JP3054952B2 - Automatic traveling device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an automatic traveling device that allows a vehicle to travel automatically while searching for a traveling road.

2. Description of the Related Art Recently, there has been developed an automatic traveling device that sets an optimal target route on a traveling route while searching for the traveling route by oneself and controls the traveling of the vehicle so that the vehicle travels on the target route. Have been.

2. Description of the Related Art Conventionally, in an automatic traveling device of this type, a continuous line segment such as a road edge is obtained by imaging an area in a traveling direction of a vehicle by an imaging device attached to the vehicle and processing the captured image. Is extracted, a travelable area in the traveling direction of the vehicle is recognized based on the extracted line segment, a target route for vehicle travel is set in the recognized travelable area, and the current detected Control of a steering angle necessary for merging a vehicle at an origin position on the XY coordinate with a target point appropriately set on a target route set on the XY coordinate plane based on a traveling state of the vehicle in the vehicle The target amount is predictively obtained by a predetermined arithmetic processing, and the traveling control of the vehicle is performed such that the vehicle follows the target route according to the control target amount of the obtained steering angle (Japanese Patent Application No. 63-199610). No. ).

However, in such a conventional automatic traveling device, a target point is set at a distance on a target route, and a control target amount of a steering angle for causing the vehicle to merge with the target point is determined so as to follow the target route. Since such vehicle running control is performed, it is inevitable that the followability to the target route is reduced due to the long-term prediction factor.

An object of the present invention is to provide an operation amount for a target point in the vicinity of the current time by repeatedly calculating an operation amount, so that the traveling control of a vehicle for following a target route can be accurately performed. It is intended to provide an automatic traveling device which is provided.

Configuration Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

In an automatic traveling apparatus according to the present invention, as shown in FIG. 1, an imaging unit 1 such as a video camera attached to a vehicle so as to be able to image an area in the traveling direction of the vehicle, and the imaging unit 1. An image processing unit 2 that performs data processing on an image captured by 1 to extract a continuous line segment such as a road edge, and recognizes a travelable area of a vehicle such as a road according to the extracted continuous line segment. A travelable area recognition unit 3, a target route setting unit 4 for setting a target route for vehicle travel in the recognized travelable area, a vehicle speed sensor 6 for detecting a travel speed v of the vehicle, a yaw accompanying the travel of the vehicle The running state of the current vehicle according to the output of each sensor such as a yaw rate sensor 7 for detecting a yaw rate あ る, which is a change in angular velocity in the direction, and a steering angle sensor 8 for detecting a tire angle δ due to steering of the vehicle. And a control unit 5 for obtaining a control target amount of a steering angle required for the vehicle to travel on a target route based on the detected driving state of the vehicle. The vehicle includes a steering control unit 9 and a steering drive unit 10 for steering the vehicle.

Actually, the image processing unit 2, the travelable area recognition unit 3, the target route setting unit 4, and the control unit 5 are replaced by a microcomputer. In addition, the steering control unit 9 can be included in the computer.

As a video camera in the imaging unit 1, a camera using a telephoto lens or a wide-angle lens is provided in addition to a camera using a standard lens so that an appropriate image can be obtained according to a vehicle speed or a road condition during traveling. Are provided with a plurality of special video cameras such as an infrared camera and an ultra-sensitive camera, and under control of a computer, the plurality of video cameras are appropriately switched and used according to a vehicle speed, a state of a captured image, and the like. It has become.

It is also possible to arrange two video cameras having the same imaging characteristics in parallel to obtain an image by binocular stereovision.

The extraction of continuous line segments such as road edges in the image processing unit 2 is performed as follows.

First, a sampled image sent from the image pickup unit 1 is sampled, and the sampled input image is differentiated to perform image edge detection processing. The automatic threshold value setting circuit automatically sets an optimum threshold value according to the density of the input image at that time, and binarizes the edge image.

At that time, the binarization of the input image may be performed first, and then the differentiation processing for edge detection may be performed. Further, instead of performing binarization, multi-value conversion expressing the shading of an image may be performed.

Next, based on the edge-detected and binarized or multi-valued processed image, information on a continuous line segment of a road edge as shown in FIG. 2, for example, is obtained.

At this time, based on the binarized processed image, an edge tracing process is performed to extract an edge portion of the image having continuity. Further, if the processing for extracting the above-described continuity image edge is performed while growing the area of the input image as the vehicle travels, more accurate road edge information can be extracted. Become.

Since the image captured by the video camera in the imaging unit 1 is based on the perspective projection, the travelable area recognition unit 3 uses the perspective projection road image shown in FIG. 2 as shown in FIG. A projection conversion process, which is a known method for converting the image into a road edge image in which the influence of the perspective projection is eliminated, is performed.

The projection conversion characteristics are set in advance in the travelable area recognition unit 3 in accordance with the characteristics of the perspective projection of the video camera.

Then, based on the image of the road edge subjected to the projective transformation processing, the travelable area recognition unit 3 connects the continuous road edges E1 and E2 between the continuous road edges E1 and E2 as shown in FIG. 11 is recognized as the vehicle travelable area RA on the XY coordinates when the traveling direction is the Y-axis direction.

In FIG. 4, the point P indicates the current position of the vehicle 11, and the center of the lower end of the imaging area of the imaging unit 1 by the video camera is set as the point P at the position of the origin on the XY coordinates. The mounting position of the video camera with respect to the vehicle is set in advance.

Next, when the road ahead of the vehicle, which is the driveable area recognized by the driveable area recognition unit 3, is recognized, the target route setting unit 4 becomes the optimal travel route of the vehicle on the recognized road. The target route is set as follows.

The target route is desirably set so as to be suitable for the running condition of the vehicle at that time, taking into account the road shape and the vehicle speed, as described later. It is set uniformly as follows depending on whether it is narrow or wide.

That is, when the target route setting unit 4 determines that the road width is a wide orbit having a certain width or more, for example, the fourth
As shown in the figure, in the case of a left-hand traffic road, a target route OC is set along the reference edge with a predetermined separation width w of, for example, about 1.5 m from the reference edge on the left side of the road.

When the target route setting unit 4 determines that the track is a narrow track with a road width smaller than a certain value, the target route is set at the center of the road, though not particularly shown.

Then, data of the position of the set target route on the XY coordinates is stored in the internal memory of the target route setting unit 4.

Note that the scale of the travelable area and the target route on the XY coordinates is determined by the magnification of the video camera in the imaging unit 1.

In FIG. 4, the trajectory from the point P to the point O indicates that the vehicle at the point P joins the target route OC by controlling the steering of the vehicle under the control of the control unit 5 as described later. It shows a traveling route up to the start. The point O is the position where the vehicle joins the target route OC at that time.

In the present invention, it is also possible to detect the running state of the vehicle and set an optimal target path of the vehicle on the road as described below according to the detected running state.

That is, the target route setting unit 4 reads, for example, the traveling speed of the vehicle detected by the vehicle speed sensor 6, and if the vehicle speed at that time is within a low speed range equal to or less than a preset threshold value, FIG. As shown in a), the target route OC is set so as to follow the shape of the road.

Similarly, if the vehicle speed at that time is within a high-speed range exceeding a preset threshold value, as shown in FIG.
When traveling on a winding road, a target route OC having a gentle curvature is set in the road so that the lateral acceleration acting on the vehicle is reduced as much as possible.

Next, when the target route on the road is set, the control unit 5 obtains the control target amount of the steering angle for causing the vehicle to travel following the target route as follows.

That is, from the data of the target route OC and the currently detected vehicle speed v, a target value Υs of the yaw rate required to cause the vehicle 11 to travel following the target route OC is determined by a predetermined calculation. The correction amount Δδ of the steering angle is calculated according to the deviation ΔΥ between the obtained target value Δs and the current yaw rate detection value Δx, and the calculated correction amount Δδ
And the currently detected steering angle δx, the control target amount δs
(= Δx + Δδ).

Now, for example, the relations shown in FIGS. 6 (a) and 6 (b), respectively, are given by the following equation as an arithmetic equation for calculating the target value Δs of the yaw rate.

Υs = v / R (1) Here, R is the radius of curvature of the target route derived from the data of the target route.

In the figure, AC is an expected route drawn by the vehicle 11 according to the currently detected yaw rate Δx.

Further, for example, in the relationship shown in FIG.
The target value Υs of the yaw rate may be obtained from the change Δθ in the angle of the traveling direction of the vehicle 11 per unit time Δt according to the following equation.

 Υs = Δθ / Δt (2) where Δθ is given by the following equation.

Here, x and y are positions on the target route OC, Is required.

Then, according to the deviation Δ の between the target value 例 え ば s of the yaw rate theoretically obtained by the calculation and the detected value Υx of the currently detected yaw rate, the correction value Δδ of the steering angle is calculated using, for example, fuzzy inference. By doing
It is derived as follows.

At this time, when a fuzzy controller is used, the input to the fuzzy controller is the target value of the yaw rate Υ
deviation ΔΥ between s and its detection value Υs, and 1
The change ΔE of the deviation ΔΥ during the sampling time is ΔE, and the output is the inference output Δu.

FIG. 8 shows a fuzzy membership function used in the antecedent part of fuzzy inference.

Here, NB, NM, and NS indicate large, medium, and small parts on the negative side, ZO indicates a zero part, and PB, PM, and PS indicate large, medium, and small parts on the positive side, respectively.

Table 1 below shows a fuzzy rule table when the fuzzy membership function shown in FIG. 8 is used.

Here, the asterisks indicate places where fuzzy inference is applied.

As the consequent part of the fuzzy inference, there are roughly two types of inference methods: a centroid type, a linear type, and a linear type. Here, it is necessary to shorten the control processing time and use the differential characteristics. For this reason, a linear inference method will be used.

Now, inference is performed according to the fuzzy rules shown in Table 1.

For example, At this time, y = f (ΔΥ, ΔE) is expressed in a linear form such as y = a · ΔΥ + b · ΔE + c. a, b, c are constants.

Then, the fitness ω ₁ , ω ₂ ,… obtained in the antecedent part is Here, a weighted average value (y ₀ ) is obtained as in the following equation.

y ⁰ = ｛ω ₁ + f ₁ (ΔΥ, ΔE ₁ ) + ω ₂ · f ₂ (ΔΥ ₂ , ΔE ₂ ) + ...｝ × 1 / (ω ₁ + ω ₂ + ...) (5) Then, the inference output Δu Is Δu = y ⁰ (6).

Therefore, the steering angle correction amount Δδ is given by the following equation.

Δδ = Δu · K (7) K is a constant coefficient for determining the steering angle ratio, and is set to about K = 35.

Then, the control target amount δs of the steering angle for joining the vehicle to the target route OC is obtained as δs = δx + Δδ (8).

The steering control unit 9 issues a drive command to the steering drive unit 10 in accordance with the steering angle control target amount δs from the control unit 5 and appropriately drives the steering by the steering drive unit 10 to move the vehicle to the target route OC. The steering that follows the vehicle is performed.

The above processing is repeatedly performed with a predetermined control cycle in consideration of the time required for the processing in advance, and as the traveling of the vehicle progresses, it is set in the travelable area sequentially recognized for each control cycle. The traveling control of the vehicle following the target route OC is continuously performed.

As described above, according to the present invention, it is possible to control the traveling of the vehicle in the same manner by focusing on the lateral load applied to the vehicle body as the vehicle travels, without depending on the yaw rate.

That is, in this case, the target value Gs of the lateral load necessary for causing the vehicle to travel following the target route OC is determined by a predetermined calculation from the data of the target route OC and the currently detected vehicle speed v. Then, using a lateral load sensor (not shown) for detecting a lateral load actually applied to the vehicle body, in accordance with the deviation ΔG from the currently detected lateral load detection value Gx, for example, a fuzzy The steering angle correction amount Δδ is calculated using inference, and the control target amount δs is determined from the calculated correction amount Δδ and the currently detected steering angle δx.

At that time, as an arithmetic expression for calculating the target value Gs of the lateral load,
For example, it is given by:

Gs = v ² / R Effect (9) As described above, in the automatic traveling device according to the present invention, the vehicle travels from image information obtained by capturing an area in the traveling direction of the vehicle with the imaging device attached to the vehicle. Means for recognizing a feasible area, means for setting a target route in the recognized feasible area, means for detecting a traveling speed, a yaw rate or a lateral load, and a steering angle of the vehicle; Means for calculating a target value of the yaw rate or the lateral load required to cause the vehicle to follow the target route from the traveling speed of the vehicle, and a deviation between the calculated target value and the current yaw rate or the detected value of the lateral load. Means for calculating a correction amount of the steering angle from, a means for obtaining a control target amount of the steering angle from the calculated correction amount and the currently detected steering angle,
Means for controlling the running of the vehicle in accordance with the obtained control target amount of the steering angle, without setting the target point far away as in the past, setting the target point near the current time, There is an excellent advantage that the traveling control of the vehicle for following the target route can be accurately performed by determining the operation amount by the repeated calculation.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of an automatic traveling apparatus according to the present invention. FIG. 2 is a view showing road segments obtained by performing data processing on an image taken by a video camera. FIG. 3 is a diagram showing an image obtained by projectively transforming the image of FIG. 2, FIG. 4 is a diagram showing an example of a target route set in a travelable area between recognized road edges, and FIG. FIGS. 6A and 6B show the target routes set on the road at low and high speeds of the vehicle, respectively. FIGS. 6A and 6B show the relationship between the target route and the vehicle, respectively. FIG. 7 is a diagram showing the relationship between the target route and the vehicle, and FIG. 8 is a characteristic diagram of the fuzzy membership function. DESCRIPTION OF SYMBOLS 1 ... Image pick-up part, 2 ... Image processing part, 3 ... Drivable area | region recognition part, 4 ... Target route setting part, 5 ... Control part, 6 ...
Vehicle speed sensor 7, yaw rate sensor 8, steering angle sensor 9, steering control unit 10, steering drive unit 11, vehicle, RA, travelable area, OC, target route

Continuation of the front page (56) References JP-A-2-48704 (JP, A) JP-A-63-273917 (JP, A) JP-A-1-260509 (JP, A) JP-A-2-151572 (JP) , A) JP-A-2-14973 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G05D 1/02 G06F 15/62

Claims (2)

(57) [Claims] 1. A means for recognizing a travelable area of a vehicle from image information obtained by capturing an area in a traveling direction of a vehicle by an imaging device attached to the vehicle, and a target route in the recognized travelable area. Means for setting the vehicle speed, yaw rate or lateral load, and steering angle,
Means for calculating a target value of a yaw rate or a lateral load necessary for causing the vehicle to follow the target path from the target path and the currently detected traveling speed of the vehicle, and a calculated target value and a current yaw rate or a lateral load. A means for calculating a steering angle correction amount from a deviation from a detected load value; a means for obtaining a steering angle control target amount from the calculated correction amount and a currently detected steering angle; Means for controlling the running of the vehicle according to the control target amount of the steering angle. 2. A steering angle correction amount is obtained by fuzzy inference using at least a deviation between a present detected value of the yaw rate and a target value of the yaw rate and a change thereof as a parameter. Automatic traveling device as described.