DE112020005630T5 - DRIVING ASSISTANCE DEVICE - Google Patents

Abstract

From the environmental information of the sensor information, a connection relationship between an entrance lane and an exit lane at the intersection is calculated, and the optimum lane for reaching the destination is selected. A driving assistance device 3 according to the present invention includes a sensor information acquisition device 103 that acquires peripheral information of a host vehicle 1 detected by a sensor S, a sensor road/lane structure generation unit 105 that generates a lane center sequence of a lane included in a road generates based on information about at least one white line or a roadside of the road contained in the surrounding information, and generates road-lane structure information based on the lane midpoint sequence, and a travel planning unit 109, which generates on the basis of the road-lane -Structure information designates a road on which the host vehicle is to travel and a lane of the road.

Description

Technical part

The present invention relates to a driving assistance device that recognizes roads and structures of lanes based on information about surroundings of a host vehicle and map information, and outputs vehicle control and notification to a driver.

State of the art

In order to avoid traffic accidents, reduce environmental pollution and create a safe and convenient mobility society, which also helps the elderly to move around, the technology development of automatic driving is actively promoted. Levels 1 through 5, defined by the Society of Automotive Engineers (SAE), are commonly used as the level of automated driving. According to the SAE definition, the system is responsible for level 3 or higher vehicle control. Therefore, in order to realize level 3 or higher automatic driving, it is necessary to perform the control with more reliable information, and it is necessary to use a map. As the map, a high-precision map prepared in advance by surveying is usually used. However, since information is required to be updated for a map used for automatic driving, and it is very expensive to keep it updated, there is a possibility that the provision area of the high-precision map is limited to an expressway or a thoroughfare. Therefore, highly reliable control is required for general roads even if a high-precision map is not made.

Therefore, for example, PTL 1 proposes that an automatic driving control system is configured as follows: "A driving assistance system that assists the driving of a vehicle, the driving assistance system having a position information acquisition unit that acquires position information of the vehicle, a route generation unit that generates a travel route indicating, which road the vehicle is passing in the simplified map, using simplified map information representing a simplified map used for route guidance and the position information, a surrounding map generation unit that generates a surrounding map showing a map of the vicinity of the vehicle is and is more accurate than the simplified map while driving the vehicle as surrounding map information using sensor information detected by a vehicle-mounted sensor and the position information, a feature extraction unit it that extracts features of a road in each of the simplified map and the surrounding map including the travel route, a positioning unit that performs positioning between the simplified map and the surrounding map based on the features of the road and calculates a position of the vehicle as a vehicle position, and a path generation unit that projects the travel route generated using the simplified map information and the position information onto the surrounding map using the vehicle position and generates a path that is a travel trajectory and a route of the vehicle on the surrounding map, the path being a path for the vehicle, to travel along the travel route based on the travel route projected on the surrounding map, the route being an input to a vehicle control mechanism of an automated vehicle.”.

citation list

patent literature

PTL 1: JP6456562B1

Summary of the Invention

Technical problem

In PTL 1, for a general road, automatic driving not deviating outward from the road is realized by surrounding map information generated by using sensor information and simplified map information. However, the simplified map information does not have a connection relationship between an entry lane and an exit lane at the intersection, and there is a problem in selecting an optimal lane for arriving at the destination. For example, when a vehicle turns right at an intersection and turns left after the intersection to reach a destination, when the driver turns right at the first intersection, he wants to move to the leftmost lane of the road after the right turn. Therefore, it is necessary to calculate the connection relationship between an entry lane and an exit lane at the intersection from the surrounding information of the sensor information and to select an optimum driving lane to reach the destination.

An object of the present invention is to provide a driving assistance device capable of selecting an optimal lane in a travel plan to a destination and performing automatic driving according to a driver's intention of a general road including an intersection.

the solution of the problem

A driving assistance device according to the present invention includes a sensor information acquisition device that acquires peripheral information of a host vehicle detected by a sensor, a sensor road/lane structure generation unit that generates a lane center point sequence of a lane included in a road based on information about at least one generates a white line or a road edge of the road included in the surrounding information, and generates road-lane structure information based on the lane midpoint sequence, and a travel planning unit which, based on the road-lane structure information, generates a road on which the host vehicle is to drive, and determines a lane of the road.

Advantageous Effects of the Invention

According to the present invention, it is possible to select an optimal lane in a travel plan to a destination, and it is possible to realize automatic driving according to a driver's intention on a general road including an intersection. Further features of the present invention will become apparent from the description of the present specification and the accompanying drawings. Problems, configurations and effects other than those described above will be apparent from the following description of the embodiments.

character list

• [ 1 ] 1 12 is a diagram for explaining a configuration of a vehicle equipped with a driving assistance device according to a first embodiment.
• [ 2 ] 2 14 is a block diagram showing an internal function of the driving assistance device according to the first embodiment.
• [ 3 ] 3 is a block diagram showing the internal workings of an in 2 shown sensor road/lane structure generation unit and a map combining unit.
• [ 4 ] 4 Fig. 12 is a flowchart showing the processing contents in a track center sequence generating unit.
• [ 5 ] 5 14 is a diagram showing an intersection and road IDs set at the intersection.
• [ 6 ] 6 is a diagram showing an example of a simplified map with a map node and a map link.
• [ 7 ] 7 Fig. 12 is a diagram schematically showing a detection range of a sensor.
• [ 8th ] 8th 12 is a diagram showing a recommended driving trajectory when passing through an intersection, and an entry point and an exit point of a non-lane portion such as an intersection.
• [ 9 ] 9 Fig. 12 is a diagram for explaining a sequence of points indicating a boundary of an intersection with the outside of a road.
• [ 10 ] 10 Fig. 14 is a diagram for explaining white line information after turning right or left at an intersection.
• [ 11 ] 11 14 is a diagram showing a trajectory in which the host vehicle turns right at an intersection.
• [ 12 ] 12 Fig. 12 is a diagram showing an example of an intersection.
• [ 13 ] 13 Fig. 12 is a diagram showing an example of an intersection.
• [ 14 ] 14 Fig. 12 is a diagram showing an example of an intersection.
• [ 15 ] 15 Fig. 12 is a diagram for explaining an example of generating a route by combining map information with road/lane information.
• [ 16 ] 16 Fig. 12 is a diagram for explaining another example of generating a route by combining map information with road/lane information.
• [ 17 ] 17 Fig. 12 is a diagram for explaining another example of creating a route by combining map information with road/lane information.
• [ 18 ] 18 14 is a block diagram for explaining an internal function of a driving assistance device according to a second embodiment.
• [ 19 ] 19 Fig. 12 is a view for describing an example of a method of generating a track center point.
• [ 20 ] 20 Fig. 12 is a diagram for explaining a method of calculating a track increase/decrease limit.
• [ 21 ] 21 is a diagram showing an example of a method for generating a vir tual lane center point sequence from an entry point and a departure point of an intersection.
• [ 22 ] 22 Fig. 12 is a schematic representation of the vicinity of a toll booth on a toll road.

Description of the embodiments

Embodiments of the present invention will be described below with reference to the drawings. In the following embodiment, a case where the road is for left-hand traffic is described as an example, but the present invention can be applied to right-hand traffic.

<First embodiment>

1 12 is a diagram for explaining a configuration of a vehicle equipped with a driving assistance device according to the present embodiment.

As in 1 1, a driving assistance device 3 is mounted on a host vehicle 1 capable of automatic driving, and is realized through the cooperation of hardware such as a CPU and a memory, and a software program executed by the hardware. The driving assistance device 3 configures a driving assistance system in cooperation with a map/locating unit U1, a sensor S, and an actuator D.

The driving assistance device 3 obtains map information, route information, host vehicle position information of the host vehicle 1, and orientation information of the host vehicle 1 from the map/locating unit U1. Then, environmental information is obtained from a camera sensor S1 and a LiDAR sensor S2 constituting the sensor S . Further, the host vehicle information is obtained from the vehicle sensor S3, and each operation target amount of a motor D1, a steering wheel D2, a brake D3, and the like, which is the actuator D, is determined.

The map/locator unit U1 includes a locator function 24 and a map transmission function 25. The locator function 24 receives GNSS (position information) and obtains host vehicle position information and heading information. The map transmission function 25 includes a communication unit U2 which receives map data 8 . The map data 8 may be a simple map (hereinafter referred to as a simplified map) for navigation or a high-accuracy map containing high-accuracy lane-level information.

2 14 is a block diagram illustrating an internal function of the driving assistance device.

The driving assistance device 3 receives data from the map/locating unit U<b>1 and the sensor S to output a calculation result of the driving assistance device 3 to a speaker 111 and the actuator D . The driving assistance device 3 includes a map/route information obtaining unit 101, a position/orientation information obtaining unit 102, a sensor information obtaining unit 103, a sensor information integrating unit 112, a self-position/orientation estimating unit 104, a sensor road/lane structure generating unit 105, a map combining unit 106, a route information adding unit 107 and a control unit 108. In addition, control unit 108 includes a trip planning unit 109 and a trajectory planning unit 110.

[data flow]

The map/route information acquisition unit 101 receives data 2001 from the map/locator unit U1 to output data 2004 and 2005. The position/orientation information acquisition unit 102 receives data 2002 from the map/locator unit U1 and outputs data 2016. The sensor information obtaining device 103 receives data 2003 from the sensor S to output data 2006 and 2017.

The sensor information integration unit 112 receives the data 2017 from the sensor information acquisition device 103 and the data 2021 from the self-position/orientation estimating unit 104 to output the data 2019 and 2020. The sensor road/lane structure generation unit 105 receives the data 2019 from the sensor information integration unit 112 and the data 2007 from the self-position/orientation estimating unit 104 to output data 2009.

The self-position/orientation estimating unit 104 receives the data 2016 from the position/orientation information obtaining unit 102, the data 2006 from the sensor information obtaining device 103, and the data 2020 from the sensor information integrating unit 112 to output the data 2007, 2008, and 2018. The map combining unit 106 obtains the data 2005 from the map/route information obtaining unit 101, the data 2018 from the self-position/orientation estimating unit 104, and the data 2009 from the sensor road/lane structure generating unit 105 to output data 2010 and 2012. The route information adding unit 107 obtains the Data 2004 from the map/route information obtaining unit 101, the data 2008 from the Self position/orientation estimation unit 104 and the data 2010 from the map combining unit 106 to output data 2011.

The control unit 108 receives the data 2011 from the route information adding unit 107 and the data 2012 from the map combining unit 106 to output the data 2014 and 2015. The trip planning unit 109 receives the data 2011 from the route information adding unit 107 and the data 2012 from the map combining unit 106 to output the data to be issued in 2013 and 2014. The trajectory planning unit 110 receives the data 2011 from the route information adding unit 107, the data 2012 from the map combining unit 106, and the data 2013 from the travel planning unit 109 to output the data 2015. The speaker 111 constitutes a notification unit that receives the data 2014 from the travel planning unit 109 to output a notification to the driver. The actuator D receives the data 2015 from the trajectory planning unit 110 and controls the vehicle.

[Description of the function blocks and data content]

The data 2001 from the map/locator unit U1 is accepted as simplified map and route information, and the map/route information unit 101 acquires the data, applies conversion if necessary, and stores the converted data in memory. 6 Figure 12 is a diagram showing an example of a simplified map with a map node and a map connection. As in 6 As shown, the simplified map includes a map node F001 and a map link F002.

A node ID is set for the map node F001, and it stores latitude and longitude as the position information of the node. In addition, map node F001 stores elevation information. In addition, the map node F001 stores information such as the connection angle F003 between map connections and the number of connections. In the example of 6 the number of connections stored in card node F001 is four. The number of connection angles F003 stored in map node F001 is also four.

In addition, the map node F001 stores a road type and, in the case of an intersection, stores information such as an intersection, a T-intersection, or a six-way intersection as an intersection type. In addition, the map link F002 stores a connection relationship between map nodes in ID, the number of lanes of the host vehicle 1 lane, the number of lanes of the opposite lane, and additionally stores a speed limit value, a position on the map link, shape information of the map link and the like.

The route information is information indicating a list of IDs of the map node F001 and through which map node the vehicle can reach the destination. Alternatively, not only the ID but also a pair of degrees of latitude and longitude can be specified as position information of the map node. In addition to the position information of the map node, guidance information about a road ahead or a travel direction of a right or left turn registered for navigation can also be recorded.

An example of the conversion by the map/route information obtaining unit 101 is the conversion of information on a global coordinate system such as latitude and longitude into a plane coordinate system. As a result, there is a Hybeny distance formula since the distance calculation is generally handled in the global coordinate system, and it has the advantage that the distance calculation at the subsequent stage can be easily performed by the rule of three. Note that the simplified map can be purchased from V2X.

The data 2002 from the map/locator unit U1 is information from the locator, assuming latitude and longitude as position information and orientation as orientation information. The position/orientation information acquisition unit 102 acquires the data 2002, applies conversion if necessary, and stores the converted data in memory. Examples of the conversion are conversion to a planar coordinate system as in the map/route information acquisition unit 101.

Data 2003 from sensor S includes data from camera sensor S1, LiDAR sensor S2, and vehicle sensor S3 in 1 . As data from the camera sensor S1, position information, size information, type information and the like of a three-dimensional object, position information of a white line, type information of a broken line or a color, position information of a roadside, position information and color information of a traffic light, position information of a stop line, position information of a crosswalk, position information and Type information of a shield and the like is accepted.

5 and 14 are diagrams showing an example of an intersection, 5 is a slide gram showing lanes of the intersection and road IDs posted at the intersection, and 14 Fig. 12 is a diagram showing an example of an intersection with a traffic light.

An example of a white line is F013 in 5 . The white line may be a series of points obtained by extracting a sequence in which end points are continuously arranged from a point group obtained by edge detection, or an approximation parameter. An example of the roadside is F010 in 5 . An example of a stop line is F014 in 5 . An example of a pedestrian crossing is F011 in 5 . Examples of the notice board are a speed limit sign, a temporary stop line, and a lane departure. An example of a traffic light is F021 in 14 .

Furthermore, it is assumed that the data of the LiDAR sensor S2 is position and size information of a three-dimensional object, position and type information such as a broken line of a white line, position information of a roadside, position information of a stop line, position information of a crosswalk, and acts like that. In addition, it is assumed that the data from the vehicle sensor S3 is information about the speed of the host vehicle, the steering angle, the yaw rate, the acceleration, and the like. Note that the sensor S can be any sensor capable of acquiring data necessary for understanding the surroundings of the host vehicle 1 as described above, and can be radar, C2X, V2X, or the like can.

The sensor information acquisition device 103 collects the data of the information from each sensor, converts the data as necessary, and stores the converted data in the memory. As an example of conversion, in a case where the unit system and the resolution of the data are different in the plurality of sensors, the data is converted into a uniform format so that it can be easily handled in the subsequent processing.

The data 2017 from the sensor information obtaining device 103 is assumed to be sensor information in a format in which the unit system and resolution of the data are unified, and the sensor information integrating unit 112 integrates the data so that it can be used in the subsequent Processing can be processed without any problems. An example of integration is that in a case where the coordinate systems of the data and the origins of the coordinate systems are different between the respective sensors, the coordinate systems or the origins are unified to the same coordinate system or origin so that the data in the subsequent Processing can be easily processed.

In the present embodiment, the present invention is described below on the assumption that the present invention is treated in a host vehicle relative coordinate system. In addition, in a case where the transmission cycles of the data in the plurality of sensors are different from each other, the transmission cycles are synchronized with the processing cycle of the driving assistance device 3, so that they can be easily treated as a uniform time in the subsequent processing. The synchronization is calculated by estimating the amount of movement from the speed information of the host vehicle and the steering angle. Then, when each sensor detects the same three-dimensional object, white line, roadside, or the like in the synchronized information, the same ID is set as the same information, and the data is integrated for output.

In addition, the past sensor data 2017 may be recorded in the memory in the form of the global coordinates based on the host vehicle position and the heading information, that is, the data 2021 from the self-position/heading estimation unit 104, and interpolation processing at the time of data loss or integration processing may be performed be performed on the basis of an empirical rule based on past information. In the case of the white line information, for example, a method is conceivable in which the position information of the past white line is compared with the position information of the newly detected white line, it is determined whether it is the same white line, and the ID is tracked . Thus, even in a situation where the number of sensor information collected is small, the information collected can be utilized and the accuracy of ID identification can be improved to enable integration processing. In contrast to data based on the relative coordinate system of the host vehicle, e.g. B. the camera sensor S1 and the LiDAR sensor S2, as is known, the information of C2X and V2X is data based on the global coordinate system and is converted to the relative coordinate system of the host vehicle based on the host vehicle position and the orientation of the data 2021 and with the data of Integrated camera sensor S1 and LiDAR sensor S2.

The data 2019 from the sensor information integrating unit 112 is three-dimensional object position information, white line position information, roadside position information, and the like for which synchronization of processing cycles and data integration and integration of coordinate systems are completed, and the sensor road/lane structure generating unit 105 understands boundary portions of roads and lanes, and calculates a connection relationship between roads and a connection relationship between lanes. Note that specific examples of the boundary portion between roads and the boundary portion between lanes include white line information F013, crosswalk information F011, stop line information F014, and roadside information F010 detected by the sensor in one situation can be, in which the carrier vehicle 1 as in 5 shown exists.

As in 5 As illustrated, the road represents a region divided by a lane increase/decrease, an intersection, a junction, and the like, such as e.g. B. the regions R01 to R06, and is managed by an ID. The track represents information in units of L01 to L13 included in the regions R01 to R06 and is managed by an ID.

The connection relationship between the roads and the connection relationship between the lanes represent a connection relationship between the respective regions R01 to R06 front and rear and a connection relationship between the respective lanes L01 to L13 front and rear, and right and left. For example, due to the position of the Host vehicle connected the front of region R01 to region R02 and the back of region R02 to region R01. Also, track L03 is connected to track L02 on the right, and track L02 is connected to track L03 on the left. In addition, the track L06 and the track L07 are connected to the front of the track L02, and the track L02 is connected to the back of the track L06 and the track L07.

As in 7 1, the sensor road/lane structure generation unit 105 generates an output based on the information detected in a detection area F015 of the sensor S and performs the output in real time every time the host vehicle 1 moves to regenerate information .

7 14 is a diagram showing the state before and after the host vehicle 1 enters the right-turn lane.

When the host vehicle 1 moves from in front of the right-turn lane at the intersection to the right-turn lane at the intersection in 7 moved, the detection range is from F015 to F020, and the sensor information in the intersection can be detected more.

The sensor road/lane structure generation unit 105 manages the data 2017 from the sensor information acquisition device 103 related to a structure of a road or a lane, and also generates additional information that can be calculated only from the sensor information.

It is conceivable that examples of management in which the data 2017 is associated with a structure of a road or a lane include associating white line information of the data 2017 with each lane and associating roadside information of the data 2017 with a road. Examples of the additional information are a center sequence of the lane, a radius of curvature, a speed limit, a length of the lane, a distance until the lane disappears or appears, and the like. As an example of the distance until the disappearance of the track, the distance until the disappearance of the track L02 in 5 from the addition of the length of the lane L01 and the length of the lane L02 when the host vehicle is at the starting position of the lane L01.

In addition, as in the detection area F015 of 7 shown, it is generally known that the detection error of the data 2017 increases with increasing distance from the limit area of the detection area or the host vehicle 1 . In addition, incorrect detection may occur depending on the driving situation. It is therefore advantageous to supplement the road/lane structure information 2009 with reliability information, taking into account the properties of these sensors S. In addition, the generation accuracy of the road and lane structure using the past information as in the sensor information integrating unit 112 can be improved by using the host vehicle position and orientation, which are the data 2007 from the self-position/orientation estimating unit 104 .

The data 2016 from the position/orientation information obtaining unit 102 is assumed to be latitude and longitude as position information and the orientation tion acts as alignment information. The data 2006 from the sensor information acquisition device 103 is assumed to be vehicle information such as speed information of the host vehicle and steering angle. The self-position/orientation estimating unit 104 estimates latitude and longitude as the host vehicle's final position information with improved accuracy for the driving assistance device 3, and estimates the orientation as the host vehicle's final orientation information.

Since the data 2016 based on the information from the locator is longer than the processing cycle of the driving assistance device 3, for example a transmission cycle between 100 milliseconds and 1 second, the host vehicle position and the orientation must be corrected in the processing cycle of the driving assistance device 3 in which the data 2016 is not be updated. The correction of the host vehicle position and orientation is performed by calculating the movement amount using the host vehicle speed information, steering angle information, yaw rate information and the like of the data 2006 from the sensor information acquisition device 103.

In addition, the accuracy of the host vehicle position and the in-lane orientation information can be improved using the white line position information of the data 2020 from the sensor information integrating unit 112 . Furthermore, the accuracy of the host vehicle position and the orientation information in the road can be improved using the roadside information of the data 2020. When the accuracy of the host vehicle position is increased, in a case where the accuracy of the map to be combined by the map combining unit 106 is high, an error in conversion to the relative coordinate system of the host vehicle can be minimized, so there is an advantage that the Control unit 108 can use more accurate information.

The data 2009 from the sensor road/lane structure generation unit 105 is the road-lane structure information described above, and the data 2018 from the self-position/orientation estimation unit 104 is that for the driving assistance device 3 corrected host vehicle position and orientation, the data 2005 from the map/route information unit 101 is map information, and the map combining unit 106 calculates the final road-lane structure information processed by the control unit 108 with high reliability.

The map combining unit 106 extracts map information of a required region from the data 2005 based on the host vehicle position and orientation of the data 2018 and checks the consistency of the information with the data 2005 based on the road-lane structure information of the data 2009. If the information do not match, the road-lane structure information with the low reliability information of the 2009 data takes over the information of the 2005 data and increases the reliability. As a result, more reliable information can be provided to the control unit 108, and control corresponding to the driver's intention can be performed. When the map information is not provided by the map/locator unit U1, the map combining unit 106 can continue driving support by using the road-lane structure information from the sensor road/lane structure generation unit 105 as the data 2010 and 2012 directly spends

The data 2004 from the map/route information unit 101 is route information, the data 2008 from the self-position/orientation estimation unit 104 is a host vehicle position and orientation corrected for the driving assistance device 3, and the data 2010 from the map combining unit 106 is one Road-lane structure information with increased reliability. The route information of the data 2004 is assumed as a route in a map-independent format so that the data 2010 from the map combining unit 106 can be matched with the coordinate-related information of the road-lane structure information expressed by the relative coordinates of the host vehicle.

As an example of a route in a map independent format, a list of latitude and longitude is set as a route, and the latitude and longitude of the route are converted to the relative coordinates of the host vehicle based on the host vehicle position and orientation of the data 2008 from the self position/orientation estimation unit carrier vehicle converted. The converted link coordinates and lane midpoint information associated with the lane structure information of data 2010 are used to determine which street the link belongs to.

As a result, the route information adding unit 107 outputs an ID list of the road-lane structure information in the data 2010. In addition, as guidance information from the map/route information acquisition unit 101, indications of an upstream road or right/left turns may be included in the data 2011 of the route information addition unit 107. Thereby there is the advantage that the guidance information is also used in an area that cannot be detected by the sensor, and helps the trip planning unit 109 of the control unit 108 to determine which lane the vehicle is traveling into.

The data 2011 from the route information adding unit 107 is assumed to be an ID list of road-lane structure information corresponding to the data 2010 and guidance information from navigation, and the data 2012 from the map combining unit 106 it is assumed that this is road-lane structure information with increased reliability. The travel planning unit 109 of the control unit 108 determines a road on which the vehicle is to travel based on the route information of the data 2011 and a lane on which the vehicle is to travel based on the lane structure information of the data 2010 . Alternatively, a lane in which the vehicle should travel is determined based on the route information of the data 2011 . The effect of the driving assistance device 3 in the present embodiment can be maximized when an optimal lane in consideration of the lane for the vehicle to travel is selected as a lane for the vehicle to travel when turning right or at the intersection turns left to get to the destination of the route.

The data 2013 from the trip planning unit 109 is a list of lane IDs on which the vehicle is to travel, the data 2012 from the map combining unit 106 is a speed limit, a curve radius, and a lane midpoint associated with lane structure information , and the trajectory planning unit 110 determines a trajectory on which the host vehicle should advance a few seconds and a target speed of the host vehicle.

The data 2014 from the travel planning unit 109 is assumed to be the lane ID where the vehicle is to travel and the lane structure information, and the speaker 111 notifies the driver of which lane the vehicle is traveling as a warning target. The data 2015 from the trajectory planning unit 110 is based on a trajectory on which the host vehicle is to travel and a target speed of the host vehicle, and the actuator D determines and controls a target steering angle, a target brake fluid pressure and a target engine torque. In order to realize the automatic driving function as a driving assistance system, it is advantageous to output the data to the loudspeaker 111 and the actuator D and to carry out the control. For the realization of the safety assistance as a driving assistance system, the data can only be output to the loudspeaker 111 in order to draw the driver's attention to it.

[Configuration]

3 12 is a block diagram illustrating the internal functions of the sensor road/lane structure generation unit 105 and the map combining unit 106. FIG.

In 3 Data from the sensor information integrating unit 112 and data from the self-position/orientation estimation unit 104 serve as inputs to the sensor road/lane structure generation unit 105, and a calculation result of the sensor road/lane structure generation unit 105 is output to the map combining unit 106. Then, data from the map/route information unit 101, data from the self-position/orientation estimation unit 104, and data from the sensor road/lane structure generation unit 105 serve as inputs to the map combining unit 106, and a calculation result of the map combining unit 106 is output.

The sensor road/lane structure generation unit 105 includes a lane midpoint sequence generation unit 302, a region division determination unit 303, a lane information generation unit 304, a road information generation unit 305 and a sensor additional information generation unit 306. In addition, the map combination unit 106 includes a map selection unit 307, an error correction unit 308, a lane information generation unit 309, a road information generation unit 310, a simplified lane midpoint sequence generation unit 311 and an additional map information combination unit 312.

[data flow]

In 3 the data 2019 from the sensor information integrating unit 112 and the data 2007 from the self-position/orientation estimating unit 104 are input to the lane center sequence generating unit 302, and the data 4005, 4006 and 4007 are output.

The data 4005 from the track center sequence generation unit 302 is input to the region division determination unit 303, and the data 4008 and 4009 are output. The data 4006 from the track midpoint sequence generation unit 302 is input to the lane information generation unit 304, and the data 4010 is output. The data 4009 from the regional division determining unit 303 and the data 4010 from the lane information generation unit 304 are input to the road information generation unit 305, and the data 4011 is output. The data 4011 from the road information generation unit 305 is input to the sensor additional information generation unit 306, and the data 2009 is output. The data 2005 from the map/route information unit 101 and the data 2018 from the self-position/orientation estimation unit 104 are input to the map selection unit 307, and the data 4015, 4023 and 4024 are output.

The data 2009 from the sensor additional information generation 306 and the data 4015 from the map selection unit 307 are input to the error correcting unit 308, and the data 4016 is output. The data 4024 from the map selection unit 307 and the data 4016 from the error correction unit 308 are input to the lane information generation unit 309, and the data 4017, 4018 and 4019 are output. The data 4017 from the lane information generation unit 309 is input to the road information generation unit 310, and the data 4022 is output.

The data 4023 from the map selection unit 307 and the data 4019 from the lane information generation unit 309 are input to the simplified lane center sequence generation unit 311, and the data 4020 is output. The data 4018 from the lane information generation unit 309, the data 4022 from the road information generation unit 310 and the data 4020 from the simplified lane center sequence generation unit 311 are input to the map additional information combining unit 312, and the data 4021 is output.

[Description of the function blocks and data content]

The data 2019 from the sensor information integration unit 112 is assumed to be the synchronization of the processing cycle, data integration, the position information of the white line in which the coordinate system has been unified, the position information of the roadside and the like, and the lane center sequence generation unit 302 calculates the positions of the track midpoint sequences of the plurality of tracks located in the sensor detection area. Note that the data 2007 from the self-position/orientation estimating unit 104 is adopted as the host vehicle position and orientation, and the history of the lane center sequence generated by the lane center sequence generating unit 302 is used to stabilize the output lane center sequence. The lane including all lanes in the sensor detection range includes, for example, the lane of an opposite lane, the lane of an adjacent parallel lane, the lane of an intersecting road, and the like, in addition to the lane where the host vehicle is located. The track midpoint is a track midpoint of the track, i.e., a midpoint in the width direction of the track in each track, and the track midpoint string is a point string in which the track midpoints of the respective tracks are connected in a direction in which the tracks run.

The lane midpoint sequence generated by the lane midpoint sequence generation unit 302 based on the host vehicle position and alignment is converted into the global coordinate system and stored in the lane midpoint sequence generation unit 302 . In the next processing cycle of the driver assistance device 3, the position of the lane center point sequence in the history is then converted from the former carrier vehicle position and alignment into the relative coordinate system of the carrier vehicle and integrated with the lane center point sequence calculated at the current time. This makes it possible, even in a processing cycle in which the lane center sequence is momentarily interrupted, to continuously supply the control unit 108 with the lane center sequence and to stabilize the entire driver assistance system.

8th 14 is a diagram showing a recommended travel trajectory when passing through an intersection, and an entry point and an exit point of a non-lane portion such as an intersection. 9 12 is a diagram illustrating a sequence of points indicating a boundary with the outside of a road at the intersection before and after entering the right-turn lane. 10 Fig. 12 is a diagram showing a sequence of points indicating a position of a white line before and after entering the right-turn lane.

In the track midpoint sequence generation unit 302 of the present embodiment, as indicated by F016 in 8th is displayed, it is assumed that a recommended traveling trajectory in an intersection is also a virtual lane, and a lane midpoint sequence is generated. As indicated by F022 in 9 indicated, an example of the position information of the roadside at the intersection includes a position of a string of points indicating a boundary with the outside of the road at the intersection. As indicated by F025 in 10 displayed, an example of the white line position information at the intersection also includes information about the white line Line after turning right or left at the intersection.

The data 4005 from the track midpoint sequence generating unit 302 is assumed to be midpoint sequences of a plurality of tracks, and the region division determining unit 303 determines a track increase/decrease limit based on the positional relationship of each track midpoint sequence. An example of track gain includes a border between regions R01 and R02 in 5 . The calculation of the track increase/decrease limit is described with reference to FIG 20 described.

20 Fig. 12 is a diagram illustrating a method of calculating a track increase/decrease limit.

In 20 the track midpoint sequences F206 and F207 are shown, and perpendicular lines are drawn from the start end point F203 and the end point F204 to the respective track midpoint sequences. Next, it is determined whether the distances F201 and F202 from the point of intersection with the vertical line have exceeded a certain threshold, and if the distances are equal to or larger than the certain threshold, the section is determined as the limit of increase/decrease of the track. In the example of 20 in a case where only the portion F201 exceeds the certain threshold, F205 is the limit of track increase/decrease.

In addition, it is assumed that the data 4005 of the lane midpoint sequence generation unit 302 is an entry point F022 of a non-track portion like that in FIG 8th shown intersection and a departure point F017 of the non-track section, and the regional division determination unit 303 determines the boundary between the non-track section having no lane and the lane section having a lane other than the non-track Section has, based on the entry point and the departure point.

Note that examples of the boundary between the non-track portion and other than the non-track portion include a boundary between regions R02 and R04, a boundary between regions R04 and R03, a boundary between regions R04 and R05 and a boundary between regions R04 and R06, as in 5 shown. In the next step, based on the boundary calculated by the region division determination unit 303, the lane information generation unit 304 and the road information generation unit 305 assign IDs to the road-lane structure information.

Moreover, the boundary between the non-lane portion including the crossing portion and other than the non-lane portion can be determined in a complex manner from the position information of the signal of the sensor information integrating unit 112, the stop line and the position of the crosswalk.

Note that an entrance and an exit of a tollbooth of a toll road in addition to an intersection correspond to the non-lane section. 22 Fig. 12 is a schematic representation of the vicinity of a toll booth of a toll road. In the case of the toll gate, the position information and the size information of a toll gate F046 can be acquired by the sensor information integration unit 112, and an entry point F041 and an exit point F042 upstream of the toll gate F046 and an entry point F044 and a departure point F045 downstream of the toll gate F046 can be calculated. By providing the boundary in this way, it is possible to determine how to change the travel plan at a point where the lane to be selected changes, e.g. B. by increasing or reducing the lane or an intersection. This has the effect that the itinerary can be easily handled by the subsequent stage logic.

The description returns 3 return. The data 4008 from the region division determination unit 303 is assumed to be track increase/decrease boundary information and information on a boundary between a non-track portion and other than the non-track portion, and the data 4006 from the track center sequence generation unit 302 is assumed to be position information of the track center sequence, and the track information generation unit 304 generates lane information. As in 5 shown, the lane information is defined as a lane where white lines exist, such as L01 to L03, L04, L05, and L11 to L13, and the non-track portion where no white line exists, such as inside an intersection such as L06 to L10 is defined as a virtual lane.

The lane information refers to information assigned with a lane ID such as L01 to L13 and indicating a connection relationship between the front and rear, and right and left lanes, respectively. In addition, a track contains information related to the track, such as a track midpoint following position and white line information associated with the a track, signal information and the like. The track includes a plurality of tracks L01 to L13. These tracks then contain information about the structure of the respective track, for example information indicating a connection relationship between the track sections front and rear and right and left, and information such as the position of the track center sequence.

The lane information generation unit 304 first generates a track ID based on the position information of the track center sequence and the boundary information from the region division determination unit 303. Then, left and right connection information is generated based on the lateral positional relationship of the track center sequences of the respective tracks. In addition, the front and rear connection information is generated based on the positional relationship of the traveling direction of the center sequences of the respective lanes. Finally, each track is assigned the track midpoint sequence position and the white line information. The lateral positional relationship of respective lane midpoint sequences represents which lane is arranged in what order in the road to indicate, for example, an arrangement order from the left.

The data 4009 from the region division determination unit 303 is adopted as track increase/decrease boundary information and information on a boundary between a non-track portion and other than the non-track portion, the data 4010 from the track information generation unit 304 is taken as track ID, assume a connection relationship between the respective lanes front and rear and right and left, and the road information generation unit 305 generates road information using these data 4009 and 4010. As in FIG 5 shown, the road information is defined as regions R01 to R06.

The road information refers to information assigned with a road ID such as R01 to R06 and indicating a connection relationship between the respective front and rear roads. In addition, a large number of lane IDs that belong to the one road are managed in a road. Furthermore, the one road contains position information of a roadside assigned to the road.

The road information generation unit 305 first generates a road ID based on the boundary information from the region division determination unit 303. Then, link information before and after each road is generated. In addition, a lane ID belonging to a street is calculated from the data 4010 and linked to the one street. Finally, the position information of the roadside belonging to one road is linked to the one road.

The data 4011 from the road information generation unit 305 is assumed to be road information and lane information associated with the road information, and the data 4007 from the lane midpoint sequence generation unit 302 is assumed to be lane information, and the Sensor additional information generation unit 306 uses this data 4011 and 4007 to generate additional information. The additional information is generated on the basis of the sequence of lane center points, and which road ID is assigned to which lane ID is determined in an integrated manner using the data 4011 .

Examples of the additional information include calculating a radius of curvature and the like based on the position of the track center sequence of the data 4007 from the track center sequence generating unit 302. An example of determining the radius of curvature includes selecting three representative points from the track center sequence and calculating one circle passing through these points.

In addition, the examples include calculating the distance until the ID of the lane information is switched based on the position of the lane midpoint sequence, and associating the signal information and the stop line from the sensor information integrating unit 112 and the position of the crosswalk information with the road information and the lane information. Considering the characteristics of the sensor S, the reliability information is associated with the road information and the lane information according to the detection end and the detection distance of the detection area F015 as shown in FIG 7 shown.

The data 2018 from the self-position/orientation estimating unit 104 is accepted as the host vehicle position, the data 2005 from the map/route information obtaining unit 101 is accepted as map information, and the map selection unit 307 selects a host vehicle position peripheral map from the data 2005 based on the host vehicle position of the data 2018. It is assumed that the selected map information includes the position information of the map node F001, the connection angle F003 between the map links, the number of links, the road type, the intersection type, the connection relationship between the map nodes of the map link F002 in ID, the Number of lanes of the driving lane and number of lanes of the opposite lane is as in 6 shown.

The data 2009 from the sensor additional information generation 306 is accepted as road-lane structure information with additional information, the data 4015 from the map selection unit 307 is accepted as map information selected based on the host vehicle position, and the error correction unit 308 compares these two pieces of information and corrects the road-lane structure information in the 2009 data that does not match the map information. Since the road-lane structure information is based on the limit of the sensor detection area and the sensor whose accuracy decreases as the distance from the host vehicle increases, the reliability of the road-lane structure information changes depending on the detected position.

On the other hand, the map information has a problem of currency and may have a different shape from the actual road due to construction work or the like. In a case where the reliability is obtained from the map information, the road-lane structure information is corrected when the map information and the road-lane structure information do not agree when the reliability of the road based on the sensor information is corrected. lane structure information is low and the reliability of the map information is high. This can reduce the possibility that only the map information that has a problem in being up-to-date is trusted and the end result is wrong.

The data 4016 from the error correction unit 308 is road-lane structure information corrected with map information, and the data 4024 from the map selection unit 307 is the number of lanes of the lane, the road type, the intersection type and the number of links, and the lane information generation unit 309 generates lane information which is not included in the data 4016. Then, a connection relation between the generated trace information and the trace information of the data 4016 is calculated. Finally, the newly generated track information and the connection information are added to the data 4016, and the data 4017 and 4019 are output.

The data 4017 from the lane information generation unit 309 is assumed to be road-lane structure information including lane information obtained by combining lane information based on sensor information and lane information based on map information, and the road information generation unit 310 generates road information to which the lane information belongs based on the combined map information. Then, a connection relationship between the generated road information and the road information of the data 4017 is calculated. Finally, the newly generated road information and link information are added to the data 4017, and the data 4022 is output.

The data 4019 from the lane information generation unit 309 is adopted as the lane information based on the sensor information and the lane information based on the map information, and the data 4023 from the map selection unit 307 uses the position information of the map node, the connection angle between the map links and the number of lanes of the driving lane, and the simplified lane midpoint sequence generating unit 311 generates a simple lane midpoint.

19 Fig. 12 is a diagram showing an example of a track center point generating method.

When generating the track midpoint, for example in a scene in which an in 19 illustrated host vehicle F100 turns right at an intersection, from a point with a distance F105 which is 1/2 the size of the intersection, ie half the size of the intersection from the map node F109, down a perpendicular F106 with respect to a map link F107 is drawn and a departure point F110 is set to a point having a mid-road estimation distance F104 from the map link F107.

Note that the mid-road estimation distance F104 is taken as (number of lanes of lane x 3.5 m)/2 from map compound F107. The use of road centers is not to generate a lane-level departure point, but rather to set the departure point as a guide, since the simplified map itself is not information that ensures accuracy. In addition, the intersection size/2 distance F105 is set as an intersection size by adding the lane width of each lane of the traffic lane where the host vehicle F100 is present. A crossing angle F103 between F101 and F107 is used as the crossing angle.

Then, an angle of intersection at a point F113 between an extension line F112 of the road to which the one contained in the data 4019 becomes Departure point F110 belongs and an extension line F111 of the lane to which the entrance point F102 belongs is set as a connecting angle F103, and a lane midpoint by approximating by an arc, a clothoid curve, a spline curve, a Bezier curve or the like using of the entry point F102, the departure point F110 and the precision F103 at the point F113 of the intersection included in the data 4019. It should be noted that the processing until the vehicle exits the intersection based on the generated lane midpoint is not controlled by the current information, but the current information is used as a guide for entering the intersection and is temporarily used until the host vehicle F100 recognizes the white line or the roadside information on the intersection exit side after entering. Note that although the number of intersection departure points is one, the number of intersection entry points is assumed to be plural, and when entering the intersection, a simplified lane midpoint is generated for each lane.

The data 4022 from the road information generation unit 310 is assumed to be road information obtained by combining the data 4015 and the road information of the data 2009, the data 4018 is assumed to be lane information obtained by combining the data 4015 and the lane information of the data 2009, and the map additional information combining unit 312 combines additional information which is not included in the road-lane structure information of the sensor road/lane structure generation unit 105 but included only in the map, i.e. additional information only included in the map are present with the road-lane structure information from the sensor road/lane structure generation unit 105 using these pieces of data. A road type of the map, a speed limit, and the like are assumed as an example of the additional information that is present only in the map. In addition, the map additional information combining unit 312 combines the data 4020 from the simplified lane center sequence generating unit 311 with the road-lane structure information. The data 4021 is then output as the final output of the card combining unit.

[Description of FIG. 4]

A detailed example of the track center sequence generation unit 302 is described with reference to the flow chart in FIG 4 described.

First, on the basis of in 10 shown information about the white line (F024) generates candidates for the center sequence for all recognized pairs of white line point sequences by the processing from S01 to S04. This processing is a process for a portion where a white line can be recognized. In this case, the white line dot sequence represents a continuously connected white line, and two of the white lines are paired. Note that a broken line is an exception and a solid line is one of the white lines.

Next, as processing for a non-track portion where a white line cannot be recognized, in S05 to S08, a center point of the virtual track is calculated based on the in 10 shown white line F024 and in 9 illustrated roadside information F022 is calculated. Finally, in S09, the track midpoint from S01 to S04 and the midpoint of the virtual track from S05 to S08 are integrated. Thereby, track structure information indicating a connection relationship between tracks with the non-track portion therebetween can be generated.

The processing in S01 to S04 in 4 is described.

S01 is a retry condition to determine whether all pairs of white line point sequences have been processed. In the portion where the white line can be recognized, one trace consists of two series of white line dots, and therefore the pair is repeatedly processed in this process.

Next, in S02, a pair of white line point sequences is selected. As in the example of 10 not only the white line point sequence continuously positioned in the direction of travel of the host vehicle, but also the white line point sequence running through the vertical intersection on the right/left turning side are set as processing targets.

In S03, a track center point sequence candidate is generated from the selected pair of white line point sequences.

Next, only necessary centers are narrowed down from the track center point candidates in S04. For example, whether the distance between the track center sequences is close to a general track width of 3.5 m is used as an index, and extremely short or long track center sequences are excluded from further processing. This can reduce the processing time in the next step. In addition, it is determined from the roadside information whether the position of a lane midpoint sequence candidate is outside the roadside as viewed from the host vehicle, and is excluded from the lane midpoint sequence candidacy.

Next, the processing from S05 to S08 will be described.

S05 is a condition for detecting a non-track portion. For example, a pattern in which the white line information F024 of 10 is broken at the entrance to the intersection and a white line appears at the exit of the intersection on the right/left turn side and the straight driving side of the intersection is designated as a non-lane section. In addition, the shape pattern of the intersection can be determined by the roadside information F020 of 9 may be determined as a non-lane portion, or the non-lane portion may be determined in a complex manner from the white line information and the stop line or crosswalk before the intersection. In addition, in order to improve the accuracy of the determination, the information of the sign of a right/left turning procedure F018 as in 12 shown, can be used as the feature point of the intersection. As in 13 shown, not only a stop line before an intersection but also information indicating the sign of a stop line F019 in the intersection can be used, and as in 14 shown, signal information F021 can be used. 12 12 is a diagram showing the right/left turn procedure F018 in an intersection, and 13 14 is a diagram showing the stop line F019 in the intersection.

In S06, an entry point of the intersection is calculated. The access point refers z. to F022 in 8th , and an end point of the track center point from S01 to S04 consisting of the white line F024 in 10 is calculated.

In S07, the departure point of the intersection is calculated. The departure point refers to F017 in 8th , and the end point of the track center point from S01 to S04 composed of the white line F024 from 10 calculated is selected as in S06. If the white line F024 from 10 cannot be recognized and the roadside information F022 from 9 can also be recognized, the distance F023 between the roadside points is assumed to be the road width, and the number of lanes is estimated from the width to generate the departure point. Furthermore, only the departure point of the leftmost lane after a right turn can be determined based on the position F021 of the signal information of 14 to be appreciated. These are determined in a complex way and a stable output can be calculated.

In S08, a combination of the entry point calculated in S06 and the departure point calculated in S07 is determined and a virtual lane center point sequence is calculated. The virtual lane center point sequence refers, for example, to a line marked by a dashed line at the intersection in 8th indicated sequence of centers. When a non-lane portion as a virtual lane is switched between a lane having an entry point and a lane having an exit point, the virtual lane midpoint sequence is a lane midpoint sequence of the virtual lane. As for the connection between the entry point and the departure point, since the L03 gauge in 5 one elevated lane and one right-turn lane, the entry point, which is on lane L03, connects to the exit point, which is on lane L12 or L13 on the right-turn side of the intersection.

At this time, by using the information of the stop line F019 in the in 13 illustrated intersection increases the possibility that it is a lane connected to a lane on the right turn side, so it is preferable to prevent erroneous connection with this information. In addition, the access point, which is on lane L02 in 5 connected to the exit points located in lanes L04 and L05 on the left turn side of the intersection and lane L11 for going straight.

21 Fig. 12 is a diagram showing an example of a method of generating a virtual lane center sequence from an entrance point and an exit point of an intersection. As in 21 shown, the entry point F022 and the departure point F017 are set as control points for the lane center point approximation in the virtual lane center point sequence at the time of the intersection. Then the intersection angle F034 of the point F031 is used, where the extension line F033 of the lane to which the departure point F017 belongs and the extension line F032 of the lane to which the arrival point F022 belongs, and it is defined by an arc, a clothoid curve, a spline curve, a Bezier curve, or the like.

Note that it can be approached with any of the right/left turning method F018 and the stop line F019 in the intersection as one of the control points so as not to be passed through the in 12 illustrated right/left turning procedure F018 or through the middle of the stop line F019 in the intersection in 13 to get lost. About that addition, as in 11 1, a center point of the virtual lane can be generated using the history of the position of a preceding vehicle F026. The preceding vehicle F026 may include the camera sensor S1 and the LiDAR sensor S2, and may receive information from C2X and V2X.

As described above, according to S05 to S08, the calculation of an entry point, an exit point and an intersection angle in a non-lane section such as an intersection or a toll booth makes a lane-level travel plan possible in the non-lane section even then , when there is little sensor information in the non-track portion.

[Description of FIG. 15 and 16]

FIG. 7 describes an example of a processing result of the present embodiment for the sensor detection area F015 before the appearance of the right turn lane at the intersection.

First, a calculation example of the sensor road/lane structure generation unit 105 in FIG 3 described.

It is assumed that the white line information from 10 and the roadside information from 9 in the in 7 shown detection area F015 can be detected. in the in 10 In the example shown in the illustration, in the white line information, an increase in the number of right-turn lanes is recognized, and a white line can be recognized on the left-turn side of the intersection. in the in 9 In the example shown, the roadside information can be used to recognize the left-turn side and the opposite right-turn lane at the intersection.

Using the above data, first, by processing S01 to S04 in 4 , as in 15 shown, the lane center sequences of lanes 1 to 3 are determined from the host vehicle position to the intersection and lane 4 and lane 5 on the left-turn side of the intersection.

Since the white line is broken in front of the intersection, then the entry point to the intersection for lane 2 and lane 3 from 15 by processing S06 in 4 be calculated. Next, the departure point from the intersection for lane 4 and lane 5 can be determined by the processing of S07 in 4 be calculated. Finally, by processing S08 in 4 a track midpoint sequence in the track 6 for connecting the track 2 and the track 4 in the non-track portion can be calculated. In addition, a track midpoint sequence for track 7 connecting track 2 and track 5 can be calculated.

If each track midpoint sequence can be generated, the left-right connection F030 of the respective tracks in 15 can be generated and the front-rear connection F031 can be generated from the positional relationship between the lane midpoint sequences. Then, the regions R01, R02, R03, and R04 are generated as road information based on the lane increase/decrease and the boundary point at the intersection, and the front-rear connection F031 can be generated. Moreover, although the white line for the right-turn side of the intersection has not been recognized, the right-turn side can be recognized as roadside information, and thus the region R06 and the link F032 can be generated.

A calculation example for the card combining unit 106 in 3 described.

It is assumed that the in 6 The information presented is a simplified map. According to the present map, the intersection type is a cross, and in the Link F002 map, there are two lanes and two opposite lanes for the right/left turn side, and one lane and two opposite lanes for the straight ahead.

Comparing the calculation example of the sensor road/lane structure generation unit 105 with the simplified map, the number of lanes and the connection of the roads in the information generated by the sensor road/lane structure generation unit 105 are correct can agree with those of the simplified map, so that a correction in the error correction unit of 3 not happened. However, since the sensor road/lane structure generation unit 105 does not generate the road and lane information for going straight at the intersection and the lane information on the right-turn side of the intersection as shown in FIG 15 shown, the map combining unit 106 combines the road information F033 and the associated lane 11 with the result of the sensor road/lane structure generation unit 105.

On the right-turn side of the intersection, from information that the number of lanes of the simplified map is 2, lane F034 (lane 12 and lane 13) is generated as the lane belonging to the region R06. Then, missing lanes are generated for the region R04 in the intersection. Here track 8 is created, which tracks track 2 and lane 11 connects. Also, track 9 connecting track 3 and track 13 and track 10 connecting track 3 and track 12 are created.

By setting the reliability of the road information and the lane information generated only with the simplified map, it is possible to ensure safety not to excessively increase the speed of the host vehicle as a control. For example, in a case where lane 13 is a target route, since it can be obtained as information generated from a simplified map, a measure such as entering an intersection at low speed and gradually driving until white line information and roadside information can be seen on the right turn side of the intersection. This configuration has the advantage that a situation in which control is not possible and the vehicle cannot be moved when the information is provided only from the sensor can be avoided. In addition, there is an advantage that it is possible to safely operate at the time of anomaly such as turning left in a case where it is difficult to turn right in a situation such as traffic jam where the left turn side can be detected by the sensor to move.

In 7 an example of a processing result of the present embodiment for the sensor detection area F020 at the time of the vehicle moving to the right-turn lane at the intersection will be described.

As the host vehicle moves into the right turn lane at the intersection, more white line information and roadside information can be captured. For example, in the white line information of 10 the white line information on the right-turn side of the intersection can be detected in the detection area F020. Further, in the roadside information of 9 the roadside information on the right/left turn side of the intersection can be detected in the detection area F020.

Therefore, the sensor road/lane structure generation unit 105 as shown in FIG 16 shown, create the lane L12 and the lane L13, which have not been recognized so far, and can create the lane L9 and the lane L10 in the intersection connected thereto. In addition, since no roadside is recognized on the straight-ahead side of the intersection, the region R05 can be generated. However, since the white line is not recognized, the tracking information cannot be generated.

On the other hand, in the processing of the map combining unit 106, since the number of lanes of the simplified map in the region R05 is one, the lane L11 and the lane L8 connected thereto can be generated.

[Description of FIG. 17]

An example of the error correcting unit 308 in the card combining unit 106 is described with reference to FIG 17 described.

In the detection area F020 from 7 it is assumed that on the straight driving side of the intersection, two lanes of lane L11 and lane L14 due to the faulty detection by the sensor at the exit of in 17 illustrated sensor road/lane structure generation unit 105 are generated. When determined only from the white line information, the number of lanes may be incorrect because there are two opposite lanes and one lane on the straight-ahead side at the intersection. At this time, the lane L14 with the information that the number of lanes of the simplified map is one is excluded as an error from the final output. This results in the advantage that very reliable information can be generated with more reliable information and the reliability can be improved.

<Second embodiment>

18 12 shows a configuration of an internal processing block of a driving assistance device in Example 2.

[Configuration]

18 12 illustrates a configuration in the case where a highly accurate map is additionally used as map information from the map/locator unit U1. The configuration of 18 differs from that of 2 in Example 1 by adding a highly accurate map information unit 131, a map road/lane structure generation unit 132, and an optimum road/lane structure selection unit 133.

[data flow]

Although only the difference too 2 will be described, the map/route information unit 101, the position/orientation information unit 102, the sensor information acquisition device 103, the sensor information integration unit 112, the self-position/orientation estimation unit 104 and the sensor road/lane structure generation processing unit 105 and the data flow to these blocks those of 2 similar.

In 18 the data 2031 from the map/locator unit U1 is input to the highly accurate map information acquisition unit 131, and the data 2032 is output. The data 2032 from the highly accurate map information acquisition unit 131 and the data 2037 from the self-position/orientation estimation unit 104 are input to the map road/lane structure generation unit 132, and the data 2033 is output.

The data 2005 from the map/route information unit 101, the data 2018 from the self-position/orientation estimating unit 104 and the data 2009 from the sensor road/lane structure generating unit 105 are input to the map combining unit, and the data 2034 is output. The output dates are similar to the 2010 and 2012 dates in 2 .

The data 2034 from the map combining unit 106 and the data 2033 from the map road/lane structure generation unit 132 are input to the optimal road/lane structure selection unit 133, and the data 2035 and 2036 are output.

Into the route information adding unit 107, the data 2004 from the map/route information adding unit 101, the data 2008 from the self-position/orientation estimating unit 104, and the data 2035 from the optimal road/lane structure selecting unit 133 are input, and the data 2011 is output.

The data 2036 from the optimum road/lane structure selecting unit 133 and the data 2011 from the route information adding unit 107 are input to the control unit 108, and the data 2014 and 2015 are output.

[Description of the function blocks and data content]

Unlike the data 2001, the data 2031 is accepted by the map/locator unit U1 as high-accuracy map information, the high-accuracy map information acquisition unit 131 acquires the data, converts the data as necessary, and stores the data in a memory. Note that the highly accurate map, in addition to the information contained in the in 6 The simplified map shown has lane-level information and contains, for example, information similar to F016, ie information indicating the shape of each lane, as in 8th shown. In addition, as for the shape, the position of the map node, and the like, information is assumed with an accuracy of several cm, unlike the simplified map. An example of the conversion by the high-accuracy map information obtaining unit 131 is the conversion of information in a global coordinate system such as latitude and longitude into a plane coordinate system. This facilitates the subsequent distance calculation.

The data 2032 from the high-accuracy map information unit 131 is assumed to be a high-accuracy map with lane-level information, the data 2037 from the self-position/orientation estimation unit 104 is assumed to be a host vehicle position and orientation information, and the map information/track structure generation unit 132 acquires the data, converts the data as necessary, and stores the data in a memory. An example of the conversion by the map information/track structure generation unit 132 includes, as a data format similar to those in FIG 5 regions R01 to R06 and tracks L01 to L13 shown in Figs. In addition, the global coordinate system data is converted into the relative coordinate system of the host vehicle by the host vehicle position and the heading information from the self-position/heading estimating unit 104 .

The data 2034 from the map combining unit 106 is assumed to be the road and lane structure combined with the simplified map, the data 2033 from the map road/lane structure generating unit 132 is assumed to be the road and lane structure based on the high-precision map, and the Optimal road/lane structure selection unit 133 merges the data considering the accuracy of the data 2034 and the optimal road and lane structure data 2033 to output data 2035 and 2036 . When the reliability of the data 2034 based on the information from the sensor is high, the data 2034 is preferentially merged with the data 2033, but when the reliability is low, the data 2033 is preferentially merged with the data 2044.

The data 2035 of the optimum road/lane structure selection unit 133 has a format similar to that of the data 2010 in FIG 2 , and the route information adding unit 107 generates a route from road and lane information in a similar method as in FIG 2 .

The data 2036 from the optimal road/lane structure selection unit 133 represents road and lane information with higher accuracy using the highly accurate map, and the control unit 108 implements the travel planning unit 109 and the trajectory planning unit 110. Consequently, even in a case where the sensor is not reliable or the simplified map is not reliable, the reliability of the Information can be increased through the use of the high-precision map and more secure control can be carried out.

It is assumed that the data 2011 from the route information adding unit 107 as an input to the control unit 108 of FIG 2 have the same format as the 2011 data from 3 , and the control unit 108 with the same logic can be used both in the configuration of 2 as well as in the configuration of 3 are operated, and there is the effect that the number of man-hours for switching the systems from 2 and 3 is reduced.

It is also assumed that the data 2012 from the card combining unit 106 as an input to the control unit 108 in 2 have the same format as the data 2036 from the optimal road/lane structure selection unit 133 in 3 , and the control unit 108 with the same logic can be used both in the configuration in 2 as well as in the configuration in 3 be operated, and there is the effect that the number of man-hours for switching the systems in 2 and 3 is reduced.

Although the embodiments of the present invention have been described in detail above, the present invention is not limited to the above embodiments, and various design changes can be made without departing from the spirit of the present invention described in the claims. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and the present invention is not necessarily limited to embodiments having all configurations described. In addition, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Further, it is possible to add, delete, and replace another configuration with respect to a part of the configuration of each embodiment.

reference list

1

carrier vehicle

3

driving assistance device

101

Maps/Route information input

102

Positions/Alignment Information Acquisition Unit

103

Sensor Information Retrieval Device

105

Sensor Road/Track Structure Generation Unit

106

card combining unit

107

route information adding unit

109

timetable unit

110

trajectory planning unit

S

sensor

U1

Maps/Locator

D

actuator

QUOTES INCLUDED IN DESCRIPTION

This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent Literature Cited

• JP 6456562 B1 [0004]

Claims (8)

Driver assistance device with: a sensor information acquisition device that acquires sensor information obtained by detecting surroundings of a host vehicle by a sensor; and a sensor road/lane structure generation unit that generates road-lane structure information based on the sensor information, wherein the sensor road/lane structure generation unit generates a lane center sequence of a lane contained in the road on the basis of the information contained in the sensor information about a road, calculates an entry point and an exit point of a non-lane portion of the road based on the lane midpoint sequence, and creates a lane midpoint sequence of a virtual lane for the host vehicle to move from the entry point to the departure point, integrates the lane midpoint sequence of the lane and the lane midpoint sequence of the virtual lane, so that lane structure information is generated showing a connection relationship between the lanes with that interposed between the lanes shows non-track section. driving assistance device claim 1 wherein the sensor road/lane structure generation unit determines a lane increase/decrease limit of the road based on a lane midpoint sequence of the road, lane structure information of the lane based on a position of the lane midpoint sequence of the road and the lane increase/decrease limit of the road, and road information of the road based on the lane increase/decrease limit and the lane structure information. driving assistance device claim 1 further comprising a map combining unit that combines map information selected based on position information of the host vehicle with the road-lane structure information generated by the sensor road/lane structure generation unit. driving assistance device claim 3 wherein the map combining unit comprises an error correction unit that compares the map information with the road-lane structure information and corrects information in the road-lane structure information that does not match the map information, a lane information generation unit that generates on the basis of the map information and the road-lane structure information corrected by the error correction unit generates lane structure information that is not included in the road-lane structure information, and which combines the generated lane structure information with the lane structure information included in the road-lane structure information, and a road information generation unit that generates road information to which the lane structure information summarized by the lane information generation unit belongs. driving assistance device claim 4 , further comprising a map/route information unit that acquires route information and information of a simplified map used for route guidance from navigation, wherein the map combining unit generates a simplified lane center sequence of the virtual lane based on the information of the simplified map, integrates a lane center sequence of the lane and a simplified lane center sequence of the virtual lane, so that the road-lane structure information is generated. The driving assistance device claim 5 further comprises a travel planning unit that determines a road on which the host vehicle is to travel and a lane of the road based on the route information acquired by the map/route information unit and the road-lane structure information combined by the map combination unit. driving assistance device claim 6 wherein the travel planning unit outputs information about a road on which the host vehicle is to travel and a lane of the road to a notification unit that notifies a driver of the host vehicle of the information. driving assistance device claim 6 wherein the travel planning unit generates a trajectory plan of the host vehicle based on information about a road on which the host vehicle is to travel and a lane of the road, and outputs the trajectory plan to an actuator that controls the host vehicle.

Images