JPH06300580A - Control device for tracking vehicle course - Google Patents

Abstract

PURPOSE:To improve tracking performance corresponding to change of a curved passage, to mitigate sudden change of steering angle against acute change of the course, and to obtain stable running performance. CONSTITUTION:The deviations d-L1, d-L2 between the extending line O of the center of a vehicle and on objective course passing through the center of the vehicle are computed at a near watching position L1 and a far watching position L2, further a estimated course of the vehicle at the respective watching positions L1, L2 are computed out of steering angle, vehicle speed, etc., and the deviations epsilon-L1, epsilon-L2 between the estimated course and the objective course are computed. Based on the computed deviations d-L1, d-L2, epsilon-L1, epsilon-L2, it is judged whether the course is a straight line or a curved passage of large curvature, or a steady curved passage, or the other. A steering angle computing equation selecting means selects a model equation at setting the steering angle. When the model equation is changed from primary steering angle estimating equation to secondary near point estimation equation and this secondary estimation equation is corrected by primary delay correcting value, the vehicle is controlled to slowly approach the objective course.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle track follow-up control device for making a vehicle follow a track along a track in order to assist a driver's driving operation.

[0002]

2. Description of the Related Art In recent years, problems concerning preventive safety of steering characteristics of a driver and maneuvering stability of a vehicle have become important issues, and the driver is freed from monotonous driving behavior to enable autonomous driving of the vehicle. Technology development is in progress.

Track-following control is a basic technique for safely and correctly driving a vehicle along a track.

Generally, the roadway is composed of straight lines and curves except for undulations, and in order to follow a curved road, it is sufficient to always give a proper steering angle according to the curvature.

As a model for following the vehicle by steering according to the curved road, 1) a program steering model, the steering pattern incorporated in advance is determined based on the characteristics of the vehicle and the pattern recognition of the course in front of the vehicle, and the vehicle is guided to the curved road. The steering angle is controlled by the above steering pattern upon entering the vehicle. 2) Forward error correction model based on primary prediction As shown in Fig. 9 (a), the vehicle 1 at the front gaze distance is
The steering angle is controlled as a steering angle δ (δ = k × d) by multiplying the deviation width d between the extension line O and the target trajectory by a predetermined proportional constant (gain) k 3) Error-correction model A model that predicts the current position, direction, and current motion state when the vehicle continues to run, compares the predicted position with the target trajectory, and follows this deviation so that it approaches zero. Then, as shown in FIG. 9B, the steering angle δ (δ = kΣε) may be obtained by multiplying a value obtained by integrating the shift width ε at a predetermined gaze distance by a proportional constant k.

The simulation of each model is described in detail, for example, in the Special Issue on Automotive Technology and Automotive Engineering (2) Vol.25, No.10, 1971, (P1058-1064).

[0007]

However, in the program steering model, the optimum follow-up performance cannot be obtained unless the setting is made so that the road surface of any shape can be dealt with. There is a problem that the responsiveness delay becomes larger as the change of γ becomes larger.

Further, in the forward error correction model based on the first-order prediction, the tracking controllability on an almost linear trajectory is good, but when traveling on a curved road with a small curvature, the gains differing according to the radius of curvature. If the (proportional constant) is not given, the trajectory cannot be followed accurately.

On the other hand, according to the forward error correction model based on the secondary prediction, even if the curvature of the road surface changes, good tracking performance can be obtained by giving a constant gain (proportional constant). Since the deviation width ε is integrated, the vehicle speed is likely to increase when traveling on a road surface having a large curvature that is almost a straight road, and control is likely to be unstable in this state. In addition, even on a curved road, when the gazing distance is long, on a trajectory having a large curvature change such as an S-shaped road, the curvature of the running road surface is different from the curvature of the gazing road surface. Therefore, it is easy to get out of the orbit.

The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a vehicle track-following control device capable of obtaining excellent tracking performance in response to any track change.

[0011]

A vehicle trajectory tracking control device according to the present invention provides a deviation width between an extension line in front of a vehicle and a target trajectory at a set short-distance gaze position and a target trajectory at a set long-distance gaze position, respectively. Primary prediction deviation width calculating means for calculating,
Calculated by the secondary prediction deviation width calculating means for calculating the deviation width between the predicted trajectory of the vehicle and the target trajectory at the set short distance gaze position and the set long distance gaze position of the vehicle traveling direction, respectively The steering angle calculation formula that selects the model formula when setting the steering angle by determining whether the trajectory is a straight road or a curved road with a large curvature, a steady curved road, or other based on the deviation amount The selecting means, the rudder angle correcting means for setting the first-order lag correction value when the model formula is switched, the corresponding deviation width described above is substituted into the selected model formula, and the first-order lag correction value is corrected. And a rudder angle determining means for determining the rudder angle.

[0012]

[Operation] In the present invention, first, the deviation amount between the extension line in front of the vehicle and the target trajectory corresponding to the set short-distance gazing position and the target trajectory corresponding to the set long-distance gazing position is calculated by the primary prediction deviation width calculating means. Are calculated respectively.

Further, the secondary predicted deviation width calculating means calculates the deviation width between the predicted trajectory at the set short distance gaze position and the set long distance gaze position in the vehicle traveling direction and the corresponding target trajectory.

The rudder angle calculation formula selecting means determines whether at least the track is a straight line or a curved road having a large curvature, a steady curved road, or other than the above based on the deviation calculated by the both deviation calculating means. Select the model formula when setting the steering angle by determining whether or not, and set the first-order lag correction value (hosei) that prevents sudden changes when the above model formula is switched by the steering angle correction means. To do.

Then, the rudder angle deciding means corresponds to the deviation range (d) corresponding to the model formula selected by the rudder angle calculation formula selecting means.
_L1, d_L2, ε_L1, or ε_L2) is substituted, or this deviation width (ε_L1) is corrected by the primary delay correction value (hosei) (ε_L
Substitute 1 ') to determine the steering angle (δ).

[0016]

Embodiments of the present invention will be described below with reference to the drawings.

1 to 8 show an embodiment of the present invention, FIG. 1 is a functional block diagram of a control means, FIG. 2 is a block diagram of a track following control device, and FIG. 3 is a track following control incorporated in a vehicle. 4 is a flow chart showing a steering angle control procedure, FIG. 5 is a schematic diagram showing a vehicle follow-up traveling state, FIG. 6 is a conceptual diagram of track following control, and FIG. 7 is a secondary diagram from primary predicted steering angle control. FIG. 8 is a conceptual diagram showing a state in which the steering angle control is switched to the predicted steering angle control, and FIG. 8 is a time chart showing variations in the steering angle control at the time of switching the steering angle control.

The track following control device is mounted on a vehicle 1 such as an automobile, and video cameras 2R and 2L fixed on both sides of the upper portion of the vehicle 1 at a predetermined interval and oriented forward. A steering actuator 4 fixed to the steering column 3 for controlling the turning angle of the front wheels via a steering shaft (not shown), and an image processing means 5 for processing image signals from the video cameras R and 2L. And a control means 6 for setting a steering angle for steering based on the image data, vehicle speed data detected by the vehicle speed sensor 7 and steering angle sensor 8, steering angle data, etc., and corresponding to the steering angle set by the control means 6. And a steer rig controller 10 that outputs a drive signal to the steer rig actuator 4. Video camera 2
Each of R and 2L uses a solid-state image sensor such as a CCD (charge coupled device) to image a predetermined area in front of the vehicle 1 as indicated by a thin line in FIG.

Further, the image processing means 5 processes the image signals from the video cameras 2R and 2L to obtain an image to be recognized (a white line indicating a width, or an object to be measured) in the luminance of the image. Select and process based on saturation, density, etc.

The control means 1 has a distance calculation means M.
1, a primary prediction deviation width calculating means M2, a secondary prediction deviation width calculating means M3, a steering angle calculation formula selecting means M4, a switching means M5, a steering angle correcting means M6, and a steering angle determining means M7.

In the distance calculation means M1, the set short-distance gaze position L1 and the set long-distance gaze position L2 on the image based on the image data from the image processing means 5 and the trajectory to be traveled (hereinafter referred to as "target trajectory") ) Is calculated. This target track recognizes the width of the running road surface from the white line, for example,
The center of this width is calculated and indexed (see Fig. 5).

The primary prediction deviation width calculating means M2 calculates deviation widths d_L1 and d_L2 between the extension line O of the center of the vehicle 1 and the target track at the short distance position L1 and the long distance position L2 (see FIG. 6). ).

In the secondary prediction deviation width calculating means M3, the target trajectories at the short distance position L1 and the long distance position L2,
The short-distance position L of the vehicle 1 calculated from the current position, direction, motion state, etc. based on the steering angle data, the vehicle speed data, etc.
Deviation width from predicted trajectory at 1 and long-distance position L2 ε_
Calculate L1 and ε_L2 (see FIG. 6).

The rudder angle calculation formula selecting means M4 selects a rudder angle calculation formula preset based on each of the deviation widths d_L1, d_L2, ε_L1 and ε_L2.

In this embodiment, the steering angle calculation formulas are the primary prediction steering angle formula (δ = k1 × d_L1) and the short range secondary prediction formula (δ
= Δ + k2 × ε_L1), long-distance quadratic prediction formula (δ = δ + k
There are three types of model formulas, 2 × ε_L2). The procedure for selecting the steering angle calculation formula will be described later in the flowchart.

In the switching means M5, the proportional constants (gains) k1, k2, k3 corresponding to the steering angle calculation formula selected by the steering angle calculation formula selection means M4 and the deviation widths d_L1, ε_L1,
Switch by inserting ε_L2.

The rudder angle correction means M6 refers to the model formula selected by the rudder angle calculation formula selection means M4, and whether this model formula is changed from the primary predicted rudder angle formula to the short distance primary predicted rudder angle formula. Immediately after switching, first-order delay correction is performed to mitigate a sudden change in steering angle due to model-based switching.

In the steering angle determining means M7, the proportional constant k1 and the deviation width d_L1, the proportional constant k2 and the deviation width ε_L1, the proportional constant k3 and the deviation width ε_L2, or the first-order lag correction value ho are taken in.
The corrected slip width ε_L1 ′ corrected by sei is substituted into the selected steering angle calculation formula to determine the steering angle δ.

Then, a signal corresponding to the steering angle δ is output to the steering controller 10.

Next, the steering angle control procedure will be described with reference to the flowchart of FIG.

In this flowchart, each calculated deviation width d_L1, d_L2, ε_L1, ε_L2 is compared with a predetermined value, and the trajectory is followed depending on whether the deviation width is larger or smaller than each set value do1, do2, εo1, εo2. Select the optimum steering angle type for control.

In the present embodiment, the primary predictive steering angle formula, the long distance secondary predictive steering angle formula, or the short range secondary predictive steering angle formula is switched according to the following criteria.

(1) Primary predicted steering angle formula (δ = k1 × d_L1) d_L1 ≦ do1, d_L2 ≦ do2 and ε_L1 ≦ εo1 and it is determined that the road is a straight road or a gentle curved road.
By using the first-order predicted steering angle formula, traveling performance becomes stable.

(2) Long-distance secondary prediction steering angle formula (δ = δ + k3 × ε_L2) d_L1> do1, d_L2> do2, ε_L1 ≦ εo1 and
In the case of ε_L2 ≤ εo2, it is determined that the road is a steady curved road, and the long-distance quadratic prediction steering angle formula is used to set the gazing point to a distant position to ensure running stability to some extent and then to the curved road. Improves trackability.

(3) Short-distance secondary prediction steering angle formula (δ = δ + k2 × ε_L1) A curved road having a deviation width other than (1) and (2) is a straight road to a curved road or an S-shaped road. It is considered that the trajectory has changed significantly, and the running accuracy is improved by setting the gazing point at a short distance.

The above-mentioned set values are obtained from experiments and the like. For example, do1 = 0.25 (m), do2 = 1.0 (m), εo1
= 0.05 (m) and εo2 = 0.5 (m).

Based on the above criteria, first, step S
At 101, it is determined whether the current trajectory following control is a primary predictive steering angle type.

Then, in the case of the primary predictive steering angle type, the process proceeds to step S102, and in the case of other than the primary predictive steering angle type, the process proceeds to step S103.

In step S102, the steering angle type switching flag flg is set, and in step S104, the steering angle δ is set based on the primary predicted steering angle formula (δ = k1 × d_L1). On the other hand, the above steps S101 to S103
If it proceeds to, it is judged whether the current trajectory following control is the short range secondary predicted steering angle type, and if it is the short range secondary predicted steering angle type, step S
105, and in the case of the long-distance secondary steering angle type, proceeds to step S106. When the process proceeds to step S105, the value of the steering angle type switching flag flg is referred to. The rudder angle switching flag flg determines whether the model formula is the first routine in which the primary predictive rudder angle formula is switched to the short distance secondary predictive rudder angle formula. When flg = 1, it is determined that the model routine is the first routine. , And proceeds to step S107, and if flg = 0, proceeds to step S108.

At step S107, the steering angle correction value h
osei is set in advance by the deviation width ε_L1 (see FIG. 6) from the predicted trajectory at the short distance position L1, and the steering angle switching flag flg is cleared, and the process proceeds to step S109.

On the other hand, when the process proceeds from step S105 to step S108, the steering angle correction value hsei is set to the set value a.
This steering angle correction value hosei is multiplied by (0 <a <1).
Is gradually decreased with each calculation cycle.

Then, when the operation proceeds from step S107 or step S108 to step S109, the correction value hosei is subtracted from the deviation width ε_L1 to set the correction deviation width ε_L1 ', and the correction deviation width ε_L1' is set in step S110. Based on ε_L1 ', the short range secondary predicted steering angle formula (δ =
δ + k2 × ε_L1 ′) is used to set the steering angle δ.

Further, when the process proceeds from step S103 to step S106, the long range secondary predicted steering angle formula (δ = δ + k
3 × ε_L2) is used to set the steering angle δ.

As a result of the above, according to the present embodiment, when the vehicle 1 enters a curved road from a straight road, as shown in FIG.
Predicted vehicle position A when switched to the secondary predicted steering angle type
7 does not change steeply toward the trajectory from the target point B to the trajectory of the target point B, and gradually shifts to the regular target trajectory D along the target trajectory C corrected by the steering angle correction value hosei. Control hunting does not occur when the model formula as indicated by the one-dot chain line is switched, and the steering angle is gently controlled as shown by the solid line in the figure. As a result, stable running performance can be obtained.

Further, when the trajectory tracking control is performed, a primary prediction formula and a secondary prediction formula depending on the curvature of the trajectory, or even if a secondary prediction formula, the gazing point is divided into the near distance side and the far distance side. Since it is selectively used, it has good followability to changes in the trajectory.

[0046]

As described above, according to the present invention,
Since the model formula for setting the steering angle is switched and used according to the change of the curved road, it is possible to accurately follow the traveling corresponding to the change of any track.

[Brief description of drawings]

FIG. 1 is a functional block diagram of control means.

FIG. 2 is a block diagram of a trajectory tracking control device.

FIG. 3 is a block diagram of a track following control device incorporated in a vehicle.

FIG. 4 is a flowchart showing a steering angle control procedure.

FIG. 5 is a schematic diagram showing a follow-up traveling state of a vehicle.

FIG. 6 is a conceptual diagram of trajectory tracking control.

FIG. 7 is a conceptual diagram showing a state in which the primary predictive steering angle control is switched to the secondary predictive steering angle control.

FIG. 8 is a time chart showing fluctuations in the steering angle control when switching the steering angle control.

FIG. 9 is a schematic diagram showing the concept of conventional trajectory following control.

[Explanation of symbols]

 1 ... Vehicle d_L1, d_L2, ε_L1, ε_L2 ... Deviation width L1 ... Set short-distance gaze position L2 ... Set long-distance gaze position M2 ... Primary predicted deviation width calculation means M3 ... Secondary predicted deviation width calculation means M4 ... Steering angle calculation formula Selection means M6 ... Rudder angle correction means M7 ... Rudder angle determination means O ... (Vehicle) extension line hosei ... First-order lag correction value δ ... Rudder angle

Front page continuation (51) Int.Cl. ⁵ Identification code Office reference number FI technical display location G06F 15/62 415 9287-5L // B62D 101: 00 113: 00 137: 00

Claims (1)

[Claims] 1. A deviation width (d_L1, d_L2) between the extension line (O) in front of the vehicle (1) and the target trajectory at the set short-distance gaze position (L1) and the target trajectory at the set long-distance gaze position (L2). Primary prediction deviation width calculating means (M2) for calculating respectively, and the predicted trajectory of the vehicle (1) at the set short-distance gaze position (L1) and the set long-distance gaze position (L2) in the vehicle traveling direction and the target trajectory Deviation width (ε_L1, ε_L2) respectively calculated by the secondary prediction deviation width calculation means (M3) and the deviation width calculated by both deviation width calculation means (M2, M3) (d_L1, d
_L2, ε_L1, ε_L2), a model formula for setting the steering angle (δ) by determining whether the trajectory is a straight road or a curved road with a large curvature, a steady curved road, or other Steering angle calculation formula selection means (M4) to be selected and first-order lag correction value (hosei) when the model formula is switched
The rudder angle correction means (M6) for setting the above-mentioned deviation width (d_L1, d_L2,
ε_L1, or ε_L2) is substituted, or this deviation width (ε_L1)
And a steering angle determining means (M7) for determining a steering angle (δ) by substituting a value (ε_L1 ′) corrected by a first-order delay correction value (hosei) with a trajectory tracking control device for a vehicle.