JP6396645B2 - Travel route generator - Google Patents

Description

  The present invention relates to a travel route generation device that generates a travel route on which a host vehicle travels.

  A technique is known in which the vehicle itself generates a travel route on which the host vehicle travels (see, for example, Patent Document 1). In Patent Document 1, a travel area defined by a forward distance and a constant left and right width is set so as to have a maximum radius of curvature and a maximum left and right width within the travelable area, and the center line of the set travel area Is described as a travel route.

JP 2012-3365 A

  However, if the travel area is set to have the maximum radius of curvature and the maximum lateral width in the travelable area, and the center line of the travel area is the travel route, the radius of curvature of the travel route is always set to the maximum. Is done.

  In this way, when the travel route is generated so that the radius of curvature of the travel route is maximized, in other words, the curvature of the travel route is minimized, when avoiding obstacles ahead or changing lanes, Since the vehicle slowly moves in the lateral direction at the earliest possible timing, the generated lateral acceleration is reduced. However, since the vehicle slowly moves in the lateral direction, there is a greater risk of contact with another vehicle that travels along the destination of the host vehicle.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a travel route generation device that sets the curvature of a travel route so as to achieve an appropriate lateral acceleration.

The travel route generation device of the present invention includes surrounding information acquisition means, travel state acquisition means, region recognition means, curvature setting means, and route generation means.
The surrounding information acquisition unit acquires the surrounding objects and road conditions of the host vehicle as the surrounding information, and the traveling state acquisition unit acquires the traveling state of the host vehicle. The area recognition means recognizes the travelable area and the travel impossible area of the host vehicle based on the surrounding information acquired by the surrounding information acquisition means, and the curvature setting means determines the vehicle speed of the host vehicle acquired as the travel state by the travel state acquisition means. And the curvature of the travel route on which the host vehicle travels is set based on the target lateral acceleration set as an allowable value for the lateral acceleration generated by the traveling of the host vehicle, and the route generation unit is set by the curvature setting unit. Based on the curvature, the travel route along which the host vehicle travels is generated in the travelable region recognized by the region recognition means.

  As a result, in the travelable region, the curvature of the travel route is set to an appropriate value based on the vehicle speed and the target lateral acceleration, and the lateral acceleration generated in the host vehicle when traveling on the travel route is within the target range. Can do. Note that the curvature of the travel route is desirably set based on at least one of the vehicle speed of the host vehicle, the target lateral acceleration, and the distance between the travel route and the untravelable region.

The functional block diagram which shows the driving | running route generation apparatus of this embodiment. The schematic diagram explaining path | route production | generation. The flowchart which shows a route production | generation process. The characteristic view which shows the relationship between the curvature of a driving | running route, and the coefficient of a kernel function. The schematic diagram explaining path | route production | generation. The schematic diagram explaining path | route production | generation.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
The travel route generation device 10 illustrated in FIG. 1 includes a surrounding information acquisition unit 12, a travel state acquisition unit 14, a travelable region recognition unit 20, a curvature parameter setting unit 22, and a route generation unit 30. It is mainly composed of a microcomputer having a CPU, RAM, ROM and the like.

  The surrounding information acquisition unit 12 includes a front sensor whose detection area is a predetermined angle range centered on a straight direction ahead of the vehicle, and a left side sensor whose detection area is a predetermined angle range centered on the vehicle width direction on the left side of the vehicle. Then, based on a signal output from the right side sensor having a predetermined angular range on the right side of the vehicle (similar to the left side sensor) as a detection area, ambient information representing objects and road conditions around the vehicle is acquired.

The front sensor and the left and right side sensors include at least one of an image sensor such as a camera, a laser radar, a millimeter wave radar, a sonar, and the like.
Objects around the vehicle include lane division protrusions, other vehicles, pedestrians, etc., and terrain, buildings, and the like that limit the traveling area of the host vehicle. The surrounding information acquisition unit 12 acquires a position, a size, a height, a length along a road, and the like as information on each object.

  In addition, the surrounding information acquisition unit 12 is configured so that the road conditions include the type of lane lines drawn on the road surface such as the lane boundary line, the lane center line, the lane outside line, and the lane line in a straight line toward the traveling direction of the host vehicle. Information on the shape of the road, which is a curve or a road, and the road width are acquired. The surrounding information acquisition unit 12 may acquire a road shape and a road width from the navigation device as a road situation.

  For example, in FIG. 2A, the surrounding information acquisition unit 12 represents the presence of the parked vehicle 102 and the presence of the roadway outer lines 200 and 202 and the roadway center line 204 as objects around the host vehicle 100 and the road conditions. Obtain as ambient information.

  The traveling state acquisition unit 14 acquires the vehicle speed of the host vehicle from the vehicle speed sensor, acquires the lateral acceleration applied to the host vehicle from the acceleration sensor, and acquires the position of the host vehicle from a satellite positioning device such as a GPS device. The traveling state acquisition unit 14 may obtain the vehicle speed of the host vehicle from the position information acquired from the GPS device.

  The travelable area recognition unit 20 recognizes an area in which the host vehicle can travel based on the surrounding information acquired by the surrounding information acquisition unit 12. For example, in FIG. 2A, the travelable area recognition unit 20 recognizes an area between the roadway outer line 200 and the roadway outer line 202 excluding the parked vehicle 102 as a travelable area. When viewed from the host vehicle 100, the left and right outer sides of the travelable area become the travel impossible area.

  The curvature parameter setting unit 22 is a target lateral acceleration that is set to a value that is allowed as a lateral acceleration that occurs when the host vehicle travels on the traveling speed that the traveling state obtaining unit 14 acquires and the traveling route that the route generating unit 30 generates. Based on the above, a parameter for determining the curvature of the travel route generated by the route generation unit 30 is set. Parameters for determining the curvature of the travel route will be described later.

  In the travelable area recognized by the travelable area recognition unit 20, the route generation unit 30 generates a travel route of the host vehicle that travels in the travelable area based on the curvature parameter set by the curvature parameter setting unit 22.

(Route generation process)
FIG. 3 shows a flowchart of route generation processing executed by the travel route generation device 10. The route generation process is always executed. In FIG. 3, “S” represents a step.

  The surrounding information acquisition unit 12 acquires surrounding information representing objects and road conditions around the host vehicle from the camera image and scanning information scanned by the radar (S400). The traveling state acquisition unit 14 acquires a traveling state such as a vehicle speed, a lateral acceleration, and a vehicle position from a vehicle speed sensor, an acceleration sensor, a GPS, and the like (S402).

  The travelable area recognition unit 20 recognizes the travelable area and the travel impossible area of the host vehicle from the road width, road shape, obstacles, and the like acquired based on the surrounding information acquired in S400 (S404).

Next, in order to explain the curvature parameter set by the curvature parameter setting unit 22 in S406 and the travel route generated by the route generation unit 30 based on the curvature parameter in S408, first, the curvature of the travel route, the vehicle speed, The relationship with acceleration will be described. When the lateral acceleration is a _lateral , the vehicle speed is v, the radius of curvature is R, the curvature is ρ, and the angular velocity is ω, the lateral acceleration a _lateral is simply expressed by the following equation (1).

a _lateral = Rω
a _lateral = v ² / R
a _lateral = ρv ² (1)
From the equation (1), the curvature ρ is expressed by the following equation (2).

ρ = a _lateral / v ² (2)
From equation (2), the curvature of the travel route decreases as the vehicle speed v of the host vehicle increases, and the curvature of the travel route increases as the lateral acceleration a _lateral increases.

  Here, for example, when the host vehicle avoids an obstacle such as a parked vehicle ahead or changes lanes, the driver predicts that a large lateral acceleration will occur if the vehicle speed is fast, and a small lateral vehicle if the vehicle speed is slow. It is predicted that acceleration will occur. If the lateral acceleration is smaller or larger than expected, the driver may feel uncomfortable.

Therefore, it is desirable that the target value of the _lateral acceleration a _lateral is set to a larger value as the vehicle speed is faster, and set to a smaller value as the vehicle speed is slower. The curvature parameter setting unit 22 stores the characteristics of the vehicle speed and the target lateral acceleration in a map or the like. However, an upper limit is set for the target lateral acceleration.

  It is also desirable to set the target lateral acceleration based on the road shape and surrounding conditions. For example, when the curvature of the road curve is small, the target lateral acceleration is decreased, and when the curvature of the curve is large, the target lateral acceleration is increased. In this case, the curvature parameter setting unit 22 stores the characteristics of the road curvature and the target lateral acceleration in a map or the like.

  When emergency avoidance is necessary, the target lateral acceleration is increased. In this case, the curvature parameter setting unit 22 stores, for example, the time and the characteristics of the vehicle speed and the target lateral acceleration that are predicted to collide with the obstacle when the host vehicle is in a steady state in a map or the like.

  It is desirable that the upper limit value of the target lateral acceleration is adjusted depending on the driving tendency of the driver or the driver's setting. For example, the target lateral acceleration is increased for a driver who normally steers large, and the target lateral acceleration is decreased for a driver who steers small. In this case, the curvature parameter setting unit 22 sets the target lateral acceleration using a map obtained by learning driving history, an operation by the driver, or the like.

  The curvature parameter setting unit 22 obtains the curvature ρ of the travel route from Expression (2) based on the vehicle speed at which the host vehicle is traveling and the target lateral acceleration acquired from the map based on the vehicle speed and the road shape.

  Next, the curvature parameter used for the route generation unit 30 to generate a travel route in S408 will be described. For example, in (A) of FIG. 2, the route generation unit 30 travels recognized by the travelable region recognition unit 20 between the roadway outer line 200 and the roadway outer line 202 while the host vehicle 100 avoids the parked vehicle 102. A travel route for traveling in the possible area is generated.

  Based on the information on the travelable area recognized by the travelable area recognition unit 20, the route generation unit 30 sets each point indicating the roadway outer line 200 and the parked vehicle 102 as one of the points as shown in FIG. A feature point belonging to the class 210 is assumed, and each point indicating the roadway outside line 202 is assumed to be a feature point belonging to the other class 212. Then, the path generation unit 30 generates an identification surface for classifying the feature points belonging to the class 210 and the feature points belonging to the class 212 using the support vector machine (SVM) as a discriminator.

  The SVM generates a discrimination plane so that the feature point closest to the discrimination plane among the feature points belonging to the classes 210 and 212 is the support vector, and the distance between the support vector and the discrimination plane is maximized. This identification surface becomes the travel route 220 of the host vehicle 100 traveling in the travelable area.

When generating an identification surface, the path generation unit 30 uses a radial basis function shown in the following equation (3) as a kernel function of the SVM.
K (x, y) = exp (−γ‖x−y‖ ² ) (3)
As shown in FIG. 4, the applicant of the present invention has a characteristic that when one increases between the coefficient γ of Equation (3) and the curvature ρ of the travel route generated by the SVM using Equation (3), the other increases. Noted that there is. The coefficient γ in Expression (3) is a curvature parameter set by the curvature parameter setting unit 22. In FIG. 4, the characteristic of the coefficient γ of the equation (3) and the curvature ρ of the travel route is shown by a straight line, but the actual characteristic is not always a straight line.

  As shown in FIG. 2B, as the coefficient γ decreases, the curvature ρ of the travel route 220 decreases and the curvature radius increases. On the other hand, as the coefficient γ increases, the curvature ρ of the travel route 220 increases and the curvature radius decreases.

  As a result, the host vehicle 100 steers to the right at an earlier timing to avoid the parked vehicle 102 as the coefficient γ is smaller, and steers to the right at a later timing to avoid the parked vehicle 102 as the coefficient γ is larger. Further, the smaller the coefficient γ, the steer to the left at a later timing after passing through the parked vehicle 102, and the greater the coefficient γ, the steer to the left at an earlier timing after passing the parked vehicle 102.

  The curvature parameter setting unit 22 stores the characteristics of the coefficient γ of equation (3) and the curvature ρ of the travel route in a map or the like, and is a curvature parameter from the map based on the curvature ρ acquired from equation (2). The coefficient γ is acquired and output to the route generation unit 30 (S406).

  When the coefficient γ corresponding to the curvature ρ is set by the curvature parameter setting unit 22 in S406, the path generation unit 30 substitutes the coefficient γ set by the curvature parameter setting unit 22 into the equation (3), and is identified by SVM. Create a face. The route generation unit 30 sets the identification surface generated by the SVM as the travel route 220 (S408). The travel route 220 generated by the route generation unit 30 has a shape corresponding to the travelable area based on the coefficient γ set by the curvature parameter setting unit 22. The curvature ρ of the travel route 220 is determined by the coefficient γ.

  As shown in (A) of FIG. 5, even when the road is bent and the travelable area recognized by the travelable area recognition unit 20 is bent, the SVM is as shown in (B) of FIG. 5. When the feature points belonging to the class 240 corresponding to the roadway outer line 230 and the feature points belonging to the class 242 corresponding to the roadway outer line 230 are classified, the feature points belonging to the classes 240 and 242 are the most on the identification plane. The identification surface is generated so that the distance between the support vector that is a close feature point and the identification surface is maximized. As a result, a travel route 250 having a shape along the road shape is generated.

  In addition, as shown in FIG. 6, when the host vehicle 100 makes a left turn at an intersection 260, the surrounding information acquisition unit 12 acquires that the front of the host vehicle 100 is an intersection, and the direction indicator instructs a left turn, When the travel state acquisition unit 14 acquires that the steering is operated to the left side, the travelable region recognition unit 20 recognizes that the travelable region is L-shaped.

  In this way, when turning left at the intersection 260, the curvature parameter setting unit 22 adds the curvature ρ obtained from the equation (2), and the shape of the straight portion of the travelable area before entering the intersection 260 and after exiting the intersection 260 Based on the shape of the corner of the travelable area when making a left turn at the intersection 260, the coefficient γ is decreased at the straight portion and the coefficient γ is increased at the corner.

  As shown in FIG. 6B, the SVM of the route generation unit 30 includes a feature point belonging to the class 270 corresponding to the roadway outer line 262 and a feature point belonging to the class 272 corresponding to the roadway center line 264. The identification surface is generated so that the distance between the support vector, which is the feature point closest to the identification surface, and the identification surface is maximized.

  In FIG. 6B, the SVM generates a straight traveling route 280 in the straight line portion based on the coefficient γ set by the curvature parameter setting unit 22 in accordance with the shape of the travelable region, and the intersection 260 in the curved portion. A curved traveling route 280 along the corner is generated.

  In FIG. 6B, a feature point is set by extending the roadway center line 264 near the center of the intersection 260. However, a feature plane near the center of the intersection 260 is left blank to generate an identification plane. Also good.

  Here, in FIGS. 2, 5, and 6, the distance between the support vector located at the boundary between the travelable area and the travel impossible area and the travel route is closer than a predetermined distance, and the host vehicle 100 is in the travel impossible area. When it is estimated that the vehicle 100 is in contact, that is, when it is estimated that the host vehicle 100 is in contact with the parked vehicle or the roadway outer line (S410: No), the route generation unit 30 returns the process to S406, and the host vehicle is in the non-running region. In order to avoid contact, the curvature parameter setting unit 22 resets the coefficient γ so as to increase the curvature.

The predetermined distance used for determining whether or not the host vehicle 100 is in contact with the untravelable region is set in consideration of the vehicle width and a safety margin for avoiding contact.
In the present embodiment described above, the curvature of the travel route on which the host vehicle travels is set so as to achieve the target lateral acceleration based on the travelable region and the vehicle speed. Thereby, the curvature of a travel route can be set so that it may become a suitable lateral acceleration according to a vehicle speed. As a result, the driver can drive without feeling uncomfortable with the lateral acceleration generated with respect to the vehicle speed.

  In addition, since the travel route is generated with an appropriate curvature based on the travelable region and the vehicle speed, the travel route is always formed with the smallest possible curvature and the largest possible curvature radius, thereby reducing the lateral acceleration generated. When the host vehicle moves sideways along the travel route, it can be avoided as much as possible from coming into contact with another vehicle that travels the destination.

[Other Embodiments]
In the above embodiment, the target lateral acceleration is variably set based on the vehicle speed and the road shape. On the other hand, the target lateral acceleration may be a fixed value regardless of the vehicle speed and the road shape.

  In addition to the SVM, a perceptron or a neural network may be used as an identifier for generating a travel route. When the actual lateral acceleration acquired from the acceleration sensor when traveling on the travel route generated by the route generation unit 30 is out of the range of the target lateral acceleration, the curvature parameter setting unit 22 acquires the actual lateral acceleration acquired from the acceleration sensor. The coefficient γ may be corrected as appropriate so that the acceleration falls within the range of the target lateral acceleration.

  In addition, when the difference between the target travel route generated by the route generation unit 30 and the actual travel route of the host vehicle acquired from the GPS device or the like is a predetermined value or more, the curvature parameter setting unit 22 The coefficient γ may be appropriately corrected so that the difference from the actual travel route is less than a predetermined value.

  In the above embodiment, in S410 of the route generation process shown in FIG. 3, it is determined whether or not the host vehicle 100 is in contact with the untravelable region based on the distance between the travel route and the untravelable region. If the curvature of the travel route generated by is set appropriately, the determination in S410 may be omitted.

Also, when setting the curvature of the travel route from the travelable area and the vehicle speed, the target vehicle speed is set from the road shape or the travelable area, and the curvature of the travel route is set from the set target vehicle speed and the travelable area. Good.
Moreover, the surrounding information acquisition part 12 may acquire surrounding information with one sensor which makes the front and the side a detection area, and the back which makes the detection area the predetermined angle range centering on the straight ahead direction of a vehicle back Ambient information may be acquired from the sensor.

  In the above embodiment, the position of the host vehicle is acquired from a satellite positioning device such as a GPS device. In addition to this, as the surrounding information acquired by the surrounding information acquisition unit 12, for example, the position of the host vehicle may be acquired by collating information such as an actual building, a pedestrian crossing, and a signal with information indicated by the map DB. .

In addition, the travelable area recognition unit 20 may recognize the shape of the travelable area that straddles the lane marking of the road based on the driver's operation such as lane change, merge, and diversion.
In the above embodiment, the coefficient γ, which is a curvature parameter, is acquired from the map based on the curvature ρ. On the other hand, the coefficient γ may be acquired from the map based on the travel pattern obtained from the surrounding information such as the obstacle and the road shape and the traveling state such as the vehicle speed and the steering angle in the surrounding information. Further, the coefficient γ may be acquired using a mathematical expression with these input information as variables.

The present invention is effective when applied to a vehicle that automatically drives based on the generated travel route, because the vehicle generates a travel route with an appropriate curvature so as to achieve the target lateral acceleration.
As described above, the present invention is not limited to the above-described embodiment, and can be applied to various embodiments without departing from the gist thereof.

  10: Traveling route generation device, 12: Ambient information acquisition unit (ambient information acquisition unit), 14: Traveling state acquisition unit (traveling state acquisition unit), 20: Travelable region recognition unit (region recognition unit), 22: Curvature parameter Setting unit (curvature setting unit, acceleration setting unit), 30: path generation unit (path generation unit)

Claims (6)

Ambient information acquisition means (12, S400) for acquiring objects and road conditions around the host vehicle as ambient information;
Traveling state acquisition means (14, S402) for acquiring the traveling state of the host vehicle;
Area recognition means (20, S404) for recognizing a travelable area and a travel impossible area of the host vehicle based on at least a road shape and an obstacle as the ambient information acquired by the ambient information acquisition means;
Travel in which the host vehicle travels based on the vehicle speed of the host vehicle acquired as the driving state by the driving state acquisition means and the target lateral acceleration set as an allowable value for the lateral acceleration generated as the host vehicle travels Curvature setting means (22, S406) for setting the curvature of the route;
Route generating means (30, S408, S410) for generating the travel route in the travelable area recognized by the area recognition means based on the curvature set by the curvature setting means;
Equipped with a,
The path generation means uses a support vector machine as a discriminator, and in the kernel function K (x, y) of the support vector machine, sets a vector representing feature points belonging to the left and right classes of the travelable area as x, and sets the left class A vector representing the discriminating surface that classifies the right class and y
Based on K (x, y) = exp (−γ‖x−y‖ ² ), the identification surface represented by the vector y of the equation is generated as the travel route,
The curvature setting means responds to the curvature based on the characteristic of γ and the curvature of the equation that the curvature decreases when γ of the equation decreases and the curvature increases when γ of the equation increases. Set γ in the above equation,
A travel route generation device (10) characterized by the above.   The travel path generation device according to claim 1, wherein the curvature setting unit decreases the curvature of the travel path as the vehicle speed of the host vehicle increases.   The travel route generation device according to claim 1 or 2, wherein the curvature setting means increases the curvature of the travel route as the target lateral acceleration increases. The route generation means (S410) causes the curvature setting means to reset the curvature so that the curvature of the travel route becomes larger when the distance between the travel route and the non-travelable region becomes shorter than a predetermined distance. The travel route generation device according to any one of claims 1 to 3, wherein   5. The apparatus according to claim 1, further comprising acceleration setting means (22, S 406) for setting the target lateral acceleration based on a shape of a road on which the host vehicle is acquired as the surrounding information. The travel route generation device according to claim 1.   The travel route generation device according to any one of claims 1 to 5, further comprising acceleration setting means (22, S406) for setting the target lateral acceleration based on a vehicle speed of the host vehicle.

Images