JP3206532B2 - Steering angle neutral learning device, curve curvature estimation device, inter-vehicle distance control device, and recording medium - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a steering angle neutral learning device, a curve curvature estimating device, an inter-vehicle distance control device, and a recording medium for realizing a steering angle neutral learning device in a computer system.

[0002]

2. Description of the Related Art A vehicular travel control device that measures the inter-vehicle distance and relative speed with a preceding vehicle to keep the inter-vehicle distance constant is well known. Such a device always includes a preceding vehicle detecting device for measuring the distance to the preceding vehicle. Conventionally, a laser radar device has been used as the preceding vehicle detection device. However, if the direction of the laser beam emitted from the laser radar is fixed, while traveling a curve,
In some cases, a vehicle traveling in another lane in addition to a signboard on the roadside, a reflector, and the like may be detected as a preceding vehicle because the vehicle cannot irradiate the vehicle on the own lane to a distant place.

To solve this problem, a scanning laser radar for scanning a laser beam within a predetermined range has been proposed. Further, there has been proposed a preceding vehicle determination that determines whether an obstacle detected by the scanning laser radar is a vehicle on the same lane as the own vehicle using a curve detection unit.
For example, in the preceding vehicle detection device disclosed in Japanese Patent Application Laid-Open No. 4-248489, it is determined whether or not the vehicle is a preceding vehicle based on a curve curvature radius R calculated from a steering angle.

[0004] However, if there is a deviation between the curve radius of curvature R and the actual curve of the road, the preceding vehicle may easily be lost, or a vehicle other than the preceding vehicle may be mistakenly recognized as the preceding vehicle. This is a practical problem. In addition to this, for example, the preceding vehicle detection device disclosed in Japanese Patent Application Laid-Open No. 6-176300 introduces a unique concept of preceding vehicle accuracy that represents the likelihood of the preceding vehicle by probability. If such a concept of the preceding vehicle accuracy is also applied to the scan type, it is expected that comfortable and safe inter-vehicle control can be performed without easily losing track of the preceding vehicle.

The preceding vehicle detecting device disclosed in Japanese Patent Application Laid-Open No. 6-176300 cannot be applied to a scanning type preceding vehicle detecting device at all, so a new concept of own lane probability is introduced. There has been proposed an inter-vehicle distance control device capable of appropriately selecting a preceding vehicle in a scan type and controlling the inter-vehicle distance (Japanese Patent Application Laid-Open No. Hei 8-279099).

[0006]

SUMMARY OF THE INVENTION However, Japanese Patent Application Laid-Open No. Hei 8-2
In the inter-vehicle distance control device disclosed in Japanese Patent No. 79099, in order to calculate a traveling path curve of the own vehicle for obtaining the own lane probability, a filtering process using a steering angle detected by a steering sensor as an input, Learning of the reference steering angle neutral position is performed.

That is, in the inter-vehicle distance control device, if the vehicle speed is 20 km / h or more, all the steering angle data is used for learning the neutral position of the steering angle regardless of the steering amount. However, the right and left curves do not always exist on the travel path, and, for example, the right curve may frequently exist for a long time. In such a case, the steering angle neutral position should be obtained as a straight traveling state by the filtering process, but may be calculated to be shifted to the right in actuality.

[0008] Such a deviation gradually converges to the true steering angle neutral position. However, since such convergence extends for a long time, in a short running time, the steering angle neutral position differs from the actual steering angle neutral position. In addition, the curvature of the traveling road and the own lane probability are calculated, and there is a possibility that accuracy may be lost.

In order to solve the above-mentioned problem of the conventional neutral position control of the steering angle, an accurate neutral position of the steering angle can be estimated at an early stage of the traveling, thereby enabling various kinds of control with high accuracy. For the purpose of providing a corner neutral learning device and the like, the present applicant has proposed the following technology in Japanese Patent Application No. 9-192813. In other words, when the vehicle turning detecting means determines that the vehicle is in the straight traveling state, and when the steering angle of the vehicle detected by the steering angle detecting means is present near the provisional steering angle neutral position obtained at that time, the steering angle is neutralized. Learning is to find the position. That is, only the steering angle when it is determined that the vehicle is in a substantially straight traveling state is used for learning.

However, according to this method, the range set near the provisional steering angle neutral position obtained at that time is further reduced on the premise that the accuracy of the straight turning state determination by the vehicle turning detecting means is improved. Is required. The reason will be described below. Generally, the driver's steering situation is
When viewed over a long period of time, the steering angle distribution is close to the normal distribution, is even in the left and right directions, and has a steering angle distribution with the true steering angle neutral position as a peak. However, when looking at the steering angle distribution in a short time of, for example, several hundred seconds, the true steering angle neutral position does not peak depending on the driving situation, and the steering angle separated by a predetermined value to the left and right from the neutral position peaks. There is. For example, a situation like a slalom that continuously steers to the left and right corresponds to this. FIG. 14 shows a specific steering angle distribution in this situation. In such a situation, when the range is set in the vicinity of the provisional steering angle neutral position (SO), it is conceivable that only one side of the left and right peaks is included as shown in FIG. This is because even if the slalom is steered continuously to the left and right, the symmetrical steering is not always performed, and the provisional steering angle neutral position serving as the reference of the originally set range is detected by the vehicle turning detection means. It is conceivable that the degree of deviation from the true steering angle neutral position is increased due to the influence of accuracy. Therefore, when the range is set in this way, the neutral steering angle position shifts to a peak at a position distant from the true steering angle neutral position due to the effect of learning.

Of course, the steering angle neutral position drifts only in a limited situation in which left and right steering is repeated with almost no straight-ahead traveling, and thereafter, in a situation in which a general straight traveling state is often present. In the meantime, the steering angle neutral position returns to the true value due to the learning effect.
However, even if temporarily, the drift of the steering angle neutral position reduces the accuracy of the preceding vehicle detection during that period.

Further, in order to set the above-mentioned "range indicating a linear state" with a margin, for example, even if the setting range in the vicinity of the tentative steering angle neutral position is expanded, the setting range is still limited in a limited situation. It is not possible to cope with the situation where the distribution of only one side of the left and right steering falls within the range.
Conversely, it is conceivable to set the right and left peaks in the case of the above-mentioned slalom running so as to be out of the range by extremely narrowing the “range indicating the linear state”. It is necessary to include the true steering angle neutral point in the range, and the accuracy of the straight-ahead state determination by the vehicle turning detection means must be considerably improved. Examples of the vehicle turning detection unit include a sensor that detects a yaw rate and a device that detects a turning state of the vehicle based on an output of a sensor that detects a vehicle lateral acceleration. However, in order to improve the accuracy of the straight-ahead state determination by the vehicle turning detecting means, the manufacturing error of each sensor and the change in temperature characteristics are further reduced, and the mounting accuracy of the acceleration sensor is further improved so as not to be affected by the gravitational acceleration. Need to be done. These are not easy, and even if the accuracy can be improved to some extent, it will increase the cost considerably,
It is not realistic to rely solely on improving accuracy.

Therefore, the present invention solves the above-mentioned problem of the conventional neutral position control of the steering angle, and estimates the accurate neutral position of the steering angle early in the start of traveling to enable various highly accurate controls. And the above-mentioned prior art (Japanese Patent Application No.
In order to realize a steering angle neutral learning device, a curve curvature estimating device, an inter-vehicle distance control device, and a steering angle neutral learning device capable of appropriately coping even in a limited situation where it is difficult to sufficiently cope with the problem, a computer system is used. The purpose of the present invention is to provide a recording medium.

[0014]

According to the steering angle neutral learning device of the present invention, the vehicle straight traveling state determining means is in a range where the turning of the vehicle detected by the vehicle turning detecting means indicates the straight traveling state. Is determined, the provisional steering angle neutral position setting means detects the steering angle of the vehicle detected by the steering angle detection means when the vehicle straight-running state determination means determines that the turning of the vehicle is in the range indicating the straight traveling state. (S: In this section, related symbols used in the embodiment are attached, but this symbol does not mean that the scope of the claims is limited.) Is set as the provisional steering angle neutral position (S0). I do.
Examples of such vehicle turning detecting means include a sensor for detecting a yaw rate or a lateral acceleration, and a device for detecting a speed difference between left and right wheels and detecting a turning of the vehicle based on the speed difference.

The steering angle neutral position learning means learns the steering angle neutral position (Sc) using the provisional steering angle neutral position (S0). Specifically, the steering angle neutral position learning means determines the difference (str_eng) between the vehicle steering angle (S) detected by the steering angle detecting means and the provisional steering angle neutral position (S0).
) Is the “predetermined range”, the angle obtained by subtracting the steering angle neutral position (Sc) from the difference (str_eng) between the detected steering angle (S) of the vehicle and the provisional steering angle neutral position (S0). (Str_eng-Sc), based on the steering angle neutral position (S
c) is corrected to obtain a new steering angle neutral position (Sc).

The "predetermined range" is based on the curvature and the vehicle speed of the traveling road on which the vehicle traveling control executed using the steering angle neutral position, such as inter-vehicle distance control, is assumed to be applied. It is set so as to include all possible angles for the curvature of the traveling road and the vehicle speed within the range. For example, when the inter-vehicle distance control is applied only on a highway, there is naturally an upper limit on the curvature of the traveling road. When the speed of the own vehicle is restricted from the distance control, the minimum speed and the maximum speed at which the control is permitted are also determined. Therefore, the range is set so as to include all possible angles when traveling in such a situation.

Specifically, for example, when the speed of the own vehicle is 40 k
If it is not less than m / h, in consideration of the case where the present invention is applied to the above-described system for permitting the inter-vehicle distance control, the steering angle that can be maximized when traveling at 40 km / h is set to a range that can be captured. Of course, the factors that determine the steering angle also affect the curvature of the traveling path, and may be taken into account. For example, if there is a curve with a minimum radius of curvature of 250 m on an expressway, for example, a range that includes the maximum possible steering angle of the curve may be set. However, if the steering angle becomes transiently large due to the lane change, it is not considered that the steering angle neutral learning should be executed, and there is no problem even if the steering angle exceeds the predetermined range.

According to the steering angle neutral learning device of the present invention, the vehicle steering angle (S) and the provisional steering angle neutral position (S
0), the steering angle neutral position (Sc) is learned when the difference (str_eng) is within such a "predetermined range". Therefore, in the prior art described in the above-mentioned Japanese Patent Application No. 9-192813. Can properly cope with the following situations that could not be dealt with.

That is, in the prior art, only when the detected steering angle (S) of the vehicle is within a set range near the provisional steering angle neutral position (S0), the steering angle neutral position (S
Since the learning of c) was performed, as described with reference to FIG. 14, for example, when a short-time steering angle distribution is viewed in a situation such as a slalom that is continuously steered left and right, only one of the left and right peaks is observed. May exist within the set range. In this case, depending on the learning effect,
The steering angle neutral position (Sc) drifts to a peak at a position distant from the true steering angle neutral position.

On the other hand, in the case of the steering angle neutral learning device according to the present invention, even if there are peaks on the left and right, and they are asymmetric with respect to the provisional steering angle neutral position (S0), the above-mentioned condition is obtained. Is set so as to include all angles including both peaks. That is, even when viewed within a limited range, even when the steering angle distribution is asymmetric, the distribution of all the steering angles is often nearly symmetric when viewed comprehensively. Therefore, even in the situation where the steering is continuously performed to the left and right as in the slalom described above, the steering angle of only one side of the left and right steering is not used for learning by deviating, and the steering angle neutral position (Sc ) Can be prevented from being inappropriately drifted.

Of course, when viewed over a long period of time, the steering angle distribution becomes close to the normal distribution and becomes equal to the left and right, so that the prior art and the present invention can be similarly applied. However, the present invention is effective in that it can appropriately cope with the above-mentioned special traveling situation and when the situation is seen in a short time.

Further, in the prior art, a straight running state is determined based on a turning state detected by a vehicle turning detecting means such as a yaw rate sensor, and the vicinity of the temporary steering angle neutral position obtained in the straight running state is determined. "Range indicating straight-ahead state" was set. Therefore, unless the detection accuracy of the vehicle turning detection means is high, the “range indicating the straight traveling state” cannot be set appropriately. Then, even if the data is originally effective for learning the steering angle neutral position (Sc), it is excluded, and the time required for the convergence to the true steering angle neutral position by the learning effect is relatively shortened. It will be long.

On the other hand, in the case of the steering angle neutral learning device of the present invention, as described above, the vehicle running control executed using the steering angle neutral position is within the range in which application of the vehicle running control is assumed. All possible angles for the curvature of the road and the vehicle speed are taken into the learning. Therefore, even if the detection accuracy of the vehicle turning detection means is not high, data effective for learning as in the above-described prior art is not excluded. In other words, the restriction on the height of the detection accuracy of the vehicle turning detection means is loose, and this is also effective in that respect.

As described above, if the steering angle neutral position (Sc) is learned by the steering angle neutral learning device of the present invention, it is possible to cope with a special driving situation such as slalom driving, and to quickly and accurately perform the operation regardless of the driving situation. A neutral steering angle neutral position (Sc) can be obtained. Therefore, the curvature of the traveling road and the own lane probability obtained based on the steering angle also have high accuracy, and various controls with high accuracy can be performed. Also, there is an effect that the restriction on the detection accuracy of the vehicle turning detection means can be relaxed.

Further, regarding the above-mentioned "predetermined range",
The following considerations can be considered. In the initial stage of learning of the steering angle neutral position by the steering angle neutral position learning means, an error existing between the learned steering angle neutral position and the true steering angle neutral position is set, and the learning of the steering angle neutral position is performed. Is advanced by a predetermined degree or more, the difference is set by removing the amount of error reduction by the learning. This is because, in the initial stage of the steering angle neutral learning, due to the error of the provisional steering angle neutral position (SO) based on the accuracy of the vehicle turning detecting means, the steering angle neutral position (S
Since there is an error between c) and the true steering angle neutral position, the “predetermined range” is set to a value including this error. On the other hand, when the steering angle neutral learning is advanced, the steering angle neutral position is set. This is because the position (Sc) is approaching the true steering angle neutral position and the error has been reduced, so that the “predetermined range” can be narrowed by the estimated reduction of the error.

At the same time, at the beginning of the steering angle neutral learning, the "predetermined range" is set around the provisional steering angle neutral position (SO), and when the steering angle neutral learning is advanced, the "predetermined range" is set to the steering angle. The neutral position (Sc) may be set as the center. The steering angle neutral position learning means learns the steering angle neutral position (Sc) when a difference (str_eng) between the vehicle steering angle (S) and the provisional steering angle neutral position (S0) is within a predetermined range. However, for example, when the vehicle speed detected by the vehicle speed detecting means is higher than a predetermined speed, the learning may be performed. In this case, the predetermined speed is set to, for example, 40 km / h. This is because steering is frequently repeated in an environment where the vehicle travels at a speed lower than 40 km / h, for example, in a city.
This is because, when traveling on a suburb or a highway where the speed is higher than m / h, steering is stable, and more accurate steering data can be obtained.

For the steering angle neutral position (Sc), for example, a provisional steering angle neutral position (S0) is set as an initial value. The provisional steering angle neutral position setting means includes a steering angle of the vehicle detected by the steering angle detection means when the vehicle straight traveling state determination means determines that the turning of the vehicle is in a range indicating the straight traveling state. S) to the provisional steering angle neutral position (S
Although the process of setting the temporary steering angle neutral position (S0) is not necessary, the process of setting the temporary steering angle neutral position (S0) does not need to be repeated every time the condition is satisfied.

That is, the provisional steering angle neutral position (S
0) is a tentative steering angle neutral position, and there is no problem even if it is performed only once at the beginning in the processing of the present steering angle neutral learning device, and the necessary steering angle neutral position (S
c) is thereafter learned and obtained with high accuracy.

More specifically, the steering angle neutral position learning means, for example, calculates the difference (st) between the vehicle steering angle (S) detected by the steering angle detecting means and the provisional steering angle neutral position (S0).
r_eng), the steering angle neutral position (Sc) is corrected at a predetermined ratio of the angle (str_eng-Sc) obtained by subtracting the steering angle neutral position (Sc) to obtain a new steering angle neutral position (Sc). , The steering angle neutral position (Sc) is learned.

The predetermined ratio is, for example, in accordance with the vehicle speed when the steering angle neutral position (Sc) is learned by the steering angle neutral position learning means, the higher the speed, the larger the speed. Is set to be This is because the higher the speed, the higher the accuracy of the straightness detection.

The predetermined ratio is made smaller as the number of times of learning is increased, for example, in accordance with the number of times of learning of the steering angle neutral position (Sc) performed by the steering angle neutral position learning means. It is set as follows. This is to stabilize the steering angle neutral position (Sc) at an early stage while taking into account long-term steering angle neutral position fluctuation.

In addition, the predetermined ratio is set, for example, in accordance with the degree of progress of learning of the steering angle neutral position performed by the steering angle neutral position learning means as the set width of the above “predetermined range” is larger. Be smaller. By the way, in the case of the steering angle neutral learning device of the present invention, as described above, the curvature of the traveling path and the vehicle in the range where the application of the vehicle traveling control executed using the steering angle neutral position is assumed. Since all possible angles in the speed are included in the learning, this may cause inconvenience. That is, since the steering angle is used for learning even if it is somewhat large, the angle used for learning (str_eng-Sc)
Of the steering angle neutral position (Sc) at the time of learning increases as a result.

Regarding this point, for example, the above-mentioned “predetermined ratio set to decrease as the number of times of learning increases”
It is possible to devise about. For example, in the embodiment of the above-mentioned Japanese Patent Application No. 9-192813, the learning degree counter Cv constituting this predetermined ratio (1 / Cv) is used.
Is set to 16384, but for example, the upper limit of the learning degree counter Cv is quadrupled to 65536, and as a result, the predetermined ratio (1 / Cv) is reduced. By doing so, the angle used for learning (str_eng-Sc)
, The fluctuation of the steering angle neutral position (Sc) during learning can be relatively small.

In addition, as described above, the absolute value of the angle (str_eng-Sc) used for learning becomes large, which is not a general problem caused by widening the angle range taken in for learning. Situations are also anticipated.
In other words, the steering situation is such that a relatively large steering angle (str_eng-Sc) continuously occurs on one of the left and right sides, such as a rampway or a loop bridge. In the steering angle neutral learning of the present invention, even if the steering distribution is asymmetrical when viewed within the limited learning range, the distribution of all the steering angles is often almost symmetric when viewed comprehensively. It is characterized in that all the steering angles within the predetermined range are taken into the learning. However, a relatively large steering angle (st
In a steering situation in which r_eng-Sc) continuously occurs on one of the left and right sides, it cannot be handled at all.

Therefore, regarding this point, in the case of a steering situation in which a relatively large steering angle (str_eng-Sc) continuously occurs on one of the left and right sides, such as the above-described rampway or loop bridge. In this case, it is determined that the data is inappropriate for learning and is not used for learning.

More specifically, a steering angle neutral learning reset means is provided so that inappropriate data is not used for learning. The steering angle neutral learning resetting means determines that the turning angle for determination (| c_int (n) |) obtained based on the turning angle of the vehicle detected by the vehicle turning detecting means exceeds a preset threshold value (π). In this case, the learning result obtained during the special turning state in which the vehicle is continuously turning in the left or right direction at the turning angle for determination (| c_int (n) |) exceeding the threshold (π) is discarded. I do. And not just discard,
The steering angle neutral position (Sc (n)) is returned to the state before the special turning state.

The steering angle neutral learning reset means not only returns the steering angle neutral position (Sc (n)) to the state before the special turning state, but also changes the predetermined ratio (Cv ( Regarding n)), the state before the special turning state may be returned.

The resetting function of the steering angle neutral learning resetting means described above makes it possible to provide a relatively large steering angle (str_eng) as when running on a rampway or a loop bridge.
In order to facilitate the function at the time of -Sc), the following measures may be added. In other words, the turning angle for determination is corrected to be larger than the actually detected turning angle when the turning of the vehicle detected by the vehicle turning detecting means is relatively large. With such a correction, it becomes easier to exceed a preset threshold value, and it is easy to eliminate the influence of data unnecessary for learning.

Conversely, when the turning of the vehicle detected by the vehicle turning detecting means is relatively small, the turning angle for determination is corrected so as to be smaller than the actually detected turning angle. Is also good. This is for the following reason. In other words, if the steering angle neutral learning is not progressing and the accuracy of the steering angle neutral position (Sc) is low, the calculated steering angle (str_en
g−Sc) increases, and the learning result may be discarded by the steering angle neutral learning reset unit. Therefore, at a relatively small steering angle (str_eng-Sc), the learning result is not discarded in a state close to straight ahead.
If the turning angle (c_int) of the vehicle is corrected so as not to increase, it is possible to prevent the turning angle (c_int) from exceeding a preset threshold value.

Further, using the above-described steering angle neutral learning device, a vehicle speed detecting means for detecting a vehicle speed, a steering angle neutral position (Sc) obtained by the steering angle neutral learning device, A steering angle (S) of the vehicle detected by the angle detecting means or a difference (str_eng) between the steering angle (S) and the tentative steering angle neutral position (S0); and a vehicle speed detected by the vehicle speed detecting means. (Vn) and
Curve curvature calculating means for calculating the curvature of the traveling path, a curve curvature estimating device can be configured,
Accurate steering angle neutral position (S
Since c) is obtained by learning at an early stage, highly accurate curve curvature estimation can be performed at an early stage.

Further, using the above-mentioned steering angle neutral learning device, a vehicle speed detecting means for detecting the vehicle speed of the own vehicle,
Based on the steering angle neutral position (Sc) obtained by the steering angle neutral learning device and the vehicle speed (Vn) of the own vehicle detected by the vehicle speed detecting means, the curvature of the traveling path of the own vehicle is calculated. And a scan curvature illuminating means that scans and radiates a transmission wave or a laser beam over a predetermined angle range in the vehicle width direction and, based on a reflected wave or a reflected light from an object, corresponds to a distance between the own vehicle and a preceding object according to the scan angle. Calculating the relative position of the object with respect to the own vehicle based on the distance detected by the distance measuring means and the corresponding scan angle, and calculating the relative speed of the object with respect to the own vehicle. Calculating the object recognition means, based on the curvature of the own vehicle travel path obtained by the curve curvature calculation means and the relative position of the object calculated by the object recognition means, Own lane probability calculating means for determining the probability that the body is in the same lane as the own vehicle, and preceding vehicle selecting means for selecting a preceding vehicle to be controlled for the following distance based on the probability obtained by the own lane probability calculating means. Inter-vehicle distance control means for controlling the inter-vehicle distance to the preceding vehicle selected by the preceding vehicle selecting means by adjusting the speed of the own vehicle detected by the vehicle speed detecting means. The control device can be configured, and the steering angle neutral learning device can learn and obtain the accurate steering angle neutral position (Sc) at an early stage, so that the accurate inter-vehicle distance control can be performed at an early stage.

The function of realizing each unit of such a steering angle neutral learning device in a computer system can be provided, for example, as a program activated on the computer system side. For such a program,
For example, floppy disk, magneto-optical disk, CD-
It can be used by recording it on a computer-readable recording medium such as a ROM or a hard disk, loading it into a computer system as needed, and starting up. Alternatively, the program may be recorded in a ROM or a backup RAM as a computer-readable recording medium, and the ROM or the backup RAM may be incorporated in a computer system and used.

The curvature obtained by the curve curvature calculating means is:
It indicates the degree of curvature of the traveling path in a broad sense, and includes a radius of curvature and a narrow meaning of curvature (reciprocal of the radius of curvature).

[0044]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an electronic control unit 2 (hereinafter, referred to as an "inter-vehicle control ECU") and a brake electronic control incorporating a steering angle neutral learning device to which the above-mentioned invention is applied. Device 4 (hereinafter referred to as "brake ECU")
Called. FIG. 2 is a block diagram illustrating a schematic configuration of various control circuits mounted on the vehicle, mainly illustrating the configuration shown in FIG.

The inter-vehicle control ECU 2 is an electronic circuit mainly composed of a microcomputer, and includes a current vehicle speed (Vn) signal, a steering angle (str-eng, S0) signal, a yaw rate signal, a target inter-vehicle time signal, and a wiper switch. Information, control state signals for idle control and brake control, and the like are received from the engine electronic control unit 6 (hereinafter, referred to as “engine ECU”). The inter-vehicle control ECU 2 calculates a steering angle neutral position by a steering angle neutral learning process described later as a steering angle neutral learning device based on the received data, or calculates a curve curvature in a process described later as a curve curvature estimating device. The radius R is estimated, and the distance control calculation is performed.

The laser radar sensor 3 is an electronic circuit composed mainly of a scanning distance measuring device using a laser and a microcomputer. The laser radar sensor 3 detects the angle and relative speed of the preceding vehicle detected by the scanning distance measuring device, as well as the distance between vehicles. Based on the current vehicle speed (Vn) signal, curve radius of curvature R, and the like received from the control ECU 2, the own lane probability of the preceding vehicle is calculated as a part of the inter-vehicle distance control device, and the preceding lane including information such as the relative speed is calculated. It is transmitted to the following distance control ECU 2 as vehicle information. Further, the diagnosis signal of the laser radar sensor 3 itself is also transmitted to the following distance control ECU 2.

The scanning distance measuring device scans and irradiates a predetermined angle range in the vehicle width direction with a transmission wave or a laser beam, and based on the reflected wave or the reflected light from the object,
It functions as a distance measuring means capable of detecting the distance between the host vehicle and the object ahead in accordance with the scan angle.

Further, the following distance control ECU 2 determines the preceding vehicle to be controlled for the following distance based on the own lane probability and the like included in the preceding vehicle information received from the laser radar sensor 3 as described above. In order to appropriately adjust the distance, a target acceleration signal, a fuel cut request signal, an OD cut request signal, a third shift down request signal, an alarm request signal, a diagnosis signal, a display data signal, and the like are transmitted to the engine ECU 6. .

The brake ECU 4 is an electronic circuit mainly composed of a microcomputer, and includes a steering sensor 8 as a steering angle detecting means for detecting a steering angle of a vehicle, and a yaw rate sensor 10 for detecting a yaw rate as a vehicle turning detecting means. A steering angle and a yaw rate are obtained from a wheel speed sensor 12 that detects the speed of each wheel, and these data are transmitted to the headway control ECU 2 via the engine ECU 6. Also brake ECU4
Sounds an alarm buzzer 14 in response to an alarm request signal from the headway control ECU 2 via the engine ECU 6.

The engine ECU 6 is an electronic circuit mainly composed of a microcomputer, and includes a vehicle speed sensor 16 as vehicle speed detecting means for detecting the vehicle speed, and a brake switch 1 for detecting the presence or absence of a brake depression.
8, cruise control switch 20, cruise main switch 22, and throttle opening sensor (not shown), detection signals from other sensors and switches, or wiper switch information and tail switch information received via body LAN 24. And further,
A steering angle (str-eng, S0) signal and a yaw rate signal from the brake ECU 4, a target acceleration signal, a fuel cut request signal, an OD cut request signal, a third speed shift down request signal, an alarm request signal from the headway control ECU 2;
A diagnosis signal, a display data signal, and the like are received.

The engine ECU 6 drives a throttle driver, a brake driver (not shown), and the like in accordance with an operation state determined from the received signal. Further, necessary display information is transmitted to and displayed on a display device (not shown) such as an LCD via the body LAN 24, or a current vehicle speed (Vn) signal, a steering angle (str-eng, S0) A signal, a yaw rate signal, a target inter-vehicle time signal, a wiper switch information signal, and a control state signal of idle control and brake control are transmitted to the inter-vehicle control ECU 2.

The steering sensor 8 is a photo-interrupter type sensor as shown in FIG. 2A, and comprises a metal disk 32 and a rotation detecting circuit 34. As shown in the plan view of FIG. 2B, the metal disk 32 has a large number of slits 32a arranged in a ring shape.
When the light output from the two laser diodes 34a and 34b provided in the metal disk 32 passes through the slit 32a of the metal disk 32, the light is transmitted to the two phototransistors 34c and 3c.
4d, the light is received, and the on / off state of the light is
The voltage output signals SS1 and SS2 of the terminals are detected by the detection circuit 34e.
Is converted and output. VIN is a power supply line, GND
Is a ground line. Since the metal disk 32 is provided on the handle shaft of an automobile (not shown), the output signals SS1 and SS2 of the two terminals are output as shown in FIG. 3 according to the rotation of the handle shaft, that is, the steering. . Due to the relationship between the two laser diodes 34a, 34b and the arrangement of the slits 32a, when the metal disk 32 rotates left (the steering wheel is turned to the left), from the falling timing of the output signal SS1 to the falling timing of the output signal SS2. Is shorter than the time from the falling timing of the output signal SS2 to the falling timing of the output signal SS1. When the metal disk 32 rotates right (the steering wheel is turned to the right), the time from the falling timing of the output signal SS1 to the falling timing of the output signal SS2 is changed from the falling timing of the output signal SS2 to the output signal SS1. Is longer than the time until the falling timing of

From this, the direction of the steering is determined. Further, the steering angle is determined by counting the falling or rising of the output signals SS1 and SS2. The steering angle from the fall of the output signal SS1 to the fall of the output signal SS2 in the left rotation, or the steering angle from the fall of the output signal SS2 to the fall of the output signal SS1 in the right rotation is 1.125 deg. .
Therefore, the measurement accuracy of the steering angle directly obtained from the pulse signal of the steering sensor 8 is up to 1.125 deg. However, the steering angle neutral position Sc obtained by the steering angle neutral position learning control described later can be estimated with higher accuracy than 1.125 deg.

Next, a description will be given of a steering angle calculation process shown as a part of the process of the brake ECU 4 shown in the flowchart of FIG. This steering angle calculation processing corresponds to processing as a vehicle straight traveling state determination means (steps S120 and S130) and a provisional steering angle neutral position setting means (steps S110 and S140).

First, it is determined whether or not the provisional steering angle neutral position S0 has not been set (S110). When the ignition is turned on, power is supplied, and the brake ECU 4 starts up, the provisional steering angle neutral position S0 is not set ("YES" in S110).
It is determined whether or not the current vehicle speed Vn obtained from the vehicle speed sensor 16 via the motor 6 is higher than a predetermined speed, here, 30 km / h (S120). If the current vehicle speed Vn is higher than 30 km / h ("YES" in S120), the yaw rate sensor 10
It is determined whether or not the absolute value of the yaw rate obtained from is smaller than 1 deg / s (S130).

Steps S120 and S130 are processes for judging whether or not the turning of the vehicle is in a range indicating a straight traveling state. If any one of these steps S120 and S130 is not satisfied ("NO" in S120) Or S1
30 ("NO"), the steering angle calculation process terminates the process without doing anything.

When the vehicle starts running, the vehicle speed is 30 km /
h and the absolute value of the yaw rate is 1d even temporarily
If the vehicle travels straight ahead so as to be smaller than eg / s (“YES” in S120 and “YE” in S130).
S "), the steering angle S at this time is set as the provisional steering angle neutral position S0 (S140).

Then, the steering angle S to be used for the control is
Is started (S150), and the measured steering angle str_en is calculated from the measured steering angle S and the tentative steering angle neutral position S0 set in step S140 according to the following equation.
g (the difference between the steering angle S and the provisional steering angle neutral position S0, that is, the steering angle with the provisional steering angle neutral position S0 as the origin) is obtained (S160).

[0059]

[Formula 1] str_eng ← S−S0 [Equation 1] Since the provisional steering angle neutral position S0 is set thereafter, “NO” is determined in the step S110, and in the present steering angle calculation processing, the steps S150 and S150 are performed. Only S160 is performed, and the process of measuring the steering angle S (S150) and the process of updating the measured steering angle str_eng (S160) are repeated.
That is, the provisional steering angle neutral position S0 in step S140.
The setting is performed only once at first.

The steering angle S is measured every time the output signals SS1 and SS2 change according to the relationship shown in FIG.
d is calculated and calculated as in the following equation.

[0061]

S ← S + Sd × 1.125 (deg) [Expression 2] The measured steering angle str_eng obtained in this way is transmitted from the brake ECU 4 to the engine together with the yaw rate and other data by a transmission process (not shown). It is transmitted to the ECU 6.

The engine ECU 6 mediates the transmission of these data, and transmits the measured steering angle str_eng to the headway control ECU 2 together with other data. Next, the headway control ECU 2
The steering angle neutral position learning process, which is a part of the process executed by, will be described. FIG. 6 shows a flowchart of the steering angle neutral position learning process.

First, the measured steering angle str_eng is stored in the engine EC.
It is determined whether it has been received from U6 (S220). If the measured steering angle str_eng has not been received ("NO" in S220), nothing is performed in this steering angle neutral position learning process. When the measured steering angle str_eng is received ("YES" in S220), it is then determined whether or not the current vehicle speed Vn is equal to or greater than 40 km / h (S230). The current vehicle speed Vn is 40 km /
If not less than h ("YES" in S230), it is next determined whether the learning degree counter Cv is less than 65536 (S240). If the learning degree counter Cv is less than 65536 ("YES" in S240), then the measured steering angle
It is determined whether the absolute value of str_eng is equal to or smaller than the steering angle determination threshold str_ref (S241). If the learning degree counter Cv is equal to or greater than 65536 ("NO" in S240), then the absolute value of the difference between the measured steering angle str_eng and the neutral steering angle Sc is calculated as the error of the provisional steering angle neutral position SO estimated in advance. It is determined whether or not 6 deg is less than or equal to a value obtained by subtracting 6 deg from the steering angle determination threshold value str_ref (S242).

If step S230 is not satisfied (S
At 230, “NO”, the steering angle neutral position learning process ends without performing anything. Step S24
If S242 is not satisfied ("NO" in S241 or "NO" in S242), nothing is performed in this steering angle neutral position learning process, and the process ends.

The steering angle determination threshold value str_ref corresponds to the boundary of the “predetermined range” described in the claims, and is set in consideration of the following points. That is, in the present embodiment, since it is assumed that the vehicle is used for inter-vehicle distance control, based on the curvature and vehicle speed of the traveling road on which the inter-vehicle distance control is applied, the traveling road within the assumed range is used. The steering angle determination threshold value str_ref is set so as to include all possible angles at the curvature and the vehicle speed.

For example, if the inter-vehicle distance control is applied only in a situation where traveling at a speed of 40 km / h or more is considered to be almost the case, such as on an expressway, etc.
The steering angle which can be maximized when the vehicle travels at m / h may be set to a value that can be taken. Further, for example, in such a highway, the minimum radius of curvature is 2
Assuming that there is a curve of 50 m, it may be set in consideration of the maximum possible steering angle of the curve.

By setting the steering angle determination threshold value str_ref as described above, the driver can take in all the steering angles steered by the driver in the case of normal cruising at a vehicle speed of 40 km / h. That is, an affirmative determination is made in step S241. However, the steering angle determination threshold str_ref includes an error of the provisional steering angle neutral position S0 obtained when the vehicle turning detection unit determines that the vehicle is in the straight traveling state. Therefore, when the learning degree has advanced, step S240 is not satisfied ("NO" in S240), and a determination is made in step S242. In this step, the advance of the learning degree corrects the error of 6 degrees of the provisional steering angle neutral position SO, which is estimated in advance, and determines that the steering angle neutral position Sc is close to the true steering angle neutral position. Judgment and comparison with the steering angle determination threshold value str_ref
eng, but the measured steering angle str_eng
From the steering angle neutral position Sc.

When it is determined that the vehicle is traveling normally in a situation where the following distance control is allowed ("YES" in S241 or "YES" in S242), the calculation of the learning degree counter Cv is performed as follows. This is performed (S250).

[0069]

Cv (n) ← Cv (n−1) + α (Equation 3) Here, n represents the current processing, and n−1 represents the previous processing. Note that the initial value Cv (0) = 0 of the learning degree counter Cv.

The increase value α is as shown in FIG. In FIG. 7, the higher the current speed Vn, the larger the increase value α is set. That is, the higher the current vehicle reverse Vn is, the faster the learning degree counter Cv increases. However, the upper limit of the learning degree counter Cv is 65536. The upper limit is set to 16384 in the learning method in the prior art (Japanese Patent Application No. 9-192813). However, in the present embodiment, the steering is performed when the vehicle is normally driven in a situation where the above-mentioned inter-vehicle distance control is allowed. Since almost all of the angles are used for learning, a relatively large steering angle is used for learning, and as a result, the value of the steering angle neutral position Sc fluctuates greatly. Therefore, as shown in Expression 4 described later, it is necessary to increase the Cv value in order to suppress the variation, and the upper limit value is increased to 65536.

Next, it is determined whether or not the conditions of steps S230 and S240 are satisfied at first after the control of the headway control ECU 2 is started (S260). If it is the first time ("YES" in S260), the provisional steering angle neutral position S0 is set as the initial value of the learning target steering angle neutral position Sc (S270). In subsequent processing, step S
Since it is determined “NO” at 260, the initialization of the steering angle neutral position Sc (S270) is not executed.

Step S270 or S260
Next, learning calculation of the steering angle neutral position Sc is performed as in the following equation (S290).

[0073]

[Equation 4] Sc (n) ← Sc (n−1) + {str_eng−Sc (n−1)} × α / Cv (n) (Expression 4) n, n−1 and α are as described above. It is.

That is, in the learning calculation of Equation 4, the steering angle neutral position Sc (n-1) is obtained from the measured steering angle str_eng.
, The steering angle neutral position Sc (n-1) is corrected based on the angle {str_eng-Sc (n-1)} to obtain a new steering angle neutral position Sc (n). The learning of the position Sc is performed.

The angle {str_eng-Sc (n-1)} is not the same, but a predetermined ratio α / Cv (n) of the angle {str_eng-Sc (n-1)} is the steering angle neutral position Sc (n-
1), the steering angle neutral position Sc (n−
1) is corrected, and a new steering angle neutral position Sc (n) is formed.

Thereafter, every time the measured steering angle str_eng is obtained, the condition of step S230 is satisfied, and
If either one of S241 and S242 is established, step S
250 and S290 are executed, and the steering angle neutral position Sc is learned. That is, when it is determined that the turning of the vehicle is in the range indicating the straight traveling state in the steering angle calculation and transfer (FIG. 4) executed by the brake ECU 4, the steering angle S of the vehicle detected by the steering sensor 8 is determined. Is set as the provisional steering angle neutral position SO. Then, in the steering angle neutral position learning process (FIG. 6) executed by the following distance control ECU 2, the temporary steering angle neutral position SO is used (str_eng ←).
S-SO), learning of the steering angle neutral position Sc is performed.

More specifically, in the steering angle neutral position learning process (FIG. 6), almost all of the steering angles when the vehicle normally travels with the inter-vehicle distance control allowed around the provisional steering angle neutral position SO. The steering sensor 8 is set within the range of the steering angle determination threshold value str_ref set to use the learning for learning.
When there is the steering angle S of the vehicle detected by the above, from the difference between the steering angle S of the vehicle detected by the steering sensor 8 and the provisional steering angle neutral position SO (measured steering angle str_eng),
Based on the angle obtained by subtracting the steering angle neutral position Sc (str_eng−Sc), the steering angle neutral position Sc is corrected to obtain a new steering angle neutral position Sc.

When the steering angle S of the vehicle detected by the steering sensor 8 is within a predetermined range around the temporary steering angle neutral position SO obtained in the straight traveling state, the steering angle is neutral. The position Sc has been learned. Since the predetermined range is set so as to include almost all of the steering angle when the vehicle runs normally in a situation where the inter-vehicle distance control is allowed, the predetermined range in the prior art described in Japanese Patent Application No. 9-192813. Can properly cope with the following situations that could not be dealt with.

That is, in the prior art, only when the detected steering angle (S) of the vehicle is within a set range near the tentative steering angle neutral position (S0), the steering angle neutral position (S
Since the learning of c) was performed, as described with reference to FIG. 14, for example, when a short-time steering angle distribution is viewed in a situation such as a slalom that is continuously steered left and right, only one of the left and right peaks is observed. May exist within the set range. In this case, depending on the learning effect,
The steering angle neutral position (Sc) drifts to a peak at a position distant from the true steering angle neutral position.

On the other hand, in the case of the steering angle neutral learning device according to the present embodiment, even if peaks exist on the left and right, and even if they are asymmetric, the above-mentioned predetermined range includes all the peaks including both peaks. Is set to include the angle.
In other words, even if the steering angle is locally asymmetric, the distribution of the total steering angle (S) is often nearly symmetric when viewed comprehensively. Therefore, even in a situation where the steering is continuously performed to the left and right as in the above-described slalom, the steering angle (S) of only one side of the left and right steering is not biased and used for learning. Position (S
The inappropriate drift of c) can be prevented.

Of course, when viewed over a long period of time, the distribution of the steering angle (S) becomes close to a normal distribution and becomes equal to the left and right, so that the case of the prior art and the case of the present embodiment can be handled similarly. However, the present invention is effective in that it can appropriately cope with the above-mentioned special traveling situation and when the situation is seen in a short time.

In the prior art, a straight running state is determined on the basis of a turning state detected by the yaw rate sensor 10 or the like, and "a straight running state is indicated near the tentative steering angle neutral position obtained in the straight running state. Range "was set. Therefore, unless the detection accuracy of the vehicle turning detection means is high, the “range indicating the straight traveling state” cannot be set appropriately. Then, even if the data is originally effective for learning the steering angle neutral position (Sc), it is excluded, and the time required for the convergence to the true steering angle neutral position by the learning effect is relatively shortened. It will be long.

On the other hand, in the case of the present embodiment, as described above, almost all of the steering angles when the vehicle normally travels in a situation where the following distance control is permitted are used for learning.
Therefore, even if the detection accuracy of the yaw rate sensor 10 is not high, data effective for learning as in the above-described prior art is not excluded. In other words, the restriction on the height of the detection accuracy of the vehicle turning detection means such as the yaw rate sensor 10 is loose, and this is also effective in that respect.

Further, in the present embodiment, in the steering angle neutral position learning process (FIG. 6), the condition that the vehicle speed detected by the vehicle speed sensor 16 is equal to or higher than a predetermined speed (40 km / h) (S
If (230) is satisfied, the learning operation (S290) is executed. That is, traveling in an urban area is 40 km
/ H is likely to be run and steering is frequently repeated, but learning is not performed in such a state. On the other hand, when the vehicle is traveling on a suburb or a highway where the speed is higher than 40 km / h, which is being learned, the steering is stable, so that more accurate steering data can be obtained. Therefore, more accurate steering angle neutral position learning can be performed early without waste.

Next, the steering angle neutral position learning reset process which is a part of the process executed by the following distance control ECU 2 will be described. FIG. 8 shows a flowchart of the steering angle neutral position learning reset process. This steering angle neutral position learning reset processing is performed in a situation where a relatively large steering angle is continuously input at a rampway or the like, despite being within the range of the steering angle determination threshold value. This is a process for suppressing drifting of. Continuously large steering angles indicate that the vehicle itself is turning significantly. That is, based on the degree of turning, it is determined whether or not the vehicle is traveling on a special road, and when it is determined that the vehicle is traveling in a special section, a portion learned in that section is discarded, thereby performing steering. The accuracy of the angular neutral position Sc is ensured.

First, the measured steering angle str_eng is set to the engine EC.
It is determined whether or not it has been received from U6 (S320). If the measured steering angle str_eng has not been received ("NO" in S320), nothing is performed in this steering angle neutral position learning reset processing. When the measured steering angle str_eng is received (S
320 "YES"), and then the current vehicle speed Vn is 40 km / h.
It is determined whether or not this is the case (S330). Current vehicle speed Vn is 4
If it is 0 km / h or more ("YES" in S330), then it is determined whether the sign of the previous value and the current value of the difference between the measured steering angle str_eng and the steering neutral position Sc matches (S340).
If the signs do not match ("NO" in S340), it is determined that the steering has been switched back from right to left or from left to right, and the steering angle neutral position Sc and the value of the learning degree counter Cv at this time are stored. It does (S350). If the signs match ("YES" in S340), it is determined that the vehicle is being steered to the right or left continuously and the vehicle is turning, and step S3 is performed.
In steps 70, S380, and S390, the turning amount c_int is estimated. If it is determined in step S340 that the codes match ("YES" in S340), the stored values of the steering angle neutral position Sc and the learning degree counter Cv are not changed as the same as the previous values (S360).

Step S370 is a turn amount c_int of the vehicle.
Is a process of calculating a correction coefficient β for estimating the steering angle, and is obtained by a map calculation using a difference between the measured steering angle str_eng and the steering neutral position Sc as an input. The map is shown in FIG. Normally, when calculating the turning amount, the correction coefficient β is unnecessary, and β = 1. However, it is modified as shown in FIG. 9 in order to make this processing function in a special situation as a target. First, in the case of a steering with a high probability of being a special traveling road, the value of the correction coefficient B is set to be larger than 1 in order to increase the apparent turning amount and to make it advantageous in the determination in the post-processing. If the probability that the vehicle is traveling normally on the traveling road to which the following distance is to be applied is high, β = 0 is set in order to prevent this reset processing from functioning carelessly.
The turning amount is determined based on the true steering neutral position and the provisional steering angle neutral position S.
Due to the difference from O, there is a possibility that integration is performed even if the vehicle is traveling straight. Therefore, considering the error of the provisional steering angle neutral position SO, β = 0 is set in a range larger than that.

After setting β in step S370,
In step S380, a curve radius R is calculated. In general, the curve radius R can be calculated by the following equation.

[0089]

R = L × (1 + K × Vn ² ) / {(str_eng−Sc) / N} (Equation 5) L is a wheel base, K is a stability factor, and N is a steering gear ratio.

Instead of the theoretical formula, an empirical formula may be used, or a value obtained by dividing the vehicle speed Vn by the yaw rate may be used. Then, the vehicle turning amount c _{_} int in step S390
Is estimated. The estimation formula is shown below.

[0091]

C_int (n) = c_int (n−1) + β × Δt × Vn / R (Equation 6) where Δt is a calculation cycle.

In step S400, the absolute value of the vehicle turning amount c_int estimated in step S390 is set to the vehicle turning amount threshold.
It is determined whether or not c_int_ref (= π) has been exceeded. If the absolute value of the vehicle turning amount c_int value is equal to or less than the vehicle turning amount threshold value c_int_ref ("NO" in S400), nothing is performed in this steering angle neutral position learning reset process.

When the absolute value of the vehicle turning amount c_int exceeds the vehicle turning amount threshold c_int_ref (“YES” in S400),
The learning value up to that point is reset (S410).
That is, when it is determined that the turning amount of the vehicle is large unlike normal running, the learning value at the time when the turning is started is returned. This learning value is the value stored in step S350. At the same time, the vehicle turning amount c_int is also reset to zero. This reset can suppress the drift of the steering angle neutral position Sc on a special traveling road such as a rampway, and as a result, can improve the accuracy of the curve curvature calculation of the traveling state of the own vehicle, and can be controlled by the following distance control. It is possible to improve the accuracy of selecting a preceding vehicle as an object.

The steering angle calculation processing of this embodiment (FIG. 4)
The steering angle neutral position Sc learning results obtained by learning the steering angle neutral position learning process (FIG. 6) and the steering angle neutral position learning reset process (FIG. 8) are compared with those of the prior art in FIG.
0, FIG. 11, FIG. 12 and FIG.

FIGS. 10 and 11 show the transition of the steering angle neutral position Sc when the vehicle speed is accelerated to about 100 km / h after the ignition is turned on on a traveling road where the straight traveling state is small and the left-right steering is continuous. When the ignition is turned on, the steering angle calculation process shown in FIG. 4 is not executed, so that the steering angle neutral position Sc is set to zero. 30 miles away
km / h, when the yaw rate value becomes close to zero, about 50 seconds, it is determined that the vehicle is traveling straight, and the provisional steering angle neutral position S is determined.
O is set. The provisional steering angle neutral position SO is displayed as zero on the vertical axis.

FIG. 10 shows a prior art (Japanese Patent Application No. 9-1928).
No. 13), and FIG. 11 shows an embodiment of the present invention. In FIG. 10, the provisional steering angle neutral position SO
The steering angle neutral learning is executed only when there is a measured steering angle str_eng near the center. However, as shown in the figure, in a situation where the straight traveling state is small and left and right steering is continuous,
The frequency at which the steering angle in the negative direction (left direction) is within the range of the steering angle determination threshold value str_ref set near the provisional steering angle neutral position SO is higher than the steering angle in the positive direction (right direction). Thereby, the true steering angle neutral position is about 6 deg in this case.
Despite being in the vicinity, the state of drifting in the negative direction without convergence continues. The steering angle neutral position S is changed to a so-called true steering angle neutral position that is about 6 deg away from the provisional steering angle neutral position SO.
There is a situation where c does not approach.

On the other hand, in FIG. 11, the provisional steering angle neutral position S
A steering angle determination threshold value str_ref is set to use all of the measured steering angles str_eng, which are considered to be normal traveling on the traveling road to which the inter-vehicle distance is applied, around O. The steering angle determination threshold value str_ref is set such that the lower the vehicle speed, the more the measured steering angle st
Taking into account the tendency of r_eng to increase, the lower the speed, the larger
It is set to be smaller at higher speeds. As shown in the figure, in order to use all the measured steering angles str_eng, the neutral steering angle position Sc converges to about 6 deg, which is the true steering angle neutral position.

FIGS. 12 and 13 show the transition of the steering angle neutral position Sc when the vehicle speed is accelerated to about 80 km / h after the ignition is turned on in a ramp way which is special from the viewpoint of neutral learning of the steering angle. I have. When the ignition is turned on, the steering angle calculation process shown in FIG. 4 is not executed, so that the steering angle neutral position Sc is set to zero. When the vehicle speed exceeds 30 km / h and the yaw rate becomes nearly zero, the vehicle is determined to be in the straight traveling state at about 50 seconds, and the provisional steering angle neutral position SO is set. This provisional steering angle neutral position SO
Is displayed as zero on the vertical axis.

FIG. 12 is based on the prior art, and FIG.
Shows an embodiment of the present invention. In FIG.
Steering angle neutral learning is executed only when the measured steering angle str_eng is near the temporary steering angle neutral position SO. As shown in the figure, in a ramp way having a large steering angle, the steering angle neutral learning position is easily drifted because the steering angle neutral learning is prohibited because the steering angle neutral learning is prohibited because the steering angle neutral learning is prohibited. Never.

On the other hand, in FIG. 13 showing the embodiment of the present invention, since the measured steering angle str_eng which is larger than the conventional one is used for learning, there is a concern about the drift of the steering angle neutral position Sc. However, it can be seen that the drift of the steering angle neutral position Sc can be suppressed by the steering angle neutral position learning reset processing shown in FIG. Specifically, in the section from 30 seconds to 50 seconds during which the vehicle is traveling on the rampway, the measured steering angle st
The difference between r_eng and the neutral steering angle position Sc increases, the correction coefficient β becomes larger than 1, and the vehicle turning amount c_int is integrated. At about 40 seconds, the vehicle turning amount c_int becomes 1π
, The steering angle neutral position Sc is returned to the value before continuous steering. At the same time, both the vehicle turning amounts c_int are returned to zero, and the integration is continued again. Once the steering angle neutral position Sc is set once, the measured steering angle str_
Since eng exceeds the steering angle determination threshold value str_ref, the steering angle neutral position Sc is maintained. The vehicle turning amount c_int continues to be integrated while the measured steering angle str_eng exceeds the steering angle determination threshold str_ref.

In FIG. 13, the dotted line indicates the fluctuation of the steering angle neutral position Sc when the steering angle neutral position learning reset processing is not performed. Measurement steering angle str_eng
In a situation where the difference between the steering angle and the steering angle neutral position Sc is large, the drift amount is correspondingly large, the convergence to the true steering angle neutral point is delayed, and the accuracy of the preceding vehicle selection in this section decreases. Would.

As described above, the steering angle neutral position Sc can be learned with high accuracy in a situation where the vehicle is left and right in a straight-ahead state and the vehicle is steered continuously, and the ramp way which is a concern that accompanies the implementation of the present invention. Even in a situation where the steering angle is large, for example, the steering angle neutral position Sc can be converged early by the learning value reset processing.

The inter-vehicle control ECU 2 performs processing as a curve curvature calculating means using the steering angle neutral position Sc and the measured steering angle str_eng learned as described above. The processing as the curve curvature calculating means includes, for example,
First, an estimated steering angle str is obtained as in the following equation.

[0104]

[Formula 7] str ← str_eng−Sc [Equation 7] Actually, the estimated steering angle str is smoothed by a first-order low-pass filter having a cutoff frequency of 1.24 Hz as shown in the following equation, and smoothing is performed. The later estimated steering angle str_filtrer is used.

[0105]

[Equation 8] str_filtrer (n) ← (176/256) × str_filtrer (n−1) + (80/256) × str (n) [Equation 8] Then, with this estimated steering angle str_filtrer, A curve radius of curvature R of the traveling path is calculated.

(1) When Vn ≧ 80 km / h

[0107]

[Expression 9] R ← Kr × (1 + 1.69 × 10∧-4 × Vn∧2-3.86 × 10∧-8 × Vn ／ 3) / str_filtrer [Equation 9] (2) When Vn <80 km / h

[0108]

[Expression 10] R ← Kr × (1 + 1.20 × 10∧-4 × Vn∧2 + 5.79 × 10∧-7 × Vn∧3) / str_filtrer [Equation 10] Here, “＾” is “＾” Means that the number before “” is raised to the power of the number after “＾”, and the same applies to other parts of this specification. Kr is a curve radius constant (for example, value 2100).

The calculated distance R of the curvature of the curve thus calculated is transmitted from the headway control ECU 2 to the laser radar sensor 3. Next, the processing performed by the laser radar sensor 3 and the processing performed by the headway control ECU 2 will be described.

FIG. 15 shows the processing of the overall headway control. The preceding vehicle detection processing in steps S1000 to S5000 is processing performed by the laser radar sensor 3, and the inter-vehicle control processing in steps S6000 to S9000 is processing performed by the inter-vehicle control ECU 2.

When the processing is started, first, the distance and distance by the scanning distance measuring device provided in the laser radar sensor 3 are calculated.
The angle measurement data is read (S1000). Next, recognition processing of a forward obstacle is performed (S2000). In the recognition processing of the front obstacle, the recognition type of the front object, the object width W, the XY coordinates of the center position of the object, and the relative speed Vr are obtained based on the current vehicle speed Vn and the result of the scanning of the front object. The recognition type can be recognized as a moving object, for example, when the relative position of the object hardly moves even though the own vehicle is running. An object that gradually moves away can also be recognized as a moving object. When the relative position of the object approaches the own vehicle at the same speed (absolute value) as the current vehicle speed Vn, it can be recognized as a stationary object. Other objects, for example, an object that has not passed for a recognizable time since it appeared, are recognized as unknown objects. The process of recognizing the forward obstacle is well known to those skilled in the art.

Next, the own lane probability calculation process (S
4000). Own lane probability calculation processing (S4000)
First, instantaneous own lane probability is calculated (S401).
0). In the instantaneous lane probability calculation, first, the center position / object width data (X0, Y0, W0) of all objects obtained in the forward obstacle recognition process (S2000) is converted into a straight path. That is, based on the curve radius of curvature R received from the following distance control ECU 2, the coordinates of the object when the curve is taken as a straight path are obtained. The conversion is represented by the following equation (11).
This is performed by performing coordinate conversion using a to 11c.

[0113]

X ← X0− (Y0 ＾ 2 / 2R) [Expression 11a] Y ← Y0 [Expression 11b] W ← W0 [Expression 11c] In other words, here, only the X coordinate is converted. I have.

The center position / object width data (X, Y, W) obtained as a result of conversion into a straight road is arranged on the own lane probability map shown in FIG. The own lane probability, that is, the probability of being in the own lane at that time is obtained. The probability exists because there is an error between the curve radius of curvature R obtained from the steering angle by the following distance control ECU 2 and the actual curve radius of curvature. To find the instantaneous lane probability of each object.

In FIG. 17, the horizontal axis is the X axis, that is, the horizontal direction of the own vehicle, and the vertical axis is the Y axis, that is, the front of the own vehicle. In this embodiment, left and right 5 m, front 10
The area up to 0 m is shown. Here, the region is a region a
(Own lane probability 80%), area b (own lane probability 60%),
Area c (own lane probability 30%), area d (own lane probability 10)
0%) and the other area (own lane probability 0%). The setting of this area is determined by actual measurement.
In particular, the area d is an area set by taking into consideration the interruption immediately before the own vehicle.

The boundary lines La, which delimit the regions a, b, c, d,
Lb, Lc, and Ld are given by, for example, the following equations 12 to 15. Note that the boundary lines La ', Lb', L
c ′ and Ld ′ are boundary lines La, Lb, Lc and L, respectively.
d has a symmetrical relationship with the Y axis.

[0117]

[Equation 12] La: X = 0.7 + (1.75-0.7) · (Y / 100) ＾ 2 [Equation 12] Lb: X = 0.7 + (3.5-0.7) · (Y / 100) ＾ 2 ... [Equation 12] 13] Lc: X = 1.0 + (5.0−1.0) · (Y / 100) ＾ 2 [Expression 14] Ld: X = 1.5 · (1-Y / 60) [Expression 15] This is represented by a general expression. And the following equations 16 to 19 are obtained.

[0118]

La: X = A1 + B1 · (Y / C1) ＾ 2 [Expression 16] Lb: X = A2 + B2 · (Y / C2) ＾ 2 [Expression 17] Lc: X = A3 + B3 · (Y / C3) ＾ 2 [Expression 18] Ld: X = A4 · (B4-Y / C4) [Expression 19] From Expressions 16 to 19, the following Expressions 20 to 22 are generally obtained. Set the area to satisfy. The actual numerical values are determined by experiments.

[0119]

A1 ≦ A2 ≦ A3 <A4 (Equation 20) B1 ≦ B2 ≦ B3 and B4 = 1 (Equation 21) C1 = C2 = C3 (C4 is not limited) (Equation 22) FIG. 17 Boundary lines La, Lb, Lc, La ', L
Although b 'and Lc' are parabolic from the viewpoint of calculation processing speed, if processing speed permits, it is better to represent them with arcs. The boundary lines Ld and Ld 'are preferably represented by parabolas or arcs bulging outward if the processing speed permits.

(6) The instantaneous lane probability P0 of each object is determined as follows. An object having any area d → P0 = 100% An object having a center in area a → P0 = 80% An object having a center in area b → P0 = 60% An object having a center in area c → P0 = 30% An object that does not satisfy the above-> P0 = 0% Next, the instantaneous own lane probability P0 of each object obtained in this way is time-averaged by the following equations 23 and 24 to obtain the own lane probability P. Ask. That is, filter processing is performed (S402).
0). However, the initial value of the own lane probability P is “0%”.

[0121]

[Expression 15] P ← P × 0.8 + P0 × 0.2 (where P0 <90%) [Expression 23] P ← P × 0.7 + P0 × 0.3 (where P0 ≧ 90%) [Expression 24] 90% or more, the ability to follow the instantaneous lane probability is high
In particular, when an interrupted vehicle is present ahead of the own vehicle, it is possible to quickly cope with the interrupted vehicle.

Next, a limit is set for the own lane probability.
The final own lane probability P is determined (S4030).
(7) The limit is set as follows. When the recognition type is a moving object, the above equation 23 or the above equation 2 is used.
It is assumed that the own lane probability P remains calculated in step 4.

When the recognition type is a stationary object, the following (a) to
(E) If any of the conditions is satisfied, the maximum value of the own lane probability P is set to 20%. (A) Y0> 40m and W0 <1.4m (b) Y0> 30m and W0 <1.2m (c) Y0> 20m and W0 <1.0m (d) Those which are recognized for less than 1 second (scan) (E.g. less than 5 times) (e) Among other moving objects, there is an object whose own lane probability P ≧ 50% and which is recognized longer than itself.

As described above, the own lane probability of each object is obtained in step S4000. Next, a preceding vehicle is selected from the objects (S5000). This preceding vehicle selection processing (S5000) is shown in FIG. First, one traveling preceding vehicle is extracted from the moving object, divided into a moving object and a stationary object (S5010), and one stationary preceding vehicle is extracted from the stationary object (S5020).

[Case of Moving Object] (S5010) A moving object satisfying the following condition and having the largest lane probability P is extracted. (A) When | R | <500m, own lane probability P> 30% (b) When 500m ≦ | R | <1000m, own lane probability P> 40% (c) When 1000m ≦ | R | The lane probability P> 50% As described above, the smaller the absolute value of the curve radius of curvature R is, the more the extraction condition is loose (something with a small own lane probability P is also extracted). This is because it is difficult to find a car.

When a plurality of moving objects are extracted as described above, (the maximum own lane probability P-1 of those moving objects)
5%) or a moving object having the own lane probability P ≧ 70% or a moving object having the own lane probability P ≧ 70% as the preceding vehicle running the moving object with the smallest Y0. Extract. If not extracted above, there is no preceding vehicle running.

[In the Case of a Stopped Object] (S5020) Among the stopped objects having the own lane probability P ≧ 70%, the stopped object with the minimum Y0 is extracted as the stopped preceding vehicle. If not extracted, there is no preceding vehicle that is stopped. In the case of a stationary object, the criterion is made stricter than that of a moving object so that a roadside object is not determined as a preceding vehicle.

[Comprehensive Judgment] (S5030) The preceding vehicle is selected as follows from the results of the extraction in the case of a moving object (S5010) and the case of a stationary object (S5020). If neither a running preceding vehicle nor a stopped preceding vehicle exists, it is determined that there is no preceding vehicle.

If there is any one of a running preceding vehicle and a stopped preceding vehicle, the preceding vehicle is regarded as the preceding vehicle. When both the traveling preceding vehicle and the stopped preceding vehicle are present, the smaller Y0 is regarded as the preceding vehicle.

(If the preceding vehicle is determined as described above, even if the preceding vehicle is lost or misidentified, the most probable object is selected as the preceding vehicle from among the plurality of detected objects each time, so there is a momentary error. The preceding vehicle detection processing (S1000 to S5000) is thus completed. The preceding vehicle information including the own lane probability of the preceding vehicle calculated in this way is transmitted from the laser radar sensor 3 to the inter-vehicle control ECU 2. The inter-vehicle control ECU 2 receives the preceding vehicle information and performs the inter-vehicle control process (S6000 to S6000).
9000).

At the beginning of the following control process, a target following calculation process shown in the flowchart of FIG. 19 is executed (S600).
0). First, it is determined whether or not the initial operation is being performed (S60).
10). “Initial” means the timing at which this processing is first executed after the power is turned on.

First, an affirmative determination is made in step S6010, and an initial value T0 is set as the target inter-vehicle time TH (S6020). As this initial value T0, for example, "2.5 seconds" is set. If a negative determination is made in step S6010, or after the processing in step S6020, it is determined whether a tap-down operation has been performed (S6030). Further, when a negative determination is made in step S6030, it is determined whether a tap-up operation has been performed (S6040).

[0133] The tap-down operation is an operation for increasing the headway by operating the tap switch in the cruise control switch 20 by the driver. In contrast to the tap-up operation, this is an operation to reduce the distance between vehicles by operating a tap switch. If the tap-down operation has been performed, an affirmative determination is made in step S6030, and a process of increasing the target inter-vehicle time TH is performed as in the following Expression 25 (S6060).

[0134]

[Expression 16] TH ← TH + 0.18 seconds [Expression 25] However, the upper limit of the target inter-vehicle time TH is set to 3.3 seconds by the processing of the following steps S6070 and S6080.

On the other hand, if the tap-up operation has been performed, an affirmative determination is made in step S6040, and the following equation (2) is obtained.
As in 6, the process of reducing the target inter-vehicle time TH is performed (S6090).

[0136]

[Expression 26] However, the lower limit of the target inter-vehicle time TH is set to 0.7 second by the processing of the following steps S6100 and S6110.

After the target inter-vehicle time TH is set in this way, the target inter-vehicle time TH is converted into the target inter-vehicle distance Dt by the current vehicle speed Vn as shown in the following Expression 27 (S6050).

[0138]

Dt ← TH × Vn (Expression 27) Next, the acceleration / deceleration rate calculation process (S7000) shown in the flowchart of FIG. 20 is executed.

First, it is determined whether the vehicle is coasting or not (S7).
010), if the vehicle is not coasting, it is determined whether the vehicle is accelerating (S7020), and it is determined whether the preceding vehicle is being recognized (S7030). Here, the coast means that when the set switch in the cruise control switch 20 is pressed during the constant speed running control, the deceleration control is performed, and then the current vehicle speed Vn when the set switch is released is set as the target speed Vm. The process shifts to the high-speed running control, and “during coasting” means a period of the deceleration control. The accelerator is controlled to increase the speed when the resume switch in the cruise control switch 20 is pressed during the constant-speed traveling control during the constant-speed traveling control, and then sets the target vehicle speed Vn when the resume switch is released. The process shifts to the constant speed traveling control as the speed Vm, and the term “acceleration” means a period of this increasing control.

Therefore, if the vehicle is coasting, an affirmative determination is made in step S7010, and the acceleration / deceleration rate At is set to “−”.
"2.6 km / h / s" is set (S7100). If the accelerator is being depressed, an affirmative determination is made in step S7020, and "2.6 km / h / s" is set as the acceleration / deceleration rate At (S7090). .

If the preceding vehicle is being recognized while the vehicle is not in the coast or the accelerator, that is, if the preceding vehicle is selected in step S5000, step S703 is performed.
0, the basic acceleration / deceleration rate calculation process (S70)
40) is executed. (8) Basic acceleration / deceleration rate calculation processing (S
7040) is performed as follows.

The headway deviation De is calculated by the following equation (28).
It is calculated from the following distance D (= Y) with the preceding vehicle and the target following distance Dt obtained in step S6050.

[0143]

[Expression 19] De ← D−Dt [Expression 28] Next, based on the inter-vehicle deviation De and the relative speed Vr, FIG.
The basic acceleration / deceleration rate MD
Find V (km / h / s). For hysteresis, a dead zone of 2 m is provided for the inter-vehicle deviation De and a dead zone of 1 km / h is provided for the relative speed Vr at each boundary between the inter-vehicle deviation De and the relative speed Vr. If the area exceeds the area of this map, the value of the closest area is set. In FIG. 21, even if the inter-vehicle deviation De is negative, when the speed of the preceding vehicle is high and the vehicle gradually moves away (Vr> 0), the speed is increased (basic acceleration / deceleration rate MD).
V> 0). This is because even if the inter-vehicle distance is short, if the preceding vehicle moves away, there is no need to decelerate the own vehicle, and if decelerated, the driver feels unnecessary deceleration.

Next, in order to correct the basic acceleration / deceleration rate MDV based on the distance, a correction coefficient KMDV is obtained from the relationship with the headway D shown in FIG. 22 (S7050). This is to prevent a sensitive reaction to a distant preceding vehicle. Next, the acceleration / deceleration rate At is obtained as in Expression 29 (S706).
0).

[0145]

[Equation 20] At ← MDV × KMDV / 100 [Equation 29] If a negative determination is made in step S7030, and if it is within 5 seconds after the end of the accelerator, step S70 is performed.
The affirmative determination is made in step 70, and the acceleration / deceleration rate At is set to "2.6 km / h / s" in step S7090. If the acceleration is not within 5 seconds after the end of the accelerator, a negative determination is made in step S7070 and a step S7080 is performed. The acceleration / deceleration rate At is set to “1.3 km / h / s”.

Here, when the affirmative determination is made in step S7070, the reason why the acceleration / deceleration rate At is set to 2.6 km / h / s is that the driver wants to accelerate in order to respect the driver's will as much as possible. This is because when expressing a will, the priority is given to this control.

Thus, the acceleration / deceleration rate calculation processing (S700)
0) is completed, and then a target vehicle speed calculation process (S8000) is performed. The target vehicle speed calculation processing (S8000) is shown in FIG.
First, the target vehicle speed Vm is calculated by the following equation (3).
It is calculated as 0 (S8010).

[0148]

Vm ← Vm + At × dt (Equation 30) Here, dt represents a time interval of the processing in step S8010, and is “0.2 seconds” in the present embodiment.

Next, the following limit is set for the target vehicle speed Vm obtained in step S8010 (S802).
0). When Vm> Vn + 2 km / h and At <0,
The target vehicle speed Vm is set as in the following Expression 31.

[0150]

Vm ← Vn + 2 km / h (Equation 31) When Vm <Vn−2 km / h and At> 0,
The target vehicle speed Vm is set as in the following Expression 32.

[0151]

Vm ← Vn−2 km / h (Expression 32) In addition to the above restrictions, the target vehicle speed Vm is further restricted as follows.

That is, (a) the target vehicle speed Vm is the set vehicle speed Vs for constant speed traveling control set by the driver.
No more. However, except during the accelerator. (B) The target vehicle speed Vm satisfies the following Expression 33.

[0153]

Vn−8 km / h ≦ Vm ≦ Vn + 3.5 km / h (Equation 33) Thus, the target vehicle speed Vm when the preceding vehicle is recognized
Is set.

After the target vehicle speed Vm is determined, it is determined whether or not the condition for fully closing the throttle is satisfied (S803).
0), if not, it is determined whether or not the throttle full-close release condition is satisfied (S8050). The throttle fully-closed condition is a condition for starting a process of rapidly decelerating when the current vehicle speed Vn is excessively higher than the target vehicle speed Vm, and is shown in the following Expression 34. The throttle fully closed release condition is a condition for stopping the deceleration process, and is shown in the following Expression 35.

[0155]

Vm <Vn−3 km / h (Equation 34) Vm ≧ Vn−2 km / h (Equation 35) If the condition of step S8030 is satisfied, the throttle fully closing control (S8040) is performed. Step S8
If the condition of 050 is satisfied, a release process of the throttle fully closed control (S8060) is performed.

The throttle fully closing control is a deceleration control in which the duty for determining the rotation speed of the motor controlling the opening of the throttle valve of the internal combustion engine is the maximum duty output (highest speed) in the direction in which the throttle valve closes. Means to do. When the target vehicle speed calculation process (S8000) is completed in this way, vehicle speed control is performed with the target vehicle speed Vm obtained in step S8000 as a target in the same manner as the conventionally known constant speed vehicle speed control (S8000). S
9000).

Since the present inter-vehicle control process is configured as described above, the coordinates of each forward object converted into a straight path based on the curve radius of curvature R are stored in the own lane probability map of the straight path set in advance. It is possible to determine the own lane probability of each object by applying, determine the preceding vehicle from the state of the own lane probability, adjust the speed of the own vehicle based on the positional relationship with the preceding vehicle, and control the inter-vehicle distance. .

In this inter-vehicle distance control process, the laser radar sensor 3 uses the data of the curve radius of curvature R estimated based on the steering angle neutral position Sc obtained early and with high accuracy by the inter-vehicle control ECU 2. The preceding vehicle can be appropriately selected by the scanning range finder.
Therefore, the following distance control ECU 2 can perform highly accurate following distance control for the preceding vehicle.

[Others] In the above embodiment, the yaw rate sensor 10 is used as the vehicle turning detecting means.
The difference between the rotational speeds of the left and right wheels may be obtained, and the magnitude of the yaw rate may be determined based on the rotational speed difference between the left and right wheels.
Alternatively, the yaw rate may be calculated from the difference between the rotational speeds of the left and right wheels, and the determination may be made in the same manner as when the yaw rate sensor 10 is used. Further, a lateral acceleration sensor may be used instead of the yaw rate sensor 10.

The above-described processing is performed by the ROM or backup RA provided in the headway control ECU 2, the laser radar sensor 3, the brake ECU 4, and the engine ECU 6.
M is executed by a program stored in M.

[Brief description of the drawings]

FIG. 1 is a system block diagram of an inter-vehicle distance control device as one embodiment.

FIG. 2 is a schematic diagram illustrating the configuration of a steering sensor.

FIG. 3 is an explanatory diagram of an output signal of a steering sensor.

FIG. 4 is a flowchart of a steering angle calculation process.

FIG. 5 is an explanatory diagram of a conversion table for calculating a steering angle based on an output of a steering sensor.

FIG. 6 is a flowchart of a steering angle neutral position learning process.

FIG. 7 is an explanatory diagram of a conversion table for obtaining an increase value α used for calculating a learning degree counter Cv from a vehicle speed.

FIG. 8 is a flowchart of a steering angle neutral position learning reset process.

FIG. 9 is a map for calculating a correction coefficient β.

FIG. 10 shows a transition of the neutral steering position Sc when the vehicle speed is increased from the ignition on to about 100 km / h after the ignition is turned on on a traveling road where the straight-ahead state is small and the left-right steering is continuous in the prior art (Japanese Patent Application No. 9-192813). FIG.

FIG. 11 is a diagram illustrating a steering angle neutral position S when the vehicle speed is increased to about 100 km / h from the ignition on on a traveling road in which straight-ahead state is small and left-right steering is continuous in the present embodiment.
It is a graph which shows transition of c.

FIG. 12 is a graph showing a transition of a steering angle neutral position Sc when a vehicle speed is accelerated to about 80 km / h from ignition on in a ramp way in the prior art (Japanese Patent Application No. 9-192813).

FIG. 13 is a graph showing a transition of the steering angle neutral position Sc when the vehicle speed is increased to about 80 km / h from the ignition on in the ramp way in the present embodiment.

FIG. 14 is a graph showing a possible steering angle distribution on a traveling road such as a slalom that is continuously steered left and right.

FIG. 15 is a flowchart of the entire process of the headway control.

FIG. 16 is a flowchart of an own lane probability calculation process.

FIG. 17 is an explanatory diagram of an own lane probability map.

FIG. 18 is a flowchart of a preceding vehicle selection process.

FIG. 19 is a flowchart of a target headway calculation process.

FIG. 20 is a flowchart of an acceleration / deceleration rate calculation process.

FIG. 21 is a map for obtaining a basic acceleration / deceleration rate MDV from an inter-vehicle deviation De and a relative speed Vr.

FIG. 22 is a graph showing the relationship between the following distance D and a correction coefficient KMDV.

FIG. 23 is a flowchart of a target vehicle speed calculation process.

[Explanation of symbols]

2: Electronic control device for inter-vehicle control (inter-vehicle control ECU) 3: Laser radar sensor 4: Electronic control device for brake (brake ECU) 6: Electronic control device for engine (engine ECU) 8: Steering sensor 10: Yaw rate sensor 12: Wheel speed Sensor 14 Alarm buzzer 16 Vehicle speed sensor 18 Brake switch 20 Cruise control switch 22 Cruise main switch 24 Body LAN 32 Metal disk 32a Slit 34 Rotation detection circuit 34a, 34b Laser diode 34c, 34d Phototransistor 34e Detection circuit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI // B62D 101: 00 113: 00 (56) References JP-A-4-161806 (JP, A) JP-A-7-110712 ( JP, A) JP-A-3-276874 (JP, A) JP-A-7-132845 (JP, A) JP-A-3-189273 (JP, A) JP-A-4-24168 (JP, A) JP JP-A-6-270829 (JP, A) JP-A-5-238403 (JP, A) JP-A-5-77627 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) B62D 6 / 00

Claims (19)

(57) [Claims] 1. A steering angle detecting means for detecting a steering angle of a vehicle, a vehicle turning detecting means for detecting a turning of the vehicle, and a turning of the vehicle detected by the vehicle turning detecting means in a range indicating a straight traveling state. A vehicle straight traveling state determining means for determining whether or not there is a vehicle, and a vehicle detected by the steering angle detecting means when the vehicle straight traveling state determining means determines that the turning of the vehicle is in a range indicating the straight traveling state Temporary steering angle neutral position setting means for setting the steering angle of the vehicle as a temporary steering angle neutral position; and a difference between the steering angle of the vehicle detected by the steering angle detecting means and the temporary steering angle neutral position is within a predetermined range. If
Calculating a new steering angle neutral position by correcting the steering angle neutral position based on an angle obtained by subtracting the steering angle neutral position from a difference between the detected steering angle of the vehicle and the provisional steering angle neutral position. And a steering angle neutral position learning means for learning a steering angle neutral position. The predetermined range is assumed to be applied to vehicle running control executed using the steering angle neutral position. A steering angle neutral learning device, which is set based on the curvature of the traveling path and the vehicle speed so as to include all possible angles of the curvature of the traveling path and the vehicle speed within the assumed range. 2. The method according to claim 1, wherein the predetermined range is an error existing between the learned steering angle neutral position and a true steering angle neutral position at an initial stage of learning of the steering angle neutral position by the steering angle neutral position learning means. The steering angle according to claim 1, wherein when the learning of the neutral position of the steering angle has advanced by a predetermined degree or more, the steering angle is set by excluding a reduction in an error due to the learning. Neutral learning device. 3. The predetermined range is set around the tentative steering angle neutral position at the beginning of learning of the steering angle neutral position by the steering angle neutral position learning means, while the learning of the steering angle neutral position is performed. 2. The steering angle neutral learning device according to claim 1, wherein when the vehicle travels a predetermined degree or more, the steering angle neutral position obtained by the learning is set as a center. 4. A vehicle speed detecting means for detecting a vehicle speed, wherein the steering angle neutral position learning means detects the steering angle neutral position when the vehicle speed detected by the vehicle speed detecting means is higher than a predetermined speed. The steering angle neutral learning device according to claim 1, wherein learning is performed. 5. The steering angle neutral learning device according to claim 1, wherein the provisional steering angle neutral position is set as an initial value of the steering angle neutral position. 6. The tentative steering angle neutral position setting means is configured to perform the steering when the vehicle straight traveling state determining means determines that the turning of the vehicle is in a range indicating the straight traveling state in the processing of the present apparatus. The steering angle neutral learning device according to any one of claims 1 to 5, wherein the process of setting the steering angle of the vehicle detected by the angle detecting means as a temporary steering angle neutral position is performed only once. 7. A steering angle neutral position learning means for determining a predetermined angle obtained by subtracting a steering angle neutral position from a difference between a steering angle of the vehicle detected by the steering angle detecting means and the temporary steering angle neutral position. The steering angle neutral learning according to any one of claims 1 to 6, wherein the learning of the neutral steering angle is performed by correcting the neutral position of the steering angle by a ratio to obtain a new neutral position of the steering angle. apparatus. 8. The predetermined ratio is set to increase as the vehicle speed increases, according to the vehicle speed when learning the steering angle neutral position is performed by the steering angle neutral position learning means. The steering angle neutral learning device according to claim 7, characterized in that: 9. The method according to claim 1, wherein the predetermined ratio is reduced as the number of times of learning increases in accordance with the number of times of learning of the steering angle neutral position performed by the steering angle neutral position learning means. Claim 7 or 8
The steering angle neutral learning device according to the above. 10. The system according to claim 9, wherein the predetermined ratio is made smaller in accordance with a degree of progress of learning of the neutral position of the steering angle as the set width of the predetermined range is larger. Steering angle neutral learning device. 11. When the turning angle for determination obtained based on the turning angle of the vehicle detected by the vehicle turning detecting means exceeds a predetermined threshold value, the turning angle for determination exceeding the threshold value is set to the threshold value. The learning result obtained by the steering angle neutral position learning means obtained during the special turning state in which the vehicle is continuously turning in one of the left and right directions is discarded, and the steering angle neutral position is changed to a value before the special turning state. The steering angle neutral learning device according to any one of claims 1 to 10, further comprising a steering angle neutral learning reset unit for returning to the state of (1). 12. The steering angle neutral learning reset means, wherein when the turning angle of the vehicle detected by the vehicle turning detecting means exceeds a predetermined threshold value, the steering angle neutral learning reset means changes the predetermined rate which varies according to the number of times of learning. The steering angle neutral learning device according to claim 11, wherein the state is returned to a state before the special turning state. 13. The turning angle for determination is corrected to be larger than the actually detected turning angle when the turning of the vehicle detected by the vehicle turning detecting means is relatively large. The steering angle neutral learning device according to claim 12, wherein: 14. The turning angle for determination is corrected to be smaller than the actually detected turning angle when the turning of the vehicle detected by the vehicle turning detecting means is relatively small. The steering angle neutral learning device according to claim 12 or 13, wherein: 15. The steering angle neutral learning device according to claim 1, wherein said vehicle turning detecting means is a sensor for detecting a yaw rate or a lateral acceleration. 16. A vehicle turning detecting device according to claim 1, wherein said vehicle turning detecting means detects a speed difference between left and right wheels and detects a turning of the vehicle based on the speed difference. The steering angle neutral learning device according to the above. 17. A steering angle neutral learning device according to claim 1, a vehicle speed detecting means for detecting a vehicle speed, a steering angle neutral position obtained by said steering angle neutral learning device, Based on the steering angle of the vehicle detected by the steering angle detection means or the difference between the steering angle and the provisional steering angle neutral position, and the vehicle speed detected by the vehicle speed detection means, the curvature of the traveling road is calculated. A curve curvature estimating device comprising: a curve curvature calculating means for calculating. 18. A steering angle neutral learning device according to any one of claims 1 to 16, vehicle speed detecting means for detecting a vehicle speed of the own vehicle, and a steering angle neutral learning device obtained by said steering angle neutral learning device. A travel path based on a position, a steering angle of the vehicle detected by the steering angle detecting means or a difference between the steering angle and the tentative steering angle neutral position, and a vehicle speed detected by the vehicle speed detecting means. A curve curvature calculating means for calculating the curvature of the vehicle, scanning and irradiating a transmission wave or a laser beam to a predetermined angle range in the vehicle width direction, and based on a reflected wave or reflected light from the object, a distance between the own vehicle and a forward object is calculated. A distance measuring unit that can be detected in accordance with a scan angle; and, based on a distance detected by the distance measuring unit and a corresponding scan angle, calculate a relative position of the object with respect to the own vehicle, and An object recognition unit that calculates a relative speed of the object; and the object based on a curvature of the own vehicle traveling path obtained by the curve curvature calculation unit and a relative position of the object calculated by the object recognition unit. The own lane probability calculating means for determining the probability of being on the same lane as the own vehicle, based on the probability obtained by the own lane probability calculating means, a preceding vehicle selecting means for selecting a preceding vehicle to be controlled inter-vehicle distance, Inter-vehicle distance control means for controlling the inter-vehicle distance to the preceding vehicle selected by the preceding vehicle selecting means by adjusting the speed of the own vehicle detected by the vehicle speed detecting means. Inter-vehicle distance control device. 19. A computer-readable recording medium in which a program for causing a computer system to function as each means of the steering angle neutral learning device according to claim 1 is recorded.