Road gradient estimation device and road gradient estimation method

Abstract

To provide a road gradient estimation apparatus which can reduce, in number, positions whose road gradient cannot be estimated.SOLUTION: An apparatus according to the present invention comprises: a coordinate extracting unit 126 for extracting coordinates of a longitudinal-direction existing object existing ahead in a traveling direction of a vehicle; a reference determining unit 127 for determining a road shaping line based on the extracted coordinates in a road width direction and determining a distance change reference value indicating a road longitudinal direction change of distance between one or more pairs of road sharing lines; a coordinate converting unit 124 for converting the coordinates of the detected road shaping lines determined based on a position of the longitudinal-direction existing object detected by an object detection sensor 110 into coordinates on candidate road surfaces of plural kinds having different road gradients; a detection value determining unit 125 for determining a distance change value, with respect to each of the candidate road surfaces, indicating the road longitudinal-direction change of the distance between the detected road shaping lines obtained after conversion of coordinates; and a road gradient estimation unit 128 for estimating the road gradient corresponding to the distance change detected value closest to the distance change reference value among the plurality of distance changed detection values, as the slope of the road on which the vehicle is traveling.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to an apparatus and method for estimating the gradient of a road using information detected by a sensor. [Background technology]

[0002] The road gradient estimation device disclosed in Patent Document 1 captures an image of the road ahead of a vehicle using a camera, and when the captured image includes a broken lane marking, detects multiple endpoints of the broken line.The device then calculates the gradient between the endpoints based on the distance between adjacent endpoints in the longitudinal direction of the road and the distance between adjacent endpoints calculated in advance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-255703 Summary of the Invention [Problem to be solved by the invention]

[0004] The technology disclosed in Patent Document 1 needs to detect the endpoints of lane markings in order to determine road gradient. Therefore, if there are no lane markings where the vehicle is traveling, or if there are lane markings but they are solid lines rather than dashed lines, the technology disclosed in Patent Document 1 cannot estimate road gradient. Therefore, the technology disclosed in Patent Document 1 is unable to estimate road gradient in many locations.

[0005] The present disclosure has been made in light of the above circumstances, and an object thereof is to provide a road gradient estimation device and a road gradient estimation method that can reduce the number of positions where the road gradient cannot be estimated. [Means for solving the problem]

[0006] The above object is achieved by the combination of features recited in the independent claims, and the subclaims define further advantageous specific examples. The reference numerals in parentheses in the claims correspond to specific aspects described in the following embodiments as one aspect, and do not limit the technical scope of the disclosure.

[0007] One disclosure relating to a road gradient estimation device for achieving the above object is: A road gradient estimation device that estimates the gradient of a road on which a vehicle is traveling, comprising: a current position acquisition unit (121) that successively acquires the current position of the vehicle; a coordinate extraction unit (126) for extracting coordinates of longitudinal objects existing in a predetermined range ahead of the current position of the vehicle in the direction of travel of the vehicle from map data including coordinates of longitudinal objects existing along the longitudinal direction of the road; a standard determination unit (127) that, when a plurality of road shape lines representing the shape of the road in the longitudinal direction can be determined in the road width direction based on the coordinates extracted by the coordinate extraction unit, determines a value indicating a change in the distance in the longitudinal direction of the road between one or more pairs of road shape lines determined from the plurality of road shape lines as a distance change standard value; a coordinate conversion unit (124) that converts coordinates of a detected road shape line (130), which is a road shape line that can be determined based on the position of a longitudinally existing object detected by an object detection sensor mounted on a vehicle, into coordinates on a plurality of types of candidate road surfaces (141) having different road gradients; a detection value determination unit (125) that, when there are one or more pairs of detected road shape lines corresponding to the road shape lines for which the distance change reference value has been determined, determines a distance change detection value, which is a value indicating a change in the distance between the detected road shape lines in the road longitudinal direction after coordinate transformation by the coordinate transformation unit, for each candidate road surface; and a road gradient estimation unit (128) that estimates that the road gradient corresponding to the distance change detection value that is closest to the distance change reference value among the plurality of distance change detection values ​​determined by the detection value determination unit is the gradient of the road on which the vehicle is traveling.

[0008] This road gradient estimation device converts the coordinates of a detected road shape line, which can be determined based on the positions of longitudinally existing objects detected by an object detection sensor, into coordinates on multiple types of candidate road surfaces, which are surfaces with different road gradients.

[0009] The degree of change in the distance between detected road shape lines in the road longitudinal direction varies depending on the road gradient. For example, even if two road shape lines extend parallel to each other, the more the road slopes upward, the closer one detected road shape line and the other detected road shape line will be to each other in the road width direction at a distance.

[0010] Therefore, if the detected road shape lines are coordinate-transformed onto a candidate road surface with the same road gradient as the road on which the vehicle is traveling, the change in the distance between the two detected road shape lines in the road longitudinal direction will match the change in the distance between the road shape lines determined from longitudinal entities on the actual road. On the other hand, if the detected road shape lines are coordinate-transformed onto a candidate road surface with a different road gradient from the road on which the vehicle is traveling, the change in the distance between the two detected road shape lines in the road longitudinal direction will not match the change in the distance between the road shape lines determined from longitudinal entities on the actual road.

[0011] This road gradient estimation device uses map data including the coordinates of longitudinally existing objects as data on actual roads, extracts the coordinates of longitudinally existing objects from the map data, and determines road shape lines based on those coordinates. If multiple road shape lines are detected, a value indicating the change in distance between the multiple road shape lines in the longitudinal direction of the road is set as the distance change reference value.

[0012] When there are one or more pairs of detected road shape lines corresponding to the road shape lines for which the distance change reference value has been determined, the detection value determination unit determines, for each candidate road surface, a distance change detection value that indicates a change in the distance between the detected road shape lines in the road longitudinal direction after coordinate transformation, and estimates the road gradient corresponding to the distance change detection value that is closest to the distance change reference value among the multiple distance change detection values ​​as the gradient of the road on which the vehicle is traveling.

[0013] Since the road gradient is estimated in this manner, this road gradient estimation device has the potential to estimate the road gradient if the object detection sensor can detect a longitudinally existing object that can determine the detected road shape line, thereby reducing the number of locations where the road gradient cannot be estimated.

[0014] One disclosure relating to a road gradient estimation method for achieving the above object includes: A road gradient estimation method for estimating a gradient of a road on which a vehicle is traveling, comprising: The current position of the vehicle is sequentially acquired (S1), The coordinates of longitudinal objects existing in a predetermined range ahead of the current position of the vehicle in the direction of travel of the vehicle are extracted from map data including the coordinates of longitudinal objects existing along the longitudinal direction of the road (S6); When a plurality of road shape lines representing the longitudinal shape of the road can be determined in the road width direction based on the extracted coordinates, a value indicating the change in the distance between one or more pairs of road shape lines determined from the plurality of road shape lines in the longitudinal direction of the road is determined as a distance change reference value (S8); The coordinates of the detected road shape line (130), which is a road shape line that can be determined based on the position of a longitudinal object detected by an object detection sensor (110) mounted on the vehicle, are converted into coordinates on a plurality of types of candidate road surfaces (141) having different road gradients (S91, S92); If there are one or more pairs of detected road shape lines corresponding to the road shape lines for which the distance change reference value has been determined, a distance change detection value, which is a value indicating the change in the distance between the detected road shape lines in the road longitudinal direction after coordinate transformation, is determined for each candidate road surface (S93); This is a road gradient estimation method in which the road gradient corresponding to the distance change detection value that is closest to the distance change reference value among a plurality of distance change detection values ​​is estimated to be the gradient of the road on which the vehicle is traveling (S10, S11). [Brief explanation of the drawings]

[0015] [Figure 1] FIG. 2 is a diagram showing the configuration of a vehicle control device. [Figure 2] FIG. [Figure 3]A block diagram showing the functions realized by the ECU. [Figure 4] FIG. 10 is a diagram showing detected road shape lines determined based on lane markings. [Figure 5] FIG. 10 is a diagram showing detected road shape lines determined based on poles. [Figure 6] Conceptual diagram of coordinate transformation. [Figure 7] A diagram showing the formulas used for coordinate conversion. [Figure 8] FIG. 10 is a diagram showing an equation for expressing the height z of the road surface. [Figure 9] FIG. 10 is a diagram showing an equation for expressing the height z of the road surface. [Figure 10] A diagram showing the difference in length in the x-axis direction due to differences in road gradient in coordinate transformation. [Figure 11] FIG. 10 is a diagram showing detected road shape lines after coordinate transformation. [Figure 12] FIG. 4 is a diagram illustrating a method for determining a distance change detection value. [Figure 13] FIG. 3 is a diagram illustrating an angle θ formed by a pair of detected road shape lines. [Figure 14] FIG. 4 is a diagram showing an example of a map road shape line. [Figure 15] FIG. 4 is a diagram showing the flow of a process for estimating a road gradient. [Figure 16] FIG. 16 is a diagram showing detailed processing of S9 in FIG. 15. DETAILED DESCRIPTION OF THE INVENTION

[0016] Hereinafter, an embodiment will be described with reference to the drawings.

[0017] [Overall structure] 1 is a diagram showing the configuration of a vehicle control device 100 including an ECU 120 that functions as a road gradient estimation device. The vehicle control device 100 is mounted on a vehicle 10.

[0018] The vehicle control device 100 includes a GNSS receiver 101 , a vehicle speed sensor 102 , an acceleration sensor 103 , a communication device 104 , a map data storage unit 105 , an object detection sensor 110 , and an ECU 120 .

[0019] The GNSS receiver 101 receives navigation signals transmitted from navigation satellites included in the GNSS (Global Navigation Satellite System), which is a satellite navigation system, and sequentially calculates the current position based on the received navigation signals. The vehicle speed sensor 102 sequentially detects the vehicle speed of the vehicle 10. The acceleration sensor 103 sequentially detects the acceleration occurring in the vehicle 10.

[0020] The communication device 104 communicates wirelessly with a communication device 210 provided in the data center 200. The communication device 104 is a device for receiving high-precision map data from the data center 200 via wireless communication.

[0021] The map data storage unit 105 is a writable non-volatile storage unit that stores high-precision map data. The high-precision map data stored in the map data storage unit 105 is updated with high-precision map data transmitted from the data center 200.

[0022] The object detection sensor 110 is a sensor that detects various objects present around the vehicle 10. The object detection sensor 110 is a sensor that can detect objects present in the longitudinal direction. In FIG. 1, a camera sensor 111, a millimeter-wave radar 112, and a LiDAR 113 are shown as the object detection sensors 110. However, any one or two of these may be provided. Furthermore, other object detection sensors 110 may also be provided.

[0023] A longitudinal object is an object that exists along the longitudinal direction of the road. The object here includes not only three-dimensional objects but also planar objects. A planar object means an object that barely protrudes from the road surface 140. An object that is both a longitudinal object and a planar object includes a lane marking 20.

[0024] Lane markings 20 are drawn on the road surface and indicate the widthwise boundaries of lanes. FIG. 2 shows an example of lane markings 20. FIG. 2 shows the lane markings 20 as follows: a solid lane marking 21 indicating the edge of the road, a dashed lane marking 22, and a solid lane marking 23 that curves in the road width direction as the number of lanes increases. Both the solid lane marking 21 and the dashed lane marking 22 run along the longitudinal direction of the road, and are therefore longitudinal entities. Furthermore, the lane marking 23, which is partially curved in the road width direction, also runs along the longitudinal direction of the road as a whole, and is therefore also a longitudinal entity. Although not shown in FIG. 2, the lane markings 20 are longitudinal entities regardless of their color.

[0025] In addition to lane markings, Figure 2 shows other longitudinally-existing, planar objects. A roadside gutter 31 located at the edge of the road or a cover 32 covering the gutter 31 is also a longitudinally-existing object if it continues in the longitudinal direction of the road to a certain extent. The extent to which it must continue in the longitudinal direction of the road can be determined appropriately based on the accuracy of the road gradient estimation. The boundary 33 of a paved road is also a longitudinally-existing object.

[0026] An example of a three-dimensional object that is a longitudinal entity will be described below. A three-dimensional object that is a longitudinal entity is, for example, a guardrail or a curb. A longitudinal entity only needs to exist along the longitudinal direction of the road, and does not need to exist connected to the longitudinal direction of the road. Poles 134 (see FIG. 5) that are placed at regular intervals to distinguish between the roadway and the sidewalk, or between two adjacent lanes, are also longitudinal entities.

[0027] The object detection sensor 110 does not need to be able to detect all specific longitudinal objects, but only needs to be able to detect at least one type of specific longitudinal object. However, it is preferable that the object detection sensor 110 be able to detect multiple types of longitudinal objects. In particular, it is preferable that the object detection sensor 110 be able to detect multiple types of longitudinal objects, including lane markings 20. This is because lane markings 20 are likely to exist in many positions.

[0028] There are no limitations on the specific configuration of the camera sensor 111, millimeter wave radar 112, and LiDAR 113 as long as they can detect an object existing in the longitudinal direction.

[0029] The ECU 120 can be realized by a configuration including at least one processor. For example, the ECU 120 can be realized by a computer including a processor, a nonvolatile memory, a RAM, an I / O, and a bus line connecting these components. The nonvolatile memory stores a road gradient estimation program for operating a general-purpose computer as the ECU 120. The processor executes the road gradient estimation program stored in the nonvolatile memory while utilizing the temporary storage function of the RAM, causing the ECU 120 to operate as each unit shown in FIG. 3 (described later). Execution of the operation of each unit shown in FIG. 3 means execution of a road gradient estimation method corresponding to the road gradient estimation program.

[0030] The data center 200 includes a communication device 210, a map data storage unit 220, and a map data acquisition unit 230. The communication device 210 wirelessly communicates with the communication device 104 provided in the vehicle 10. The communication device 210 is a device for transmitting high-precision map data to the vehicle 10.

[0031] The map data storage unit 220 is a storage unit that stores high-precision map data. The high-precision map data includes, in addition to data indicating the road shape, one type of coordinates, preferably two or more types of coordinates, of specific longitudinal entities. For example, the high-precision map data includes coordinates of lane markings 20 at predetermined intervals to the extent that the lane markings 20 can be reproduced. The high-precision map data may include three-dimensional point cloud data for at least some areas. Furthermore, at least some areas may be represented by vector data including coordinates and connecting lines connecting the coordinates.

[0032] The high-precision map data may include a plurality of coordinates indicating the positions of gutters 31, a plurality of coordinates indicating the positions of gutters 31 covers 32, and a plurality of coordinates indicating the positions of road boundaries 33 at predetermined intervals. The high-precision map data may also include the coordinates of some three-dimensional objects that exist in the real world. Three-dimensional objects are also called ground objects.

[0033] The map data acquisition unit 230 successively updates the high-precision map data stored in the map data storage unit 220. In addition, the map data acquisition unit 230 acquires high-precision map data of the area around the current position of the vehicle 10 from the map data storage unit 220, and transmits the acquired high-precision map data from the communication device 210 to the communication device 104 mounted on the vehicle 10.

[0034] [Functions of ECU120] 3 shows the functions realized by the ECU 120. A current position acquisition unit 121 acquires the current position detected by the GNSS receiver 101 successively.

[0035] The map data processing unit 122 refers to the current position acquired by the current position acquisition unit 121 and determines whether high-precision map data of a predetermined range determined based on the current position is stored in the map data storage unit 105. If high-precision map data of the required range is not stored in the map data storage unit 105, a signal requesting high-precision map data of the required range is transmitted from the communication device 104 to the data center 200. If high-precision map data is transmitted from the data center 200, the map data processing unit 122 updates the high-precision map data stored in the map data storage unit 105.

[0036] The object recognition unit 123 acquires a sensor signal from the object detection sensor 110, and based on the sensor signal, recognizes various objects that exist around the vehicle 10. The objects to be recognized include obstacles that hinder the travel of the vehicle 10 as well as longitudinally existing objects.

[0037] The coordinate transformation unit 124 determines a road shape line based on the position of the longitudinal object recognized by the object recognition unit 123. The road shape line determined here is set as the detected road shape line 130. If the longitudinal object is a lane marking 20, a line along the recognized lane marking 20, for example, a line passing through the center of the lane marking 20 in the width direction, is set as the detected road shape line 130. FIG. 4 shows the detected road shape line 130. FIG. 4 is a diagram conceptually showing an image captured by the camera sensor 111 (hereinafter, referred to as a camera image). FIG. 4 shows the detected road shape line 130 to show the relationship between the detected road shape line 130 and the lane marking 20. The detected road shape line 130 is not actually displayed in the camera image.

[0038] 4, the detected road shape line 130 is shown by road shape points 131 and line segments connecting adjacent road shape points 131. The detected road shape line 130 can be expressed by points and lines in this way. Alternatively, the detected road shape line 130 may be expressed by a set of road shape points 131 that are denser than the road shape points 131 shown in FIG.

[0039] Fig. 5 shows a detected road shape line 130 and road shape points 131 different from those shown in Fig. 4. In Fig. 5, the detected road shape line 130 is determined using a pole 134 as a longitudinal object. Fig. 5 shows a detected road shape line 130 in which the lower end of the pole 134 is the road shape point 131, and a detected road shape line 130 in which the upper end of the pole 134 is the road shape point 131. If the longitudinal object is a three-dimensional object, the detected road shape line 130 may pass through any height of the three-dimensional object.

[0040] The coordinate conversion unit 124 further converts the coordinates of the detected road shape line 130 into coordinates on multiple types of candidate road surfaces 141 with different road gradients. The candidate road surfaces 141 are candidates that are estimated to be the road surface 140 on which the vehicle 10 is traveling. FIG. 6 shows a conceptual diagram of coordinate conversion. When the detected road shape line 130 is detected by the camera sensor 111, the coordinates of the detected road shape line 130 become the camera coordinate system. The camera coordinate system can be expressed as (u, v, 1). The coordinate conversion unit 124 converts the coordinates of the camera coordinate system, i.e., the object detection sensor coordinate system, into the vehicle coordinate system (x, y, z). The vehicle coordinate system is a coordinate system whose origin is a predetermined part of the vehicle 10, for example, the front end of the vehicle, and the up-down direction of the vehicle 10 is the z-axis direction, the fore-aft direction of the vehicle 10 is the x-axis direction, and the width direction of the vehicle 10 is the y-axis direction.

[0041] The plane including the x-axis and y-axis is defined as the vehicle horizontal plane. The road surface 140 on which the vehicle 10 is traveling is not necessarily parallel to the vehicle horizontal plane. In the example shown in Fig. 6, the height z of the road surface 140 is assumed to change according to z = ax + b, where a is a coefficient representing the road surface gradient, and b is a constant representing the height from the road surface 140 to the origin of the vehicle coordinate system.

[0042] The equation shown in Figure 7 shows the relationship between the coordinates (u, v, 1) in the camera coordinate system and the coordinates (x, y, z) in the vehicle coordinate system. In this equation, the first matrix on the right side is the camera internal parameter, and f x is the focal length in the x-axis direction, f y is the focal length in the y-axis direction, c x , c y are the image center coordinates. The second matrix on the right side is a translation and rotation parameter for converting the camera coordinate system into the vehicle coordinate system. The coordinate conversion unit 124 converts the coordinates of the detected road shape line 130 into coordinates on a plurality of types of candidate road surfaces 141 with different road gradients, using a conversion formula set in advance as exemplified in FIG. 7. The road gradient can be represented by a in FIG. 7. Therefore, the coordinate conversion unit 124 converts the coordinates of the detected road shape line 130 into coordinates in the vehicle coordinate system while changing a by a predetermined value set in advance.

[0043] 6 and 7, the road surface 140 and the candidate road surface 141 are assumed to be a single plane with a constant road gradient, and z = ax + b is used. However, the equation representing the height z of the road surface 140 and the candidate road surface 141 does not have to be a single linear equation. For example, the height z of the candidate road surface 141 may be expressed by a polynomial equation as shown in FIG. 8. Furthermore, as shown in FIG. 9, the height z of the candidate road surface 141 may be expressed by a different equation for each section.

[0044] FIG. 10 shows the difference in length in the x-axis direction after coordinate transformation due to differences in road gradient. FIG. 10 shows three candidate road surfaces 141a, 141b, and 141c with different road gradients. The portion of each candidate road surface 141 shown by a solid line represents the detected road shape line 130 after coordinate transformation. As can be seen from FIG. 10, the steeper the road gradient is, the shorter the length of the detected road shape line 130 in the x-axis direction, and the steeper the road gradient is, the more the detected road shape line 130 extends in the x-axis direction. FIG. 10 shows the xz plane to make it easier to show the relationship between the road gradient and the length of the detected road shape line 130. However, the same applies to the y-coordinate, as the steeper the road gradient is, the more the detected road shape line 130 extends after coordinate transformation.

[0045] 11 shows the detected road shape lines 130 after coordinate transformation for each candidate road surface 141. As shown in Fig. 11, even if the detected road shape lines 130 before transformation are the same, the degree to which the distance in the road width direction between a pair of detected road shape lines 130 after coordinate transformation increases with increasing distance from the vehicle 10 in the road longitudinal direction varies depending on the road surface gradient.

[0046] Returning to Fig. 3 for the explanation, when there are one or more pairs of detected road shape lines 130, the detection value determination unit 125 determines a value indicating the degree to which the distance in the road width direction between the detected road shape lines 130 changes in the road longitudinal direction for one or more pairs of detected road shape lines 130 after coordinate transformation. Hereinafter, this value is referred to as the distance change detection value.

[0047] The distance change detection value will be explained using Figure 12. Figure 12 shows a pair of detected road shape lines 130 shown in the candidate road plane 141c of Figure 11. However, the pair of detected road shape lines 130 are extended so that they intersect. The distance between the detected road shape lines 130 in the road width direction is the distance dy between the pair of detected road shape lines 130 in the y-axis direction. When the x-axis coordinate changes, the value indicating how this distance dy changes is the distance change detection value.

[0048] An example of a distance change detection value is the angle θ formed by a pair of detected road shape lines 130, as shown in Figure 12. If the shape of the detected road shape lines 130 and the angle θ formed by them are known, the change in distance dy can be determined. Therefore, the angle θ formed is the distance change detection value.

[0049] As shown in FIG. 13 , the angle θ is the slope of a relational line that indicates the relationship between the x-coordinate and the distance dy. Real roads are not always straight. Therefore, the detected road shape line 130 may be curved or a combination of multiple straight lines and curves. Even if the detected road shape line 130 is curved or a combination of multiple straight lines and curves, it is possible to discretely extract combinations of the x-coordinate and the distance dy. By plotting the discretely extracted combinations of the x-coordinate and the distance dy on an x-dy plane and approximating the plotted points with a straight line, the slope of the line can be determined. Therefore, even if the detected road shape line 130 is curved or a combination of multiple straight lines and curves, it is possible to determine the angle θ formed by a pair of detected road shape lines 130. Note that the distance change detection value is not limited to the angle θ formed by a pair of detected road shape lines 130. A group of values ​​of the distance dy for each x-coordinate may also be used as the distance change detection value.

[0050] The distance change detection value is a value to be compared with a distance change reference value, which will be described later. Therefore, the detection value determination unit 125 may determine the distance change detection value only if the corresponding distance change reference value can be calculated. The correspondence between the distance change detection value and the distance change reference value will be described later.

[0051] Returning to Fig. 3 for the explanation, the coordinate extraction unit 126 extracts the coordinates of longitudinally existing objects that exist within a predetermined range ahead of the current position of the vehicle 10 in the traveling direction of the vehicle 10 from the high-precision map data stored in the map data storage unit 105. The predetermined range includes at least a part of the range in which the object detection sensor 110 can recognize objects. It is preferable that the predetermined range includes as much of the range in which the object detection sensor 110 can recognize objects as possible. The predetermined range is set in advance in consideration of the processing load while including as much of the range in which the object detection sensor 110 can recognize objects as possible.

[0052] The longitudinal objects whose coordinates are to be extracted preferably include longitudinal objects recognized by the object recognition unit 123 and whose detected road shape line 130 and distance change detection value were determined by the longitudinal objects. Therefore, for example, the coordinate extraction unit 126 can acquire information indicating the specific location of the recognized longitudinal object from the object recognition unit 123, and determine the longitudinal object whose coordinates are to be extracted based on that information. Alternatively, the coordinates of all longitudinal objects included in the predetermined range may be extracted.

[0053] The reference determination unit 127 determines a road shape line based on the coordinates extracted by the coordinate extraction unit 126. The road shape line determined here is referred to as a map road shape line 150. Fig. 14 shows an example of the map road shape line 150. The map road shape line 150 shown in Fig. 14 uses lane dividing lines 20 as the road shape line.

[0054] Furthermore, when a plurality of map road shape lines 150 can be determined in the road width direction, the reference determination unit 127 determines one or more pairs of combinations of map road shape lines 150 from the plurality of map road shape lines 150. Then, for each determined pair of map road shape lines 150, it determines a value (hereinafter referred to as distance change reference value) that indicates the degree to which the distance between the map road shape lines 150 in the road width direction changes in the road longitudinal direction.

[0055] The distance change reference value is determined in the same manner as the distance change detection value. Therefore, one example of the distance change reference value is the angle θ formed by a pair of map road shape lines 150. In the example of FIG. 14, there are three map road shape lines 150a, 150b, and 150c. These three map road shape lines 150a, 150b, and 150c can form three pairs of map road shape lines 150. The first pair is the map road shape lines 150a and 150b. The second pair is the map road shape lines 150a and 150c. The third pair is the map road shape lines 150b and 150c. As shown in the third pair, the pair of map road shape lines 150 does not need to be adjacent to each other. The three map road shape lines 150 in FIG. 14 are parallel to each other. In this case, the angle θ formed by the pair of map road shape lines 150 is 0 degrees.

[0056] Also, as shown in this example, the reference determination unit 127 can use one map road shape line 150 for two or more pairs. In other words, one map road shape line 150 can be used to calculate two or more distance change reference values. Like the reference determination unit 127, the detection value determination unit 125 can also use one detected road shape line 130 for two or more pairs. In other words, one detected road shape line 130 can be used to calculate two or more distance change detection values.

[0057] Returning to Figure 3, the road gradient estimation unit 128 estimates that the road gradient corresponding to the distance change detection value that is closest to the distance change reference value among the multiple distance change detection values ​​determined by the detection value determination unit 125 is the gradient of the road on which the vehicle 10 is traveling.

[0058] The road gradient estimation unit 128 compares the corresponding distance change detection value and distance change reference value. Both the distance change detection value and the distance change reference value are values ​​determined based on longitudinal objects such as lane markings 20. The distance change detection value and the distance change reference value determined from the same longitudinal object correspond to each other. In addition, the distance change detection value and the distance change reference value based on road shape lines that can be determined from longitudinal objects arranged approximately parallel to each other also correspond to each other.

[0059] 5, for example, the direction in which the multiple poles 134 are arranged is parallel to the direction in which the lane markings 21 extend. Therefore, the road shape lines determined from the multiple poles 134 and the road shape lines determined based on the lane markings 21 correspond to each other. When one of the distance change detection value and the distance change reference value is a value determined based on the road shape lines determined from the multiple poles 134 and the other is a value determined based on the road shape lines determined from the lane markings 21, the distance change detection value and the distance change reference value correspond to each other. Which longitudinally-directed objects correspond to which longitudinally-directed objects can be determined based on high-precision map data.

[0060] Note that if the detected road shape line 130 and the map road shape line 150 have a constant curvature, such as a straight line, the positions and ranges in the vehicle fore-and-aft direction of the detected road shape line 130 and the map road shape line 150 do not need to match. This is because, if the detected road shape line 130 and the map road shape line 150 have a constant curvature, even if their positions and ranges in the vehicle fore-and-aft direction change slightly, the distance change detection value and the distance change reference value do not change.

[0061] The vehicle control unit 129 at least temporarily controls one or both of the speed and steering direction of the vehicle 10. To control the vehicle 10, the vehicle control unit 129 uses the vehicle speed detected by the vehicle speed sensor 102, the acceleration detected by the acceleration sensor 103, the high-precision map data stored in the map data storage unit 105, and the road gradient estimated by the road gradient estimation unit 128.

[0062] [Process flow for estimating road gradient] Fig. 15 shows the flow of the process by which ECU 120 estimates the road gradient. ECU 120 periodically executes the process shown in Fig. 15 while vehicle 10 is traveling. In S1, current position acquisition unit 121 acquires the current position of vehicle 10. In S2, object recognition unit 123 acquires a sensor signal and recognizes objects present around vehicle 10 based on the sensor signal.

[0063] In S3, the coordinate extraction unit 126 acquires high-precision map data of a predetermined range determined by the current position of the vehicle 10. S4 is executed by the coordinate extraction unit 126. In S4, it is determined whether there are one or more pairs of longitudinally existing objects that are the same or correspond to the object whose presence is recognized based on the sensor signal and the high-precision map data acquired in S3. If the determination result in S4 is NO, the process returns to S1. If the determination result in S4 is YES, the process proceeds to S5.

[0064] In S5, the coordinate transformation unit 124 determines one or more pairs of detected road shape lines 130 based on the longitudinal detected objects recognized from the sensor signals.

[0065] In S6, the coordinate extraction unit 126 extracts the coordinates of longitudinal objects from the high-precision map data acquired in S3 to determine the map road shape line 150 corresponding to the detected road shape line 130 determined in S5.

[0066] In S7, the reference determination unit 127 determines one or more pairs of map road shape lines 150 based on the coordinates extracted in S6. In S8, the reference determination unit 127 determines a distance change reference value for each pair based on the one or more pairs of map road shape lines 150 determined in S7.

[0067] In S9, a distance change detection value is determined for each gradient candidate. In S9, the process shown in FIG. 16 is executed in detail. In FIG. 16, in S91, the coordinate conversion unit 124 determines the gradient of the candidate road surface 141. S91 to S94 are repeated multiple times. When S91 is executed for the first time, a preset initial value is used as the gradient of the candidate road surface 141.

[0068] In S92, the coordinate conversion unit 124 converts the coordinates of the detected road shape lines 130 determined in S5 onto the surface of the candidate road surface 141 whose gradient was determined in S91. In S93, the detection value determination unit 125 determines a distance change detection value based on each pair of detected road shape lines 130 after the coordinate conversion in S92.

[0069] In S94, the detection value determination unit 125 determines whether or not distance change detection values ​​have been determined for all candidate road surfaces 141. If the determination result in S94 is NO, the process returns to S91. When S91 is executed for the second or subsequent times, a candidate road surface 141 is determined whose road gradient has changed from the previous candidate road surface 141 by a preset angle. Then, S92, S93, and S94 are executed for the newly determined candidate road surface 141. If the determination result in S94 is YES, the process proceeds to S10 in FIG. 15.

[0070] In S10, the road gradient estimation unit 128 determines the distance change detection value that is closest to the distance change reference value determined in S8 from among the plurality of distance change detection values ​​determined in S9.

[0071] In S11, the road gradient estimation unit 128 estimates that the road gradient represented by the candidate road surface 141 at the time when the distance change detection value determined in S10 was determined is the road gradient of the road surface 140 along which the vehicle 10 is about to travel.

[0072] Summary of the embodiment In the embodiment described above, the coordinate conversion unit 124 converts the coordinates of the detected road shape line 130, which can be determined based on the position of a longitudinal object detected by the object detection sensor 110, into coordinates on multiple types of candidate road surfaces 141.

[0073] The degree of change in the distance between detected road shape lines 130 in the road longitudinal direction varies depending on the road gradient. Therefore, if the detected road shape lines 130 are coordinate transformed onto a candidate road surface 141 with the same road gradient as the road on which the vehicle 10 is traveling, the change in the distance between two detected road shape lines 130 in the road longitudinal direction after the coordinate transformation will match the change determined from longitudinal entities on the actual road. On the other hand, if the detected road shape lines 130 are coordinate transformed onto a candidate road surface 141 with a road gradient different from that of the road on which the vehicle 10 is traveling, the change in the distance between two detected road shape lines 130 in the road longitudinal direction after the coordinate transformation will not match the change determined from longitudinal entities on the actual road.

[0074] Therefore, the reference determination unit 127 uses high precision map data including the coordinates of longitudinally existing objects as data on the actual road, extracts the coordinates of the longitudinally existing objects from the high precision map data (S6), and determines the map road shape lines 150 based on the coordinates (S7). If multiple map road shape lines 150 are detected, a value indicating the change in distance between the multiple map road shape lines 150 in the longitudinal direction of the road is set as the distance change reference value (S8).

[0075] When there are one or more pairs of detected road shape lines 130 corresponding to the map road shape line 150 for which the distance change reference value has been determined, the detection value determination unit 125 determines a distance change detection value for each candidate road surface 141. Then, it estimates that the road gradient corresponding to the distance change detection value that is closest to the distance change reference value among the multiple distance change detection values ​​is the gradient of the road on which the vehicle 10 is traveling.

[0076] Since the road gradient is estimated in this manner, there is a possibility that the road gradient can be estimated if the object detection sensor 110 can detect an object existing in the longitudinal direction that can determine the detected road shape line 130. Therefore, it is possible to reduce the number of positions where the road gradient cannot be estimated.

[0077] In addition, the method of this embodiment also has the following effect: In the method of this embodiment, if the curvature of the detected road shape line 130 and the map road shape line 150 determined from longitudinal entities is constant, the positions and ranges of the detected road shape line 130 and the map road shape line 150 in the vehicle longitudinal direction do not need to match.

[0078] In the method disclosed in Patent Document 1, when the endpoints of the dashed line are far from the vehicle 10 or on a downhill slope, the detection distance between the endpoints of the dashed line becomes short, resulting in a decrease in the accuracy of estimating the road gradient.

[0079] Even in this embodiment, there may be cases where the detected road shape line 130 and the map road shape line 150 are far from the vehicle 10. Furthermore, there is a possibility that the positional accuracy of the detected road shape line 130 in the vehicle's fore-and-aft direction may decrease on a downhill slope. However, with the method of this embodiment, even if the positions and ranges of the detected road shape line 130 and the map road shape line 150 in the vehicle's fore-and-aft direction are misaligned, the distance change detection value and distance change reference value do not change as long as the detected road shape line 130 and the map road shape line 150 have a constant curvature. Therefore, with the method of this embodiment, it is possible to prevent a decrease in the estimation accuracy of the road gradient.

[0080] The longitudinal objects only need to be able to determine the longitudinal shape of the road. Therefore, both the solid lane markings 21 and 23 and the dashed lane markings 22 can be included in the longitudinal objects. Three-dimensional objects arranged along the longitudinal direction of the road, such as guardrails, curbs, and poles 134, can also be included in the longitudinal objects. Planar objects that exist along the longitudinal direction of the road outside the lane markings 20 in the road width direction, such as gutters 31 and gutter covers 32, can also be included in the longitudinal objects. The more types of longitudinal objects that can be used to determine distance change detection values, the fewer locations where the road gradient cannot be estimated.

[0081] The reference determination unit 127 can determine two or more distance change reference values ​​using two or more pairs of map road shape lines 150. Furthermore, the detection value determination unit 125 can determine two or more distance change detection values ​​corresponding to the distance change reference values ​​for each candidate road surface 141. When there are two or more distance change reference values ​​and distance change detection values, the road gradient estimation unit 128 estimates the gradient of the road on which the vehicle 10 is traveling based on the two or more distance change reference values ​​and distance change detection values. In this way, the accuracy of estimating the gradient of the road on which the vehicle 10 is traveling is improved compared to estimating the gradient of the road on which the vehicle 10 is traveling based on one distance change reference value and one distance change detection value.

[0082] Furthermore, the reference determination unit 127 can use at least one map road shape line 150 to determine two or more distance change reference values. The detection value determination unit 125 can also use at least one detected road shape line 130 to determine two or more distance change detection values.

[0083] When the reference determination unit 127 uses at least one map road shape line 150 to determine two or more distance change reference values, the number of map road shape lines 150 to be determined in order to determine two or more distance change reference values ​​can be reduced. Also, when the detection value determination unit 125 uses at least one detected road shape line 130 to calculate two or more distance change detection values, the number of 130 to be determined in order to determine two or more distance change detection values ​​can be reduced. Therefore, the amount of calculation by the ECU 120 can be reduced.

[0084] Although the embodiments have been described above, the disclosed technology is not limited to the above-described embodiments, and the following modifications are also included in the scope of the disclosure. Furthermore, various modifications other than those described below can be made without departing from the spirit of the invention.

[0085] <Variation 1> The ECU 120 and methods described herein may be implemented by a special-purpose computer comprising a processor programmed to perform one or more functions embodied in a computer program. Alternatively, the ECU 120 and methods described herein may be implemented by a special-purpose hardware logic circuit. Alternatively, the ECU 120 and methods described herein may be implemented by one or more special-purpose computers comprising a processor executing a computer program in combination with one or more hardware logic circuits, such as an ASIC or FPGA.

[0086] Furthermore, the storage medium for storing the computer program is not limited to a ROM, and the program may be stored in any computer-readable, non-transitory storage medium as instructions to be executed by a computer. For example, the program may be stored in a flash memory. [Explanation of symbols]

[0087] 10: Vehicle 20, 21, 22, 23: Lane markings (longitudinal objects) 31: Gutter (longitudinal object) 32: Cover (longitudinal object) 33: Boundary (longitudinal object) 100: Vehicle control device 105: Map data storage unit 110: Object detection sensor 111: Camera sensor 112: Millimeter wave radar 120: ECU 121: Current position acquisition unit 122: Map data processing unit 123: Object recognition unit 124: Coordinate conversion unit 125: Detection value determination unit 126: Coordinate extraction unit 127: Reference determination unit 128: Road gradient estimation unit 129: Vehicle control unit 130: Detected road shape line 131: Road shape point 134: Pole (longitudinal object) 140: Road surface 141: Candidate road surface 150: Map road shape line 200: Data center 210: Communication device 220: Map data storage unit 230: Map data acquisition unit

Claims

1. A road gradient estimation device that estimates the gradient of a road on which a vehicle is traveling, comprising: a current position acquisition unit (121) for successively acquiring the current position of the vehicle; a coordinate extraction unit (126) for extracting coordinates of longitudinally existing objects existing in a predetermined range ahead of the current position of the vehicle in the traveling direction of the vehicle from map data including coordinates of longitudinally existing objects existing along the longitudinal direction of the road; a reference determination unit (127) that, when a plurality of road shape lines representing the shape of the road in the longitudinal direction can be determined in the road width direction based on the coordinates extracted by the coordinate extraction unit, determines a value indicating a change in the distance in the road longitudinal direction between one or more pairs of road shape lines determined from the plurality of road shape lines as a distance change reference value; a coordinate conversion unit (124) that converts the coordinates of the detected road shape line (130), which is the road shape line that can be determined based on the position of the longitudinal object detected by the object detection sensor mounted on the vehicle, into coordinates on multiple types of candidate road surfaces (141) with different road gradients; a detection value determination unit (125) that, when there are one or more pairs of detected road shape lines corresponding to the road shape line for which the distance change reference value has been determined, determines a distance change detection value, which is a value indicating a change in the distance between the detected road shape lines in the road longitudinal direction after coordinate transformation by the coordinate transformation unit, for each of the candidate road surfaces; a road gradient estimation unit (128) that estimates, among the plurality of distance change detection values ​​determined by the detection value determination unit, the road gradient corresponding to the distance change detection value that is closest to the distance change reference value as the gradient of the road on which the vehicle is traveling.

2. 2. The road gradient estimation device according to claim 1, The road gradient estimation device, wherein the longitudinal entities include solid lane markings and dashed lane markings.

3. 3. The road gradient estimation device according to claim 2, A road gradient estimation device, wherein the longitudinal objects include three-dimensional objects arranged along the longitudinal direction of the road.

4. 4. The road gradient estimation device according to claim 2 or 3, A road gradient estimation device, wherein the longitudinal objects include planar objects that exist along the longitudinal direction of the road and are located outside the lane markings in the width direction of the road.

5. The road gradient estimation device according to any one of claims 1 to 4, the reference determination unit determines two or more distance change reference values ​​using two or more pairs of the road shape lines; the detection value determination unit determines two or more distance change detection values ​​corresponding to the distance change reference value for each of the candidate road surfaces; The road gradient estimation unit estimates the gradient of the road on which the vehicle is traveling based on two or more of the distance change reference values ​​and the distance change detection values.

6. 6. The road gradient estimation device according to claim 5, the reference determination unit uses at least one of the road shape lines to determine two or more of the distance change reference values; The road gradient estimation device, wherein the detection value determination unit uses at least one of the detected road shape lines to determine two or more of the distance change detection values.

7. A road gradient estimation method for estimating a gradient of a road on which a vehicle is traveling, comprising: The current position of the vehicle is sequentially acquired (S1); coordinates of longitudinal objects existing along the longitudinal direction of the road that are located within a predetermined range ahead of the current position of the vehicle in the traveling direction of the vehicle are extracted from map data including coordinates of longitudinal objects existing along the longitudinal direction of the road (S6); When a plurality of road shape lines representing the longitudinal shape of the road can be determined in the road width direction based on the extracted coordinates, a value indicating a change in the distance between one or more pairs of road shape lines determined from the plurality of road shape lines in the longitudinal direction of the road is determined as a distance change reference value (S8); The coordinates of the detected road shape line (130), which is the road shape line that can be determined based on the position of the longitudinal object detected by the object detection sensor (110) mounted on the vehicle, are converted into coordinates on a plurality of types of candidate road surfaces (141) having different road gradients (S91, S92); When there are one or more pairs of detected road shape lines corresponding to the road shape lines for which the distance change reference value has been determined, a distance change detection value, which is a value indicating a change in the distance between the detected road shape lines in the road longitudinal direction after coordinate transformation, is determined for each of the candidate road surfaces (S93). a road gradient corresponding to a distance change detection value that is closest to the distance change reference value among the plurality of distance change detection values, the road gradient being estimated as the gradient of the road on which the vehicle is traveling (S10, S11).

Images