A → result prediction 11 → motor system prediction error 2A → motor system internal model 1A the muscle system loop 4, and the part (loop) consisting of motor system internal model 1A → driving operation 12 → vehicle behavior 13 → perception cognition 14 → motor system internal model 1A the vehicle behavior recognition loop 5, it is thought that experienced drivers do not make many corrections to predictions in vehicle behavior recognition loop 5 (corrections to driving operation 12 to obtain result prediction 11).In this sense, it can be said that the function of muscle system loop 4 is dominant in the driver internal model 1 of experienced drivers.

[0042] Therefore, in driving assistance for experienced drivers, driving assistance (coaching) for unconscious driving operations is not very useful for reducing the motor system prediction error 2A, and may even interfere with the unconscious (refined) driving operations of experienced drivers. Furthermore, the internal model of the body system 1B predicts that there will be annoying driving assistance, which may result in negative emotions based on this prediction (inference). For example, anticipating annoying coaching may cause irritation.

[0043] On the other hand, information about vehicle behavior is also useful for experienced drivers. Furthermore, if an experienced driver can perform driving operations unconsciously, even if detailed information about vehicle behavior is given, the experienced driver will be able to process the information appropriately, which is thought to contribute to comfortable independent driving. Therefore, as driving assistance for experienced drivers, it is appropriate to provide only information about vehicle behavior without coaching on driving operations themselves.

[0044] On the other hand, in the case of a novice driver (a driver who is not familiar with driving operations in the target driving scene or who has only recently started driving the vehicle), it is thought that the driver is unable to accurately predict the outcome of the driving operation, resulting in a large motor system prediction error 2A. For this reason, it is necessary to make significant changes to the driving operation 12 in order to reduce the motor system prediction error 2A.

[0045] That is, in the case of a novice driver, unconscious driving is not sufficient to perform appropriate driving, and it is considered that the weight of the processing in the vehicle behavior recognition loop 5 (i.e., the processing of adjusting the driving operation 12 based on the actual vehicle behavior 13 so as to reduce the motor system prediction error 2A) in the motor system processing loop is large. In this sense, it can be said that the function of the vehicle behavior recognition loop 5 is dominant in the driver internal model 1 of the novice driver.

[0046] For this reason, when providing driving assistance to novice drivers, it is considered effective to provide driving assistance for the driving operation itself (for example, appropriate coaching by HMI or intervention in the driving operation (for example, steering assistance)) so that the driving operation itself can be performed appropriately. On the other hand, even if detailed information about the vehicle's behavior is provided to the novice driver, there is a possibility that the novice driver will not be able to fully digest it and will become confused. Therefore, when providing driving assistance to novice drivers, it is appropriate to provide driving assistance that mainly consists of coaching for the driving operation itself. This makes it possible to guide the internal driver model 1 of the novice driver to a state in which the muscle force loop 4 is dominant.

[0047] The driving assistance device of the present invention has been developed based on the above-mentioned viewpoints, and is characterized by being able to effectively achieve appropriate improvement of the driver's driving skills and improvement of the driver's emotions (improvement of happiness) by providing appropriate driving assistance according to the driver's characteristics (level of proficiency in the target driving scene). Specific configuration examples of the driving assistance device will be described in detail below.

[0048] An example of a control system in the driving assistance device of the present invention is shown in a block diagram in Figure 2. The driving assistance device is installed in a vehicle such as an automobile or in a device that simulates vehicle driving (driving simulator), and is, for example, an advanced driving assistance system (ADAS).

[0049] As shown in the figure, the control system includes a control unit U, a driving information acquisition means 21, a driving information acquisition means 22, a vehicle information acquisition means 23, a driver information acquisition means 24, a memory means 25, an HMI (human-machine interface) 26, and a driving operation intervention means 27.

[0050] The driving information acquisition means 21 is a means for acquiring information outside the vehicle (the external environment of the vehicle, such as roads and other vehicles) related to the vehicle's driving state, and is composed of, for example, an exterior camera that captures images of the outside of the vehicle and a sensor that detects the situation outside the vehicle.

[0051] The driving information acquisition means 22 is a means for detecting the state of driving operation, and includes, for example, a steering angle sensor, an accelerator sensor, a brake sensor, etc. The vehicle information acquisition means 23 is a means for acquiring (detecting) the vehicle state such as the vehicle speed, and includes, for example, a vehicle speed sensor, an acceleration sensor, etc.

[0052] The driver information acquisition means 24 is a means for acquiring information relating to the state of the driver of the vehicle, and is configured, for example, by an in-vehicle camera that captures an image of the driver.

[0053] The storage means 25 is a means for storing various data, and is, for example, an external storage device. The storage means 25 stores various information such as road information, information on a standard driver model, and information on an individual vehicle model, as well as the driving history of the vehicle (the history of driving operations and vehicle behavior for each driver). In this embodiment, the storage means 25 corresponds to the "driving history storage means" in the claims.

[0054] The HMI 26 is a device that provides visual and auditory information to the driver of the vehicle during driving assistance, and includes, for example, a display 26A (e.g., an in-vehicle monitor or a head-up display (HDU)) that can display images and text, and a notification means such as a speaker 26B that can output sound. In this embodiment, the HMI 26 corresponds to the "interface means" in the claims.

[0055] The driving operation intervention means 27 is a means for directly intervening (for example, assisting steering) in the driver's driving operations (steering, accelerator operation, braking operation) in driving assistance, and performing part or all of the driving operations on behalf of the driver.

[0056] The control unit U is a control device configured by, for example, a microcomputer, and includes a driving scene recognition means 31, a driving target calculation means 32, a driving operation detection means 33, a vehicle behavior detection means 34, a driver monitoring means 35, a driver determination means 36, and an assistance mode setting means 37. These means are provided as control programs in the control unit U.

[0057] The driving scene recognition means 21 is a means for recognizing (identifying) the driving scene of the vehicle (the situation in which the vehicle is driving) based on information outside the vehicle acquired by the driving information acquisition means 21. The driving scenes are grouped according to their contents (for example, lane changes, starting on a slope, etc.), and driving scenes belonging to the same group are treated as the same driving scene.

[0058] The driving target calculation means 32 is a means for calculating, as a target driving operation and a target vehicle behavior, (appropriate) driving operations and vehicle behaviors to be performed in the specific driving scene recognized by the driving scene recognition means 21. The calculation of the target driving operation and the target vehicle behavior is performed based on the content of the specific driving scene recognized, with reference to various information (road information, information on the standard driver model, information on the individual vehicle model) stored in the storage means.

[0059] The driving operation detection means 33 is a means for detecting the driving operation of the driver for each driving scene based on the information (detection signal) acquired by the driving information acquisition means 22. The detected driving operation data for each driving scene is stored in the storage means 25 as data for each driver.

[0060] The vehicle behavior detection means 34 is a means for detecting (calculating) the behavior of the vehicle in a driving scene based on the information (detection signal) acquired by the vehicle information acquisition means 23. Data on the detected vehicle behavior for each driving scene is stored in the storage means 25 as data for each driver. In this embodiment, the driving operation detection means 33 and the vehicle behavior detection means 34 correspond to the "driving detection means" in the claims.

[0061] The driver monitoring means 35 is a means for monitoring the driver based on the information acquired by the driver information acquisition means 24. For example, by analyzing an image of the driver taken by an in-vehicle camera, the driver's identity is identified, and the driver's physiological state (heart rate, amount of sweating) is detected (estimated) to detect the driver's behavior. Note that the driver's physiological state can also be detected directly by using a means for detecting the physiological state that is directly worn by the driver (for example, a heart rate monitor that measures the heart rate or a sweat sensor that detects the amount of sweating) as the driver information acquisition means 24.

[0062] The driver determination means 36 is a means for determining the characteristics of a driver to be determined based on the driving operation and vehicle behavior history data stored in the storage means 35. Specifically, the means 36 compares the driving operation data of the driver in the driving scene to be determined with a target driving operation, and also compares the vehicle behavior data of the driver in the driving scene to a target vehicle behavior, and determines whether the driver to be determined is skilled in driving (the vehicle) in the driving scene to be determined (driving proficiency) based on the results of these comparisons.

[0063] More specifically, the system determines whether the data on the driver's driving operation in the driving scene (e.g., all data within a predetermined period) is within a predetermined variation range (target driving operation range) from the target driving operation, and whether the data on the driver's vehicle behavior in the driving scene (e.g., all data within a certain period) is within a predetermined variation range (target vehicle behavior range) from the target vehicle behavior.If both the data on the driving operation and the data on the vehicle behavior are within the target driving operation range and the target vehicle behavior range, respectively, the system determines that the driver is proficient (highly proficient) in the driving scene.Conversely, if either the data on the driving operation or the data on the vehicle behavior is not within the target driving operation range or the target vehicle behavior range, the system determines that the driver is not proficient (lowly proficient) in the driving scene.

[0064] Note that the determination of a driver's proficiency using the target driving operation and the target vehicle behavior is not limited to the above. For example, the driver's proficiency may be determined to be high if either the driving operation or the vehicle behavior of the driver in the driving scene is within a predetermined variation range from the target value. Alternatively, the driver's proficiency may be determined to be high if the average value of the data on the driving operation and / or the vehicle behavior of the driver in the driving scene is within a predetermined variation range from the target driving operation and / or the target vehicle behavior.

[0065] The determination result by the driver determination means 36 is stored in the storage means 25 as data (driver data) representing the characteristics of each driver.

[0066] The assistance mode setting means 37 is a means for determining whether the current driving scene requires driving assistance (whether to provide driving assistance) based on the recognition of the driving scene by the driving scene recognition means 21, and if it is a driving scene in which driving assistance should be provided, for setting the mode of driving assistance by the HMI 26 and the driving operation intervention means 27 based on the determination (driver characteristics) by the driver determination means 36. A command signal based on this setting is sent to the HMI 26 and the driving operation intervention means 27, and driving assistance according to the setting is performed.

[0067] Specifically, in this embodiment, when the driving assistance by the HMI 26 determines that the driver is proficient in the driving scene during driving (in the case of an experienced driver), information regarding the driving operation itself is not provided (coaching), and only information related to the behavior of the vehicle at that time (i.e., information related to the content of the driving scene and the behavior of the vehicle itself) is notified to the driver by image or sound. On the other hand, when the driving assistance by the HMI 26 determines that the driver is not proficient in the driving scene during driving (in the case of a novice driver), information regarding the driving operation itself, rather than information regarding the behavior of the vehicle, is provided (coaching) by image or sound.

[0068] 3 and 4 show specific examples of displays that are displayed on the display 26A of the HMI 26 when the driving scene is a hill start for a manual transmission vehicle (MT vehicle). Fig. 3 shows a display for a highly skilled driver (display for an experienced driver), and Fig. 4 shows a display for a less skilled driver (display for a novice driver).

[0069] In the display for experienced drivers, as shown in Fig. 3, a display that provides information about the vehicle behavior itself is displayed on display 26A of HMI 26. To explain in more detail, display 26A displays an image of host vehicle 42 and a following vehicle 43 traveling on a slope 41, and also displays information such as the gradient of slope 42, current vehicle weight, driving force and gravity, and the distance between host vehicle 42 and the following vehicle 43. In addition, information about the vehicle behavior is also announced by voice from a speaker as necessary.

[0070] Furthermore, if the difference between the vehicle behavior resulting from the driver's driving operation and the target vehicle behavior is equal to or greater than a predetermined value, information indicating the difference is displayed on display 26A. In the example of Fig. 3, when the vehicle's driving force is insufficient, a graph 44 is displayed in which the driving force and gravity are represented by opposing arrows, and when the accelerator pedal is revved, the arrow of the driving force in graph 44 extends (as shown by the dashed line in the figure). This allows experienced drivers to recognize the vehicle behavior required at that time (the difference from the target value) rather than the driving operation, and encourages the driver to respond autonomously.

[0071] In this way, driving assistance for experienced drivers involves only presenting information about the vehicle's behavior itself, since there is little need for coaching on the driving operation itself. As a result, for experienced drivers in whom the processing of the muscle system loop 4 is dominant in the driver's internal model 1 (who are able to perform unconscious driving operations), assistance is provided to enable appropriate processing in the vehicle behavior recognition loop 5 without causing any bother (without adversely affecting the internal body model 1B).

[0072] On the other hand, in the display for novice drivers, as shown in Fig. 4, displays directly related to the driving operation of the vehicle are displayed on display 26A of HMI 26. To explain in more detail, display 26A displays a graph 51 of accelerator opening, a graph 52 of handbrake force, and a graph 53 of clutch depression. Each graph shows the current operation amount (shown by a solid line in the figure) and the target value (shown by a dashed line in the figure), and also indicates the direction in which driving operation should be performed.

[0073] Furthermore, the display 26A displays an image 54 of the driver, and specific suggestions regarding the operation of the parking brake 55, shift lever 56, accelerator pedal 57, brake pedal 58, and clutch pedal 59 are displayed by the driver's movements, color changes, etc. Also, if necessary, specific suggestions regarding driving operations (such as "Step on the accelerator and increase the speed to 1500 rom") are given by voice from the speaker (and / or text displayed on the display).

[0074] In this way, since there is a high need to assist the driving operation itself in the driving assistance for a novice driver, information related to the driving operation itself (coaching on the timing and amount of driving operation) is presented. In other words, in the notification to the novice driver, the driving operation that the driver should perform is specifically shown, so the driver can perform the driving operation by directly referring to the notification content.

[0075] This effectively encourages novice drivers, in whom the processing of the vehicle behavior recognition loop 5 is dominant in the driver's internal model 1 (who are unable to perform unconscious driving operations), to improve their driving skills (reduce the motor system prediction error 1B), thereby appropriately supporting the formation of the motor system internal model 1A in which the muscle system loop 4 is dominant, and ultimately effectively achieving an improvement in the driver's emotions (improvement of happiness).

[0076] As described above, the driving assistance provided by the HMI 26 in this embodiment is appropriately set depending on whether the driver is an experienced driver or a novice driver, and therefore, it is possible to effectively improve the driver's driving skills and improve their emotional state depending on the driver's characteristics.

[0077] In addition to the above-described notifications by the HMI 26, the driving assistance may also involve direct intervention in driving operations (such as assisting with driving operations) by the driving operation intervention means 27 as necessary. In this case, the manner of driving assistance by the driving operation intervention means 27 is also changed appropriately depending on whether the driver is a novice driver or an experienced driver. For example, assistance for a novice driver who is highly in need of assistance with the driving operation itself may be increased in the degree of intervention in driving (increasing the amount and strength of driving assistance) compared to assistance for an experienced driver.

[0078] In addition, the driving assistance provided by the driving operation intervention means 27 can change the mode of assistance depending on the state of the motor system prediction error 2A. For example, when traveling around a left corner on a gentle uphill slope, steering assistance is provided. However, if the actual steering is lighter than the driver's prediction (prediction that the front wheels will not be loaded and the vehicle will be difficult to turn due to acceleration on the uphill slope), and there is no tendency for the steering to understeer, the steering assistance is weakened and changed to a heavier steering that the driver is accustomed to. This eliminates the need for the driver to learn how to adjust the steering force, and reduces the prediction error in the operation input without any effort on the part of the driver. This allows the driver to learn the relationship between the operation input and vehicle behavior while performing driving operations as intended, thereby accelerating the reduction of prediction errors related to vehicle behavior (however, intervention in driving operations is not provided when the operation input is inconsistent, in order to avoid adversely affecting the learning of driving operations).

[0079] Furthermore, the assistance mode set by the assistance mode setting means 37 can be adjusted depending on the driver's condition at that time. Specifically, in this embodiment, if the physiological values ​​(e.g., heart rate, amount of sweat) or behavioral values ​​(e.g., frequency of characteristic gestures) indicating the driver's tension monitored by the driver monitoring means 35 are lower in the current driving scene than in the previous driving scene, the level of driving assistance in the same driving scene (this time or the next time onwards) is reduced from the previous level. For example, the amount of information displayed on the image of the HMI 26 is reduced.

[0080] In other words, a decrease in the feature quantities (physiological quantities and behavior quantities) indicating tension suggests that the motor system internal model 1A and the body system internal model 1B in the driver internal model 1 have been positively updated, allowing for smoother driving in that driving situation and reducing the need for driving assistance. Therefore, the level of driving assistance (e.g., the amount of information) is reduced to encourage more autonomous driving by the driver. For example, in the novice driver display shown in Figure 4, the display of the image 54 of the driver is stopped and switched to displaying only graphs 51 to 53. This promotes improvement in the driver's driving skills and improved emotional well-being.

[0081] Furthermore, the assistance mode may be set based on characteristics other than the driver's proficiency. For example, if the driver is elderly or if the driver's fatigue is apparent from observation by the driver monitoring means 35, the visual display by the HMI 26 may be simplified (the amount of information may be reduced) or audio coaching may be increased to prevent the information provided from becoming excessive and placing a burden on the driver.

[0082] Next, an example of control in the present invention will be described with reference to Figures 5 and 6. Figure 5 is a flowchart showing the overall control of the present invention. In the control, first, in step S1, the driving scene during driving is recognized.

[0083] In the next step S2, it is determined whether the recognized driving scene is a target for driving assistance, and if it is not a target, the series of processes is terminated. On the other hand, if the driving scene is a target for driving assistance, the process proceeds to step S3, where data related to the driver (determination result related to proficiency) is acquired (read from the storage means 25).

[0084] In the next step S4, it is determined whether the driver has a high level of proficiency for the driving scene based on the data acquired in step S3, and if the driver has a high level of proficiency, the process proceeds to step S5, where driving assistance for an experienced driver is performed, and the series of processes ends.On the other hand, if the driver has a low level of proficiency, the process proceeds to step S6, where driving assistance for a novice driver is performed, and the series of processes ends.

[0085] An example of control in driver determination (determination of driver proficiency) is shown in a flowchart in Fig. 6. In driver determination, first, in step S11, the driving scene during driving is recognized.

[0086] In the next step S12, it is determined whether the recognized driving scene is a target for driving assistance, and if it is not a target, the series of processes ends. On the other hand, if the driving scene is a target for driving assistance, the process proceeds to step S13.

[0087] In steps S13 and S14, the driving history data is updated based on the driving operations and vehicle behavior in the current driving scene. That is, in step S13, the driving operations and vehicle behavior in the current driving scene are detected, and in the subsequent step S14, the driving history data is updated based on the detected driving operations and vehicle behavior.

[0088] In subsequent steps S15 to S21, the driver's proficiency determination for the driving scene is updated based on the updated driving history data. First, in step S15, a target driving operation and a target vehicle behavior for the driving scene are calculated. In step S16, driving history data related to the driving operation and vehicle behavior of the driver in the driving scene is read.

[0089] In the following step S17, it is determined whether the driving operation data acquired in step S16 is within the driving operation target range from the target driving operation. If it is not within the driving operation target range, the process proceeds to step S20, where it is determined that the driver is a novice driver, and in the following step S21, the driver data (data related to the driver's judgment) is updated, and the series of processes is terminated.

[0090] On the other hand, if it is determined in step S17 that the vehicle behavior is within the total driving target range, the process proceeds to step S18, where it is determined whether the vehicle behavior data acquired in step S16 is within the vehicle behavior target range from the target vehicle behavior. If it is determined that the vehicle behavior is not within the vehicle behavior target range, the process proceeds to step S20, where it is determined that the driver is a novice driver, and in the following step S21, the driver data is updated, and the series of processes ends.

[0091] If it is determined in step S18 that the vehicle behavior is within the target range, the process proceeds to step S19, where it is determined that the driver is an experienced driver, and in the following step S21, the driver data is updated, and the series of processes ends.

[0092] Next, application of the present invention to a specific driving scenario will be described in relation to the driver internal model 1. When the driving scenario is starting a manual transmission (MT) vehicle on a slope, the driver needs to execute a series of driving operations required for starting on a slope (i.e., a series of operations in the following sequence: step on the brake (with the right foot) → step on the clutch with the left foot → lightly press the accelerator with the right foot → gradually release the clutch with the left foot → unlock the handbrake with the left hand and begin to release it after forcibly pulling → press the accelerator with the right foot → fully release the clutch with the left foot).

[0093] In such continuous operations, a novice driver is unable to make accurate predictions using the motor system internal model 1, and therefore is unable to perform driving operations unconsciously (as predicted); instead, he or she individually imagines the timing and operating muscle force of each driving operation for starting on a slope, and executes them; in order to reduce motor system prediction error 2A, processing (correction of predictions) by vehicle behavior perception loop 5 is thought to be particularly necessary (a state in which vehicle behavior perception loop 5 is dominant).

[0094] In other words, when a novice driver perceives the actual results of their driving operations (for example, hearing the sound of an engine speed increasing, feeling the change in vehicle body posture when driving force is greater than or equal to gravity, feeling the vehicle start to move forward, etc.), they do not have an evaluation index for this, so they are likely to vaguely recognize the occurrence of a motion system prediction error 2A as abnormal behavior (recognizing the entire series of operations as a failure). Also, when a problem occurs in the vehicle's behavior (for example, an engine stall), they are unable to identify the driving operation that caused it. Therefore, in order to reduce the motion system prediction error 2A, novice drivers need to challenge themselves on slopes by changing the timing and degree of each driving operation and master each driving operation that reduces turbulence (learning through trial and error).

[0095] Therefore, when providing driving assistance to novice drivers, the HMI 26 selects and executes assistance such as providing specific instructions regarding the level and timing of each driving operation (for example, coaching on the driving operation itself, such as voice instructions such as "Step on the accelerator and increase the speed to 1500 rom") or assisting in adjusting the engine speed (see Figure 4).

[0096] On the other hand, in the case of an experienced driver, accurate predictions can be made using the motor system internal model 1, and the motor system prediction error 2A is small, so predictions in the vehicle behavior recognition loop 5 are not corrected very often, and it is thought that smooth driving can be performed based on the resulting predictions (a state in which the muscle system loop 4 is dominant).

[0097] In other words, when an experienced driver performs a series of driving operations for starting on a slope, they can perform the operation as a series of unconscious hand and foot movements without thinking about each operation individually, and the resulting vehicle behavior is stable (for example, the vehicle is quiet and shake-free, just like starting on flat ground, the duration and degree of half-clutch is short, accelerator jerking is minimal, and no shock occurs).

[0098] Furthermore, motor system prediction error 2A can also be broken down as an everyday error into the results of each operation without much thought, and corrections can be made immediately (for example, you can sit deeper in the seat and stabilize your hips so that you can press the heavy clutch accurately). Also, if a problem occurs, you can immediately determine the specific cause and how to deal with it. For example, if the engine stalls (perceived through physical G-forces and hearing), you can determine that the clutch was released too early (you were tired and the heavy clutch gave in). Also, if the vehicle rolls backward (perceived through physical G-forces), you can determine that the handbrake was released too early. Also, if the engine was revved too much (perceived through hearing), you can determine that the clutch was engaged too late (you were careless in finding the engagement point).

[0099] In this way, for experienced drivers, each driving operation can be said to be a limb operation that is ingrained in the brain, and therefore does not require assistance. Therefore, assistance is provided by providing information about the vehicle's behavior (for example, graphs of slope, gravity, and driving force) via the HMI 26 (see Figure 3), which promotes autonomous corrections.

[0100] Next, we will explain the case where the driving scene is heel-and-toe in a manual transmission vehicle. In this case, the driver must perform a series of heel-and-toe driving operations (continuous operations of heel-and-toe: pressing the brake with the toe of the right foot → pressing the clutch pedal with the left foot → moving the shift knob (for example, from 3 to N) → engaging the clutch at N → pushing the accelerator pedal with the heel of the right foot → moving the shift knob (for example, from N to 2) → engaging the clutch with the left foot).

[0101] Novice drivers will perform the above-mentioned heel-and-toe sequence by individually imagining the timing and muscle force of each operation. However, because they do not have an evaluation index for driving operations, if a motor system prediction error 2A occurs, they will likely vaguely recognize it as abnormal behavior (they will recognize it as a failure of the entire series of operations). Furthermore, even if a problem occurs, they will not be able to identify the driving operation that caused it. Therefore, in order to reduce motor system prediction error 2A, novice drivers need to practice driving by changing the timing and degree of each driving operation, and master driving operations that minimize disruption of vehicle behavior (learning through trial and error).

[0102] Therefore, in providing driving assistance to novice drivers, the HMI 26 provides notifications showing tips on how to move the body (exercise muscle strength) for each driving operation (for example, notifications instructing the driver to concentrate on the toe brake and guidance on how to twist the ankle when stepping on the heel), and assists in adjusting the engine speed.

[0103] On the other hand, experienced drivers can perform the series of driving operations during heel-and-toe as unconscious hand and foot operations without breaking down the operations into individual operations and thinking deeply about them, and the vehicle's behavior is stable (for example, the vehicle deceleration remains constant even while performing right-foot heel-toe operations, and no shock occurs at the final clutch engagement). Also, the motor system prediction error 2A can be broken down into the results of each operation as an everyday error, and corrections can be made immediately. For example, to prevent the engine speed from dropping too much when engaging the clutch after a downshift, the driver can press the heel more deeply to increase the EG speed or advance the timing of the final clutch engagement.

[0104] In this way, when providing driving assistance to experienced drivers, no special assistance is required for the driving operation itself, which is a limb operation that is ingrained in the brain. Therefore, the HMI 26 provides assistance by providing information on the vehicle's behavior (for example, a G graph during heel-and-toe maneuvers or an engine RPM graph) to encourage the driver to make autonomous corrections.

[0105] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and appropriate modifications are possible within the scope of the claims. [Industrial Applicability]

[0106] The present invention can be used in a driving assistance device that assists a driver of a vehicle. [Explanation of symbols]

[0107] 1 Driver internal model 1A Internal model of the motor system 1B Internal model of the body system 2A Motor system prediction error 2B Prediction error of the body system 3A Motor System Processing Loop 3B Internal body processing loop 4 Muscle Strength Loop 5 Vehicle behavior recognition loop 21 Means of obtaining driving information 22 Driving information acquisition means 23 Vehicle information acquisition means 24 Driver information acquisition method 25 Memory means 26 HMI 26A Display 26B Speaker 27 Driving operation intervention measures 31 Driving scene recognition means 32 Operation target calculation means 33 Driving operation detection means 34 Vehicle behavior detection means 35 Driver Monitoring Measures 36 Driver determination means 37 Support mode setting means U Controller

Claims

1. A driving assistance device that assists a driver of a vehicle, a driving scene recognition means for recognizing a driving scene of the vehicle; a driving detection means for detecting a driving operation of the driver and / or a behavior of the vehicle; a driving target calculation means for calculating a target driving operation and / or a target vehicle behavior that is suitable for the driving scene; an interface means for notifying the driver of driving information; a driving history storage means for storing the driving operation of the driver and / or the behavior of the vehicle for each driving scene; a driver determination means for determining the driver's proficiency in a driving scene based on a comparison between the driver's past driving operations and / or vehicle behavior in the driving scene to be determined and the target driving operations and / or target vehicle behavior; an assistance mode setting means for setting a mode of notification by the interface means to the driver based on the determination of the driver's proficiency by the driver determination means; A driving assistance device equipped with the above.

2. The driving assistance device according to claim 1, The assistance mode setting means sets the notification mode by the interface means for a driver who is determined to have a high level of proficiency with respect to a driving scene during driving to a mode in which notifications are not provided regarding the driving operation itself, but rather regarding vehicle behavior.

3. The driving assistance device according to claim 2, The notification regarding the vehicle behavior includes a notification of a difference between the vehicle behavior and the target vehicle behavior when the difference between the vehicle behavior and the target vehicle behavior is equal to or greater than a predetermined value.

4. The driving assistance device according to claim 1, The driving assistance device wherein the assistance mode setting means sets the notification mode by the interface means to a mode that notifies a driver who is determined to have a low level of proficiency with respect to a driving scene during driving, to a mode that notifies the driver regarding the driving operation itself.

5. The driving assistance device according to claim 1, The driver determination means determines that the driver has a high level of proficiency with respect to the driving scene being determined if the driver's driving operation over a predetermined period of time is within a predetermined driving operation target range from the target driving operation, and the vehicle behavior over the predetermined period of time is within a predetermined vehicle behavior target range from the target vehicle behavior.

6. The driving assistance device according to claim 1, a driver monitoring means capable of detecting a physiological quantity and / or a behavior of the driver; The assistance mode setting means is a driving assistance device that reduces the level of notification by the interface means when the physiological values ​​and / or behaviors indicating the driver's tension detected by the driver monitoring means in the same driving scene decrease in the current driving scene compared to the previous driving scene.

7. The driving assistance device according to claim 1, a driving operation intervention means for intervening in the driving operation of the driver, The driving assistance device includes an assistance mode setting means for changing the mode of driving assistance provided by the driving operation intervention means based on the determination by the driver determination means.

Images