JP7298483B2 - Driving support control device, driving support control program - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a driving support control device and a driving support control program for supporting driving of a vehicle when an abnormality occurs in a vehicle-mounted sensor in automatic driving.

In recent years, developments related to automatic driving of automobiles have been advanced. In particular, the development of self-driving vehicles, in which the on-board system automatically performs all acceleration, steering, and braking of the vehicle and the driver manually drives the vehicle only when requested by the system, is being actively developed.

Furthermore, along with the development of such self-driving vehicles, progress is also being made in the development of fail-operational technology to ensure the safety of vehicles in the event of an abnormality or failure in an in-vehicle system during self-driving.

Japanese Patent Application Laid-Open No. 2002-200001 describes an electronic control device that, when an abnormality occurs in a vehicle, changes the travel plan of the vehicle based on the cause of the abnormality and the surrounding conditions of the vehicle at the time of the occurrence of the abnormality.

In Patent Literature 1, the causes of abnormalities are classified into contact accidents, disasters other than contact accidents, and natural failures of in-vehicle systems, and an operation plan is set according to each. Measures are taken to stop the vehicle (hereinafter referred to as "stop") or to evacuate the vehicle from the location where the abnormality occurred and stop the vehicle at a safe position (hereinafter referred to as "retraction").

Japanese Patent Application Laid-Open No. 2019-8540

However, depending on the situation (for example, spontaneous failure of the in-vehicle system, etc.), even if it is possible to continue autonomous driving, it is unnecessarily biased toward safety and is limited to either stop or evacuation. there is

SUMMARY OF THE INVENTION It is an object of the present invention to obtain a driving support control device and a driving support control program that can find the possibility of continuation of automatic driving, not limited to stopping or retreating, when an on-vehicle sensor necessary for automatic driving malfunctions. .

A driving support control device according to the present invention is a driving support control device that supports driving when an abnormality occurs in a sensor necessary for automatic driving, and sets a destination candidate when an abnormality is received from a vehicle. Equipped with a setting unit and a travel hindrance factor management database storing a plurality of types of travel impediment factors, the occurrence frequency of the travel impediment factors when traveling from the current location where the abnormality of the sensor occurs to a candidate destination is determined. Based on an extraction unit that extracts each of a plurality of obstacles, past cases stored in a past case database, and the frequency of occurrence of driving obstacles extracted by the extraction unit, a predetermined limit is imposed during automatic driving. a calculation unit that calculates the probability of arrival at a candidate destination in a plurality of patterns of degenerate driving modes; a notification unit that notifies the plurality of patterns of degenerate driving modes and the arrival probabilities in each of the degenerate driving modes; have.

A driving support control device according to the present invention is characterized by operating a computer as each part of the driving support control device.

ADVANTAGE OF THE INVENTION According to this invention, the possibility of continuation of automatic driving|operation can be found not only in a stop or evacuation at the time of abnormality of the vehicle-mounted sensor required for automatic driving|operating.

1 is a schematic diagram of a driving support system including a driving support control device that supports driving in automatic driving of a vehicle according to the present embodiment; FIG. In the functional block diagram for executing the control for guiding the availability of the degenerate driving mode and the type of the degenerate driving mode executed between the automatic driving control device and the driving support control device according to the present embodiment be. FIG. 3 is a front view of a map around a point where an abnormality has occurred in a vehicle, as an example of destination candidate search in the setting unit shown in FIG. 2 ; FIG. 4 is a characteristic diagram showing the relationship between each hindrance factor and the occurrence frequency of the hindrance factor on a specific route; FIG. 4 is a front view of a map image showing a route from the same current position to the same destination candidate; FIG. 10 is a front view of a list showing arrival probabilities for each destination candidate for each degenerate driving mode k; FIG. 10 is a flow chart showing a procedure for extracting teacher driving data from the degree of similarity to the teacher driving data calculated by formula (3) for each future driving data and obtaining an evaluation; FIG. 8 is a diagram clearly showing that the evaluation of each supervised travel data by a calculation formula depends on the similarity of obstacles (corresponding to FIG. 7). It is an example of the image shown on the monitor provided in the user interface, (A) is a main screen, (B) is a sub-screen. 4 is a flow chart showing a data accumulation control routine executed by the driving support control device according to the present embodiment; 5 is a flow chart showing a driving support control routine that is started when an abnormality is received in the driving support control device according to the present embodiment.

FIG. 1 is a schematic diagram of a driving support system including a driving support control device 12 that supports driving in automatic driving of a vehicle 10 according to the present embodiment.

The vehicle 10 is equipped with a vehicle control device 14 and an automatic driving control device 20 .

The vehicle control device 14 executes control including a drive system (engine control, etc.) and an electrical system (failure diagnosis, etc. by state detection sensors of each part) while the vehicle 10 is running.

The vehicle control device 14 includes a group of cameras for photographing the surroundings of the vehicle 10 (in FIG. 1, as an example, a front camera 16A, a left front camera 16B, a left rear camera 16C, a right front camera 16D, and a right rear camera 16E). , and a rear camera 16F) are connected (generically referred to as a “camera group 16”). A radar group 18 including a plurality of millimeter wave radars and LIDAR is connected to the vehicle control device 14 .

The automatic driving control device 20 determines the driving operation to the destination based on the information necessary for automatic driving from the vehicle control device 14 (for example, the detection information from the camera group 16 and the radar group 18), and the vehicle control device 14.

The automatic driving control device 20 can communicate with the driving support control device 12 via a wireless communication device 22A of the network 22 .

In the driving support control device 12, the driving history information by the automatic driving from each vehicle 10 is aggregated, and the operator instructs the driving support as necessary. Therefore, the driving support control device 12 acquires real-time road information from the real-time map management system 24 via the network 22 .

The real-time map management system 24 aggregates information from infrastructure such as cameras installed on roads and road information transmission devices installed on vehicles 10, etc., and displays current road conditions (including road restrictions, on-street parking, etc.). It has a function to analyze

Here, when a part of the camera group 16 or the radar group 18 or the like malfunctions while the vehicle 10 is traveling by automatic driving, the driving support control device 12 instructs appropriate driving support to prevent an accident or the like. We need to avoid unforeseen circumstances.

At this time, conventionally, the vehicle 10 is immediately stopped (hereinafter referred to as a stop mode), or the vehicle 10 is evacuated from the site to a safe place and stopped (hereinafter referred to as an avoidance mode) to prevent an unexpected situation. was avoiding.

However, for example, depending on the situation of the failure that occurred, it may be possible to travel by automatic driving. Therefore, in the present embodiment, as in the stop mode or the avoidance mode, while avoiding unforeseen situations, automatic driving is continued under restrictions according to the abnormal situation (hereinafter referred to as degenerate running mode). I made it

FIG. 2 is a functional block diagram for executing control for guiding whether or not a degenerate driving mode can be executed between the automatic driving control device 20 and the driving support control device 12, and the type of the degenerate driving mode. be.

(Automatic driving control device 20)

A diagnostic code (CAN (Controller Area Network) signal) from the vehicle control device 14 is input to the abnormality information management unit 50 to manage abnormality information of the camera group 16 or the radar group 18 . The abnormality information includes current position information, current time information, diagnostic code information at the time of abnormality occurrence, and specific information of the abnormality sensor.

The abnormality information management unit 50 is connected to the abnormality information transmission unit 52A of the communication I/F 52, and transmits the acquired abnormality information to the abnormality information reception unit 56A of the communication I/F 56 of the driving support control device 12.

Communication I/F52 of automatic operation control device 20 is provided with automatic operation information receiving part 52B.

The automatic driving information receiving unit 52B receives the automatic driving information from the automatic driving information transmitting unit 56B of the communication I/F 56 of the driving support control device 12 .

The received automatic driving information is sent to the driving plan unit 54, and based on the driving plan planned by the driving planning unit 54, the drive system and the electric system of the vehicle control device 14 are controlled to perform automatic driving (stop mode , avoidance mode, and degenerate driving mode).

(Driving support control device 12)

As shown in FIG. 2 , the driving support control device 12 has a setting section 58 and acquires the abnormality information received by the abnormality information receiving section 56A of the communication I/F 56 .

The setting unit 58 has a function of searching for destination candidates in the reduced travel mode.

The setting unit 58 is connected to the destination candidate management database 60 . The destination candidate management database 60 stores in advance all locations (garages, stops, etc.) that are candidates for the destination of the vehicle 10 .

The setting unit 58 searches the destination candidate management database 60 for destination candidates that meet predetermined conditions based on the current position information in the abnormality information.

FIG. 3 schematically shows a front view of a map 62 around a point where an abnormality has occurred in the vehicle 10 as an example of destination candidate search in the setting unit 58 (see FIG. 2).

In FIG. 3, as a predetermined condition, destination candidates within a radius Rm (dotted line circle in FIG. 3) from the current location are selected. In FIG. 3, there are four destination candidates, and the ○△ garage 64, the XX bus stop 66, and the ◇□ bus stop 68 are selected, while the □■ bus stop 69 is excluded from the destination candidates.

As shown in FIG. 2, the setting unit 58 is connected to the extracting unit 70 and notifies the extracting unit 70 of information regarding all the selected destination candidates.

The extracting unit 70 has a function of calculating a driving impeding factor.

The extraction unit 70 is connected to an obstacle database 72 . The hindrance factor database 72 stores the frequency of occurrence for each of a plurality of preset types of hindrance factors, using destination candidates, current position, and current time as factors.

Obstacle factors are requirements necessary for traveling to a destination candidate, and include distance (factor A), number of lane changes including on-street parking (factor B), number of right turns (factor C), number of left turns (factor D), Presence or absence of traffic congestion (factor E), the number of pedestrian crossings (factor F), and the like. It should be noted that the types of inhibiting factors are not limited to factor A to factor F. For example, factor F (the number of pedestrian crossings) may be omitted in the case of expressways, and the number of schools in the case of school commuting routes may be used as a hindrance factor.

The extraction unit 70 acquires information from the hindrance factor database 72 based on the current position information and current time information in the abnormality information and the destination candidate information acquired from the setting unit 58, and extracts a plurality of destination candidates. Calculate multiple routes to and obstacles for each route.

FIG. 4 is a characteristic diagram showing the relationship between each hindrance factor (here, factor A to factor F) and the frequency of occurrence of the hindrance factor in a specific route.

In the characteristic example of FIG. 4, although the distance to the destination candidate is long, it can be predicted that the vehicle will travel relatively straight ahead.

On the other hand, as shown in FIG. 5, there may be two types of routes from the same current position S to the same destination candidate G. In FIG. In this case, the hindrance factor is calculated for each route.

In FIG. 5, the first route (see the dotted line in FIG. 5) has 0 lane changes, one right turn, and one left turn, and the second route (see the dashed line in FIG. 5) has , lane changes three times, right turns twice, and left turns twice. Even if the destination candidate is the same, the obstacles are different for each route. In addition, parking on the street is counted as a lane change, but it is a factor that can change depending on the time of day.

As shown in FIG. 2, the extraction unit 70 is connected to the derivation unit 74, and notifies the derivation unit 74 of the extracted blocking factors for each route.

The derivation unit 74 has a function of calculating the probability of arrival.

The derivation unit 74 is connected to the past case database 76 . In the past case database, situation information (route, obstruction factor, abnormality information (diagnostic code information at the time of abnormality occurrence, specific information of abnormal sensor) is used as a factor, result information (degenerate driving mode type, arrival probability, overriding image etc.) are stored as so-called big data.

Various types of degenerate driving modes are set based on the location and number of sensor abnormalities, and the arrival probability for each destination candidate is calculated for each type (details will be described later, see FIG. 6). .

Details of information processing (obtaining result information for situation information) in the derivation unit 74 will be described below.

"Calculation of features of obstructive factors"

In the following, for example, it is assumed that the diagnostic code of the failure is Z number. The feature quantity f _F of the obstruction factor of each route acquired from the extraction unit 70 is calculated by the following equation (1).

・・・（１）

... (1)

Here, N is a serial number that identifies the type of hindrance factor, and x is the numerical value of the frequency of the hindrance factor.

Next, assuming that one degenerate driving mode is k, the feature quantity fs,k,M of the hindrance factor of the teacher driving data stored in the past case database 76 in the degenerate driving mode k is given by the following equation (2): is expressed as

As shown in FIG. 6, there are a number of degenerate running modes, depending on the number and location of sensor failures. In FIG. was set as the identification number.

・・・（２）

... (2)

Here, k is the type of degenerate driving mode (k-th), and M is the number of samples. Each sample corresponds to the Z-number of the diagnostic code.

Next, the degree of similarity between each feature value fs,k,M and the teacher driving data stored in the past case database 76 is obtained.

The calculation for obtaining this degree of similarity, for example, compares each feature dimension and calculates the degree of similarity for the number m of teacher travel data. For example, by taking the reciprocal of the mean square error or the like, the similarity dk,m with respect to the travel data heading for m destination candidates in the teacher travel data can be obtained by the following equation (3).

・・・（３）

... (3)

FIG. 7 shows a threshold value set in advance (here, similarity=0.5) based on the degree of similarity to teacher traveling data calculated by equation (3) for each future traveling data for traveling to a specific destination candidate. 7) A procedure for extracting the above-mentioned supervised driving data and obtaining an evaluation for each of the extracted supervised driving data (O if degeneracy is possible, x if degeneracy is not possible) is shown.

FIG. 8 is similar to FIG. 7, but is a diagram clearly showing that the evaluation of each supervised travel data by a calculation formula depends on the similarity of impeding factors.

The left column 74A in FIG. 8 is, for example, the similarity calculated value by the reciprocal of the mean squared error of the equation (3). A central column 74B in FIG. 8 is a characteristic diagram of hindrance factors for each supervised travel data. The right column 74C in FIG. 8 shows the evaluation results (“◯” or “X” and whether or not there is an override image) in the teacher driving data with a similarity of 0.7 or more.

As shown in FIG. 8 (and FIG. 7), "○" or "X" (calculated to be classified as "1" or "0") is given to each teacher travel data. become.

In the present embodiment, it is assumed that the greater the number of teaching travel data marked with "○", the higher the probability of arrival, and the probability of arrival is calculated by the following equation (4).

・・・（４）

... (4)

FIG. 6 is a list showing the arrival probability for each destination candidate for each degenerate driving mode k.

Based on the list shown in FIG. 6, the derivation unit 74 shown in FIG. The override video (if any) is read and sent to notification manager 78 .

Here, in this embodiment, the notification manager 78 is connected to the user interface 80 . The setting unit 58 displays the derived result on a user interface 80 that is visible to the operator who manages the driving support control device 12 .

FIG. 9 is an example of an image displayed on a monitor 80M provided in the user interface 80. As shown in FIG.

FIG. 9A shows a main screen 80A. For example, in an illustration style, a vehicle image 10G in which sensor A has a Δ□ failure and destination candidate images as destination candidates set by the setting unit 58 are displayed. 64G, 66G, 68G, and route images 64R, 66R, 68R (arrows) from the vehicle image 10G to destination candidate images 64G, 66G, 68G are displayed.

Each destination candidate image 64G, 66G, 68G also displays a text image indicating the type of degenerate driving mode and the probability of arrival.

FIG. 9B shows a sub-screen 80B that can be displayed by switching from the main screen 80A. A sub-screen is prepared for each of the destination candidate images 64G, 66G, and 68G, and each sub-screen can be displayed by a switching operation. . Note that a plurality of sub-screens 80B may be displayed simultaneously.

As an example, FIG. 9(B) is a sub-screen 80B as detailed information when ○△ garage 64 is selected as a destination candidate. A formula image 82A is shown, and here, details such as how the probability of arrival became 98% are displayed.

・・・（５）

... (5)

Also, in the lower column of the sub-screen 80B in FIG. 9(B), an image (regardless of whether it is a still image or a moving image) at the time of past override and a situation image 82B at that time are displayed.

As shown in FIG. 2, the operator selects the most suitable destination candidate from among all the destination candidates, and notifies (replies) to the notification management unit 78 the information regarding the route and degenerate driving mode as operator instruction information. do.

The notification management unit 78 transmits the instruction information received from the operator to the automatic operation control device 20 via the automatic operation information transmission unit 56B of the communication I/F 56 .

In FIG. 9A, the destination candidate with the highest arrival probability (○△ garage, arrival probability 98%) may be simply selected. It is possible to select a destination other than the one with the highest probability of arrival (stop XX, probability of arrival 43%) for various reasons, such as being able to contact a repairman who can repair the vehicle, or being close to the final destination.

The operation of this embodiment will be described below with reference to the flow charts of FIGS. 10 and 11. FIG.

FIG. 10 is a flow chart showing a data accumulation control routine executed by the driving support control device 12 according to this embodiment.

At step 100, it is determined whether or not the destination has been determined, and if the determination is negative, the process proceeds to step 114 (described later). If the determination in step 100 is affirmative, that is, if it is determined that the destination has been determined, the process proceeds to step 102 .

At step 102, the obstacles to the destination are calculated, then the process proceeds to step 104 to set a simulated state of failure and malfunction of the vehicle 10, and the process proceeds to step 106 to switch to the degenerate running mode.

In the next step 108, running control in the degenerate running mode switched in step 106 is executed.

In the next step 110, after the degenerate driving mode ends (including arrival at the destination and stoppage of traveling), the quality of the degenerate driving mode is determined. For example, the pass/fail determination may be made based on whether or not an override has occurred. In addition, weighting may be applied to the overriding situation, and even if there is an overriding, the weighting is light, and if the destination is reached, the determination may be good.

In the next step 112 , the running data in the degenerate running mode this time is stored in the past example database 76 (see FIG. 2 ) as teacher running data (including pass/fail judgment), and the process proceeds to step 114 .

At step 114, it is determined whether or not to continue accumulating the teaching travel data. Also, if the determination in step 114 is negative, this routine ends.

FIG. 11 is a flow chart showing a driving support control routine that is activated when an abnormality (failure, malfunction) occurrence is received in the driving support control device 12 according to the present embodiment.

In step 150, the abnormality information is acquired from the vehicle 10, and then the process proceeds to step 152, in which destination candidates are selected from the destination candidate management database 60 (see FIG. 2) based on the current position in the abnormality information. Explore.

In the next step 154, obstacles to each destination candidate are calculated from the obstacles database 72 (see FIG. 2) based on the current position and current time in the abnormality information.

In the next step 156, based on the abnormality information, the degenerate driving mode and the arrival probability are calculated from the past case database 76 (see FIG. 2), and the process proceeds to step 158.

In step 158, an image at the time of overriding, which is one of the factors for the arrival probability not being 100%, is acquired, and then the process proceeds to step 160, and the result information is sent to the notification management section 78 (see FIG. 2). .

In step 162, the result information is displayed on the monitor 80A (see FIG. 9) by sending the result information to the user interface 80 (see FIG. 2).

Here, for example, the operator determines whether degeneracy running is possible and notifies the notification management unit 78 of it.

Here, if degenerate driving is not possible, a driving support stop instruction to stop or evacuate the vehicle 10 is notified, and if degenerate driving is possible, detailed information (destination, driving support information including driving route and degenerate driving mode).

In the next step 164 , the notification manager 78 determines whether or not instruction information has been received from the user interface 80 .

If an affirmative determination is made in step 164, the process proceeds to step 166 to determine whether or not the command is for driving support in the degenerate driving mode.

If a negative determination is made in step 166, the routine proceeds to step 168, where the vehicle 10 is instructed to stop or evacuate without executing the degeneracy mode, and this routine ends.

Further, when the determination in step 166 is affirmative, the process proceeds to step 170, and based on the instruction information, the driving support information including the destination, the driving route, and the degenerate driving mode is transmitted to the automatic driving control device 20 of the vehicle 10. and go to step 172 .

At step 172, the result of determination as to whether or not the degenerate running mode can be executed, obstructive factors, etc. are stored in the past case database 76 (see FIG. 2), and this routine ends.

In the present embodiment, the automatic driving control device 20 of the vehicle 10 is notified of a plurality of destination candidates with their respective arrival probabilities (see FIG. 9). After the selection, only one of the destination candidates may be preferentially notified based on a predetermined condition.

For example, if the predetermined condition is "arrival probability priority", only the destination candidate with the highest arrival probability is notified, and if the predetermined condition is "distance priority", only the closest destination candidate is notified. Notice. The predetermined condition may be set on the automatic driving control device 20 side of the vehicle 10 or may be set on the driving support control device 12 side.

In addition, in the present embodiment, the route, destination, degenerate driving mode, etc. are notified with emphasis on the operator's judgment result. Based on the received information, the optimum information may be selected by machine learning and notified. Also, the AI function may be provided on the side of the automatic driving control device 20 of the vehicle 10 .

In the present embodiment, evaluation is performed based on the teacher driving data based on the presence or absence of overriding, but other than overriding may be used as long as it is possible to determine whether the driving is good or bad. For example, when the distance to a wall or an obstacle does not become extremely small during running, and when it does, it is evaluated as "O" and "X", respectively.

10 vehicle, 12 driving support control device, 14 vehicle control device, 20 automatic driving control device, 16 camera group, 16A front camera, 16B left front camera, 16C left rear camera, 16D right front camera, 16E right rear camera , 16F rear camera, 18 radar group, 50 abnormality information management unit, 52 communication I / F, 52A abnormality information transmission unit, 56 communication I / F, 56A abnormality information reception unit, 52B automatic operation information reception unit, 56B automatic operation Information transmission unit 58 Setting unit 60 Destination candidate management database 62 Map 64 ◯△ garage 66 XX bus stop 68 ◇□ bus stop 69 □■ bus stop 70 Extraction unit 72 Impediment factor database 74 Derivation unit , 76 past case database, 78 notification management unit, 80 user interface, 80M monitor, 10G vehicle image, 64G, 66G, 68G destination candidate image, 64R, 66R, 68R route image, 80B sub-screen

Claims (10)

A driving support control device (10) for supporting driving when an abnormality occurs in a sensor necessary for automatic driving,
a setting unit (58) for setting destination candidates when an abnormality is received from the vehicle;
A running impediment factor management database is provided in which types of a plurality of travel impediment factors are stored, and the frequency of occurrence of the travel impediment factor when traveling from the current location where the abnormality of the sensor occurs to a candidate destination is determined by the plurality of impediment factors. an extraction unit (70) for extracting each
Based on the past cases stored in the past case database and the frequency of occurrence of driving impediment factors extracted by the extraction unit, multiple patterns of degenerate driving mode candidates with predetermined restrictions during automatic driving Purpose a calculation unit (74) for calculating the probability of arrival to the ground;
a notification unit (78) that notifies the multiple patterns of degenerate driving modes and the arrival probability in each of the degenerate driving modes;
A driving support control device having 2. The driving support control device according to claim 1, wherein the notification destination of said notification unit is a user interface (80) provided in an operation system that manages driving of a vehicle in which an abnormality has occurred in said sensor. 3. The driving support control device according to claim 2, wherein the user interface notifies a driving failure case in which the arrival probability of the degenerate driving mode cannot reach 100%. A plurality of patterns of degenerate driving modes notified to the user interface and information on the running state of the vehicle after occurrence of an abnormality determined based on the probability of arrival in each degenerate driving mode are added as past cases to the past case database. The driving support control device according to claim 2 or 3. 2. The driving support control device according to claim 1, wherein a destination of said notification is an automatic driving control device (20) mounted on a vehicle in which an abnormality has occurred in said sensor. In the past case database, the sensor type in which the abnormality occurred and the abnormality information are stored as past cases in association with each other, and each time an abnormality is received from the vehicle, the past case is added as a past case. The driving support control device according to any one of the above. 7. The calculation unit according to any one of claims 1 to 6, wherein when there are a plurality of travel routes to the destination, the calculation unit calculates the probability of arrival in the degenerate travel mode for each travel route. driving support control device. The driving support control according to any one of claims 1 to 7, wherein at least a degenerate driving mode with a high arrival probability is preferentially selected from among a plurality of degenerate driving modes as a recommended degenerate driving mode and notified. Device. The driving support control device according to any one of claims 1 to 8, wherein the calculating section calculates the probability of arrival in consideration of real-time road information. the computer,
Operate as each part of the driving support control device according to any one of claims 1 to 9,
Driving support control program.

Images