JP2010079689A - Vehicle - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable a vehicle to efficiently and safely perform follow-up traveling in back and forth directions. <P>SOLUTION: There are a method (control method 1) for deciding a follow-up target point on the basis of the traveling vector of a following vehicle, and a method (control method 2) for deciding the follow-up target point on the basis of the traveling vector of a preceding vehicle, in normal backward and forward follow-up control. In the control method 1, since the vehicle travels inside the orbit of a preceding vehicle when turning, it is difficult to avoid an obstacle existing inside the turning. In the control method 2, since the following vehicle steers to the opposite side of the turning in the initial turning according as the turning radius becomes small, stability is reduced, and it becomes difficult to avoid an obstacle existing outside the turning at that time. A new control method is provided by using the control method 1 and the control method 2. The states in the control method 1 and the control method 2 including an intermediate state between them are dynamically changed, so that it is possible to achieve the optimal follow-up traveling in each case without requiring any complicated arithmetic operation. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a follow-up vehicle control device, for example, to follow-up running that follows another vehicle based on a command from another vehicle or the host vehicle.

When a vehicle travels, a vehicle that auto-cruises while following a preceding vehicle (hereinafter referred to as a preceding vehicle and a following vehicle as a following vehicle) has been proposed.
For example, Patent Document 1 proposes a vehicle including an inter-vehicle distance control type constant speed traveling device that performs constant speed traveling while keeping a preset following inter-vehicle distance setting value following a preceding vehicle.
Moreover, in patent document 2, while making a follow-up driving | running | working with respect to a preceding vehicle, using the inter-vehicle communication, the possibility of crossing of the own vehicle and a preceding vehicle is judged, and when this crossing possibility becomes more than a predetermined value. A technique for stopping the follow-up running has been proposed.
JP-A-10-67254 JP 2007-126146 A

Incidentally, Patent Document 1 and Patent Document 2 describe “longitudinal following traveling” in the case where the following vehicle travels backward with respect to the preceding vehicle. There is no description on how to make it follow efficiently.
In the case of following control including turning in the preceding vehicle, there are conventionally two control methods.
Hereinafter, this control method will be described with reference to FIGS.

(1) First control method This is based on the travel vector of the following vehicle.
FIG. 10A shows a case where the preceding vehicle 80 is traveling straight in the first control method.
As shown in the figure, in the first control method, the follower vehicle 1 has a side in the direction of the travel vector of the follower vehicle 1 (the arrow line starting from the follower vehicle representative point 5 in the figure). A rectangle with the preceding vehicle representative point 2 of the vehicle 80 and the following vehicle representative point 5 of the following vehicle 1 as a diagonal is set, and a target is set at a distance d from the preceding vehicle representative point 2 on the side of the rectangle. Point 4 is set.
Then, the following vehicle 1 controls traveling so as to move toward the target point 4. For this reason, when the following vehicle representative point 5 deviates from the target point 4, the following vehicle 1 is controlled to return to the target point 4, so that the following vehicle 1 follows the position d behind the preceding vehicle 80. .

FIG. 10B shows a case where the preceding vehicle 80 turns by the first control method.
As shown in the figure, in the case of turning using the first control method, the follower vehicle 1 uses the target vector 4 based on the travel vector of the follower vehicle (the arrow line starting from the follower vehicle representative point 5 in the figure). Set.
For this reason, in the first control method, the following vehicle 1 travels inside the locus of the preceding vehicle 80 when turning.

(2) Second control method This is based on the travel vector of the preceding vehicle.
FIG. 10C shows a case where the preceding vehicle 80 is traveling straight in the second control method.
As shown in the figure, in the second control method, the follower vehicle 1 uses the traveling vector of the preceding vehicle of the preceding vehicle 80 (the arrow line starting from the preceding vehicle representative point 2 in the figure) behind the inter-vehicle distance d. A target point 4 is set at a location that is extended only by this amount.
Then, the following vehicle 1 controls traveling so as to move toward the target point 4. For this reason, when the following vehicle representative point 5 deviates from the target point 4, the following vehicle 1 is controlled to return to the target point 4, so that the following vehicle 1 follows the position d behind the preceding vehicle 80. .

FIG. 10D shows a case where the preceding vehicle 80 turns by the second control method.
As shown in the figure, in the case of turning using the second control method, the follower vehicle 2 sets the target point 4 with reference to the travel vector of the preceding vehicle 80.
For this reason, in the second control method, the following vehicle 1 may travel outside the locus of the preceding vehicle 80 when turning.

In the first control method, since the target point 4 is determined based on the travel vector of the following vehicle, the travel distance is shortened. Therefore, the energy consumption is less than that of the preceding vehicle 80 and the calculation is simple. There is a problem that it is difficult to avoid the obstacle when there is an obstacle inside.
In the second control method, since the target point 4 is determined based on the travel vector of the preceding vehicle 80, there is an advantage that the following vehicle 1 is always located behind the preceding vehicle 80 at the time of turning. When there is an obstacle in the vehicle, the traveling locus of the preceding vehicle 80 swells outside, which makes it difficult to avoid the obstacle.
Furthermore, in both control methods, when the turning radius is large, such as on an expressway, the following vehicle 1 can follow the preceding vehicle 80 well, but when the turning radius is small, the traveling locus of the following vehicle 1 becomes unstable. There is.
This problem is particularly noticeable in a small vehicle having a high turning ability, for example, an inverted pendulum vehicle (such as a single-shaft two-wheel vehicle) proposed in Patent Document 1.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to enable efficient and safe following when a vehicle follows and runs in the front-rear direction.

In the first aspect of the present invention, a tracking reference point setting unit that sets a tracking reference point behind a preceding vehicle, a moving target point setting unit that sets a moving target point using the set tracking reference point, There is provided a vehicle comprising travel control means for traveling toward a set movement target point, and change means for changing a distance from the preceding vehicle to the follow-up reference point.
According to a second aspect of the present invention, there is provided an obstacle detection means for detecting an obstacle existing within a range of a predetermined width with respect to the traveling direction, wherein the changing means is the obstacle detection means, and the turning direction of the vehicle 2. The vehicle according to claim 1, wherein when the obstacle is detected, the distance is extended, and when the obstacle is detected in a direction opposite to the turning direction, the distance is shortened. .
According to a third aspect of the present invention, when the obstacle detecting means detects an obstacle, the obstacle is detected within the range of the predetermined width by changing the distance of the tracking reference point by the changing means. The vehicle according to claim 2, further comprising search means for searching for a tracking reference point that is not performed.
According to a fourth aspect of the present invention, there is provided a vehicle stop means for stopping the vehicle when the search means does not search for a tracking reference point where the obstacle is not detected in the range of the predetermined width. A vehicle according to claim 3 is provided.
According to a fifth aspect of the present invention, the vehicle stopping means includes linear path detecting means for detecting a straight path to the preceding vehicle after the vehicle stops running, and has a predetermined width from the detected straight path. The vehicle according to claim 4, wherein when the obstacle is not detected in the range, the traveling with respect to the preceding vehicle is resumed.

  According to the present invention, by providing a virtual joint behind the preceding vehicle and setting a target point of the vehicle that follows this as a reference, it is possible to track efficiently and safely.

Hereinafter, a preferred embodiment of a vehicle of the present invention will be described in detail with reference to FIGS. 1 to 9.
(1) Outline of Embodiment As described in the conventional example, in normal front-rear tracking control, a method for determining a tracking target point based on a travel vector of a tracking vehicle (control target vehicle) (first control method) ) And a method (second control method) for determining a tracking target point based on a travel vector of a preceding vehicle (a vehicle ahead of the vehicle to be controlled).
In the first control method, the traveling of the following vehicle is more stable and energy efficient than the second control method. However, since the vehicle travels inside the trajectory of the preceding vehicle when turning, it exists inside the turn. It is difficult to avoid obstacles.
In the second control method, if the turning radius of the preceding vehicle is large, the following vehicle can follow well, but if the turning radius becomes small, the following vehicle steers to the opposite side of the turn at the beginning of turning, and the stability decreases. In this case, it is difficult to avoid obstacles existing outside the turn.
Therefore, in this embodiment, while realizing a new control method using the first control method and the second control method, including an intermediate state between the first control method and the second control method, By dynamically changing the state, the best follow-up traveling is performed at that time without performing complicated calculations.

(2) Details of Embodiment The vehicle according to the present embodiment can be applied to a normal four-wheel vehicle, but is particularly effective when applied to a small vehicle having a high degree of lateral turning freedom. It is.
Therefore, in this embodiment, a case where the present invention is applied to an inverted pendulum vehicle (such as a single-shaft two-wheel vehicle) that can turn with a small radius will be described as an example. Although the preceding vehicle will be described as an inverted pendulum vehicle, other four-wheel vehicles may be used.

  The inverted pendulum vehicle travels while sensing the posture of the riding section and performing posture control so as to maintain the balance in the front-rear direction in the driving direction of the drive wheels in accordance with the posture. As the attitude control method, for example, various controls disclosed in US Pat. No. 6,302,230, JP-A 63-35082, JP-A 2004-129435, and JP-A 2004-276727 are disclosed. The method can be used.

FIG. 1 shows a configuration related to the follow-up control of the vehicle (follow-up vehicle) of the present embodiment.
As shown in FIG. 1, the vehicle includes an ECU (Electronic Control Unit) 10, a laser radar 20, a vehicle speed sensor 30, a yaw rate sensor 40, a wireless communication device 50, and drive motors 60a and 60b.
Although not shown, a battery is also provided that supplies driving power to the drive motors 60a and 60b and the like and supplies a low-voltage power source for control to the main control unit ECU10.

The ECU 10 includes a computer system that includes a CPU (Central Processing Unit) that performs various calculations, a ROM that stores various programs and data (not shown), a RAM that is used as a work area, an external storage device, an interface unit, and the like. Yes.
In the present embodiment applied to an inverted pendulum vehicle, a posture control program for maintaining the posture, a travel control program for controlling travel based on various instruction signals from the control device, and following the preceding vehicle 80 in the present embodiment. Various programs such as a follow-up traveling program for traveling are stored in the ROM, and the ECU 10 performs corresponding processing by executing these various programs.

  Further, the ECU 10 includes a preceding vehicle relative position detection unit 11, a follow-up control unit 12, and a vehicle control unit 13, and performs follow-up control on the preceding vehicle 80 including the case where the preceding vehicle 80 turns. .

The preceding vehicle relative position detection unit 11 detects the relative position and orientation of the preceding vehicle 80 based on the position and distance of the preceding vehicle 80 measured by the laser radar 20 and supplies the detected position to the tracking control unit 12.
Note that the vehicle speed and yaw rate information may be received from the preceding vehicle 80 by communication.
Further, in the case of performing dead reckoning by receiving position coordinate data from the preceding vehicle 80, a current position detecting device such as GPS may be provided to detect the absolute position of the host vehicle.

The follow-up control unit 12 supplies control command values (front-rear direction command value, left-right direction command value) during straight traveling and turning to the vehicle control unit 13 as follow-up running in the present embodiment.
The follow-up control unit 12 acquires the vehicle speed from the vehicle speed sensor 30 and the yaw rate from the yaw rate sensor 40. Further, the target command value of the preceding vehicle 80 received by the wireless communication device 50 is acquired. The wireless communication device 50 transmits and receives data to and from the preceding vehicle 80 and other vehicles (including following vehicles) by inter-vehicle communication.
The follow-up control unit 12 includes a front-rear direction control circuit and a left-right direction control circuit by feedback control and feedforward control.
The follow-up control unit 12 is based on each input from the preceding vehicle relative position detection unit 11, the vehicle speed sensor 30, the yaw rate sensor 40, and the wireless communication device 50, and a front-rear control command value (target speed or target Acceleration) and a left / right control command value (target rotational angular velocity) by the left / right control circuit are supplied to the vehicle control unit 13.

The vehicle control unit 13 controls the drive motors 60a and 60b.
That is, the vehicle control unit 13 controls the drive motors 60a and 60b so that the target speed (or target acceleration) supplied from the follow-up control unit 12 is achieved. Specifically, the vehicle control unit 13 includes a speed (or acceleration, torque) -current map for the drive motors 60a, 60b, and a target speed supplied from the follow-up control unit 12 according to the speed-current map. Current control is performed so that a current corresponding to (or target acceleration) is output to the drive motors 60a and 60b.

  In addition, the vehicle control unit 13 outputs different currents to both the drive motors 60a and 60b so that the vehicle turns at the target rotational angular velocity (left / right control command value) supplied from the follow-up control unit 12 to drive both. The motors 60a and 60b are differentiated.

  In the present embodiment, an inverted pendulum vehicle has been described. However, the present invention can also be applied to other vehicles such as a four-wheel vehicle, and in this case, the vehicle 1 has a steering motor. Prepare. And the vehicle control part 13 controls a steering motor so that it may become the steering angle corresponding to the target rotational angular velocity u supplied from the follow-up control part 12. FIG.

Each diagram in FIG. 2 is a diagram for explaining the concept of follow-up control according to the present embodiment.
As shown in FIG. 2A, the following vehicle 1 sets the virtual joint 3 behind the preceding vehicle 80.
The virtual joint 3 is a point at a distance d1 in the reverse direction of the traveling vector of the preceding vehicle 80 from the preceding vehicle representative point 2 (arrow line starting from the preceding vehicle representative point 2 in the figure).
The preceding vehicle representative point 2 is a point that represents the position of the preceding vehicle 80, and is defined from the shape of the preceding vehicle 80, for example.
The virtual joint 3 is a point corresponding to a target point in the second control method, and serves as a reference point for the following vehicle 1 to set the target point 4.
The virtual joint 3 functions as a follow-up reference point when the follow-up vehicle 1 follows the preceding vehicle 80. Therefore, the follow-up vehicle 1 includes follow-up reference point setting means for setting a follow-up reference point behind the preceding vehicle. I have.

Further, the following vehicle 1 is a rectangle whose diagonal is the virtual joint 3 and the following vehicle representative point 5 and whose side is the direction of the traveling vector of the following vehicle 1 (an arrow line starting from the following vehicle representative point 5 in the figure). Among the sides parallel to the traveling direction of the following vehicle 1, the target point 4 is set on the side where the preceding vehicle representative point 2 is located with respect to the following vehicle representative point 5. The distance between the virtual joint 3 and the target point 4 is d2.
The following vehicle representative point 5 is a point representing the position of the following vehicle 1 and is defined from the shape of the following vehicle 1, for example.
As shown in FIG. 2A, the following vehicle 1 is arranged such that when the following vehicle representative point 5 and the target point 4 do not coincide with each other, the target point 4 and the following vehicle representative point 5 coincide with each other (following vehicle representative). The following vehicle 1 is moved toward the target point 4 (so that the point 5 converges to the target point 4).
Here, the target point 4 functions as a movement index for moving the follower vehicle 1, and the follower vehicle 1 travels toward the target point 4. Therefore, the follower vehicle 1 uses the follow reference point to move to the target point. And a travel control unit that travels toward the set travel target point.

FIG. 2 (b) shows that the following vehicle 1 has moved toward the target point 4. The dotted line represents the following vehicle 1 before the movement, and the solid line represents the following vehicle 1 after the movement. . In the following vehicle 1 indicated by the solid line, the following vehicle representative point 5 and the target point 4 coincide.
Thus, when the following vehicle representative point 5 deviates from the target point 4, the following vehicle 1 feedback-controls traveling so that the following point is returned to the position of the target point 4, and the preceding vehicle 80 and the following vehicle 1 are as if The virtual link on the preceding vehicle 80 (the line connecting the preceding vehicle representative point 2 and the virtual joint 3) and the virtual link on the following vehicle 1 (the line connecting the target point 4 and the virtual joint 3) are connected by the virtual joint 3. Travel as you were.

As described above, the following vehicle 1 calculates the virtual link on the preceding vehicle 80 side from the traveling vector of the preceding vehicle 80, calculates the virtual link on the following vehicle 1 side from the traveling vector of the following vehicle 1, and these virtual links Follow the preceding vehicle 80 according to the link.
Furthermore, the following vehicle 1 makes the length of d <b> 1 variable by calculation and adjusts the position of the virtual joint 3 to perform finer follow-up control.
For example, if there is an obstacle on the side on which the preceding vehicle 80 turns, this can be avoided by extending the position of the virtual joint 3, and if there is an obstacle outside the turn, This can be avoided by shortening the position of the virtual joint 3.
Hereinafter, the following distance L between the preceding vehicle 80 and the following vehicle 1 is defined by d1 + d2.
In the present embodiment, d1 is variable in the range of 0 ≦ d1 ≦ L. However, other restrictions such as 0 ≦ d1 <L / 2 may be imposed.
Thus, the following vehicle 1 is provided with changing means for changing the distance from the preceding vehicle 80 to the following reference point (virtual joint 3).

Each drawing in FIG. 3 is a diagram for explaining an example of adjusting the position of the virtual joint 3.
These figures show after the following vehicle representative point 5 has converged to the target point 4, and the target point 4 and the following vehicle representative point 5 coincide.

Normally, the following vehicle 1 follows the vehicle at d1 = 0 as shown in FIG. This is the same as the control method 1 described above.
When the preceding vehicle 80 turns, the follower vehicle 1 performs feedback control to extend d1 so as not to contact an obstacle existing in the turning direction, and in parallel, exists in a direction opposite to the turning direction. Feedback control is performed to shorten d1 so as not to come into contact with the obstacle.

First, the control when an obstacle exists on the side of the turning direction will be described.
Now, as shown in FIG. 3A, it is assumed that the preceding vehicle 80 turns right beyond the obstacle 100.
In this case, the following vehicle 1 measures the lateral distance dR to the obstacle 100 on the side of the turning direction (the traveling direction of the preceding vehicle 80). dR is a distance in a direction perpendicular to the traveling direction of the following vehicle 1, and is a distance from the line segment including the travel vector of the following vehicle 1 to the obstacle 100.
The measurement of dR is performed using the laser radar 20.

When dR is measured, the following vehicle 1 determines the magnitude relationship between dR and w / 2 + α. Here, w represents the vehicle width of the following vehicle 1 and α represents the margin distance between the following vehicle 1 and the obstacle.
If dR <w / 2 + α, the following vehicle 1 may come into contact with the obstacle 100 if the vehicle follows straight ahead (in the calculation, the following vehicle 1 passes through a position away from the obstacle 100 by α). As shown in FIG. 3 (b), d 1 is gradually extended until dR ≧ w / 2 + α, and the position of the virtual joint 3 is extended to the rear of the preceding vehicle 80.
The above processing is feedback control in which d1 is a control target and dR is an observation target.

Incidentally, the following vehicle 1 controls d1 so that d1 + d2 = L (constant), and therefore, d1 ≦ L.
Therefore, when the conditional expression dR ≧ w / 2 + α is not satisfied even if d1 = L, the following vehicle 1 determines that the obstacle 100 cannot be avoided, stops traveling, and notifies the preceding vehicle 80 to that effect.
In the above processing, L is constant and d2 is shortened by the amount of extension of d1, but this is an example, and L is variable, and d1 is extended while d2 is constant. Alternatively, d1 may be extended and d2 may be extended.

The following vehicle 1 performs the following processing on the obstacle present on the opposite side of the turning direction in parallel with performing the above processing on the obstacle 100 in the turning direction of the preceding vehicle 80.
As shown in FIG. 4, the following vehicle 1 also measures the lateral distance dL to the obstacle 101 for the obstacle 101 existing outside the turn, similarly to dR.
When dL <w / 2 + α, the following vehicle 1 gradually shortens d1 until dL ≧ w / 2 + α, and moves the position of the virtual joint 3 toward the preceding vehicle 80.
If the conditional expression dL ≧ w / 2 + α is not satisfied even when d1 = 0, the following vehicle 1 determines that the obstacle 101 cannot be avoided, stops traveling, and notifies the preceding vehicle 80 to that effect.
The above processing is feedback control in which d1 is a control target and dL is an observation target.

When d1 satisfying the conditional expression d (R · L) ≧ w / 2 + α can be searched for by the above processing, the following vehicle 1 follows the preceding vehicle 80 without contacting any of the obstacles 100 and 101. You can turn. Here, d (R · L) means dR or dL.
The following vehicle 1 repeats the above processing while turning, and returns d1 to 0 when there is no obstacle in the turning direction. When d1 = 0 is returned, the following vehicle 1 performs normal following traveling.

Further, when d1 satisfying the conditional expression d (R · L) ≧ w / 2 + α cannot be searched and stopped, the following vehicle 1 continues to monitor obstacles ahead from the target point 4 under the condition of d1 = 0. When the tracking becomes possible due to the movement of the obstacle or the change of the position of the preceding vehicle, the preceding vehicle 80 is notified and the tracking is resumed. This can be performed by searching for a direct connection route (a straight route from the following vehicle 1 to the preceding vehicle 80), which will be described later.
Further, if d1 satisfying d (R · L) ≧ w / 2 + α cannot be searched, the following vehicle 1 can be configured to travel between the obstacle 100 and the obstacle 101.

Thus, the following vehicle 1 detects an obstacle existing in the range of a predetermined width in the traveling direction in order to detect an obstacle in the range of w / 2 + α ahead using the laser radar 20. Obstacle detection means is provided.
In the present embodiment, the range of the predetermined width is the area of the predetermined width ahead including the following vehicle 1 and extending in the direction of the travel vector of the following vehicle 1. In addition to this, for example, the target point 4 (movement target point) ) And the virtual joint 3 (following reference point) and other forms are possible, such as a range of a predetermined width.

Further, when the following vehicle 1 detects an obstacle in the turning direction of the following vehicle 1, the distance between the preceding vehicle representative point 2 and the virtual joint 3 is extended, and the obstacle is detected in the direction opposite to the turning direction. To shorten the distance.
When the following vehicle 1 detects an obstacle, the distance from the preceding vehicle 80 to the following reference point (preceding vehicle representative point) such that the obstacle is not detected within a predetermined width range that is an obstacle monitoring area. Search means for searching for a distance d1) from 2 to the virtual joint 3 is provided.
Furthermore, the following vehicle 1 is provided with a vehicle stop unit that stops traveling of the vehicle when the distance from the preceding vehicle 80 to the following reference point (d1 at which no obstacle is detected) is not searched.
In addition, the following vehicle 1 includes a straight route detection unit that detects a straight route (directly connected route) to the preceding vehicle after stopping, and an obstacle is not detected within a predetermined width range on the detected straight route. Then, the traveling with respect to the preceding vehicle 80 is resumed.

The following effects can be obtained by the above tracking control.
First, when there are no obstacles in the vicinity, regardless of the route of the preceding vehicle 80, the following vehicle 1 travels on the optimum route at that time (the effect of the control method 1), so that the energy used can be minimized. it can.
Further, even when there are obstacles in the vicinity, it is possible to follow the preceding vehicle 80 while avoiding contact with the obstacle at a minimum cost without following the route of the preceding vehicle 80.
Furthermore, by adjusting d1 by feedback control, the traveling of the following vehicle 1 can be adjusted in accordance with the possibility of contact with an obstacle.

Next, the procedure of the tracking process performed by the tracking vehicle 1 will be described using the flowchart of FIG.
In the following processing, the CPU of the ECU 10 performs information processing such as calculation and controls each part of the vehicle according to a predetermined program.

When the following vehicle 1 detects a turn of the preceding vehicle 80 using communication with the preceding vehicle 80 or a detected value from the laser radar 20, the variable W is set as W = w / 2 + α (step 5). ), D1 = 0 and d2 = L, and variables d1 and d2 are set (step 10).
Next, the tracking vehicle 1 calculates the target point 4 and performs tracking control to converge to the target point 4 (step 15).

Next, the following vehicle 1 drives the laser radar 20 to search for an obstacle ahead, and analyzes the value of the laser radar 20 to detect dR and dL (step 20).
Next, the following vehicle 1 determines in which direction the preceding vehicle 80 turns, and the lateral distance d (R · L) of the obstacle in the turning direction (dR for a right turn, dL for a left turn) is d. It is determined whether (R · L)> W is satisfied (step 25).
When d (R · L)> W is satisfied (step 25; Y), the following vehicle 1 sets the flag e to e = 0 (step 30), and the lateral distance of the obstacle in the direction opposite to the turning direction. It is determined whether d (L · R) (dL for right turn, dR for left turn) satisfies d (L · R)> W (step 35). The flag e is a flag indicating whether or not an obstacle in the turning direction is processed, and is 1 when processed and 0 when not processed. Then, the following vehicle 1 proceeds to Step 35. Step 35 will be described later.

On the other hand, if d (R · L)> W is not satisfied in step 25 (step 25; N), the following vehicle 1 sets the flag e to e = 1 because there is a possibility of contact with an obstacle in the turning direction. (Step 55), d1 = d1 + b, d1 is extended by b, and d2 = L−d1, and the tracking distance L is kept constant (step 60). b is a length when d1 is gradually extended / reduced.
Then, the following vehicle 1 checks whether or not d1 ≦ L, that is, whether or not the result of repeatedly extending d1 has reached L (step 65).

When d1 ≦ L is not satisfied (step 65; N), that is, when d1 = L, it is difficult to extend d1 further to avoid contact with an obstacle in the turning direction. Stops traveling and notifies the preceding vehicle 80 to that effect (step 70).
Then, the following vehicle 1 stops on the spot, turns, and searches for a directly connected route to the preceding vehicle 80 (step 75).

When the directly connected route can be searched (step 80; Y), the following vehicle 1 resumes the following traveling with respect to the preceding vehicle 80, notifies the preceding vehicle 80 to that effect, and returns to step 5.
If the directly connected route is not a search (step 80; N), the following vehicle 1 repeats step 75.
On the other hand, if d1 ≦ L in step 65 (step 65; Y), contact with the obstacle in the turning direction can be avoided, so the following vehicle 1 moves to step 35 and moves in the direction opposite to the turning direction. Confirmation of existing obstacles is performed (step 35).

In step 35, the following vehicle 1 determines whether or not d (L · R)> W is satisfied. If d (L · R)> W is satisfied (step 35; Y), e = 0. It is determined whether or not (step 90).
If e = 0 is not satisfied (step 90; N), the following vehicle 1 returns to step 15. If e = 0 (step 90; Y), the following vehicle 1 further determines whether or not d1>. (Step 93).
If d1> 0 is not satisfied (step 90; N), the following vehicle 1 returns to step 15. If d1> 0 (step 90; Y), the following vehicle 1 sets d1 to d1-b and d2 to L−. It is set as d1 (step 97).
In this way, it is determined whether or not e = 0 in step 90 because d1 may be positive even when e = 0, and feedback control is performed so that d1 is returned to 0 when d1 is positive. It is.
Such a case may occur, for example, immediately after avoiding an obstacle and on the contrary when there is no obstacle.
Further, in the case where the obstacle in the turning direction is avoided, the following vehicle 1 repeats Y and N for each control cycle in step 25.

In step 35, when d (LR)> W is not satisfied (step 35; N), the following vehicle 1 checks whether or not the flag e is e = 0 (step 40).
If e = 0 is not satisfied (step 40; N), that is, if e = 1, the following vehicle 1 performs steps 55 to 65 to set a minimum d1 that avoids contact with an obstacle. Therefore, d1 cannot be shortened any more. Therefore, in this case, the following vehicle 1 shifts the processing to Step 70, and stops traveling (Step 70), searches for a directly connected route (Step 75), and the like.

On the other hand, when the flag e is e = 0 (step 40; Y), there is a possibility that d1 can be shortened.
Therefore, the following vehicle 1 determines whether d1 satisfies d1> 0 (step 45).
If d1> 0 is not satisfied (step 45; N), that is, if d1 = 0, d1 cannot be shortened, so the following vehicle 1 proceeds to the processing of step 70.
On the other hand, when d1> 0, there is room for shortening d1, so that the following vehicle 1 shortens d1 by b, d1 = d1-b, and d2 = L-d1, and keeps the following distance L constant. (Step 50), and returns to the follow-up control in step 15.

Although b used above can be set appropriately, in this embodiment, b is set as follows as an example.
For example, for an obstacle present on the right side with respect to the traveling direction of the following vehicle 1, first, the absolute value of the time derivative of dR is calculated. This can be approximately calculated by Δd = (| dR (n) −dR (n−1) |) / Δt. Here, Δt is a sampling period of dR (n).
Further, since Δt is constant, Δd = (| dR (n) −dR (n−1) |) may be used as a time derivative.

Next, the following vehicle 1 uses this Δd to calculate b by b = Δd / sin (θ) + β.
Here, θ is a relative angle formed by the traveling vector of the following vehicle 1 and the obstacle (that is, an angle formed by the traveling vector of the following vehicle 1 and the direction in which the obstacle exists as viewed from the following vehicle 1), and β is It is an appropriate constant.
The following vehicle 1 similarly calculates b for dL.
If b is defined in this way, a feedback gain of at least β can be ensured even when the rate of change in the lateral distance between the following vehicle 1 and the obstacle is small.
When the possibility of contact with an obstacle is large, b can be set large, and when the possibility is small, b can be set small.

FIG. 6 is a flowchart for explaining the follow-up control procedure in step 15 (FIG. 5).
First, the following vehicle 1 transmits the position of the preceding vehicle representative point 2 from the preceding vehicle 80 or analyzes the detection value of the laser radar 20, for example. The relative position (the relative position of the preceding vehicle representative point 2) is detected (step 100).

Next, the following vehicle 1 acquires the relative angle of the preceding vehicle 80, that is, the angle formed by the traveling vector of the preceding vehicle 80 and the traveling vector of the following vehicle 1 (step 105).
This can be detected, for example, by detecting the preceding vehicle 80 by the laser radar 20 and analyzing it, or by receiving a travel vector from the preceding vehicle 80.

Next, the following vehicle 1 sets the virtual joint 3 from the preceding vehicle representative point 2 to a position at a distance d1 in the direction opposite to the travel vector of the preceding vehicle 80, and the virtual joint 3 and the following vehicle representative point 5 are diagonally set. The rectangle to be set is set, and the coordinate value of the target point 4 existing at the position of the distance d2 from the virtual joint 3 in the return of the rectangular shape is calculated (step 110).
Next, the following vehicle 1 performs target point convergence control by moving the following vehicle 1 so that the following vehicle representative point 5 coincides with the calculated target point 4 (step 115).

FIG. 7 is a flowchart for explaining the procedure of the direct route search process in step 75 (FIG. 5).
First, following vehicle 1 stops traveling, d1 is set to d1 = 0, and d2 is set to d2 = L (step 150). This is the same parameter setting as normal follow-up running.
Next, the following vehicle 1 turns so as to rotate on the spot where it is stopped, and tracks the preceding vehicle 80 using, for example, dead reckoning (step 155).
Here, the reason why dead reckoning is used is that the laser radar 20 may not be able to track if there is an obstacle.

  When an obstacle is detected by the laser radar 20 on the straight path from the following vehicle 1 to the preceding vehicle 80 (step 160; Y), the following vehicle 1 determines that there is no direct connection route to the preceding vehicle 80. If no obstacle is detected (step 160; N), the following vehicle 1 determines that there is a directly connected route (step 170).

FIG. 8 is a diagram for explaining a method in which the following vehicle 1 monitors the front with the laser radar 20.
The following vehicle 1 monitors the front monitoring area having a width of 2W + γ (γ is a constant) by scanning the laser beam from the laser radar 20 in the horizontal direction.
The following vehicle 1 can specify the relative position of the preceding vehicle representative point 2 by dead reckoning and pattern recognition by the laser radar 20. By communicating with the preceding vehicle 80, the current position or travel vector of the preceding vehicle 80 may be notified, and these may be used.

The following vehicle 1 recognizes the preceding vehicle 80 based on the relative angle of the preceding vehicle 80, the distance to the preceding vehicle 80, and the known outer shape of the preceding vehicle 80, and an object other than the preceding vehicle 80, that is, the obstacle 100. Or the obstacle 101 can be recognized as an obstacle.
The following vehicle 1 can calculate dR from the relative angle r (θb) of the obstacle 100 and the distance to the obstacle 100, and similarly can calculate dL with the obstacle 101.

FIG. 9 is a diagram for explaining a method in which the following vehicle 1 searches for a directly connected route by the laser radar 20.
The laser radar 20 is installed at the right end in the traveling direction in front of the vehicle.
θm is an irradiation angle of the laser beam scanned by the laser radar 20, and a distance to an obstacle existing in the direction of θm is r (θm).
In FIG. 9, as an example, a case where the irradiation angle of the laser beam is θm = θa and θm = θb is illustrated.
The following vehicle 1 monitors the front area of the width Wc when searching for a directly connected route.
When r (θm) cos (θm)> Wc, since the object existing in the direction of θm exists outside the monitoring area Wc, the distance d (θm) to the object existing on the direct connection route is d (Θm) = ∞, that is, it can be considered that there is no obstacle.

On the other hand, when r (θm) cos (θm) ≦ Wc, the distance to the object is d (θm) = r (θm) sin (θm).
Here, assuming that the minimum value of d (θm) is D = min (d (θm)) and the distance to the front object measured by the laser radar 20 is D_LRF, D_LRF = D + dR + dF.
Here, dR is ½ of the vehicle diagonal of the preceding vehicle 80. Regardless of the relative angle θrel of the preceding vehicle 80, a value that most likely approaches the following vehicle 1 was used. DF is a distance from the following vehicle representative point 5 to the front end of the following vehicle 1.

On the other hand, when the front-rear direction inter-vehicle distance in the preceding vehicle recognition is D_LL, the following vehicle 1 determines that there is an obstacle 101 when D_LL> D_LRF, and determines that there is no obstacle 101 when D_LL ≦ D_LRF. .
Then, the following vehicle 1 performs turn feedback control on the spot so that θdir = 0, and if θdir = 0 and D_LL ≦ D_LRF, it is determined that there is a direct connection route.

  As described above, the following vehicle 1 normally follows the preceding vehicle 80 by the first control method (that is, when there is no obstacle), and can save energy required for traveling. By moving the virtual joint 3 to the rear of the preceding vehicle 80 as necessary, obstacles can be avoided and stable running can be performed.

It is the figure which showed the structure relevant to the tracking control of the vehicle of this Embodiment. It is a figure for demonstrating the follow-up control by this Embodiment. It is a figure for demonstrating the example which adjusts the position of a virtual joint. It is a figure for demonstrating the case where an obstruction exists in the outer side of turning. It is a flowchart for demonstrating the procedure of a tracking process. It is a flowchart for demonstrating the procedure of follow-up control. It is a flowchart for demonstrating the procedure of a direct route search process. It is a figure for demonstrating the method in which a following vehicle monitors the front. It is a figure for demonstrating the method for a following vehicle to search. It is a figure for demonstrating a prior art example.

Explanation of symbols

1 following vehicle 2 preceding vehicle representative point 3 virtual joint 4 target point 5 following vehicle representative point 10 ECU
DESCRIPTION OF SYMBOLS 11 Leading vehicle relative position detection part 12 Follow-up control part 13 Vehicle control part 20 Laser radar 30 Vehicle speed sensor 40 Yaw rate sensor 50 Wireless communication apparatus 60a Drive motor a
60b Drive motor b
80 Leading vehicle 100 Obstacle 101 Obstacle

Claims (5)

Tracking reference point setting means for setting a tracking reference point behind the preceding vehicle;
A moving target point setting means for setting a moving target point using the set tracking reference point;
Traveling control means for traveling toward the set movement target point;
Change means for changing a distance from the preceding vehicle to the tracking reference point;
A vehicle characterized by comprising: Comprising obstacle detection means for detecting an obstacle present in a range of a predetermined width with respect to the traveling direction;
The changing means extends the distance when the obstacle detecting means detects an obstacle in the turning direction of the vehicle, and shortens the distance when detecting the obstacle in a direction opposite to the turning direction. The vehicle according to claim 1.   Search means for searching for a tracking reference point where the obstacle is not detected in the range of the predetermined width by changing the distance of the tracking reference point by the changing means when the obstacle detecting means detects an obstacle. The vehicle according to claim 2, further comprising: 4. The vehicle stop unit according to claim 3, further comprising a vehicle stop unit that stops traveling of the vehicle when the search unit does not search a tracking reference point where the obstacle is not detected within the range of the predetermined width. vehicle. A straight path detecting means for detecting a straight path leading to the preceding vehicle after the vehicle stop means stops running of the vehicle;
The vehicle according to claim 4, wherein when an obstacle is not detected within a predetermined width range from the detected straight route, the vehicle resumes traveling with respect to the preceding vehicle.

Images