WO2021229671A1 - Travel assistance device and travel assistance method - Google Patents

Abstract

The present disclosure relates to a travel assistance device for assisting the travel of a mobile object. The travel assistance device determines a travel difficulty representing the difficulty of travel of the mobile object in at least three levels, on the basis of mobile object information including at least the location, speed, and orientation of the mobile object acquired by a mobile object sensor mounted on the mobile object, and sensing information including at least the location and speed of an obstacle around the mobile object acquired by a surroundings recognition sensor mounted on the mobile object, and sets assistance information, which is used for assisting the travel of the mobile object in accordance with the determined travel difficulty.

Description

Driving support device and driving support method

The present disclosure relates to a driving support device for an automobile, and particularly to a driving support device having a reduced processing load.

In recent years, based on cameras mounted on automobiles and peripheral monitoring sensors such as millimeter waves, autonomous automatic braking such as collision avoidance braking that autonomously avoids or reduces danger and adaptive cruise control (ACC) that runs at a constant speed Driving systems are beginning to spread. In addition, using wireless communication, information is exchanged with peripheral vehicles and roadside vehicles, and information that cannot be detected by the peripheral monitoring sensor mounted on the vehicle and real-time dynamic map information such as accidents and regulations possessed by the infrastructure are acquired. Therefore, studies are being conducted on a road-vehicle cooperative automatic driving system that realizes the sophistication of automatic driving.

In the automatic driving system, it is not possible to properly select autonomous driving or cooperative automatic driving according to the surrounding situation, and a method of selecting cooperative automatic driving only when information can be obtained by wireless communication is being considered. There is.

For example, Patent Document 1 describes a technique for switching the implementation, non-execution, and automation level of autonomous driving according to the traffic volume of the autonomous driving vehicle.

Japanese Unexamined Patent Publication No. 2017-151041

In Patent Document 1, the implementation, non-execution, or switching of the automation level of autonomous driving is switched according to the traffic volume of autonomous driving, but it depends on the degree of congestion of the autonomous driving vehicle, and even in a situation where automatic driving is difficult. , Self-driving vehicles need to make their own judgments and drive. In addition, since the amount of information used for determining automatic driving uses information that can be acquired by the own vehicle and communication, the processing load may increase.

In addition, in general, as the number of autonomous driving vehicles increases, the number of communication vehicles increases, and the amount of information to be communicated increases, the amount of information to be processed in order to realize driving support increases, and there is a problem that the processing load increases.

The present disclosure has been made to solve the above-mentioned problems, and an object of the present disclosure is to provide a driving support device having a reduced processing load for realizing driving support.

The present disclosure relates to a travel support device that supports the travel of a mobile body, and includes moving body information including at least the position, speed, and orientation of the moving body acquired by a moving body sensor mounted on the moving body, and the movement. Based on the sensing information including at least the position and speed of obstacles around the moving body acquired by the peripheral recognition sensor mounted on the body, at least the running difficulty level indicating the difficulty of running the moving body is determined. Judgment is made in three or more stages, and the support information used for the running support of the moving body is set according to the determined running difficulty level.

According to the driving support device according to the present disclosure, the support information used for the driving support of the moving object is set according to the driving difficulty level, so that the information to be processed is reduced in the situation where the driving difficulty level is low, and the safety and comfort are achieved. It is possible to reduce the processing load while ensuring the performance, and in a situation where the driving difficulty is low, it is possible to increase the information to be processed and realize efficient processing while ensuring safety and comfort.

It is a figure explaining the structure of the driving support system which concerns on Embodiment 1. FIG. It is a block diagram which shows the structure of the server which concerns on Embodiment 1. FIG. It is a figure which shows the example of the support target and support information for a support level. It is a figure which shows an example of the support level management table. It is a figure which shows an example of the state transition of a support level. It is a figure which shows an example which manages the switching of a support level by a tree structure. It is a block diagram which shows the structure of the moving body which concerns on Embodiment 1. FIG. It is a flowchart which shows the whole processing of the server which concerns on Embodiment 1. It is a figure which shows an example of the support level determination condition in the server which concerns on Embodiment 1. FIG. It is a figure explaining an example of the determination of the driving difficulty level and the determination of a support level. It is a figure explaining an example of the determination of the driving difficulty level and the determination of a support level. It is a flowchart which shows the transmission process of the moving body which concerns on Embodiment 1. It is a flowchart which shows the reception process of the moving body which concerns on Embodiment 1. It is a block diagram which shows the structure of the server which concerns on Embodiment 2. It is a block diagram which shows the structure of the moving body which concerns on Embodiment 2. It is a block diagram which shows the structure of the server which concerns on Embodiment 3. FIG. It is a figure which shows the application example of the group determination in the server which concerns on Embodiment 3. FIG. It is a flowchart which shows the whole processing of the server which concerns on Embodiment 3. It is a block diagram which shows the structure of the server which concerns on Embodiment 4. FIG. It is a block diagram which shows the structure of the server which concerns on Embodiment 5.

<Embodiment 1>
<Structure>
FIG. 1 is a diagram illustrating a configuration of a traveling support system 1000 according to the first embodiment. It should be noted that the reference numerals of the respective configurations in FIG. 1 are the same or the same reference numerals are given to the configurations corresponding to the same in the other drawings, and duplicate description will be omitted. The configuration of the travel support system 1000 is also common to the embodiments 2 to 5, and will be described as the travel support systems 2000 to 5000, respectively.

The travel support system 1000 shown in FIG. 1 includes a server 101, a roadside communication device 102, a mobile body 103, and a roadside sensor 104. The travel support system 1000 is composed of one or more units, the server 101 is connected to a plurality of roadside communication devices 102, and the roadside communication device 102 is connected to a plurality of mobile bodies 103 and a plurality of roadside sensors 104. .. In the first embodiment, the moving body 103 will be described assuming a vehicle. The server 101 and the roadside communication device 102 may be connected via the Internet network.

The server 101 may be integrated with the roadside communication device 102 and the roadside sensor 104. Further, in the first embodiment, the server 101 is provided separately from the mobile body 103 and will be described as a travel support device for supporting the travel of the mobile body 103, but the server 101 is mounted on one of the mobile body 103. It may have been done.

By providing the server 101 separately from the mobile body 103, there are few restrictions on the size of the server 101, and it is easy to increase the processing capacity of the server 101.

FIG. 2 is a block diagram showing the configuration of the server 101 of the driving support system 1000 according to the first embodiment. As shown in FIG. 2, the server 101 includes a processor 100, a communication interface 10, a communication device 30, and a storage device 40. Since the server 101 is a computer server and is realized by an edge server, a cloud server, or the like, the server 101 may be displayed as an edge server or a cloud server.

The processor 100 communicates with the communication device 30 via the communication interface 10 and also acquires information such as map information from the storage device 40.

The processor 100 is composed of an IC (Integrated Circuit) for executing a process described in a program to execute processes such as data transfer, calculation, processing, control, and management, and functions by executing the instruction. Each function of the surrounding situation recognition unit 1, the driving difficulty level determination unit 2, the support level determination unit 3, and the recommended action generation unit 4 represented as blocks is realized.

The processor 100 has an arithmetic circuit, a register for storing instructions and information, and a cache memory. A processor such as a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or a GPU (Graphics Processing Unit) is applied to the processor 100.

The support level management table 5 and the map data 6 are stored in the storage device 40, which will be described later.

The storage device 40 includes RAM (RandomAccessMemory), ROM (ReadOnlyMemory), flash memory, EPROM (ErasableProgrammableReadOnlyMemory), EEPROM (ElectricallyErasableProgrammableReadOnlyMemory), SD (SecureDigital: registered trademark), Non-volatile or volatile semiconductor memories such as memory cards, CF (Compact Flash) memories, and NAND flash memories can be applied. In addition, portable storage media such as HDD (Hard Disk Drive), SSD (Solid State Drive), magnetic disk, flexible disk, Blu-ray (registered trademark) disk, DVD (Digital Versatile Disc), optical disk, compact disk, and mini disk can be used. May be applied.

The communication device 30 is a device including a receiver for receiving data from the mobile body 103 or the roadside sensor 104 via the roadside communication device 102 and a transmitter for transmitting the data.

For example, a communication chip or a NIC (Network Interface Card) can be applied to the communication interface 10.

Further, the communication interface 10 can use DSRC (Dedicated Short Range Communication) dedicated to vehicle communication and a communication protocol such as IEEE802.11p.

Further, the communication interface 10 may use a communication line such as LTE (Long Term Evolution: registered trademark) or a 5th generation mobile communication system (5G).

Further, the communication interface 10 may use a wireless LAN such as Bluetooth (registered trademark) or IEEE802.11a / b / g / n / ac.

Next, the processor 100 will be further described with reference to FIGS. 1 and 2. The peripheral situation recognition unit 1 determines the position, speed, direction, traveling route, driver state, driver viewpoint, number of occupants, occupant state, etc. detected by the moving body sensor mounted on the moving body 103 via the communication interface 10. Get moving object information. In addition, sensing information such as the position, speed, attributes, detection accuracy, video, and point cloud of obstacles around the moving body 103 detected by the peripheral recognition sensor mounted on the moving body 103, and the moving body 103 automatically operates. Acquires the travel path information composed of the travel locus and the speed indicating the action plan to be generated in order to perform the above.

Further, the peripheral situation recognition unit 1 includes sensing information such as the position, speed, attribute, and accuracy of obstacles around the roadside sensor 104 detected by the roadside sensor 104, and the roadside machine such as the position and installation angle of the roadside sensor 104. Get information. The roadside unit equipped with the roadside sensor 104 is also treated as a moving body, and the sensing information is integrated.

The traveling difficulty determination unit 2 aggregates the information of the moving object 103 based on the moving object information, the sensing information, and the traveling path information acquired by the surrounding situation recognition unit 1, and the map data 6 stored in the storage device 40. Integrate with map information obtained from. Further, the traveling difficulty determination unit 2 analyzes the traveling situation and the traveling scenario in consideration of the surrounding situation of each moving body based on the moving body information and the sensing information, and is based on the conflict index between the moving bodies. Calculate the risk of surrounding conditions, including the possibility of contact with surrounding moving objects. Further, the traveling difficulty determination unit 2 analyzes the traffic condition from the number of moving bodies 103, the relative distance, the relative speed, and the like.

Here, the driving difficulty level indicates the difficulty of driving in manual driving by the driver or automatic driving by the automatic driving system, and the vehicle condition, the road environment, the weather, and the automatic driving system existing around the moving body 103. It changes depending on the configuration, sensor performance, etc. The driving difficulty level changes in stages and is defined in at least three stages, and is classified into, for example, easy, normal, and difficult.

The driving difficulty level is different from the automatic driving level defined by SAE (Society of Automotive Engineers) International. The automatic driving level is level 0 operated by the driver, level 1 where the system supports steering, acceleration and deceleration, level 2 where the system supports steering and acceleration and deceleration, and the system in a specific place. There are level 3 where the driver operates everything in an emergency, level 4 where the system operates everything in a specific place, and level 5 where the system operates everything regardless of the location.

The driving difficulty determination unit 2 can calculate the driving difficulty based on the driving situation, the driving scenario, the risk of the surrounding situation, and the traffic situation.

In addition, since driving situations, driving scenarios, etc. are taken into consideration in order to calculate (determine) the difficulty level, it is possible to accurately estimate the driving difficulty level of the vehicle to be supported.

The driving difficulty may be calculated by a rule-based algorithm determined from a table prepared in advance or a conditional expression, may be calculated as a cost function from each parameter, or may be calculated based on a probability model. However, it may be calculated using artificial intelligence such as machine learning or decision tree.

In the first embodiment, a rule-based method of calculating the cost from the driving situations prepared in advance and determining the support level from the total cost will be described. The calculation of the driving difficulty level will be further described later.

The support level determination unit 3 determines the support level based on the driving difficulty calculated by the driving difficulty determination unit 2.

The support level determination unit 3 controls by manual driving or control control as the driving difficulty level increases, and provides support by cooperative automatic driving that reduces the amount of information to be processed as the driving difficulty level decreases. In addition, if the driving difficulty is extremely high or if only manual driving is supported, it is judged that the support is provided by manual driving. The method of determining the support level from the driving difficulty level can be determined based on the support level management table 5 stored in the storage device 40. The support level determination unit 3 notifies the recommended action generation unit 4 of the determined support level.

The support level is different from the automatic driving level defined by SAE International. The support level is roughly divided into autonomous automatic driving in which the moving body 103 autonomously and automatically travels, cooperative automatic driving in which the moving body 103 automatically travels in cooperation with other moving bodies 103, and control by the server 101. It can be classified into four stages: control control for automatic driving and manual operation.

Cooperative autonomous driving can be classified into three stages when vehicle information, sensing information, and driving path information are used, depending on the type of information to be supported. In addition, control control can be classified into two stages: recommended behavior and vehicle control.

By setting multiple levels as the support level, if the difficulty level is low, it can be judged by a method with a low processing load. If the number of connected cars and self-driving vehicles increases in the future, the amount of information handled will increase dramatically, so it is effective to perform the minimum necessary data processing.

To switch the support level, you can use the conflict index, the degree of duplication of driving paths, the degree of congestion, and other traffic conditions and vehicle action plans. This makes it possible to increase the amount of information handled in situations where driving is difficult and to continue driving in automatic driving.

The recommended action generation unit 4 generates support information according to the support level based on the support level determined by the support level determination unit 3, the moving object information, the sensing information, and the travel path information acquired by the surrounding situation recognition unit 1. Then, it is transmitted to the mobile body 103 via the communication interface 10.

Here, the support information includes, for example, support information for manual driving, support information for cooperative automatic driving, and support information for control control. Autonomous automatic driving does not exist as support information because it determines driving by the moving body alone, but the recommended action generation unit 4 notifies that it is determined to be autonomous automatic driving.

Support information for manual driving includes contact warnings with surrounding moving objects, recommended speeds, recommended lanes, recommended lane change timings, recommended right / left turn timings, and emergency vehicle approach warnings. In addition, the support information for manual driving includes additional information such as the generation point of the warning information and the effective time. Assistance information for manual driving is information displayed to the driver.

As support information for cooperative autonomous driving, in addition to moving object information, sensing information, and driving path information, there is recommended behavior information such as recommended speed, recommended driving lane, recommended lane change timing, and recommended right / left turn timing.

The recommended action is the action to be taken next between the moving bodies in anticipation and consideration of the action between the moving bodies. It is an action to make the vehicle run smoothly, such as decelerating and changing the lane of the other vehicle.

The support information for control control is the vehicle control information itself that controls the vehicle, such as the traveling speed and the traveling path information. By transmitting vehicle control information to the moving body 103 to control the moving body 103, control control or remote control of the moving body 103 can be performed. The vehicle control information determines the traveling speed and the traveling lane for the smooth flow of the traffic flow, and by controlling the surrounding vehicles, the lane change and the right / left turn can be efficiently realized. For example, by setting the speeds of peripheral vehicles to the same speed, unnecessary acceleration and deceleration can be avoided, and the inter-vehicle distance can be adjusted to smoothly change lanes.

Figure 3 shows an example of support targets and support information for support levels. As shown in FIG. 3, in the autonomous automatic driving, since only the in-vehicle sensor of the mobile body 103 to be distributed is used, there is no support information to be distributed from the server.

In cooperative automatic driving that utilizes information on other vehicles (first cooperative automatic driving). Among the mobile information collected by the server, the mobile information (information on other vehicles in the vicinity) of the mobile 103 existing in the vicinity of the mobile 103 to be distributed is distributed as support information.

In the cooperative automatic driving (second cooperative automatic driving) that utilizes the sensor information of other vehicles, among the sensing information collected by the server, the sensing information of the moving body 103 existing in the vicinity of the moving body 103 to be distributed (the sensing information of the moving body 103). Distributes sensing information of surrounding vehicles).

In cooperative automatic driving (third cooperative automatic driving) that utilizes the path information of another vehicle, among the path information collected by the server, the traveling path of the moving body 103 existing around the moving body 103 to be distributed. Information (pass information of surrounding vehicles) is distributed as support information.

In the control control (first control control) that utilizes the recommended behavior information, the server 101 considers the driving scenario of the mobile body 103 to be supported and the driving scenario of the surrounding moving body, and the recommended speed, recommended lane, and recommended lane. Deliver recommended action information such as change timing as support information.

In control control (second control control) that utilizes vehicle control information, the server 101 distributes vehicle control information that controls traffic by controlling not only the mobile body 103 to be supported but also all surrounding mobile bodies as support information. do.

In manual operation, the server distributes warning information and recommended action information to the driver of the mobile 103 to be supported as support information.

The support target for control control utilizing vehicle control information is the vehicle control unit of the mobile body 103 to be supported, and the support target for manual driving is the display unit that can be visually recognized by the driver of the mobile body 103 to be supported. The support target at the other support level is the automatic driving determination unit of the mobile body 103 to be supported.

The support level management table 5 is composed of, for example, a support level and a table that defines a driving difficulty level and a cost as shown in FIG. As shown in FIG. 3, the support level can be broadly classified into autonomous automatic driving, cooperative automatic driving, control control, and manual driving, and cooperative automatic driving can be further classified into three stages according to the information to be supported. Control control can be classified into two stages of distributing recommended behavior information and vehicle control information. As the cooperative automatic driving, the case where the other vehicle information, the other vehicle sensor information, and the other vehicle path information are distributed to the moving body 103 is shown.

As shown in FIG. 4, the driving difficulty levels are autonomous automatic driving, cooperative automatic driving that distributes other vehicle information, cooperative automatic driving that distributes other vehicle sensing information, and cooperative type that distributes other vehicle driving path information. Driving difficulty levels 1 to 7 are specified in the order of automatic driving, control control for distributing recommended action information, control control for distributing vehicle control information, and manual driving.

As explained above, the cost is prepared in advance according to the driving difficulty level. For example, in the autonomous automatic driving with the driving difficulty level 1, the cost (c) is less than 5, and the driving difficulty level is 2 cooperative type. 5 or more and less than 10 for automatic driving, 10 or more and less than 15 for cooperative automatic driving with driving difficulty 3, 15 or more and less than 20 for cooperative automatic driving with driving difficulty 4, and 20 or more and less than 25 for control control with driving difficulty 5. In the control control of driving difficulty level 6, it is 25 or more and less than 30, and in the control control of driving difficulty level 7, it is 30 or more.

Although FIG. 4 shows an example of switching the driving difficulty level in 7 stages, it may be switched in at least 3 stages of autonomous automatic driving, cooperative automatic driving, and control control, and it may be further subdivided into 8 or more stages. You may switch with.

FIG. 4 shows an example in which the support level management table is composed of a table, but the present invention is not limited to this, and for example, as shown in FIG. 5, the support level is switched by a finite state machine (Finite State Machine). May be good.

FIG. 5 is a state transition diagram showing switching of support levels in a finite state machine. In FIG. 5, arrows indicate an increase in driving difficulty (Up) and a decrease in driving difficulty (Down) between the four states of autonomous automatic driving, cooperative automatic driving, control control, and manual driving. .. Further, in the cooperative automatic driving, the driving difficulty level Up and the driving difficulty level Down are shown between the three states of the cooperative automatic driving utilizing the vehicle information, the sensing information, and the traveling path information. Further, in the control control, the driving difficulty level Up and the driving difficulty level Down are shown between the two states of the recommended action and the vehicle control.

Although the state is changed according to the change in the driving difficulty level in FIG. 5, the state is not limited to this, and the transition may be made based on a specific driving situation or driving scenario.

Furthermore, switching of support levels may be managed by a tree structure such as the decision tree shown in FIG. As shown in FIG. 6, in contrast to the four states of autonomous automatic driving, cooperative automatic driving, control control, and manual driving, cooperative automatic driving utilizes vehicle information, sensing information, and travel path information. The three states of autonomous driving are branched, and the control control shows a tree-like classification in which the two states of recommended behavior and vehicle control are branched.

In addition to the driving difficulty and cost, the support level management table may be defined based on the risk of the surrounding situation and the action plan. In addition, the driving difficulty and cost, and the judgment conditions of the risk of the surrounding situation may be updated by machine learning or deep learning.

Map data 6 has map information related to the map. Map information is composed of a plurality of maps corresponding to a predetermined scale in a layered manner, and constitutes road information which is information about a road, lane information which is information about lanes constituting a road, and lanes. Contains constituent line information, which is information about the constituent lines to be used.

Road information includes, for example, road shape, road latitude, longitude, road curvature, road slope, road identifier, number of road lanes and road line types, as well as general roads, highways and priority roads. Contains information about road attributes.

The lane information includes, for example, the identifier of the lane constituting the road, the latitude, longitude and the center line of the lane.

The constituent line information includes the identifier of each line constituting the lane, the latitude and longitude of each line constituting the lane, and the line type and curvature of each line constituting the lane.

Road information is managed for each road, and lane information and constituent line information are managed for each lane. Map information is used for navigation, driving support, automatic driving, and the like.

In addition, road information includes traffic regulation information (lane regulation, speed regulation, traffic regulation, chain regulation, etc.) that changes with time, tollgate regulation information (entrance / exit, tollgate), and traffic congestion information (presence / absence of congestion, section, lane). ), Traffic accident information (stopped vehicle, low speed vehicle), obstacle information (falling object, animal), road abnormality information (road damage, road surface abnormality), peripheral vehicle information, etc. may be defined as a dynamic map.

FIG. 7 is a block diagram showing the configuration of the moving body 103 of the traveling support system 1000 according to the first embodiment. As shown in FIG. 7, the mobile body 103 includes a mobile body system 300 having a processor 200, a communication interface 20, a peripheral recognition interface 21, a vehicle sensor interface 22, and a vehicle control interface 23, a peripheral recognition sensor 31, and a mobile sensor 32. It includes a vehicle control ECU 33 and a communication device 34. The moving body 103 also includes an actuator for automatic operation and the like, but a configuration that is not closely related to the embodiment is omitted.

The processor 200 communicates with the peripheral recognition sensor 31, the mobile sensor 32, the vehicle control ECU 33, and the communication device 34 via the communication interface 20, the peripheral recognition interface 21, the vehicle sensor interface 22, and the vehicle control interface 23, and acquires information. do.

Similar to the processor 100 mounted on the server 101 shown in FIG. 2, the processor 200 executes instructions described in the program to execute processes such as data transfer, calculation, processing, control, and management. By executing commands, which are composed of ICs, the functions of the peripheral situation recognition unit 41, the automatic driving judgment unit 42, the vehicle control unit 43, and the support level switching unit 44, which are represented as functional blocks, are realized, and the display unit 45 is realized. Notify the driver of driving support information via.

The peripheral situation recognition unit 41 has the same function as the peripheral situation recognition unit 1 mounted on the server 101, and has sensing information of obstacles around the moving body 103 detected by the peripheral recognition sensor 31 and a moving body sensor. The moving body information detected by the 32, the moving body information of the peripheral moving body received by the communication device 34, and the sensing information are acquired via the vehicle sensor interface 22, the peripheral recognition interface 21, and the communication interface 20, respectively, and are internally stored in advance. Recognize the travelable area by integrating with the map information managed in.

Based on the moving object information, sensing information, and map information integrated by the surrounding situation recognition unit 41, the automatic driving determination unit 42 maintains a lane and changes lanes to drive safely without contacting obstacles and surrounding vehicles. , Acceleration, deceleration, etc. Further, the automatic driving determination unit 42 generates vehicle control information of the traveling path and speed for the vehicle control unit 43 to control the steering, the accelerator, and the brake, and notifies the vehicle control unit 43.

The vehicle control unit 43 controls the accelerator, brake, and steering of the moving body 103 via the vehicle control interface 23 according to the vehicle control information notified from the automatic driving determination unit 42.

The support level switching unit 44 receives support information from the server 101, determines the type of the received support information, notifies the display unit 45 if the support information is support information for manual driving, and the support information is cooperatively automatic. In the case of support information for driving, the automatic driving determination unit 42 is notified, and in the case of the support information is vehicle control information for control control, the vehicle control unit 43 is notified. The support information for cooperative automatic driving is notified to the surrounding situation recognition unit 41 instead of the automatic driving judgment unit 42, and after the peripheral situation recognition unit 41 integrates all the moving body information and the sensing information, the automatic driving is performed. It may be configured to notify the determination unit 42.

The display unit 45 displays the support information for manual driving notified from the support level switching unit 44, and notifies the driver. As the display unit 45, a display such as a car navigation system, a head-up display, an AR (augmented reality) system, a display panel such as a cockpit may be used, or a voice may be combined to notify the driver. Further, the display unit 45 may change the display position according to the position of the support information. Further, the display unit 45 may display only when the driver does not recognize the driving support information.

The communication interface 20 is a device including a receiver that receives positioning data from a peripheral mobile body and a base station, detects the arrival angle of radio waves and the time required for transmission / reception, and a transmitter that transmits data.

For example, a communication chip or NIC can be applied to the communication interface 20. Further, the communication interface 20 can use a DSRC dedicated to vehicle communication and a communication protocol such as IEEE802.11p.

Further, the communication interface 20 may use a mobile phone network such as LTE (registered trademark) or a 5th generation mobile communication system (5G).

Further, the communication interface 20 may use a wireless LAN such as Bluetooth (registered trademark) or IEEE802.11a / b / g / n.

Further, the communication device 34 is a communication device compatible with DSRC, LTE, and 5G, and is configured to give the received information to the mobile system 300 in FIG. 7, but the information from the mobile system 300 is sent to the outside. It also has a function to send.

The peripheral recognition interface 21 is an interface for acquiring data from the peripheral recognition sensor 31 mounted on the mobile body 103. Specific examples are sensor data acquisition LSI (Large Scale Integration), USB (Universal Serial Bus), and CAN (Controller Area Network) ports.

The peripheral recognition sensor 31 is a sensor capable of positioning such as millimeter wave radar, monocular camera, stereo camera, LiDAR (Light Detection and Ringing, Laser Imaging Detection and Ringing), sonar, GPS (Global Positioning System) and the like. The peripheral recognition sensor 31 also includes a DMS (Driver Monitoring System) and a drive recorder that monitor the driver inside the mobile body 103.

The vehicle sensor interface 22 is a device for connecting a moving body sensor 32 such as a GPS, a speed sensor, an acceleration sensor, and an orientation sensor to the processor 200. Further, the vehicle sensor interface 22 is a device for connecting an in-vehicle device such as an in-vehicle ECU, an EPS (Electric Power Steering), a car navigation system, and a cockpit to the processor 200. As a specific example, the vehicle sensor interface 22 is a sensor ECU (Electronic Control Unit).

The vehicle control interface 23 is a device for connecting the processor 200 to the vehicle control ECU 33 that controls the accelerator, brake, and steering.

The server 101 or the mobile body 103 of the travel support system may be mounted in an integrated form or an inseparable form with other components shown in the figure, or may be in a removable form or a separable form. It may be implemented.

Further, in FIGS. 2 and 7, only one processor 100 and 200 are shown, respectively. However, the number of processors 100 and 200 may be plural, and the plurality of processors 100 and 200 and programs that realize each function may be executed in cooperation with each other.

<Operation>
Hereinafter, the operation of the traveling support system 1000 according to the first embodiment will be described with reference to FIGS. 8 to 13. The operation of the travel support system 1000 of the first embodiment also includes the operation of the travel support device and the travel support method.

First, the overall operation of the server 101 in the driving support system 1000 according to the first embodiment will be described with reference to the flowchart shown in FIG.

The peripheral situation recognition unit 1 of the server 101 acquires the moving body information, the sensing information, and the traveling path information from the plurality of moving bodies 103 via the roadside machine, for example, in a cycle of 100 milliseconds (step S101).

In step S101, the information type acquired from the moving body 103 is changed, the acquisition cycle is changed, or the acquisition route is changed based on the driving difficulty level determined in step S103 or the support level determined in step S104. do. For example, when the driving difficulty level or the support level is low, only the moving object information is acquired, and when the traveling difficulty level or the support level is high, the moving object information, the sensing information, and the traveling path information are acquired. Further, for example, when the driving difficulty level or the support level is low, the information is acquired in a cycle of 200 milliseconds, and when the driving difficulty level or the support level is high, the information is acquired in a cycle of 20 milliseconds.

Although an example in which the server 101 acquires information from the mobile body 103 is shown here, the information is notified from the mobile body 103 to the server 101 after the server 101 notifies the mobile body 103 of the traveling difficulty level. You may do so. Further, the roadside sensor information may be received from the roadside machine and processed in combination with the information of the moving body 103.

Further, when the driving difficulty level or the support level is low, sensing information such as the relative distance and the relative speed primary processed by the moving body 103 is acquired, and when the traveling difficulty level or the support level is high, the periphery of the moving body 103 is acquired. The raw data such as the image and the point cloud acquired by the recognition sensor are acquired.

The driving difficulty determination unit 2 of the server 101 calculates the driving situation, the driving scenario, the risk of the surrounding situation, and the degree of congestion based on the received mobile body information, sensing information, and driving path information (step S102).

When the driving situation is based on a certain vehicle (moving body), the number and frequency of lane changes of vehicles existing around the vehicle, the relative distance and relative speed of the vehicle from the front, rear, left and right vehicles, or the surroundings are used. It is defined by situation information with surrounding vehicles such as acceleration of existing vehicles, amount of change in deceleration, degree of overlap of driving paths, and road situation information such as weather, accident occurrence section, and traffic jam section. It should be noted that a certain vehicle here is all the moving bodies 103 managed by the server 101, and the calculation is performed for each vehicle.

The driving scenario shows the behavior of the vehicle such as lane change, right turn, left turn, merging, branching, temporary stop, signal stop, and driving, and is judged from the route information, driving path information, and map information of the moving body information.

The risk of the surrounding situation is, as a conflict index, the time required for two vehicles to make contact when traveling in the same direction at the current speed TTC (Time-to-Collision), the time for a vehicle to enter the contact point. PET (Post Encroachment Time), which indicates the difference between the time when another vehicle reaches this point, and DRAC, which defines the minimum required deceleration rate at which the following vehicle must decelerate in order to avoid contact with the preceding vehicle. Judgment is made using a value such as (Deceleration Rate to Avoid a Crash).

The degree of congestion is determined based on the number of vehicles passing within a certain time and the inter-vehicle time specified by the passing time interval of vehicles for a certain point. In addition, a certain point here is each point where a vehicle exists, and is any arbitrary place.

The driving difficulty determination unit 2 of the server 101 determines the driving difficulty based on the driving situation, the driving scenario, the risk of the surrounding situation, and the congestion degree (step S103).

For example, here, the total cost is calculated based on the driving situation, the driving scenario, the risk of the surrounding situation, and the degree of congestion, and the driving difficulty is determined from the total cost. As a result, the determination of the running difficulty level can be determined numerically.

FIG. 9 shows an example of the cost for the driving situation, the driving scenario, the risk of the surrounding situation, and the degree of congestion. In addition, in FIG. 9, in addition to the above-exemplified driving situations, driver information such as configuration and performance information of various sensors, driver status, driver crisis recognition, and driver line of sight is also included. Sensor configuration and performance information, for example, represents sensor performance in high, medium, and low, driver state is represented by whether the driver is awake or sleeping, and driver crisis awareness is a driver's danger. It is represented by whether it is recognized (yes) or not aware of danger (no), and the driver's line of sight is represented by whether the direction of the driver's line of sight is normal or looking aside.

In FIG. 9, the cost is set in three stages of 0, 1, and 2. When the driver is awake, the cost is set to the lowest (0) so that the driving difficulty is lowered, and the driver is in a sleeping state. In such a case, the cost is set to the highest (2) so that the driving difficulty level increases. The number of floors of the cost may be less than or greater than the three levels. The driving difficulty may be determined by summing up the costs of all items, or may be determined by the cost of at least one item.

Further, in addition to the information shown in FIG. 9, the driving difficulty determination unit 2 may determine the driving difficulty according to the restrictions of the server 101 in the driving support system or the automatic driving system and the restrictions of the in-vehicle system.

Here, the restrictions of the server 101 are, for example, the processing performance of the server 101, the number of vehicles being processed, and the like, and if the processing performance is low or the number of processed vehicles is large, the driving difficulty is increased. The restrictions of the in-vehicle system are, for example, the detection accuracy of the peripheral recognition sensor 31 mounted on the moving body 103 and the recognizable distance of the peripheral recognition sensor 31, and it is difficult to drive when the sensing distance is short or the accuracy is low. The degree will be high.

Further, the conditions and costs shown in FIG. 9 may be updated by learning the judgment conditions by machine learning and deep learning.

Further, when the traveling difficulty level of a certain moving body 103 is low, the next determination timing is delayed or the determination cycle is lengthened, and when the traveling difficulty level is high, the next determination timing is advanced or the determination cycle is set. shorten. Further, when the driving difficulty level is low, the range of information used for the determination is narrowed, and when the driving difficulty level is high, the range of information used for the determination is widened.

Furthermore, the driving difficulty may be weighted based on the results determined in the past, or may be determined based on the results within a certain period of time.

Although this running difficulty level is determined for the moving body 103, the running difficulty level may be assigned to the position where the moving body 103 exists.

The support level determination unit 3 of the server 101 determines the support level according to the driving difficulty level notified from the driving difficulty level determination unit 2 and the support level management table (step S104). For example, in the support level management table shown in FIG. 4, when the cost is 12, the driving difficulty level is 3, and the support level indicates cooperative automatic driving (other vehicle sensor information). Here, the driving difficulty level is determined based on the support level management table, but the determination may be made using a cost function or a probability model instead of the table.

Similar to the driving difficulty level, the support level may be weighted based on the results determined in the past, or may be determined based on the results within a certain period of time.

Further, although the support level is determined for the moving body 103, the support level may be assigned to the position where the moving body 103 exists.

The support level determination unit 3 of the server 101 requests the recommended action generation unit 4 to generate support information based on the determined support level (step S105). Examples of support levels, support targets, and support information are as described with reference to FIG.

In the example of FIG. 3, another vehicle information, another vehicle sensor information, and another vehicle path information are distributed separately, but when the other vehicle sensor information is distributed, the other vehicle information is integrated. , When distributing the other vehicle path information, the other vehicle information and the other vehicle sensor information may be integrated and distributed.

Further, instead of the support level, the recommended action generation unit 4 may be requested to generate support information based on the driving difficulty level. The generation cycle of the support information may be changed based on the support level or the driving difficulty level.

Here, an example of determining the driving difficulty level and determining the support level will be described with reference to FIGS. 10 and 11.

For example, as shown in FIG. 10, when the road is not congested and the relative distance between the vehicle VA to be supported and the surrounding vehicles is large, the cost is small, so that the driving difficulty is low and the vehicle VA is autonomous. Type automatic operation is recommended. On the other hand, as shown in FIG. 11, when there are many peripheral vehicles and the vehicles are close to each other and the vehicle VA and the vehicle VB running in parallel are trying to change lanes, the cost becomes high and the driving difficulty level is high. Is high, and control control that utilizes vehicle control information is recommended. In this control control, the server 101 controls not only the vehicle VA to be supported but also the vehicle VB. In the example of FIG. 11, the vehicle VA is changed to the lane first, and the vehicle VB is changed to the lane later.

The recommended action generation unit 4 of the server 101 distributes the generated support information to the mobile body 103 to be supported. The generated support information may be distributed not only to the mobile body 103 to be supported but also to the mobile body 103 in the vicinity of the support target.

If the driving difficulty level or the support level is associated with the position where the moving body 103 exists, the information may be distributed to the point.

The processes of steps S101 to S106 described above may be sequentially processed each time data is received, or may be processed at regular time intervals.

Further, the traveling difficulty level in steps S102 and S103 may be determined by the moving body, the traveling difficulty level may be notified from the moving body 103 to the server 101, and steps S104 and subsequent steps may be performed on the server 101.

Next, the operation of transmission and reception by the mobile body 103 in the travel support system 1000 according to the first embodiment will be described with reference to FIGS. 12 and 13. Specifically, FIG. 12 shows a processing sequence when information is transmitted from the mobile body 103 to the server 101, and FIG. 13 shows a processing sequence when the mobile body 103 receives information from the server 101. ..

First, a processing sequence when information is transmitted from the mobile body 103 to the server 101 will be described using the flowchart shown in FIG. As shown in FIG. 12, the peripheral situational awareness unit 41 of the mobile body 103 acquires the mobile body information from the mobile body sensor 32 and acquires the sensing (sensor) information from the peripheral recognition sensor 31 (step S201).

The moving body information is moving body information such as the position, speed, direction, acceleration, and traveling path of the moving body 103. Further, the sensing information is information such as the relative position, relative speed, relative angle, and type of the detected object of other moving objects and obstacles existing in the vicinity of the moving object 103 as a starting point. Further, the sensing information may handle raw data such as a video before processing and a point cloud.

The peripheral situation awareness unit 41 of the mobile body 103 transmits the acquired mobile body information and sensing information to the server 101 (step S202). Since step S202 is a process corresponding to the process of step S101 in the peripheral situation awareness unit 1 of the server 101, the information may be transmitted after receiving the information acquisition request from the server 101.

Further, although the information is transmitted to the server 101 in the above, the mobile body information and the sensing information may also be transmitted to the peripheral mobile body. This is a configuration corresponding to the case where the peripheral mobile unit has a function of executing the processing of the server 101.

The automatic driving judgment unit 42 of the moving body 103 creates an action plan for automatic driving such as lane keeping and lane change based on the moving body information and sensing information acquired by the surrounding situation recognition unit 41, and realizes the action plan. (Step S203).

Here, in addition to the moving body information and sensing information that can be acquired by the vehicle alone, map information may be used. Further, here, an example of making a judgment without using the support information notified from the server 101 is shown.

In the creation of an action plan for autonomous driving, behavior judgment is made by observing traffic rules, driving judgment along the road shape, state estimation of surrounding vehicles and risk estimation, and rule-based method, optimization method, probabilistic method, Action plans can be created using learning-based methods and methods that integrate these methods.

The rule-based method is a method that makes decisions based on defined conditions and rules, such as FiniteStateMachine and DecisionTree, and is a robust process for simple scenarios where it is easy to handle defined situations. Is possible.

The Optimization Based method is a method that defines a function for data and finds the minimum or maximum value of the function, like CostBasedFunction, and is easy to implement, maintain, and test.

The probabilistic based method is a method of probabilistically judging the result by inputting each condition, such as the Markov decision process (Partially Observable Markov Decision Process: POMDP) and Bayesian estimation, and formulates a reliable scenario. It is possible.

The learning based method is a method of learning from past data and correctly predicting new inputs, such as reinforcement learning and machine learning, and can handle unknown situations that cannot be defined.

Also, as an example of a method that integrates multiple methods, there is a method that combines a rule-based method and a learning-based method, and the parts that do not require learning are dealt with on a rule-based basis, and operation is guaranteed by restrictions based on the rules. The test is easy and the scope of responsibility can be divided.

In addition, the generation of driving path information for autonomous driving can be realized by using model predictive control (Model Predictive Control), route planning algorithm (Rapidly exploring random tree), and the like. Model predictive control is a control method that optimizes while predicting future responses at each time, and route planning algorithm is a method that randomly searches for a route and repeats the search until the goal is reached. ..

The automatic driving determination unit 42 of the mobile body 103 transmits the generated travel path information to the server 101 (step S204). The travel path information may also be transmitted to surrounding moving objects. This is a configuration corresponding to the case where the peripheral mobile unit has a function of executing the processing of the server 101.

The vehicle control unit 43 of the mobile body 103 controls the travel of the mobile body 103 (vehicle) according to the travel path information generated by the automatic driving determination unit 42 (step S205). Although an example in which the automatic driving determination unit 42 controls the vehicle is shown here, the manual driving may be maintained when the driver is manually driving.

Note that the sequence shown in FIG. 12 is a sequence in which data is transmitted to the server 101, so processing when data is received from the server 101 is not considered.

Next, using the flowchart shown in FIG. 13, a processing sequence when the mobile body 103 receives information from the server 101 will be described.

The support level switching unit 44 of the mobile body 103 waits until the support information is received from the server 101 (step S301), and when the support information is received, the process proceeds to step S302.

In step S302, the support level switching unit 44 determines the support target of the received support information. If the support target is the vehicle control unit 43, the process proceeds to step S303. If the support target is the automatic driving determination unit 42, the process proceeds to step S305, and if the support target is the display unit 45, the process proceeds to step S308.

In step S303, the support level switching unit 44 notifies the vehicle control unit 43 of the vehicle control information received as the support information.

The vehicle control unit 43 performs vehicle control via the vehicle control interface 23 according to the notified vehicle control information (step S304). After that, the process of step S301 and the like is repeated.

Although the vehicle control unit 43 uses the notified vehicle control information, the vehicle is notified by determining safety and comfort by comparing with the vehicle control information generated by the automatic driving determination unit 42. It may be determined whether or not to use the control information, and if the vehicle control information generated by the automatic driving judgment unit 42 is superior, the vehicle is autonomous based on the vehicle control information generated by the automatic driving judgment unit 42. Carry out automatic driving by.

In step S305, the support level switching unit 44 notifies the automatic driving determination unit 42 of the automatic driving support information received as the support information. Although the automatic driving support information is notified to the automatic driving determination unit 42 here, it may be notified to the surrounding situation recognition unit 41.

The automatic driving determination unit 42 generates a traveling path by using the notified automatic driving support information in addition to the moving body information and the sensing information, and notifies the vehicle control unit 43 (step S306).

When other vehicle information or other vehicle sensing information is received as support information, the peripheral recognition sensor detects it by treating it as moving object information existing in the vicinity, similar to the peripheral vehicle or obstacle detected by the peripheral recognition sensor 31. Information that cannot be used can also be used.

Further, when the other vehicle path information is received as the support information, the traveling path of the moving body 103 existing in the vicinity can be recognized, so that the safety can be enhanced by the judgment considering the behavior of the surrounding moving body.

The vehicle control unit 43 performs vehicle control via the vehicle control interface 23 according to the travel path information notified from the automatic driving determination unit 42 (step S307). After that, the process of step S301 and the like is repeated.

In step S308, the support level switching unit 44 notifies the display unit 45 of the alarm information and the recommended action information received as the support information.

The display unit 45 displays the support information and notifies the driver (step S309). After that, the process of step S301 and the like is repeated.

<Effect>
Since the amount of information used for automatic driving changes based on the driving difficulty level of the driving support system 1000 of the first embodiment described above, the amount of information to be processed is reduced in a situation where driving is easy to improve safety and comfort. It is possible to reduce the processing load while ensuring, and in situations where driving is difficult, it is possible to increase the information to be processed and realize efficient processing while ensuring safety and comfort.

Also. Since the amount of information to be processed is changed according to the difficulty of driving, it is not necessary to perform unnecessarily complicated processing, the processing load can be reduced, and the processing speed can be increased.

Further, since the information type to be used in the server 101 is determined based on the driving difficulty level instead of determining the information type to be used in each mobile body 103, the efficiency in consideration of the distant situation that cannot be grasped by the mobile body alone. Support can be realized.

In addition, the amount of communication can be changed by changing the type of information acquired from the mobile body 103, the frequency of acquiring information, or the communication route to be acquired, depending on the driving difficulty level or the support level. It is possible to reduce the processing load.

<Modification example>
In the travel support system 1000 of the first embodiment described above, the server 101 determines the travel difficulty level and determines the information type to be provided to the mobile body 103, but the mobile body 103 determines the travel difficulty level. Then, the configuration may be such that the server is requested to provide information.

<Embodiment 2>
<Structure>
Next, the second embodiment will be described with reference to FIGS. 14 and 15. In the driving support system 2000 (FIG. 1) according to the second embodiment, the same reference numerals are given to the same parts as the driving support system 1000 of the first embodiment, and detailed duplication will be omitted.

FIG. 14 is a block diagram showing the configuration of the server 101 of the driving support system 2000 according to the second embodiment. As shown in FIG. 14, the server 101 includes a processor 100A, a communication interface 10, a communication device 30, and a storage device 40.

FIG. 15 is a block diagram showing the configuration of the moving body 103 of the traveling support system 2000 according to the second embodiment. As shown in FIG. 15, the mobile body 103 includes a mobile body system 300 having a processor 200A, a communication interface 20, a peripheral recognition interface 21, a vehicle sensor interface 22, and a vehicle control interface 23, a peripheral recognition sensor 31, and a mobile sensor 32. It includes a vehicle control ECU 33 and a communication device 34.

The processor 200A communicates with the peripheral recognition sensor 31, the mobile sensor 32, the vehicle control ECU 33, and the communication device 34 via the communication interface 20, the peripheral recognition interface 21, the vehicle sensor interface 22, and the vehicle control interface 23, and acquires information. do.

The processor 200A has a server communication unit 46 and a vehicle control unit 43 as functional blocks. The server communication unit 46 receives the sensing information of obstacles around the moving body 103 detected by the peripheral recognition sensor 31 and the moving body information detected by the moving body sensor 32, respectively, in the vehicle sensor interface 22 and the peripheral recognition interface 21. The acquired information is notified to the vehicle control unit 43 and transmitted to the server 101 via the communication interface 20.

In the mobile body 103 of the second embodiment, the server communication unit 46 uses the server 101 to transmit the sensing information of obstacles around the mobile body 103 detected by the peripheral recognition sensor 31 and the mobile body information detected by the mobile body sensor 32. Is sent periodically to.

In the driving support system 2000 of the second embodiment, the processor 100A of the server 101 determines the automatic driving and transmits the vehicle control information to the moving body 103.

That is, as shown in FIG. 14, the processor 100A of the server 101 has, in addition to the configuration of the processor 100 shown in FIG. 2, an automatic operation determination unit 7 that receives the output of the recommended action generation unit 4, and has a recommended action. The generation unit 4 notifies the automatic driving determination unit 7 of the support information generated according to the support level.

The automatic driving determination unit 7 uses the automatic driving support information notified from the recommended action generation unit 4 in addition to the sensing information and the moving body information of obstacles around the moving body 103 transmitted from the moving body 103. Generates vehicle control information such as a travel path. Then, the generated vehicle control information is transmitted to the mobile body 103 via the communication interface 10.

Then, the server communication unit 46 of the moving body 103 notifies the vehicle control unit 43 of the vehicle control information received from the server 101, and the vehicle control unit 43 passes through the vehicle control interface 23 according to the notified vehicle control information. The accelerator, brake, and steering of the moving body 103 are controlled to perform automatic traveling.

<Effect>
In the driving support system 2000 of the second embodiment described above, since the server 101 generates vehicle control information for automatic driving based on the driving difficulty level, the information to be processed is reduced in a situation where driving is easy, and safety and comfort are achieved. It is possible to reduce the processing load while ensuring the performance, and in situations where driving is difficult, it is possible to increase the information to be processed and realize efficient processing while ensuring safety and comfort.

Further, since the server 101 generates vehicle control information for automatic driving and remotely controls the moving body 103, the system configuration mounted on the moving body 103 can be simplified.

<Embodiment 3>
<Structure>
Next, the third embodiment will be described with reference to FIGS. 16 to 18. In the driving support system 3000 according to the third embodiment, the same reference numerals are given to the same parts as the driving support system 1000 of the first embodiment, and detailed description of duplication will be omitted.

FIG. 16 is a block diagram showing the configuration of the server 101 of the driving support system 3000 according to the third embodiment. As shown in FIG. 16, the server 101 includes a processor 100B, a communication interface 10, a communication device 30, and a storage device 40. As the configuration of the moving body 103, the configuration shown in FIG. 7 may be adopted.

The processor 100B of the server 101 has a group determination unit 13 connected to the support level determination unit 3 in addition to the configuration of the processor 100 shown in FIG. 2, and the group determination unit 13 is based on the mobile body information and the sensing information. , A plurality of moving bodies 103 moving in the same direction are determined as a group. As a method for determining the group, the determination is made based on the inter-vehicle time, the relative distance, and the relative speed between the moving bodies.

In the traveling support system 3000 of the third embodiment, the group determination unit 13 of the processor 100B determines a plurality of moving objects moving in the same direction as one group, and the information of the determined moving body group is used as a support level determination unit. Notify 3. The support level determination unit 3 applies the same support level to the same group based on the information of the mobile group.

FIG. 17 shows an example in which a plurality of moving objects are determined as a group and the same support level is applied. FIG. 17 shows a state in which a plurality of vehicles including the vehicle VA to be supported are about to turn right at an intersection, and the group determination unit 13 determines that the plurality of vehicles are vehicle group G1 and vehicles in the group. Apply the same level of support to. In addition, there are multiple vehicles trying to turn right at the intersection in the opposite lane across the intersection, and the multiple vehicles are judged to be vehicle group G2, and the same support level is applied to the vehicles in the group. do.

Further, when the light color and cycle of the signal at the intersection can be acquired by the roadside sensor 104 (FIG. 1) or the moving body sensor 32 mounted on the moving body 103, when the light color changes from red to blue, and when the arrow signal. In that case, the control control is used to control the vehicle group to start at the same time, and when the color changes from blue to red, the control control is used to simultaneously decelerate the vehicle group.

<Operation>
Next, the overall operation of the server 101 in the driving support system 3000 according to the third embodiment will be described with reference to the flowchart shown in FIG. Since steps S401 to S404 and step S407 shown in FIG. 18 are the same as steps S101 to S104 and step S106 in the flowchart shown in FIG. 8, the description thereof will be omitted.

After the support level determination unit 3 determines the support level in step S404, if there are a plurality of moving bodies 103 that the group determination unit 13 moves in the same direction, they are determined as one group, and the information of the moving body group is determined. Is notified to the support level determination unit 3 (step S405).

The support level determination unit 3 extracts the support level of the mobile body 103 having the highest running difficulty in the group based on the information of the mobile body group notified from the group determination unit 13, and the mobile body 103 of the same group. Requests the recommended action generation unit 4 to generate support information based on the same driving difficulty and support level (step S406).

When notifying recommended behavior information and vehicle control information by matching the support level within the same group, smooth the traffic flow by matching the traveling speed and lane change timing of each moving object in the same group. can do.

The peripheral situation recognition unit 1 of the server 101 acquires the mobile body information, the sensing information, and the travel path information from the plurality of mobile bodies 103 via the roadside unit (step S401), and the group determination unit 13 By associating these information with the information of the determined mobile body group, each mobile body in the mobile body group can be specified, and the information can be distributed to each mobile body via the communication interface 10.

Further, since each moving object is notified of the traveling difficulty level or the support level from the server 101, it is possible to know its own traveling difficulty level or the support level, and the traveling of a plurality of moving objects 103 included in one group. When the difficulty level or the support level is low, one mobile body 103 in the group collects information on the mobile body 103 and sends it to the server 101, and a plurality of mobile bodies included in one group. If the traveling difficulty level or the support level of 103 is high, each moving object in the group may transmit to the server.

As shown in FIG. 5, the state transition diagram showing the switching of the support level in the third embodiment may change the state according to the driving difficulty level, or cooperates from the autonomous automatic driving by forming and leaving the group. The transition from type automatic operation and cooperative automatic operation to control control may be performed.

<Effect>
The traveling support system 3000 of the third embodiment described above can smoothly travel the moving bodies 103 in the same group by treating a plurality of moving bodies 103 moving in the same direction as a group.

<Embodiment 4>
<Structure>
Next, the fourth embodiment will be described with reference to FIG. In the driving support system 4000 according to the fourth embodiment, the same reference numerals are given to the same parts as the driving support system 1000 of the first embodiment, and the overlapping detailed description will be omitted.

FIG. 19 is a block diagram showing the configuration of the server 101 of the driving support system 4000 according to the fourth embodiment. As shown in FIG. 19, the server 101 includes a processor 100C, a communication interface 10, a communication device 30, and a storage device 40A. As the configuration of the moving body 103, the configuration shown in FIG. 7 may be adopted.

In addition to the configuration of the processor 100 shown in FIG. 2, the processor 100C of the server 101 has a driving difficulty learning unit 14 connected to the traveling difficulty determination unit 2, and the traveling difficulty learning unit 14 is used for moving body information. Based on the sensing information, the travel path information, and the travel difficulty determined by the travel difficulty determination unit 2, the travel difficulty determination model 15 is generated and stored in the storage device 40A.

The driving difficulty determination unit 2 reads the driving difficulty determination model 15 from the storage device 40A, and determines the driving difficulty using the driving difficulty determination model 15.

By learning the driving difficulty level on the server 101 to generate the driving difficulty level determination model 15 and updating the driving difficulty level determination model 15 stored in the storage device 40A, the driving difficulty level determination can be improved and improved. can do.

The driving difficulty determination model 15 determines and evaluates the input data and sets the difficulty level. This includes finite state machines, decision trees and logistic regression analysis.

In the generation of the driving difficulty determination model 15, a sample of input data is prepared, and data analysis, model creation and model evaluation are repeatedly carried out. If the result of model evaluation is bad, the cause is analyzed and the model is corrected. For example, the cost is calculated based on the conditions shown in FIG. 9, a model for determining the driving difficulty level is generated based on the cost range shown in FIG. 4, and if the generated model is not valid, the conditions in FIG. 9 are generated. And modify the cost range in Figure 4.

The driving difficulty learning unit 14 may generate the driving difficulty determination model 15 based on the moving body information, the sensing information, the traveling path information, the driving situation, the driving scenario, the risk of the surrounding situation, and the traffic situation. ..

<Effect>
The driving support system 4000 of the fourth embodiment described above learns the driving difficulty level on the server 101, generates a driving difficulty level determination model 15, updates it, and can feed back to the next determination. It is possible to improve the judgment accuracy and speed up the processing.

<Embodiment 5>
<Structure>
Next, the fifth embodiment will be described with reference to FIG. In the driving support system 5000 according to the fifth embodiment, the same reference numerals are given to the same parts as the driving support system 1000 of the first embodiment, and detailed description of duplication will be omitted.

FIG. 20 is a block diagram showing the configuration of the moving body 103 of the traveling support system 4000 according to the fifth embodiment. As shown in FIG. 20, the mobile body 103 includes a storage device 40, and the processor 200A has a running difficulty level determination unit 2, a support level determination unit 3, and a recommended action generation unit 4.

The driving difficulty determination unit 2, the support level determination unit 3, and the recommended action generation unit 4 have the same functions as the server 101 of the first embodiment shown in FIG. That is, the traveling difficulty determination unit 2 aggregates the information of the moving object 103 based on the moving object information, the sensing information, and the traveling path information acquired by the peripheral situation recognition unit 1, and the map stored in the storage device 40. It is integrated with the map information acquired from the data 6. The support level determination unit 3 determines the support level based on the driving difficulty calculated by the driving difficulty determination unit 2.

The recommended action generation unit 4 has support information according to the support level based on the support level determined by the support level determination unit 3, the moving object information, the sensing information, and the traveling path information acquired by the surrounding situation recognition unit 1. Is generated, and if the support information is support information for manual driving, the display unit 45 is notified, and if the support information is support information for cooperative automatic driving, the automatic driving judgment unit 42 is notified, and the support information is for control control. In the case of the vehicle control information of the above, the vehicle control unit 43 is notified.

<Effect>
The driving support system 5000 of the fifth embodiment described above determines the driving difficulty level in the moving body 103 and determines the support level, thereby determining which information is used for automatic driving. The load can be expected to be reduced. For example, when autonomous automatic driving or cooperative automatic driving using vehicle information is selected, the processing load can be reduced because it is not necessary to process the sensor information and the path information received from the surrounding moving body.

Although the present disclosure has been described in detail, the above description is exemplary in all aspects and the present disclosure is not limited thereto. It is understood that a myriad of variants not illustrated can be envisioned without departing from the scope of the present disclosure.

It should be noted that, within the scope of the disclosure, each embodiment can be freely combined, and each embodiment can be appropriately modified or omitted.

Claims (14)

1. It is a running support device that supports the running of moving objects.
   The driving support device is
   The moving body information including at least the position, velocity, and direction of the moving body acquired by the moving body sensor mounted on the moving body, and
   Based on the sensing information including at least the position and speed of obstacles around the moving body acquired by the peripheral recognition sensor mounted on the moving body.
   A running support device that determines the running difficulty level indicating the running difficulty of the moving body in at least three stages, and sets support information used for running support of the moving body according to the determined running difficulty level. ..
2. The driving support device is
   The driving support device according to claim 1, wherein the type of information acquired from the moving body information and the sensing information is changed according to the driving difficulty level.
3. The driving support device is
   The traveling support device according to claim 1, wherein the frequency of acquiring the moving body information and the sensing information is changed according to the traveling difficulty level.
4. The driving support device is
   The traveling path information including the traveling locus when the moving body performs automatic driving is acquired from the moving body, and the traveling body is obtained.
   The traveling support device according to claim 1, wherein the traveling difficulty level is determined based on the moving body information, the sensing information, and the traveling path information.
5. The driving support device is
   A peripheral situational awareness unit that acquires the moving body information, the sensing information, and the traveling path information, and the peripheral situation recognition unit.
   A traveling difficulty determination unit that determines the traveling difficulty level based on the moving body information, the sensing information, and the traveling path information acquired by the peripheral situation awareness unit.
   Autonomous automatic driving in which the moving body autonomously and automatically runs on the support level of the moving body based on the running difficulty level, cooperative automatic driving in which the moving body automatically runs in cooperation with peripheral moving bodies, and the above. A support level determination unit that determines at least four stages of control control and manual driving that automatically travels under the control of a travel support device.
   A claim including a recommended action generation unit that generates the support information according to the support level based on the support level, the moving body information, the sensing information, and the traveling path information, and transmits the support information to the moving body. Item 4. The traveling support device according to item 4.
6. The driving difficulty determination unit is
   Based on the moving body information, the sensing information, and the traveling path information,
   Driving situations including positional relationships between vehicles, lane changes, changes in acceleration and deceleration,
   Driving scenarios including lane changes, left / right turns, merging,
   Risk of surrounding conditions, including the possibility of contact of the moving object with the peripheral moving object,
   4. Driving support device.
7. The driving difficulty determination unit is
   The driving support device according to claim 6, wherein the driving difficulty level is determined based on the total value of the driving situation, the driving scenario, the risk of the surrounding situation, and the cost set according to the congestion degree.
8. The support level determination unit
   The cooperative automatic driving
   A first cooperative automatic driving that uses the moving body information acquired from the peripheral moving body, and a second cooperative automatic driving that uses the sensing information acquired from the peripheral moving body.
   Judgment is made by classifying into three stages of the third cooperative automatic driving using the travel path information acquired from the peripheral moving body.
   The control control,
   A first control control that notifies the moving body of recommended actions including recommended speed, recommended lane, and recommended lane change timing,
   The traveling support device according to claim 5, wherein the traveling support device according to claim 5 is classified into two stages of a second control control for controlling the speed and traveling path of the moving body.
9. The driving support device is
   Based on the moving body information and the sensing information, an action plan including lane keeping, lane change, acceleration and deceleration for the moving body to travel without contact with obstacles and surrounding moving bodies is created. It is further equipped with an automatic driving judgment unit that generates driving path information that realizes the action plan.
   The automatic driving determination unit notifies the moving body of the traveling path information, and the automatic driving determination unit notifies the moving body.
   The travel support device according to claim 5, wherein the moving body is controlled according to the travel path information.
10. The driving support device is
    Further, a group determination unit for determining a plurality of moving bodies moving in the same direction as one group based on the moving body information and the sensing information is provided.
    The driving support device according to claim 5, wherein the traveling difficulty level and the support level are the same for a plurality of moving objects included in the one group.
11. The driving support device is
    The vehicle further includes a travel difficulty learning unit that inputs the moving body information, the sensing information, the travel path information, and the travel difficulty level, and generates a travel difficulty level determination model that determines and evaluates the travel difficulty level.
    The driving difficulty determination unit is
    The driving support device according to claim 5, wherein the driving difficulty determination model is used to determine the driving difficulty.
12. The driving support device is
    The traveling support device according to claim 1, which is provided separately from the moving body.
13. The driving support device is
    The traveling support device according to claim 1, which is mounted on the moving body.
14. It is a running support method that supports the running of moving objects.
    The moving body information including at least the position, velocity, and direction of the moving body acquired by the moving body sensor mounted on the moving body, and
    Based on the sensing information including at least the position and speed of obstacles around the moving body acquired by the peripheral recognition sensor mounted on the moving body.
    The running difficulty level indicating the running difficulty of the moving body is determined in at least three stages, and the support information used for the running support of the moving body is set according to the determined running difficulty level.
    A running support method for supporting the running of the moving body by using the support information.

Images