Driving assistance systems

Abstract

This driver assistance system provides the ability to prematurely stop deceleration assistance, which is a driver assistance function based on the driving environment before the change, when the driving environment changes. [Solution] When the driver assistance ECU detects a deceleration operation performed by the vehicle's driver to slow down the vehicle, it identifies the object of the deceleration request that caused the deceleration operation. It also learns deceleration support start position information, which represents the "deceleration support start position, which is the position of the vehicle when the deceleration operation started," and associates it with the object of the deceleration request identified for that deceleration operation. Furthermore, after learning the deceleration support start position information, when the vehicle reaches the deceleration support start position represented by the learned deceleration support start position information, it executes deceleration support. In addition, after learning the deceleration support start position information, if it is determined that the object of the deceleration request learned in association with the deceleration support start position represented by that deceleration support start position information has changed, the deceleration support start position information is discarded.

Description

Technical Field

[0001] The present invention relates to a driving support device that learns a position where a driver of a vehicle performs a deceleration operation and performs deceleration support to automatically decelerate the vehicle when the vehicle passes through that position after learning.

Background Art

[0002] One of the conventional driving support devices (hereinafter simply referred to as "conventional device") creates map information by associating changes in the driving operation of a driver of a vehicle with position information when such a driving operation change occurred. The conventional device uses the created map information for vehicle control (see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

[0004] However, for example, when a driver performs a deceleration operation, it is difficult to determine whether the deceleration operation is an accidental operation or an operation caused by a driving environment (for example, a deceleration requirement target such as a stop sign and a curve ahead) that requires the vehicle to decelerate. Therefore, in order to create map information as in the prior art, it is necessary to statistically determine that the deceleration operation is an operation caused by the driving environment, and for this purpose, a large amount of data must be collected. Furthermore, even after the map information is once created, when the driving environment changes, it is necessary to statistically determine that the driving environment has changed, and for this purpose, a large amount of data must be collected. As a result, there is a risk that unnecessary vehicle control (for example, deceleration support) may be executed during a long period from the time when the driving environment changes until the map information is actually changed.

[0005] This invention was made to address the above-mentioned problems. Specifically, one of the objectives of this invention is to provide a driver assistance device that can stop deceleration assistance, which is a driver assistance function based on the driving environment before the change, at an early stage when the driving environment changes. In the following, "step" may be written as "S".

[0006] One aspect of the driver assistance device of the present invention is The vehicle is equipped with a controller (10) configured to perform deceleration assistance, which automatically slows down the vehicle.

[0007] The aforementioned controller, The object that caused the deceleration request, which is a driving operation performed by the driver of the vehicle to slow down the vehicle, is identified (S420), The deceleration support start position information, which represents the position of the vehicle when the deceleration operation is initiated, is learned together with the deceleration request target identified for the deceleration operation (S455), After learning the deceleration support start position information, the vehicle will perform the deceleration support when it reaches the deceleration support start position represented by the learned deceleration support start position information (S530). It is configured in such a way, Furthermore, the controller, After learning the deceleration support start position information, if it is determined that the deceleration request target, which has been learned in association with the deceleration support start position represented by the deceleration support start position information, has changed (S520: Yes), the deceleration support start position information for the deceleration request target that has been determined to have changed is discarded (S540). It is structured in this way.

[0008] According to this, if it is determined that the target of the deceleration request has changed after learning the deceleration support start position information, the deceleration support start position information for the changed deceleration request target is immediately discarded. Therefore, when the vehicle passes through a driving environment where the deceleration request target has changed, deceleration support corresponding to the deceleration operation performed based on the deceleration request target before the change will no longer be executed. Thus, the frequency of unnecessary deceleration support being executed can be reduced.

[0009] In the above description, to aid in understanding the present invention, the names and / or reference numerals used in the embodiments described later are indicated in parentheses for the components of the invention corresponding to those embodiments. However, the components of the present invention are not limited to the embodiments defined by the above names and / or reference numerals. The present invention also extends to the driving assistance method realized by the driving assistance device, and to the program that causes the computer mounted on the vehicle to execute each step of the driving assistance method. [Brief explanation of the drawing]

[0010] [Figure 1] This is a schematic diagram of a driver assistance device according to an embodiment of the present invention. [Figure 2] This is a plan view of an intersection and vehicles to explain deceleration assistance. [Figure 3] (A) is a table showing an example of "deceleration request targets and driving condition index values" stored in non-volatile memory, and (B) is a table showing an example of "deceleration request targets and learned values" stored (learned) in non-volatile memory. [Figure 4] Figure 1 is a flowchart showing the routines executed by the CPU of the driver assistance ECU. [Figure 5] Figure 1 is a flowchart showing the routines executed by the CPU of the driver assistance ECU. [Figure 6] Figure 1 is a flowchart showing the routines executed by the CPU of the driver assistance ECU. [Modes for carrying out the invention]

[0011] (composition) The driver assistance device according to an embodiment of the present invention (hereinafter referred to as "device DS") comprises the components shown in Figure 1. Device DS is applied to (mounted on) a vehicle HV. The vehicle HV may be any of the following: a vehicle powered by an internal combustion engine, a vehicle powered by an electric motor (i.e., an electric vehicle), and a hybrid vehicle.

[0012] In this specification, "ECU" refers to an electronic control unit equipped with a microcomputer. The microcomputer includes a CPU (processor), ROM, RAM, data-writable non-volatile memory, and interfaces, etc. An ECU is also referred to as a controller or computer.

[0013] The driver assistance ECU 10 performs deceleration support control, which will be described later, as part of the driver assistance control. The driver assistance ECU 10 is connected to the components described below and transmits and receives information or signals between them. The driver assistance ECU 10 may be composed of multiple ECUs.

[0014] The camera sensor 21 captures images of the scene around the vehicle HV, including the area in front of the vehicle HV, at predetermined intervals, and acquires image data. Based on the image data, the driver assistance ECU 10 acquires information representing the positional relationship between the vehicle HV and objects and various lines drawn on the road surface (e.g., lane markings) present around the vehicle HV (i.e., positional relationship information). Furthermore, based on the image data, the driver assistance ECU 10 acquires information about the driving environment of the vehicle HV. The information about the driving environment includes information on whether or not there is a vehicle requiring deceleration and information identifying the type of vehicle requiring deceleration. This information obtained based on the image data acquired by the camera sensor is sometimes referred to as "camera information".

[0015] The deceleration requirement target is a driving environment indicating that the vehicle HV needs to decelerate before reaching the deceleration requirement target. In other words, the deceleration target requirement is the target that causes the "deceleration operation, which is a driving operation to decelerate the vehicle HV" by the driver of the vehicle HV. For example, it includes at least one of a stop sign, a stop line, a curve (curved road) existing in front of the vehicle HV, a level crossing signal, a gate provided at the entrance to a toll road, a priority road intersecting the lane of the vehicle HV, and other signs.

[0016] The radar sensor 22 is a well-known sensor that acquires information about an object existing in front of the vehicle HV using radio waves in the millimeter wave band. The radar sensor 22 transmits millimeter wave information about the transmitted and received millimeter waves to the driving support ECU 10. The driving support ECU 10 acquires "radar information" based on the millimeter wave information. The radar information includes the distance to the object, the azimuth of the object, the relative speed of the object, and the like.

[0017] Note that the driving support ECU 10 generates fusion information by integrating the camera information and the radar information.

[0018] The vehicle speed sensor 23 outputs a signal indicating the speed (i.e., vehicle speed) Vh of the vehicle HV. The acceleration sensor 24 outputs a signal indicating the acceleration G in the front-rear direction of the vehicle HV. The accelerator pedal operation amount sensor 25 outputs a signal indicating the operation amount AP of an accelerator pedal of the vehicle HV (not shown). The brake pedal operation amount sensor 26 outputs a signal indicating the operation amount BP of a brake pedal of the vehicle HV (not shown).

[0019] The first shift paddle 27 is disposed on a steering wheel (not shown) and outputs a signal Sup for shifting up the gear position of a transmission (not shown) when pushed by the driver. When the driving support ECU 10 receives the signal Sup, it shifts up the gear position of the transmission (not shown) by one gear. The second shift paddle 28 is located on the steering wheel (not shown) and outputs a signal Sdn when pressed by the driver to downshift the transmission (not shown) by one gear. When the driver assistance ECU 10 receives the signal Sdn, it downshifts the transmission (not shown) by one gear.

[0020] The driver assistance ECU 10 is further connected to the powertrain actuator 30, brake actuator 40, steering motor (i.e., steering actuator) 50, warning device 60, and navigation ECU 70.

[0021] The powertrain actuator 30 drives a "drive system including a power source and transmission" (not shown) of the vehicle hybrid to adjust the driving force of the vehicle hybrid. The brake actuator 40 drives a braking system (not shown) of the vehicle hybrid to adjust the braking force applied to the vehicle hybrid. Therefore, the driver assistance ECU 10 can automatically decelerate the vehicle HV by driving the powertrain actuator 30 and the brake actuator 40.

[0022] The steering motor 50 applies torque to the steering mechanism (not shown) of the vehicle HV, thereby changing the steering angle of the steering wheels (not shown) of the vehicle HV.

[0023] The alarm device 60 includes a display device and an alarm sound generator. The alarm device 60 causes the display device to display a predetermined alarm and the alarm sound generator to generate a predetermined alarm sound in response to an instruction signal from the driver assistance ECU 10.

[0024] The navigation ECU 70 is connected to the GPS receiver 71 and the map information storage device 72. The navigation ECU 70 acquires the current position of the vehicle HV based on the GPS signal received by the GPS receiver 71. Based on the acquired current position and the map information stored in the map information storage device 72, the navigation ECU 70 can provide the driver assistance ECU 10 with information about the vehicle HV's driving environment (i.e., whether or not there is a vehicle requiring deceleration and its type). The navigation ECU 70 may also communicate with an external device to acquire the latest map information and update the map information stored in the map information storage device 72 based on the acquired latest map information.

[0025] (Summary of operation) The device DS detects (recognizes / identifies) targets for deceleration requests based, for example, on image data. Figure 2 shows an example where a stop sign ST is detected as a target for deceleration requests.

[0026] The DS device determines whether the driving environment in front of the HV vehicle includes a vehicle requiring deceleration when the HV driver initiates deceleration. Deceleration operations include, for example, pressing the brake pedal and shifting down the transmission. If the driving environment in front of the HV vehicle includes a vehicle requiring deceleration, the DS device stores a driving condition index value associated with the type of vehicle requiring deceleration, as shown in Figure 3(A) as an example. The driving condition index value includes the deceleration operation start position (latitude and longitude), the deceleration operation end position (latitude and route), the vehicle speed at the start of the deceleration operation (i.e., the vehicle speed at the start of the deceleration operation), and the vehicle speed at the end of the deceleration operation (i.e., the vehicle speed at the end of the deceleration operation), etc.

[0027] When the device DS acquires more than the data threshold Cth of stored data (driving condition index values) that can be considered to represent substantially the same starting position for deceleration operation, it generates a learned value based on that stored data set and learns (i.e., stores in non-volatile memory) the generated learned value in association with the vehicle subject to the deceleration request.

[0028] The learned values ​​include, for example, deceleration support start position information representing the deceleration support start position (latitude and route), deceleration support end position information representing the deceleration support end position (latitude and route), and deceleration support end speed information representing the deceleration support end speed, as shown in Figure 3(B). The deceleration support start position is the average position of the deceleration operation start positions in the set of stored data. The deceleration support end position is the average position of the deceleration operation end positions in the set of stored data. The deceleration support end speed is the average speed of the deceleration operation end speeds in the set of stored data. However, the learned values ​​only need to include the deceleration support start position information and are not limited to the above example; they just need to include parameters that can specify the content of the deceleration support (for example, target deceleration).

[0029] Subsequently, when the vehicle HV reaches the deceleration support start position represented by its learned deceleration support start position information, the device DS performs deceleration support to decelerate the vehicle HV at a constant deceleration rate so that the vehicle speed at the deceleration support end position matches the deceleration support end vehicle speed. Therefore, the learned value can be said to be a parameter that identifies the content of the deceleration support.

[0030] Incidentally, conventional devices do not associate learned values ​​with the target of deceleration requests. Therefore, if the driver accelerates after the vehicle HV has reached the learned deceleration support start point, the conventional device cannot immediately determine whether the acceleration is accidental or due to a change in the driving environment. Consequently, the conventional device discards the learned values ​​only when such acceleration is repeated a very large number of times and it is statistically confirmed that the driving environment has changed. As a result, the period before the learned values ​​are discarded becomes long, and unnecessary deceleration support may be provided during that period.

[0031] In response to this, when the vehicle HV reaches a learned deceleration support start point, if the device DS determines that the deceleration request target associated with that deceleration support start point is different from the deceleration request target recognized (identified) at that point, it immediately discards the learned values ​​associated with that deceleration support start point (i.e., deceleration support start position information, deceleration support end position information, and deceleration support end vehicle speed information, etc.). This reduces the frequency of unnecessary deceleration support.

[0032] (Specific operation) The CPU of the driver assistance ECU 10 executes the routines shown in Figures 4 to 6 at predetermined intervals.

[0033] At a predetermined timing, the CPU starts processing from S400 in Figure 4 and proceeds to S405 to determine whether or not the driver has performed a deceleration operation. A deceleration operation is either a brake operation that increases the brake pedal operation amount BP from "0", or a downshift operation that generates a signal Sdn on the second shift paddle 28.

[0034] If a deceleration operation occurs, the CPU proceeds from S405 to S410 to determine whether the average value Dav of the actual deceleration (acceleration G) during the period from the start to the end of the deceleration operation is greater than or equal to the threshold Dth. Alternatively, the CPU may determine in S410 whether the actual deceleration immediately after the start of the deceleration operation is greater than or equal to a predetermined threshold. The CPU may also omit the processing in S410. In this case, if the CPU determines "Yes" in S405, it proceeds to S415, which will be described later.

[0035] If the average value Dav is less than the threshold Dth, the CPU proceeds directly from S410 to S495 and terminates this routine. On the other hand, if the average value Dav is greater than or equal to the threshold Dth, the CPU proceeds from S410 to S415 and determines whether a learned value has been saved that has a position P that substantially corresponds to the position of the vehicle HV at the start of the current deceleration operation as the deceleration support start position. The position P that substantially corresponds to the position of the vehicle HV at the start of the current deceleration operation is a position within a predetermined distance (for example, 10m) from the position of the vehicle HV at the start of the current deceleration operation, and will be hereinafter referred to as the "corresponding position P". That is, in S415, the CPU determines whether learning for deceleration support for the corresponding position P has been completed by determining whether the value of the learning completion flag XG(P) corresponding to the corresponding point P is "1". If the value of the learning completion flag XG(P) is "1", the CPU proceeds from S415 to S495 and terminates this routine.

[0036] In contrast, if the value of the learning completion flag XG(P) is "0" (i.e., learning for deceleration assistance for the corresponding position P has not been completed), the CPU proceeds from S415 to S420 and determines whether one of the predetermined types of deceleration request targets has been recognized and identified based on the image data. If a deceleration request target has been recognized and identified, the CPU proceeds from S420 to S425 and determines whether the deceleration request target recognized at this moment is the same as the deceleration request target that was recognized when the vehicle HV previously passed a position that can be considered substantially the same as the corresponding position P (i.e., the previous deceleration request target).

[0037] If the vehicle currently recognized as requiring deceleration is the same as the vehicle requiring deceleration in the previous deceleration request, the CPU proceeds from S425 to S430 and stores the driving status index value in non-volatile memory in association with the vehicle requiring deceleration, as shown in Figure 3(A). Next, in S435, the CPU increases the value of the number of stored data C(P) for the corresponding point P by "1" and proceeds to S450.

[0038] In response, if the CPU determines "No" in either step S420 or S425, it proceeds from the step where it determined "No" to S440 and discards all driving status index values ​​stored in association with the corresponding point P. Next, the CPU proceeds to S445 and sets the value of the number of stored data C(P) to "0". After that, the CPU proceeds to S450.

[0039] In S450, the CPU determines whether the value of the number of stored data C(P) is greater than or equal to the data threshold Cth. That is, it determines whether a sufficient amount of data has been stored for learning to assist with deceleration at the corresponding position P. If the value of the number of stored data C(P) is less than the data threshold Cth, the CPU proceeds directly from S450 to S495 and terminates this routine.

[0040] In response to this, if the value of the number of stored data C(P) is greater than or equal to the data threshold Cth, the CPU proceeds from S450 to S455, generates a learned value (i.e., a parameter that can identify the content of deceleration support) based on the driving condition index value for the corresponding point P, and stores the generated learned value in non-volatile memory in association with the type of vehicle subject to deceleration request (see Figure 3(B)). Next, in S460, the CPU sets the value of the learning completion flag XG(P) to "1", and in S465, it sets the value of the number of stored data C(P) to "0". After that, the CPU proceeds to S495. In this way, the content of deceleration support is learned in association with the vehicle subject to deceleration request.

[0041] At a predetermined timing, the CPU starts processing from S500 in Figure 5 and proceeds to S510, where it determines whether the current position of the vehicle HV matches any of the deceleration support start positions P represented by the learned deceleration support start position information. If the current position of the vehicle HV does not match any of the learned deceleration support start positions, the CPU proceeds directly from S510 to S595 and terminates this routine. In other words, in this case, deceleration support is not performed.

[0042] In response to this, if the current position of the vehicle HV matches any of the "learned deceleration support start positions" P, the CPU proceeds from S510 to S520 and determines whether the driving environment corresponding to position P (i.e., the target of the deceleration request) has changed by checking whether the value of the environment change flag XCh(P) corresponding to position P is "1". The environment change flag XCh(P) is set to "1" when it is determined that the driving environment corresponding to position P (i.e., the target of the deceleration request) has changed, as will be described later (see S615 in Figure 6).

[0043] If the value of the environmental change flag XCh(P) is "0" (i.e., the target of the deceleration request has not changed), the CPU proceeds from S520 to S530 and executes deceleration support as described above based on the learned values ​​for the deceleration support start position P (i.e., deceleration support start position information, deceleration support end position information, and deceleration support end vehicle speed information, etc., which indicate the content of the learned deceleration support). After that, the CPU proceeds to S595.

[0044] In contrast, if the value of the environmental change flag XCh(P) is "1" (i.e., the target of the deceleration request has changed), the CPU proceeds from S520 to S540 and discards the learned values ​​for the deceleration support start point P (i.e., deceleration support start position information, deceleration support end position information, and deceleration support end vehicle speed information, etc.). After that, the CPU proceeds to S550 and sets the learning completion flag XG(P) to "0", and proceeds to S560 and sets the environmental change flag XCh(P) to "0". After that, the CPU proceeds to S595.

[0045] At a predetermined timing, the CPU starts processing from S600 in Figure 6 and proceeds to S605, where it determines whether the current position of the vehicle HV matches any of the deceleration support start positions P represented by the learned deceleration support start position information. If the current position of the vehicle HV does not match any of the learned deceleration support start positions, the CPU proceeds directly from S605 to S695 and terminates this routine.

[0046] In response to this, if the current position of the vehicle HV matches any of the learned deceleration support start positions P, the CPU proceeds from S605 to S610 and determines whether the driver has performed a specific operation to inform the system that the driving environment (i.e., the target of the deceleration request) has changed. The specific operation may be, for example, an operation in which either or both of the first shift paddle 27 and the second shift paddle 28 are continuously pressed for a predetermined time (e.g., 2 seconds) or longer. Alternatively, the specific operation may be, for example, an operation in which either or both of the first shift paddle 27 and the second shift paddle 28 are pressed again a predetermined number of times (e.g., 4 times) or more within a predetermined time (e.g., 3 seconds).

[0047] Furthermore, the specific operation used by the driver to inform the system of a change in the driving environment (the object requiring deceleration) is not limited to operations on the first shift paddle 27 and the second shift paddle 28. For example, if the device DS is connected to a specific switch, the device DS may treat an operation on that specific switch as the above-mentioned specific operation.

[0048] If such a specific operation is detected, the CPU proceeds from S610 to S615 and sets the value of the environment change flag XCh(P) for position P to "1". Next, the CPU proceeds to S620 and sets the value of the total score TSC(P) to "0", and proceeds to step 695.

[0049] If no specific operation is detected in S610, the CPU proceeds from S610 to S625 and determines whether a preceding vehicle is recognized based on the image data. If a preceding vehicle is recognized, the CPU proceeds from S625 to S630 and sets the value of the added score SC to "0". Then, the CPU proceeds to S635 and increases the value of the total score TSC(P) by the amount of the added score SC. In other words, the CPU updates the value of the total score TSC(P) by accumulating the added score SC. Note that if a preceding vehicle is recognized, the value of the added score SC is "0", so the total score TSC(P) does not increase. This is because if a preceding vehicle is recognized, it is difficult to determine whether the driving environment has changed based on the image data.

[0050] Next, the CPU proceeds to S640 to determine whether the total score TSC(P) is equal to or greater than the total score threshold TSCth. If the total score TSC(P) is not equal to or greater than the total score threshold TSCth, the CPU proceeds directly from S640 to S695.

[0051] In response to this, if the total score TSC(P) is equal to or greater than the total score threshold TSCth, the CPU determines that the driving environment at position P (i.e., the target of the deceleration request) has changed. Therefore, the CPU executes the "processing in S615 and S620" described above. After that, the CPU proceeds to S695.

[0052] When the CPU proceeds to S625, if no preceding vehicle is recognized, the CPU proceeds from S625 to S645 and determines whether the deceleration request target associated with the deceleration support start position P corresponding to the current position is different from the deceleration request target recognized and identified at the present time based on image data. The CPU also determines that the deceleration request target associated with the deceleration support start position P corresponding to the current position is different from the deceleration request target recognized and identified at the present time based on image data, even if no deceleration request target is recognized at the present time. In other words, even in this case, the CPU determines that the driving environment (i.e., the deceleration request target) at the deceleration support start position P has changed.

[0053] If the deceleration request target associated with the deceleration support start position P corresponding to the current position and the deceleration request target recognized and identified based on the image data at this point are the same (i.e., identical), the CPU proceeds directly from S645 to S695.

[0054] In contrast, if the deceleration request target associated with the deceleration support start position P corresponding to the current position differs from the deceleration request target recognized based on the image data at the current time, the CPU proceeds from S645 to S650 and determines the addition score SC based on the type of deceleration request target associated with the deceleration support start position P. For example, stop signs or level crossing signals are deceleration request targets that are recognized with relatively high accuracy based on the image data, so the addition score SC is set to a relatively high value A1. In contrast, curves are deceleration request targets that are often not accurately determined by the image data, so the addition score SC is set to a relatively low value A2 (i.e., a value A2 smaller than value A1). Thus, the addition score SC for deceleration request targets that are recognized with high accuracy based on the image data is set to a higher value than the addition score SC for deceleration request targets that are recognized with low accuracy based on the image data. After that, the CPU proceeds to processing from S635 onwards.

[0055] Therefore, the total score TSC(P), which increases as the likelihood of "the deceleration request target recognized at position P being different from the deceleration request target learned for position P" increases, becomes greater than or equal to the total score threshold TSCth, and it is determined that the driving environment (i.e., the deceleration request target) at position P has changed, and the value of the environment change flag XCh(P) is set to "1".

[0056] As explained above, when the driver performs a deceleration operation, the DS device identifies the target of the deceleration request and learns the details of the deceleration support based on the deceleration operation (deceleration support start position information, deceleration support end position information, and deceleration support end vehicle speed information, etc.) along with the target of the deceleration request. If the DS device determines that the target of the deceleration request at the already learned deceleration support start position has changed, it discards the learned value for that deceleration support start position. Therefore, deceleration support based on the deceleration request target before the change will not be performed immediately.

[0057] The present invention is not limited to the embodiments described above, and various modifications can be adopted within the scope of the present invention, as described below. For example, the present invention is applicable to a vehicle in which the driving mode has transitioned from autonomous driving to driver-operated driving in an autonomous vehicle.

[0058] The device DS may identify the deceleration target request based on the map information stored in the map information storage device 72. Furthermore, when the map information stored in the map information storage device 72 is updated, the device DS may detect that the target of the deceleration request has changed based on that map information and immediately discard the learned value corresponding to the position of the changed deceleration target request. [Explanation of Symbols]

[0059] 10...Driver assistance ECU, 21...Camera sensor, 26...Brake pedal operation amount sensor, 40...Brake actuator.

Claims

1. In a driver assistance system equipped with a controller configured to perform deceleration assistance that automatically slows down a vehicle, The aforementioned controller, Identify the object of the deceleration request that caused the deceleration operation, which is a driving operation performed by the driver of the vehicle to slow down the vehicle. The deceleration support start position information, which represents the position of the vehicle when the deceleration operation is initiated, is learned together with the deceleration request target identified for the deceleration operation, in association with the deceleration request target. After learning the deceleration support start position information, the vehicle will perform the deceleration support when it reaches the deceleration support start position represented by the learned deceleration support start position information. It is configured in such a way, Furthermore, the controller, After learning the deceleration support start position information, if it is determined that the deceleration request target, which has been learned in association with the deceleration support start position represented by the deceleration support start position information, has changed, the deceleration support start position information for the deceleration request target that has been determined to have changed is discarded. It is configured in such a way. Driving assistance system.

2. In the driving support device according to claim 1, The target of the deceleration request is, Located in front of the vehicle, including at least one of the following: a stop sign, a stop line, a curve, a level crossing signal, a gate at the entrance to a toll road, and a priority road that intersects with the vehicle's path. Driving assistance system.

3. In the driving support device according to claim 2, The aforementioned controller, The content of the deceleration support is determined based on the driving condition index value representing the driving condition of the vehicle when the deceleration operation occurred, and the content of the deceleration support is learned in association with the deceleration support start position information. The deceleration support is performed according to the learned deceleration support content. It is configured in such a way. Driving assistance system.

4. In the driving support device according to claim 3, The aforementioned controller, The vehicle to which the deceleration request is to be made is identified based on image data acquired by a camera mounted on the vehicle. When the vehicle reaches the deceleration support start position represented by the learned deceleration support start position information, if the deceleration request target identified based on the image data differs from the deceleration request target learned in association with the deceleration support start position reached by the vehicle, an additional score is determined based on the type of the learned deceleration request target. When the vehicle reaches the deceleration support start position, the total score is calculated by accumulating the determined addition scores. When the total score exceeds a predetermined score threshold, it is determined that the target of the deceleration request, which has been learned in association with the deceleration support start position reached by the vehicle, has changed. It is configured in such a way. Driving assistance system.

5. In the driving support device according to claim 4, The aforementioned controller, When the vehicle reaches the deceleration support start position represented by the learned deceleration support start position information, if the image data includes a vehicle located in front of the vehicle, the added score is set to zero. It is configured in such a way. Driving assistance system.

Images