JP2008044561A - Lane following controller and vehicle with the controller mounted thereon - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To release a lane following control operation at appropriate timing so as to prevent a driver from feeling discomfort when the driver changes a lane in operating a lane following controller, and to secure vehicle travelling stability in changing the lane. <P>SOLUTION: A lane following controller 1 comprises a driving mode determination unit 11 for obtaining a steering angle δ, detecting whether or not a temporal change state of the angle δ satisfies a prescribed condition and determining whether or not the temporal change state of the angle δ is caused by the lane change operation; a control unit 12 for generating a target yaw rate γ<SB>c</SB>for the lane following control when the determination unit 11 determines that the change state of the angle δ is not caused by the lane change operation, and for generating a target yaw rate γ<SB>δ</SB>for the lane following control when the determination unit 11 determines that the temporal change state of the angle δ is caused by the lane change operation; and a vehicle drive control unit 13 for generating a vehicle control signal M with the target yaw rate obtained from the control unit 12 as a control parameter. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention provides a lane tracking control device capable of accurately performing the estimation by using the steering angle information or the steering angular velocity information for estimation of the driver's operation intention and further performing vehicle motion control adapted to the estimation result. It relates to a car equipped with this.

In recent years, vehicles (automobiles) equipped with a lane tracking control device have been developed. In this type of automobile, when a lane change operation is performed by a driver, the lane tracking control is canceled at an appropriate timing (that is, at an appropriate timing that does not make the driver feel uncomfortable during driving). For example, in the technique of Patent Literature 1, when a lane change is performed during lane tracking control, the driver's intention to change lane is recognized based on the driver's driving behavior pattern stored in advance, and the lane tracking control is canceled. .
JP-A-11-321690

  However, with the technique of Patent Document 1 described above, it is impossible to recognize the intention to change lanes when different drivers drive. In addition, if control for all driving operations is canceled at the time of lane change, a serious accident may occur if a slip or the like occurs. For this reason, when there is a lane change when the lane tracking control is being performed, it is possible to switch the control to a control corresponding to the lane change.

  If the lane change operation is performed when the lane tracking control is being performed, the lane tracking control and the control for changing the lane interfere with each other, causing the following problems. In other words, even if the driver performs a lane change operation based on the intention of changing the lane, the lane tracking control device continues to perform the lane tracking control, so that the lane change cannot be performed smoothly. For example, when the handle is slightly rotated due to carelessness of the drive, the control may be switched to the control for the lane change operation even though the driver does not intend to change the lane.

  Furthermore, it is conceivable to estimate the driver's intention to change the lane in consideration of the driving history, and to perform control based on the estimation result, but no control technology that can respond to high safety has been provided.

  An object of the present invention is to accurately perform the estimation by using history information such as a steering angle to estimate the driver's operation intention, and further to perform vehicle motion control (maintenance of lane tracking control or lane change operation) adapted to the estimation result. To provide a lane tracking control device and a vehicle equipped with the same.

The lane tracking control device of the present invention is summarized as (1) to (3).
(1) In a lane tracking control device that performs lane tracking control using the yaw rate as one of the control information,
From the steering angle and / or the steering angular velocity, whether the vehicle is currently in the lane tracking mode, the lane change mode, or the middle of the transition from one of the lane tracking mode or the lane change mode to the other An operation mode determination unit that determines whether the mode is under;
When the operation mode determination unit determines each mode, a control unit that performs control corresponding to each mode;
With
The controller is
When the driving mode determination unit determines that the vehicle is in the lane following mode, a control signal for a yaw moment or lateral biasing force for lane following driving is generated,
When the driving mode determination unit determines that the vehicle is in the lane change mode, a control signal is not generated, or a control signal of a yaw moment or a lateral urging force for lane change operation is generated,
When the driving mode determination unit determines that the vehicle is in the intermediate mode, when the control signal for the lane following operation is generated immediately before, the control signal for the lane following operation is When the control signal for the lane change operation is generated, the control signal for the lane change operation is generated.
A lane tracking control device characterized by that.

(2) The driving mode determination unit is capable of observing the current lane following mode, the lane change mode, or the intermediate mode under which an observable amount (typically, the steering The lane tracking control device according to (1), wherein the lane tracking control device is determined using an angle or a rotational speed of the steering angle (steering angular speed).

(3) The lane following operation mode, the lane change mode, and the intermediate mode are transition states in a hidden Markov model, and the observable amount is an output symbol in the hidden Markov model (2) Lane tracking control device according to).

The automobile of the present invention is summarized as (4) and (5).
(4) An automobile equipped with the lane tracking control device according to any one of (1) to (3), including a motor that is independently driven on each of the left and right wheels or all the wheels on the rear side or the front side. The vehicle controls the motor. For example, in the case where all the wheels are provided with independently driven motors, the control unit can control the motors of both the left and right wheels of the rear wheels.

(5) A vehicle equipped with the lane tracking control device according to any one of (1) to (3), and capable of independently applying a braking force to the left and right wheels or all the wheels on the rear side or the front side. And the control unit controls the brake. When the automobile is an engine automobile, the brake is a brake mechanism. When the vehicle is an electric vehicle, the brake may be a brake mechanism or a motor braking system.

By using the steering angle (steering angular velocity) or the like for the estimation of the driver's operation intention, the estimation can be performed accurately, whereby the lane tracking control device and the driver can be coordinated.
In addition, since there is no interference between the lane tracking control and the lane change operation, the driver who intends to change the lane does not feel uncomfortable during steering.
Further, by switching to the skid zero control during the lane change, the vehicle can stably travel with respect to a sudden lane change.

  The lane tracking control and the vehicle of the present invention are typically applied to an in-wheel motor type electric vehicle, but can also be applied to a wheel shaft drive type vehicle such as an engine vehicle.

  In the present invention, normally, the steered wheel and the wheel controlled by the lane tracking control device are different, but may be the same. For example, normally, when the front wheels are steered, the rear wheels (drive wheels) are controlled by the lane tracking control device. However, in the case of all-wheel drive, the front wheels are steered and controlled by the lane tracking control device. You may do it.

  FIG. 1 is a functional block diagram showing an embodiment of a lane tracking control device of the present invention. In FIG. 1, a lane tracking control device 1 mounted on an automobile 5 (including a steering wheel 101, front wheels 102, and rear wheels 103) performs lane tracking control using yaw rate γ as one of control information. 11, a control unit 12, and a vehicle motion control unit 13.

Based on the steering angle δ (including the steering angular velocity) or the steering angle history (including the steering angular velocity history), the driving mode determination unit 11 determines whether the vehicle 5 is currently under the lane tracking mode S ₁ or changes to the lane change mode S ₃ . located in or can be determined whether the middle mode under S ₂ in the course of transition from one to the other of these modes. In this embodiment, this determination is made based on the steering angle δ or the steering angle history. However, as shown by the dotted line n in FIG. 1, the lateral deviation Δy (which can be measured by a camera or the like) is referred to (steering). (Refer to the lateral deviation Δy when determining based on the angle or steering angle history). Note that the driving mode can be determined based on the lateral deviation Δy without using the steering angle or steering angle history. However, in this case, since the accuracy of the driving mode determination is reduced, the driving mode is usually based only on the lateral deviation Δy. The mode is not judged.

The control unit 12 includes a lane tracking control yaw rate generation unit 121, a lane change control yaw rate generation unit 122, and a switching unit 123. In FIG. 1, yaw rate generation and yaw rate switching are performed by software.
The lane tracking control yaw rate generating means 121 generates the lane tracking control yaw rate γ _δ when the driving mode determination unit 11 determines that the automobile 5 is under the lane tracking mode S ₁ .

Further, the lane change control yaw rate generating means 122 generates the lane change control yaw rate γ _c when the driving mode determination unit 11 determines that the automobile 5 is under the lane change mode S ₃ .
Further, the control unit 12, the operation mode determination unit 11, the lane when the vehicle 5 when it is determined that the underlying intermediate mode S _2, is to generate the lane line follow-up control yaw rate gamma _[delta] immediately before When the lane change control yaw rate γ _c is generated immediately before the tracking control yaw rate γ _δ , the lane change control yaw rate γ _c is generated.

As described above, the driving mode determination unit 11 acquires the steering angle δ from the steering wheel 101, detects whether or not the history of the steering angle δ meets a predetermined condition, and the history of the steering angle δ determines the lane change operation. (This determination can be made with reference to the lateral deviation Δy as shown by the dotted line n in FIG. 1 as described above).
For example, the operation mode determination unit 11 acquires the time change of the steering angle δ, and determines whether or not the time change state of the steering angle δ satisfies a predetermined condition. Typically, the “time change state of the steering angle δ” is the steering angular velocity dδ / dt, but is not limited thereto, and may include, for example, d ² δ / dt ² .
Thereby, the driver's intention to change lanes can be estimated.
That is, the driving mode determination unit 11 has set three modes, (a) lane tracking mode S ₁ , (b) intermediate mode S _2, and (c) lane change mode S ₃ , as the driving mode. Can be determined by a hidden Markov model based on the observable output symbol η.
In this case, the output symbol can be determined by the steering angle δ and the steering angular velocity dδ / dt.

When the operation mode determination unit 11 determines that the time change state of the steering angle δ is not due to the lane change operation, the control unit 12 determines the lane tracking control yaw rate (first target yaw rate) γ _c (t). When it is determined that the time change of the steering angle δ is due to the lane change operation, a lane change control yaw rate (second target yaw rate) γ _δ (s) is generated. The switching means 123 switches between γ _c (t) and γ _δ (s) in software and outputs the result to the vehicle motion control unit 13. The vehicle motion control unit 13 acquires the first target yaw rate γ _c (t) or the second target yaw rate γ _δ (s), and uses the moment control using γ _c (t) and γ _δ (s) as control parameters. A signal M is generated, and a motor provided in the rear wheel 103 is controlled.

FIG. 2 is a more specific block diagram of the lane tracking control device 1 of FIG. A technique is employed in which the first target yaw rate γ _c (t) and the second target yaw rate γ _δ (s) are switched by the weighting factor w. In the lane tracking control device 1 of FIG. 2, the deviation between the information on the vehicle body lateral position y ^* and the actual vehicle body lateral position y is sent to the driver 21 and the lane tracking control means 231. Based on information on the steering angle δ (including information on the steering angular velocity) by the steering operation of the driver 21, the driving intention determination means (corresponding to the driving mode determination unit 11 in FIG. 1) 22 changes the lane. Either the lane tracking control means 231 or the control by the lane change control means 232 is selected according to the determination by the driving intention determination means 22. Which of the lane following control means 231 and the lane change control means 232 performs control is determined by the probability (w) and the probability (1-w). That is, either the output of the lane tracking control means 231 or the lane change control means 232 is selected according to the probability (reference numerals 241 and 242) given by the driving intention determination means 22 (corresponding to the switching means 123 in FIG. 1) 25. Select. Based on the selection of the selector 25, the vehicle movement control unit 26 (corresponding to the vehicle movement control unit 13 in FIG. 1) generates a moment in the electric vehicle 27 (corresponding to the motor provided on the wheel of the rear wheel 103 in FIG. 1). A control signal M is transmitted.

Hereinafter, the operation of the lane tracking control device 1 of FIG. 2 will be described in detail. In FIG. 2, the first target yaw rate γ _c (t) is a target yaw rate for zero front lateral deviation using the in-vehicle camera system, and a quadratic prediction formula (1) using the vehicle body front lateral deviation y _sr. ) To calculate in real time.
γ _c (t) = − (2 V / l _c ² ) y _sr (1)
V: body speed, l _c : forward gaze distance of camera, y _sr : forward lateral deviation The second target yaw rate γ _δ (s) in FIG. 2 is a target yaw rate for zero body slip angle, and is a first order lag Calculated by equation (2). Note that the second target yaw rate γ _δ (s) used in the present embodiment does not zero out the side slip angle in a steady state.
γ _δ (s) = [k / (τ ₀ s + 1)] × δ _sw (s) (2)
δ _sw : steering angle, k: yaw rate gain, τ ₀ : time constant, s: Laplace operator,
k: 2C _f V / [mV ² +2 (l _f C _f −l _r C _r )] × (1 / N),
τ ₀ : IV / 2 (C _f l _f ² + C _r l _r ² )
Incidentally, m is vehicle weight, I is the body of the finished moments, C _f, C _r are the front wheels, the rear wheels of the cornering power, l _f, l _r is the center of gravity front wheels, to the rear wheels length, N is the steering gear Is the ratio. A weighting factor is applied to the target yaw rate shown in equations (1) and (2), and switching is performed according to equation (3).
γ _d (t) = w (t) · γ _c (t) + [1−w (t)] × γ _δ (t) (3)
The weighting factor is determined by the driving behavior of the driver, and the recognition method will be described later.

In order to make the actual yaw rate follow the target yaw rate, the yaw moment M is calculated from the equation (4) by considering an inverse model of vehicle dynamics. However, since the first term of equation (4) is not a proper control system, a first-order low-pass filter with τ as a time constant is added.
M (s) = _{^{[s 2 - (a 11 + a}} 22) s + a 11 × a 22 -a 12 × a 21 ]
/ [(H ₂ s + a ₂₁ h ₁ −a ₁₁ h ₂ ) (τs + 1)] × γ _d (s)
-[H ₂ s + a ₂₁₁ -a ₁₁ h ₂ ) / (b ₂ s + a ₂₁ b ₁ -a ₁₁ b ₂ )] × δ _f (s)
... (4)
a ₁₁ = −2 (C _f + C _r ) / mV,
a ₁₂ = − [1+ (2C _f l _f −2C _r l _r ) / mV ² ],
a ₂₁ = −2 (C _f l _f −C _r l _r ) / I,
a ₂₂ = −2 (C _f l _f ² + C _r l _r ² ) / IV,
_{b 1 = 0, b 2 =} 1 / I, h 1 = 2C f l f / mV, h 2 = 2C f l f / I
δ _f is the actual steering angle of the front wheels, but cannot be calculated directly, so it is obtained from equation (5) using the driver steering angle.
δ _f = δ _sw / N (5)

In order to realize the yaw moment input determined by the equation (4), in this embodiment, the yaw moment is generated by the driving force control system of the electric vehicle of the rear-wheel drive system as a target. The driving force of each of the left and right wheels can be calculated as follows, assuming that the same driving force command value is increased or decreased from the driving force F _x0 during straight traveling.
[Left wheel driving force] F _xrl = F _x0 -M / d (6)
[Driving force of right wheel] F _xrr = F _x0 + M / d (7)
d: Rear wheel tread The above driving force is generated by the in-wheel motor torque of each of the left and right wheels. If transmission loss and wheel slip are ignored by the motor deceleration mechanism, the left and right drive torque command values are as shown in equations (8) and (9).
[Left wheel drive torque] T _ml = F _xrl x r _w (8)
[Right Wheel Drive Torque] T _mr = F _xrr × r _w (9)
r _w : Effective radius of tire Driving control using a hidden Markov model as a driving behavior recognition means of the driver will be described.
The hidden Markov model is an established model based on time series data and is a general action recognition method.
By using this, the driver's lane change behavior can be recognized.

In this example, as shown in the state transition model of FIG. 3, three patterns of S ₁ : lane following state, S ₂ : standby state, and S ₃ : lane change state are installed as the driving state of the driver.
Since the driving states of these drivers are not in an observable state (visible state), the current driving state is estimated by replacing these driving states with a visible symbol η _d .
The visible symbol η _d is determined by the steering angle and steering angular velocity of the driver. For this purpose, a membership function as shown in FIGS. 4A and 4B is set, and this is used to output the output symbol η _d. To decide.
The threshold values of the membership functions in FIGS. 4A and 4B are δ ₁ = 0.25 rad, δ ₂ = 0.5 rad, dδ ₁ /dt=1.0 rad / sec, dδ ₂ /dt=2.0 rad / sec.
By using this membership function, at a certain step time t _n , three values μ _δl , μ _δ′m (1, m = 1, 2, 3, respectively) from the steering angle δ and the steering angular velocity dδ / dt, δ ′ is the derivative of δ). Next, nine values H _k (k = 1, 2,..., 9) are obtained from these values according to the equation (10). Finally, according to equation (11), the subscript with the largest value of the nine products H _{k is taken as} the value of the output symbol μ _δ′m at time t _n .
H _k (t _n ) = μ _δl × μ _δ′m (10)
μ _d (t _n ) = arg max [H _k (t _n )] (11)
Each state that changes with time changes with a transition probability <a _y> (hereinafter, the probability is indicated by a triangle bracket “<>”), and an output symbol is output from each state with an output probability <b _y >. The These probabilities are expressed as a state transition probability matrix << A _y >> (hereinafter, the matrix is indicated by double triangle brackets "<<>>") and an output probability matrix << B _y >> using the Baum-Welch algorithm. Learning is performed, and an output distribution probability << B _d >> corresponding to each output symbol is set as in Expression (12).
<< B _d >> (y _d (t _n )) _{[3 × 3]} = diag {<b _j > y _d (t _n )} (12)

A method for recognizing the current driving behavior of the driver from the hidden Markov model described above is described below.
First, a matrix vector such as Equation (13) is defined.
<< P >> _{n | n [3 × 1]} = [P _{n | n} (1) P _{n | n} (2) P _{n | n} (3)] ^T (13)
However, P _{n | n} (i) is the probability that the driving state of the driver at the current step time t _n is S _i (i = 1, 2, 3), and these are the following equations (14), ( 15).
<< P >> _{n | n-1 [3 × 1]} = << A >> ^T _{[3 × 3]} << P >> _{n-1 | n-1 [3 × 1]} (14)
<< P >> _{n | n [3 × 1]} = << B >> _d (y _d (t _n )) _{[3 × 3]} << P >> _{n | n−1 [3 × 1]} }
/ [111] << B >> _d (y _d (t _n )) _{[3 × 3]} << P >> _{n | n−1 [3 × 1]} (15)
With the above method, the driving behavior of the driver is determined from the obtained values according to the rules of Expression (16), Expression (17), and Expression (18). In order to apply the recognition result to the control, the weighting coefficient of the switching system described in the previous section is determined based on the weighting coefficient of the switching system described in the previous section based on the determined driving behavior. In the present embodiment, “0” is output if the lane change action is recognized, and “1” is output if the lane change action is not recognized.

That is,
When arg max {P _{n | n} (i)} = 1, w (t _n ) = 1 (16)
When arg max {P _{n | n} (i)} = 3, w (t _n ) = 0 (17)
In other cases, w (t _n ) = w (t _n-1 )
The effectiveness of the driving behavior recognition system using the hidden Markov model described above was verified. The system was installed in an actual vehicle, and a single lane change driving experiment shown in FIG. 5 was conducted to verify whether the action recognition algorithm for lane tracking and lane change was effective by the driver's steering angle input and steering angular velocity input. The results are shown in FIG. 6, (A) is a steering angle of a driver, (B) is a steering angular velocity, (C) is an output symbol, and (D) is a diagram showing an action recognition output.
From the result, it can be confirmed that the action recognition output becomes 0 after the driver's steering wheel operation, and the lane change action is recognized. In addition, after the driver finished changing the lane and finished the steering wheel operation, the action recognition output returned to 1 again, and the driver was able to recognize the driving intention of following the lane, confirming the effectiveness of the system.

The effect of support based on driving behavior recognition was examined by simulation. As shown in FIG. 7, the simulation conditions were that the vehicle traveled at a vehicle speed of 35 km / h, and the lane change with a width of 2 m was performed within 10 m after 5 seconds. In this simulation, the upper limit value of DYC (Dynamic Yawmoment Control) input is set to 400 Nm, and the lower limit value is set to −400 Nm.
The simulation results are shown in FIGS. FIG. 8 shows a result when the vehicle travels only by lane tracking control without applying the switching system. Moreover, FIG. 9 is a figure which shows the result at the time of applying this invention.
8 and 9, (A) is the front lateral deviation, (B) is the steering angle, (C) is the weighting factor and DYC input, (D) is the actual yaw rate and target yaw rate, and (E) is the skid. It is a graph which shows each corner.
Further, FIG. 10 shows a vehicle trajectory for lane tracking control only, and FIG. 11 shows a vehicle trajectory for control using the action recognition system.
From FIG. 8, when the target yaw rate is not switched and the vehicle travels only with the lane tracking control, a yaw moment against the driver's action works when the lane is changed, and a large steering angle is required in the initial lane change. I can confirm. On the other hand, the proposed system can confirm that the action recognition of the lane change is performed from FIG. Thereby, during the lane change, the side yaw rate is switched to the target yaw rate according to the steering angle in order to nullify the side slip angle, and the actual yaw rate follows the target yaw rate, and as a result, the side slip angle is suppressed. Furthermore, it can be seen from FIGS. 10 and 11 that the vehicle trajectory in the case of traveling only by lane tracking control has a large deviation from the target course, whereas the system of the present invention has improved tracking performance.

The actual vehicle NOVEL-1 is equipped with a yaw moment control that applies a driving behavior recognition system using a hidden Markov model, and the single lane change driving experiment of Fig. 5 is performed in the same way as the simulation, and the recognition accuracy of the driving behavior and the proposed DYC The support effect was compared with the case where the vehicle traveled only by lane tracking control without switching the target yaw rate. The time series data of the experimental results are shown in FIGS. 12 and 13, (A) is the front lateral deviation, (B) is the steering angle, (C) is the weighting factor and DYC input, (D) is the driving torque and yaw rate for the left and right rear wheels, (E ) Is a graph showing the sideslip angle.
FIG. 14 shows a driver's steering angular velocity, and FIG. 15 shows a Lissajous curve based on a vehicle body side slip angle and a driver's steering angle. The Lissajous curve in FIG. 15 means that the smaller the area on the steering angle-slip angle diagram, the better the stability.
As can be seen from FIG. 12, when the vehicle travels only in the lane follow-up control mode in which the target yaw rate is not switched, a yaw moment for returning to the original lane is added during the lane change, which is a driver's steering burden.
In addition, during the lane change, the target yaw rate proportional to the driver's steering input is calculated from FIG. 13 by switching the weighting coefficient to 0 and switching to the side slip angle zeroing control during the lane change. Moreover, since the weighting coefficient has returned to 1 again after the lane change is completed, it has been confirmed that driving action recognition is being performed.
As shown in FIG. 14, in the initial lane change, the yaw moment control to which the present invention is applied has a small steering angular velocity, which makes it easier for the driver to intervene in the steering angle in an emergency. Further, from FIG. 15, it is confirmed that the skid angle is simultaneously reduced together with the steering angle, and it is confirmed that the vehicle handling stability is in the factory, and the effectiveness of the yaw moment control system based on the driving behavior recognition can be verified by an actual vehicle experiment. It was.

As described above, the lane tracking control device 1 of the present invention has a function of performing lane tracking control using the yaw rate γ as one of the control parameters, and (a) obtains the steering angle δ and changes the steering angle over time. Driving mode determination unit 11 that detects whether or not the time change state of the steering angle δ is due to a lane change operation, and (b) the driving mode determination unit 11 performs steering. When it is determined that the time change state of the angle is not due to the lane change operation, a target yaw rate (γ _c ) for lane tracking control is generated, and the driving mode determination unit 11 determines that the time change state of the steering angle δ is the lane change operation. If the control unit 12 determines that the target yaw rate (γ _δ ) for lane change control is generated, and (c) at least a vehicle control signal M using the target yaw rate acquired by the control unit 12 as a control parameter. And a vehicle motion control unit 13 to be generated. As a result, when a lane change operation is performed by the driver while the lane tracking control device is in operation, the lane tracking control operation is canceled at an appropriate timing so as not to feel a sense of incongruity, and vehicle running stability at the time of lane change is ensured can do.

It is a functional block diagram showing one embodiment of a lane following control device of the present invention. FIG. 2 is a more specific block diagram of the lane tracking control device of FIG. 1. Three patterns were installed: lane tracking state, standby state, and lane change state. (A), (B) is a figure which shows a membership function. It is explanatory drawing which shows the driving | running | working experiment of a single lane change. It is a figure for verifying whether the action recognition algorithm of lane following and lane change is effective, (A) is a steering angle of a driver, (B) is a steering angular velocity, (C) is an output symbol, (D) is It is a figure which shows action recognition output. It is a figure which shows the conditions of the simulation which shows the assistance effect based on driving action recognition. It is a figure which shows the result at the time of drive | working only by lane tracking control, without applying a switching system, (A) is a front lateral deviation, (B) is a steering angle, (C) is a weighting coefficient and DYC input. (D) is a graph showing an actual yaw rate and a target yaw rate, and (E) is a graph showing a skid angle. It is a figure which shows the result at the time of applying this invention, (A) is a front lateral deviation, (B) is a steering angle, (C) is a weighting factor and DYC input, (D) is an actual yaw rate and target (E) is a graph showing the side slip angle. It is a figure which shows the vehicle locus only of lane following control. It is a figure which shows the vehicle locus | trajectory of the control which applied the action recognition system in this invention. It is a figure which shows the time series data of the experimental result at the time of drive | working only by the lane tracking control mode which does not switch a target yaw rate, (A) is a front lateral deviation, (B) is a steering angle, (C) is The weighting coefficient and DYC input, (D) is a graph showing the left and right driving torque and yaw rate, and (E) is a graph showing the side slip angle. It is a figure which shows the time series data of the experimental result at the time of driving | running | working controlling by the action recognition system by this invention, (A) is a front lateral deviation, (B) is a steering angle, (C) is a weighting coefficient. And (D) is a graph showing the left and right driving torque and yaw rate, and (E) is a graph showing the side slip angle. It is a figure which shows the steering angular velocity of a driver. It is a Lissajous curve by a body side slip angle and a driver's steering angle.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Lane tracking control apparatus 5 Vehicle 11 Operation mode judgment part 12 Control part 13 Vehicle motion control part 14 Deviation detection part 101 Handle 102 Front wheel 103 Rear wheel 121 Yaw rate generation means for lane tracking control 122 Yaw rate generation means for lane change control 123 Switching means

Claims (5)

In a lane tracking control device that performs lane tracking control using the yaw rate as one of the control information,
From the steering angle and / or the steering angular velocity, whether the vehicle is currently in the lane tracking mode, the lane change mode, or the middle of the transition from one of the lane tracking mode or the lane change mode to the other An operation mode determination unit that determines whether the mode is under;
When the operation mode determination unit determines each mode, a control unit that performs control corresponding to each mode;
With
The controller is
When the driving mode determination unit determines that the vehicle is in the lane following mode, a control signal for a yaw moment or lateral biasing force for lane following driving is generated,
When the driving mode determination unit determines that the vehicle is in the lane change mode, a control signal is not generated, or a control signal of a yaw moment or a lateral urging force for lane change operation is generated,
When the driving mode determination unit determines that the vehicle is in the intermediate mode, when the control signal for the lane following operation is generated immediately before, the control signal for the lane following operation is When the control signal for the lane change operation is generated, the control signal for the lane change operation is generated.
A lane tracking control device characterized by that.   The operation mode determination unit determines, using an observable amount, which mode is currently under the lane following operation mode, the lane change mode, or the intermediate mode. Item 2. A lane tracking control device according to item 1.   3. The lane following operation mode, the lane change mode, and the intermediate mode are transition states in a hidden Markov model, and the observable amount is an output symbol in the hidden Markov model. Lane tracking control device.   An automobile equipped with the lane tracking control device according to claim 1, further comprising a motor that is independently driven on each of the left and right wheels or all the wheels on the rear side or the front side, and the control unit includes the motor. An automobile characterized by controlling. An automobile equipped with the lane tracking control device according to claim 1, further comprising a brake that independently applies a braking force to the left and right wheels or all the wheels on the rear side or the front side. An automobile characterized by controlling the brake.