JPWO2007015288A1 - Axis deviation amount estimation method and axis deviation amount estimation device - Google Patents

Abstract

Accurately estimate the radar axis deviation. In a misalignment estimation apparatus for estimating the misalignment of a radar that detects a target in front of a traveling vehicle, a stationary object extraction unit 21 that selects a stationary object echo from a transmission wave echo based on a distance and a speed. An observation area dividing unit 221 that divides the coverage area of the radar into a plurality of areas, and a reflection point of an echo of a stationary object selected by the stationary object extracting unit 21, which is divided by the observation area dividing unit 221. A reference straight line calculation unit 222 that obtains a reference straight line parallel to the traveling track of the vehicle by linearly approximating the distribution of reflection points belonging to a part of the area, and obtains the amount of axis deviation of the radar based on the reference straight line. It was.

Description

  The present invention relates to an apparatus for detecting a target using a radar, and particularly to a technique for increasing the detection accuracy of a target by calculating an axis deviation amount of the radar.

  Conventional radar systems have been widely applied in fields such as military / defense and weather, but mass production costs and installation costs were not a major problem in these fields. However, in recent years, an attempt has been made to increase the safety in driving a vehicle by mounting a radar device on the vehicle, detecting an obstacle while the vehicle is running. In radar mounted on automobiles, it is not only a problem to keep the mass production cost low, but it is also required to mount a low-priced radar with high accuracy with a small number of man-hours.

  Particularly problematic here is the simplification and high accuracy of the adjustment of the radar axis. In-car radars are required to detect the distance and speed of a target distant from about 200 meters with a resolution of about 1 meter, but even if the radar axis is only 0.5 degrees away from the original forward direction, The detection error is 200 meters × tan 0.5 ° = 1.7 meters, and the required resolution cannot be achieved.

  Furthermore, even if the radar apparatus can be installed with high accuracy before shipment from the factory, the installation of the radar apparatus may be distorted through the use of the automobile. For example, it is almost impossible to prevent radar axis errors caused by long-term vibrations caused by running on rough roads or deformation of the vehicle body due to minor accidents. Therefore, it is necessary to incorporate a process for automatically detecting and correcting the axis deviation while using the radar apparatus.

  Conventionally, as this kind of axis deviation estimation method, it is assumed that the vehicle on which the radar is mounted is traveling on a straight road, and a stationary object such as a signboard or a guardrail existing on the road side in the echo detected by the radar is used. As a method in which many echoes are distributed, a method is known in which the echoes distributed in parallel to the road are linearly approximated, and the deviation of the radar axis is estimated based on the direction of the straight line (for example, Patent Document 1 and Non-Patent Documents). 1).

JP-A-7-120555

W. Kederer, J. Detlefsen, Sensor-based determination of angular misalignment and lane configuration of a radar sensor for ACC-applications, Proceedings of 30th European Microwave Conference, pp.313-316, 2000.

  The conventional technology described above can estimate the amount of radar axis deviation well when many stationary object echoes are distributed on a single line and the approximation accuracy of the line is high. However, since obstacles observed on an actual road have various shapes, it is well known that the echo distribution cannot be approximated by such a single straight line alone. Under such circumstances, it is difficult to accurately estimate the amount of radar axis deviation even if the spatial distribution of the stationary object echo is approximated by a single straight line.

  The present invention has been made to solve the above-described problems, and it is an object of the present invention to make it possible to accurately estimate the amount of misalignment of the radar axis and to improve the measurement performance of the radar.

In order to solve such a problem, the axial deviation estimation method according to the present invention,
An axis deviation amount estimation method for obtaining an axis deviation amount of a radar that detects a target in front of a traveling vehicle,
An echo selection step of selecting a stationary object echo to which a reflection point belongs to a predetermined partial region of the radar coverage from the echo of the radar transmission wave;
A reference straight line calculating step for calculating a reference straight line parallel to the traveling track of the vehicle by linearly approximating the distribution of reflection points of the stationary object echo selected in the echo selection step;
An axis deviation amount estimation step for obtaining an axis deviation amount of the radar based on the direction of the reference line calculated in the reference line calculation step;
It is supposed to have.

  As described above, in the axial deviation amount estimation method according to the present invention, the linear approximation is not performed on the echo of the stationary object existing in the entire radar coverage, but the stationary motion existing in a partial area of the coverage. Since linear approximation is performed only for object echoes, it is possible to increase the accuracy of linear approximation by rejecting stationary object echoes that degrade the approximation accuracy, and to accurately calculate the amount of radar axis misalignment. Become.

1 is a block diagram showing a configuration of a radar apparatus according to Embodiment 1 of the present invention. 1 is a block diagram showing a configuration of a radar apparatus according to Embodiment 1 of the present invention. 1 is a block diagram showing a configuration of a radar apparatus according to Embodiment 1 of the present invention. Explanatory drawing which shows the structure of the observation area | region of Embodiment 1 of this invention. The flowchart of the axial deviation | shift amount estimation process of Embodiment 1 of this invention. Explanatory drawing which shows the operation principle of Embodiment 1 of this invention. The block diagram which shows the structure of Embodiment 2 of this invention.

Explanation of symbols

4 Echo detection unit,
5 position detector,
6 Speed detector
7 Axis deviation estimation part,
8 Position correction unit,
21 stationary object extraction unit,
22 Straight line approximation,
24 orbital curvature acquisition unit,
25 reference straight line storage unit,
26 orbital curvature storage unit,
221 Observation area division,
222 Reference straight line calculation unit.

Embodiment 1 FIG.
The radar apparatus according to Embodiment 1 of the present invention is assumed to be mounted on, for example, an automobile, radiates a radar transmission wave in the traveling direction of the automobile (direction perpendicular to the axle), and targets (many of which are It is an obstacle in driving). Here, the radar apparatus according to the first embodiment of the present invention has a function of correcting a deviation between the reference azimuth in detecting the position of the target and the front direction of the automobile, that is, a radar axis deviation. .

  First, the principle of the method of correcting the axis deviation angle of the radar used in the first embodiment of the present invention will be roughly described as follows. That is, many objects having a plane parallel to the traveling track of the automobile, such as guardrails, sidewalks, and median strips, are seen on the road. Therefore, the reflection points on the object parallel to the traveling track of the automobile and the reflection points of the radar wave (echo reflection points) estimated to be extracted are extracted, and a straight line is approximated to the distribution of these reflection points.

  Assuming that the straight line thus obtained is parallel to the traveling track of the vehicle, the difference between the traveling track of the vehicle and the reference direction of the radar itself is obtained. In this way, even if a deviation (unknown amount) occurs in the reference direction of the radar (the direction of the radar axis), the amount of deviation is calculated appropriately. By calculating the position and orientation of the reflector in consideration of the amount of axis deviation, the accuracy of the position, orientation, speed, etc. of the reflector is improved.

  FIG. 1 is a block diagram showing a configuration of a radar apparatus according to Embodiment 1 of the present invention. As shown in the figure, this radar apparatus radiates a radar wave from the antenna 1 in the traveling direction of the automobile and receives a radar wave reflected by a radar wave reflector such as a target or an obstacle. The transmission / reception unit 2 supplies a transmission signal 11 that is a source of radar waves radiated from the antenna 1 to the outside. At the same time, when the antenna 1 receives the radar wave reflected by the radar wave reflector, the transmitting / receiving unit 2 converts the received signal 12 in the RF (Radio Frequency) band output from the antenna 1 with the internal signal of the intermediate frequency. The difference signal is used to down-convert the received signal in the video band. Down-conversion is realized by outputting the frequency difference between the received signal 12 and the reference signal which is the internal signal of the transmission / reception unit 2. Hereinafter, the received signal in the video band is referred to as a difference signal. As a result, the transmission / reception unit 2 outputs a difference signal 13. The difference signal 13 is input to the signal processing unit 3 and used for detecting the position of the object.

  The echo detector 4 detects a reflected wave using the difference signal 13 and outputs a frequency 14 of the detected difference signal of the reflected echo. As a general detection method of such a reflected echo, a method of assuming that there is a reflected wave when power is larger than unnecessary wave components such as receiver noise and clutter is usually used. For this purpose, a certain margin is added to the receiver noise or the power of the clutter, and a signal component having a power exceeding the margin is regarded as a reflected echo and detected.

  The position detector 5 obtains the position 15 of the reflection point (echo) for each predetermined period using the frequency 14 of the difference signal output from the echo detector 4. For this purpose, first, the relative distance between the reflection point and the radar apparatus and the direction (angle) of the reflection point are obtained, and the position of the reflector is determined from the obtained relative distance and direction.

  The distance measurement method can be obtained by pulse-modulating the transmission wave and measuring the distance based on the transmission / reception time difference, or by transmitting a frequency-modulated continuous wave and mixing the transmission signal and the reception signal. An FMCW (Frequency Modulation Continuous Wave) method or a multi-frequency CW method that obtains a distance from the frequency of a difference signal is known. In addition, as a method of obtaining the direction, there is a method of mechanically driving and observing an antenna having a pencil beam, or by arranging a plurality of element antennas and performing phase control at each element during transmission or reception. A method of performing beam scanning is also known. If these known methods are used, those skilled in the art can easily configure the position detector 5.

  In the following description, it is assumed that the position detection unit 5 expresses and outputs the position 15 of the reflection point in a Cartesian coordinate system composed of an x coordinate and a y coordinate. Here, the y-axis which is the coordinate axis of the y-coordinate is set so that the forward direction of the radar axis direction is a positive direction and the opposite direction is a negative direction. In addition, the x-axis and the y-axis, which are coordinate axes of the x-coordinate, are assumed to be orthogonal, and the x-axis is set with the right direction toward the front in the radar axis direction as the positive direction and the left direction as the negative direction.

  On the other hand, the speed detection unit 6 calculates the relative speed of the reflection point using the frequency 14 of the difference signal. It is well known that the relative velocity of the reflection point is calculated using Doppler modulation of the received wave frequency. When the position detecting unit 5 uses the FMCW method or the multi-frequency CW method, the distance and the speed can be calculated at the same time, so that the position detecting unit 5 and the speed detecting unit 6 are configured by the same part or element. You may do it.

  The axis deviation amount estimation unit 7 extracts reflection points estimated to be distributed on a straight line parallel to the traveling track of the automobile from the reflection points detected by the echo detection unit 4 and extracts the extracted reflections. An axis deviation amount is calculated based on the distribution of points. FIG. 2 is a block diagram illustrating a detailed configuration of the axis deviation amount estimation unit 7. In the figure, when the axial deviation amount estimation unit 7 obtains the reflection point position 15 and the reflection point relative velocity 16 from the outside, the reflection point position 15 and the reflection point relative velocity 16 are input to the stationary object extraction unit 21. To do.

  The stationary object extraction unit 21 selects an echo at a reflection point on the stationary object (hereinafter referred to as a stationary object echo) based on the position 15 or the relative velocity 16 of the reflection point, and outputs it as a stationary object echo 31. The contents of the stationary object echo 31 include at least the position of a reflection point on the stationary object (Cartesian coordinates or coordinates obtained by converting Cartesian coordinates to a polar coordinate system, etc., and equivalent to Cartesian coordinates). . As a method for determining whether or not the reflection point to which the stationary object extraction unit 21 is input is a reflection point on the stationary object, for example, there are several methods as shown in the following method 1 to method 3.

  (Method 1) A method for discriminating whether or not the reflection point is a reflection point on a stationary object based on the relative speed of the reflection point. As a method of using the relative speed of the reflection point, first, vehicle speed data is acquired from the outside and collated with the relative speed of the reflection point to identify whether the reflection point is a reflection point on a stationary object (Method 1). -1) may be considered. More specifically, for example, vehicle speed data is obtained by a vehicle speed sensor attached to the host vehicle, and a reflector having a Doppler speed that is approximately the same size as the vehicle speed and that approaches is regarded as a stationary object.

  Similarly, as a method of using the relative speed of the reflection point, a method of discriminating whether or not the reflection point is a reflection point on a stationary object from only the relative speed of the reflection point without taking in information or signals such as vehicle speed from the outside. There is also (Method 1-2). This method is based on the fact that there are many stationary object echoes in the radar observation area and the relative velocities are substantially the same for a plurality of stationary object echoes. That is, this is a method in which a plurality of reflection points are classified with respect to relative velocities, and reflection points where the distribution of relative velocities is concentrated are regarded as stationary object echoes.

  (Method 2) This is a method in which an object (reflection point) existing at a position off the lane is regarded as a stationary object. In order to determine whether or not the reflection point is out of the lane, a predetermined reference (reference x coordinate) is set for the x coordinate, and the difference between the x coordinate of each reflection point and the reference x coordinate becomes a certain level or more. In this case, the reflection point may be regarded as deviating from the lane.

  In this case, the reference x coordinate is calculated by statistically processing the x coordinate of the reflection point obtained so far. For example, an average of x coordinates of a plurality of reflection points input immediately before is obtained, and this average value is set as a reference x coordinate. Furthermore, it goes without saying that the reference x coordinate can be determined based on the radar axis deviation finally obtained by the radar apparatus according to the first embodiment of the present invention.

  According to this method 2, since it is possible to determine whether or not the reflection point is a point on a stationary object only by the x coordinate of the reflection point, the reflection point speed 16 is set to the axis deviation amount estimation unit 7. No need to enter. Therefore, if it is sufficient to obtain only the purpose of obtaining the radar axis deviation amount or correcting the axis deviation amount to improve the reflection point position calculation accuracy, the speed detection unit 6 is configured as a radar device. It may be omitted from the element.

  In the method 1-1, since it is necessary to acquire a signal and information from an external speed sensor, it is necessary to connect between the radar apparatus and the speed sensor of the moving platform (vehicle or the like) so that the speed data can be transmitted. is there. However, there is a feature that whether each reflection point is a stationary object echo can be identified independently for each reflection point.

  On the other hand, both the method 1-2 and the method 2 do not require information from the external speed sensor, and the method 2 has a feature that it is not necessary to calculate the speed of the reflection point. However, it is necessary to have a configuration in which the position information of a plurality of reflection points can be accessed simultaneously. That is, it is necessary to store position information of a plurality of reflection points in a storage device (not shown).

  In addition, since these methods do not contradict each other in principle, there is no problem in using them in combination. For example, as in method 2, the stationary object echo may be narrowed down based on the x coordinate, and the stationary object echo may be narrowed down again using the method 1-1 or method 1-2.

Based on the stationary object echo 31 thus calculated, the straight line approximation unit 22 obtains a straight line that approximates the distribution of the stationary object echo. Then, the approximate straight line slope and intercept are output as a straight line 32. If a straight line approximating the distribution of stationary object echoes is expressed as in equation (1), the slope is given as p, and the intercept (x intercept: the x coordinate of the intersection of the straight line and the x axis) is given as q.

  Hereinafter, a method in which the straight line approximation unit 22 approximates a straight line with respect to the distribution of the stationary object echo will be described in detail. FIG. 3 is a detailed block diagram of the straight line approximation unit 22. The stationary object echo 31 in the figure is input to the observation area dividing unit 221. In the observation area dividing unit 221, an observation area divided based on a predetermined boundary line is set in advance, and an observation area to which a stationary object echo used for linear approximation belongs is determined.

  FIG. 4 is a diagram showing the state of such an observation area (covering area of the radar device). This figure is an example in which the boundary line is set so as to be parallel to the traveling track of the automobile 41, and the area A on the left side of the boundary line and the area B on the right side of the boundary line in the traveling direction. The x (cross mark or alphabet X) in the figure indicates the position of the stationary object echo. The stationary object echoes in the region A are distributed almost along the guardrail 42. Further, the stationary object echoes in the region B are distributed substantially along the central separation band 43.

  In addition, the expression “the boundary line is parallel to the traveling track of the automobile 41” used here means that the angle formed by the traveling track and the boundary line is “approximately 0 °”. It does not require that boundaries do not intersect. That is, the traveling track and the boundary line may exist on the same straight line or may have an intersection.

  The observation area dividing unit 221 selects only a stationary object echo in one of the observation areas of the area A and the area B and outputs it as an in-observation area stationary object echo 32 (hereinafter simply referred to as a stationary object echo 32). At the same time, the stationary object echo in the other observation area not selected is rejected. For example, only the stationary object echo in the region A is selected and output as the stationary object echo 32, and the stationary object echo in the region B is rejected. Alternatively, it goes without saying that the stationary object echo in the region B may be selected and output as the stationary object echo 32 and the stationary object echo in the region A may be rejected.

  Note that the data of the stationary object echo 32 includes the coordinates of the position of the reflection point (Cartesian coordinates, or coordinates obtained by converting Cartesian coordinates to a polar coordinate system, etc. equivalent to the Cartesian coordinates, etc. ) Is included.

  In this way, it is possible to obtain a straight line that approximates the distribution of stationary object echoes belonging to only one of the regions A and B. In general, on roads and vehicle tracks, stationary objects are distributed on both sides of the vehicle track. If a stationary object echo is extracted in such a situation, the stationary object echo is scattered left and right in the traveling direction. That is, a straight line that simultaneously approximates both the stationary object echo on the guardrail 42 and the stationary object echo on the central separation band 43 must be obtained.

  As a result, the straight line obtained as the approximate straight line is greatly deviated from the track of the vehicle and cannot be used as a reference for obtaining the amount of radar axis deviation. To solve this problem, if the observation area dividing unit 221 is used to divide the area in advance, and linear approximation is performed by using only the stationary object echo belonging to the predetermined area, the accuracy of the linear approximation is greatly improved. .

  In order to determine whether the stationary object echo belongs to the region A or the region B, it is sufficient to compare the x coordinate of the stationary object echo with the x coordinate of the boundary line. In addition, it is desirable that the boundary line be set in parallel to the traveling track (traveling direction) of the automobile 41. However, in the case of the example of FIG. 4, it is sufficient to reject either the stationary object echo on the guardrail 42 or the stationary object echo on the central separation band 43. Therefore, the boundary line may be a straight line that forms substantially 0 ° with the traveling direction of the automobile. In addition, normally, a certain degree of direction adjustment should be performed when the radar device is attached to the moving platform (vehicle), and even the direction accuracy of the boundary line before correcting the axis deviation functions sufficiently. Therefore, a very precise accuracy is not required for setting the boundary line.

In addition to dividing the observed area in the traveling direction lateral, x-coordinate x _j stationary object echo (i = 1, ..., N , N is a natural number) calculates an average value x _ave of the average value x _ave The x coordinate of the boundary line may be used. Also, the absolute value | x _j −x _ave | of the difference between the x-coordinate x _{ave of the} boundary line and the x-coordinate x _j of the stationary object echo is calculated, and only the stationary object whose absolute value is smaller than a preset value. The stationary object echo may be selected. In this case, the observation area is divided by the set value to be compared with the absolute value of the difference and the x coordinate of the boundary line.

  In addition to the roadside, there is a possibility that a reflective object such as a traffic sign may be present on the upper part of the road in front of the traveling direction. In order to extract only the roadside data, the x coordinate of the stationary object may be limited and extracted. On the other hand, for example, since a wall surface located parallel to the vehicle track exists above the road in the tunnel, the observation area may be divided so as to select a stationary object echo on the upper wall surface.

  Next, the stationary object echo 32 is input to the reference line calculation unit 222 in FIG. 3, and the reference line calculation unit 222 obtains a straight line that approximates the distribution of the stationary object echo 32. FIG. 5 is a flowchart of the process of the reference straight line calculation unit 222. The reference straight line calculation unit 222 follows this flowchart so as to be synchronized with a predetermined cycle, more specifically, the timing at which the echo detection unit 4 outputs the frequency 14 of the difference signal or the timing at which the position detection unit 5 outputs the reflection point echo. The indicated processing is executed, and the amount of axis deviation is calculated individually in accordance with each timing.

In step ST101 in FIG. 5, first, the number of repetitions is set to zero. This number of repetitions is a counter for measuring the number of straight line approximation processes. Thereafter, the reference straight line calculation unit 222 approximates a straight line (step ST102). Assume that there are N stationary object echoes 32 (N is a natural number), and the position of the i-th stationary object echo among the N stationary object echoes is (x _i , y _i ) (i = 1,..., N ). On the other hand, the approximated straight line is given by equation (1). Here, the slope p and the x-intercept q are calculated by the equations (2) and (3) by the least square method.

After obtaining the slope p and the x-intercept q, the reference straight line calculation unit 222 evaluates the approximate accuracy of the straight line obtained as an approximate straight line, and if the approximate accuracy is inferior to a predetermined one, the stationary object echo Some of them are rejected, and the process shifts again to obtaining an approximate straight line. Specifically, first, the distance between the stationary object echo and the approximate line is calculated (step ST103). The distance d _i between the stationary object echo (x _i , y _i ) and the straight line x = py + q is calculated by equation (4).

Next, the stationary object echo in which d _i calculated through the equation (4) is a predetermined value or more is rejected from the set of stationary object echoes 32 (step ST104). By doing so, it is possible to efficiently remove the stationary object echo that is a factor that degrades the accuracy of the linear approximation from the stationary object echo to be approximated.

  FIG. 6 is a diagram for explaining that the approximation accuracy of the straight line, and thus the accuracy of estimating the amount of axis deviation is improved by step ST104. In the figure, a set 51 of stationary object echoes exists along the straight line 52 on the left front side of the automobile 41. A set 53 of stationary object echoes exists along a straight line 54 different from the straight line 52. If the stationary object echo sets 51 and 53 are approximated by the same straight line in step ST102, an approximate straight line such as a straight line 55 is obtained.

  Therefore, the distance between the straight line 55 and each stationary object echo belonging to the stationary object echo sets 51 and 53 is calculated in step ST103, and the stationary object echo belonging to the stationary object echo set 53 is rejected in step ST104. It can be expected that the echo set 51 is approximated to a straight line that is close to the original straight line 52 along which it is located.

  As a result, when the axis deviation amount is obtained based on the straight line 55, the angle 56 is obtained, and the straight line 52 is used as a reference straight line for calculating the axis deviation amount, and the axis deviation amount is obtained based on the reference straight line. Therefore, the angle 57 is obtained, and it can be seen that the estimation accuracy of the axis deviation amount is improved.

  Here, when the number of rejected stationary object echoes is 0, or when the number of stationary object echoes remaining in the set of stationary object echoes 32 is less than a predetermined number (step ST105: Yes). The estimated axis deviation amount is set as a missing value (step ST106). Conversely, if any stationary object echo is rejected in step ST104 and the number of stationary object echoes still remaining in the set of stationary object echoes 32 is greater than or equal to a predetermined number (step ST105: No). Then, a straight line that approximates the distribution of the remaining stationary object echo is obtained again (step ST107). The calculation performed here is the same as in step ST103.

  Thereafter, 1 is added to the number of repetitions (step ST108), and the distance between the stationary object echo and the straight line is calculated by equation (4) (step ST110). After calculating the distance, the reference straight line calculation unit 222 further calculates the average of the calculated distance, and tests whether the average value exceeds a predetermined value (step ST110). By such a test, it is possible to evaluate whether or not the approximation accuracy of the straight line has reached a sufficient level.

  Here, when the average value exceeds the predetermined value (step ST110: Yes), it is determined that the approximation accuracy is not yet sufficient, so the approximation process is performed again. However, if the number of repetitions exceeds a predetermined value, no further approximation processing is performed, and the estimated axis deviation with respect to the stationary object echo 32 is set as a missing value (step ST106). In the case where the radar device is mounted on an automobile, the response performance deteriorates if it takes too much time to calculate the amount of axis deviation estimation. Then, when the radar apparatus is applied to a collision avoidance / prevention system or the like, there may be a serious problem in safety during traveling. Therefore, response performance is not deteriorated by controlling so that the number of times of the linear approximation process does not exceed a predetermined number.

  On the other hand, if the number of repetitions does not exceed the predetermined number, the linear approximation process is performed again. For this purpose, first, a stationary object echo rejection process is performed based on the distance between the stationary object echo and the straight line calculated in step ST109 (step ST104). Thereafter, the straight line approximation process in steps ST105 to ST111 is sequentially repeated.

  As described above, the radar apparatus according to the first embodiment of the present invention not only obtains a straight line approximating the distribution of stationary object echoes, but also evaluates whether the obtained straight line conforms to the distribution of each stationary object echo. That is, the degree of coincidence and convergence of the stationary object echo distribution with respect to the approximate straight line are evaluated. When it is determined that the distribution of stationary object echoes is not sufficiently approximated, after rejecting stationary object echoes that cause approximate accuracy degradation, a straight line that approximates the distribution of stationary object echoes again is used. Ask. As a result, the accuracy is greatly improved as compared with the estimation of the amount of axis deviation by simple linear approximation.

  On the other hand, if the average value does not exceed the predetermined value in step ST110 (step ST110: No), it is determined that the approximation accuracy of the straight line is already sufficiently high, and thus the obtained approximate straight line is output as the axis deviation amount 17. (Step ST111), the process ends.

When outputting the axis deviation amount 17, the reference straight line calculation unit 222 may output the obtained inclination and x-intercept of the reference line as the axis deviation amount 17 as it is, or in an expression format adapted to the requirements of the system. You may make it convert into the amount 17 of axis deviations. For example, if an angle expression (radian value or the like) is appropriate as the amount of axis deviation, the angle α formed by the straight line 32 and the radar axis is calculated as the axis deviation amount 17 of the radar axis by Expression (5).

  Finally, the position correction unit 8 corrects the position 15 of the reflection point output from the position detection unit 5 using the axis deviation amount 17 and outputs the corrected position 18 to the outside of the radar apparatus.

  Thus, according to the radar apparatus of Embodiment 1 of the present invention, a straight line that approximates the distribution of reflection points is obtained in order to estimate the amount of radar axis deviation. Since the division is performed based on the line and only the reflection points included in a part of the observation area are used, the factor that deteriorates the accuracy of the linear approximation is removed, and the accuracy of the linear approximation is greatly improved.

  Further, by removing a reflection point that is far from the straight line in the linear approximation, the factor that degrades the accuracy of the linear approximation is further removed, and the accuracy of the linear approximation is greatly improved.

Embodiment 2. FIG.
The radar apparatus according to the first embodiment obtains a straight line that approximates the distribution of stationary object echoes on the premise that the vehicle as a moving platform is moving in a straight line and the trajectory is also a straight line, and calculates the slope of the straight line. It was used to estimate the amount of radar axis misalignment.

  However, when the radar apparatus is mounted on an automobile, the ratio of the time when the steering angle is not 0 to the total operation time cannot be ignored. Further, the track (road) of the vehicle is not always a straight line. In such a case, a correct result cannot be obtained by simply obtaining a straight line approximating the distribution of stationary object echoes.

  Further, the radar apparatus according to the first embodiment generates a loss of the axis deviation amount when the desired approximation accuracy cannot be obtained, or when the number of approximation processes exceeds a predetermined number. The handling method for the loss of deviation is not clarified.

  Therefore, in the second embodiment of the present invention, a step of smoothing the shaft misalignment amount in the shaft misalignment amount estimation processing of the radar apparatus according to the first embodiment is further provided, so that the variation in the track shape of the vehicle or the steering angle of the vehicle It will be described that a method for estimating the amount of misalignment that can cope with fluctuations and loss of the amount of misalignment is obtained.

  The radar apparatus according to the second embodiment of the present invention is an improvement of the axis deviation estimation unit 7 of the radar apparatus according to the first embodiment. Therefore, unless otherwise specified, the configuration of the radar apparatus according to the second embodiment is the same as that of the radar apparatus according to the first embodiment.

  FIG. 7 is a block diagram showing the configuration of the straight line approximation unit 22 which is a characteristic part of the radar apparatus according to the second embodiment of the present invention. In the figure,

  The track curvature acquisition unit 24 is a part that acquires the track curvature 34 at the position where the vehicle is currently traveling. As a specific method of acquiring the curvature 34, for example, a measurement value of a yaw rate sensor that acquires an angular velocity of a running vehicle may be used, or steering angle information from a vehicle steering device is converted into a curvature. You may make it do.

  The reference straight line storage unit 25 is a part that stores the axis deviation amount 35 obtained for each observation period having a predetermined length for a plurality of periods. The track curvature storage unit 26 is a part that stores the curvature of the vehicle track acquired from the track curvature acquisition unit 24 by the reference line calculation unit 222 at the timing when the reference line is calculated.

  In the following description, a time zone in which one axis deviation amount 33 is calculated is referred to as a section, and each section is identified by a number such as section 1, section 2,. The sections in which the numbers assigned to the sections are continuous are adjacent to each other. Further, the axis deviation amount 33 acquired in a certain section N (N is a natural number) is represented as an axis deviation amount 33 (N), and the curvature 34 is represented as a curvature 34 (N).

  Next, the process of the reference straight line calculation unit 222 will be described. The reference straight line calculation unit 222 acquires the curvature 34 (N) from the trajectory curvature acquisition unit 24. Further, the reference straight line calculation unit 222 acquires the curvature 36 from the trajectory curvature storage unit 26. This curvature 36 is the curvature 34 (k) (k <N) that the reference straight line calculation unit 222 previously acquired from the orbital curvature acquisition unit 24. Thereafter, the reference straight line calculation unit 222 stores the curvature 34 (N) in the trajectory curvature storage unit 26. As a result, the curvature stored in the trajectory curvature storage unit 26 is replaced from the curvature 36 to the curvature 34 (N), and the curvature 34 (N) is used as the curvature 36 in the next processing.

  Next, the reference straight line calculation unit 222 examines whether the curvature 36 (curvature 34 (k), k <N) and the curvature 34 (N) are both equal to or smaller than a predetermined value. As a result, when either the curvature 36 or the curvature 34 (N) exceeds a predetermined value, the reference line calculation unit 222 replaces the process of calculating the reference line from the current stationary object echo 32 and the reference line storage unit 25. The reference straight line 35 stored in FIG. For example, if the reference line storage unit 25 has stored the reference lines of the last few times, the current reference line is calculated using the average value of these reference lines (the average value of the slope and the average value of the x-intercept). calculate.

  Further, when both the curvature 36 and the curvature 34 (N) are equal to or less than a predetermined value, it is determined that the automobile is in a straight traveling state. Then, in the same manner as in the first embodiment, the reference straight line is obtained and the estimated axis deviation amount is calculated. When the estimated axis deviation amount is correctly calculated, the reference straight line is stored in the reference line storage unit 25, and the calculated axis deviation amount is output.

  In addition, as a result of calculating the amount of axial deviation by the method of the first embodiment, when the estimated amount of axial deviation is a missing value, when either the curvature 36 or the curvature 34 (N) exceeds a predetermined value, Similarly, the current reference line is calculated using the reference line 35 stored in the reference line storage unit 25.

  As described above, according to the radar apparatus of the second embodiment of the present invention, when the curvature of the trajectory is acquired from the trajectory curvature acquisition unit 24 and the calculated axis deviation amount is determined to be invalid, The substitute value is calculated and output based on the effective axis deviation amount. For this reason, it is possible to prevent the position from being corrected based on the axis deviation amount calculated in the curved trajectory, and to improve the reliability of the axis deviation amount estimation process.

  In addition, even if a loss of axis deviation occurs due to the fact that linear approximation cannot be performed with sufficient accuracy, or linear approximation cannot be performed within the desired response time, Since the recovery process is performed using a highly reliable axis deviation amount, the robustness of the axis deviation amount estimation process can be enhanced, and the safety of the collision prediction / avoidance system, which is the main application of such a radar device, is further improved. Contributes to improving sex.

The present invention is useful for improving the accuracy of radar technology and can be applied particularly to a radar device mounted on an automobile.

Claims (10)

An axis deviation amount estimation method for obtaining an axis deviation amount of a radar that detects a target in front of a traveling vehicle,
An echo selection step of selecting a stationary object echo to which a reflection point belongs to a predetermined partial region of the radar coverage from the echo of the radar transmission wave;
A reference straight line calculating step for calculating a reference straight line parallel to the traveling track of the vehicle by linearly approximating the distribution of reflection points of the stationary object echo selected in the echo selection step;
An axis deviation amount estimation step for obtaining an axis deviation amount of the radar based on the direction of the reference line calculated in the reference line calculation step;
A method of estimating an axis deviation amount, comprising: In the axial deviation amount estimation method according to claim 1,
The echo selection step is characterized by selecting a stationary object echo to which a reflection point belongs to a predetermined partial region of the radar coverage divided into a boundary line where the angle formed with the traveling track of the vehicle is substantially 0 °. To estimate the amount of misalignment. In the axial deviation amount estimation method according to claim 2,
The echo selection step selects a stationary object echo belonging to any one of the covered areas divided right and left by a boundary line in which the angle formed with the traveling trajectory of the vehicle is approximately 0 °. Estimation method. In the axial deviation amount estimation method according to claim 1,
The reference straight line calculation step obtains a first straight line that approximates the distribution of the reflection points of the stationary object echo selected in the echo selection step, and calculates the distribution of the reflection points whose distance from the first straight line is a predetermined value or less. A method for estimating an axis deviation amount, wherein the second straight line to be approximated is calculated as a reference straight line. A radar for detecting a target in front of a traveling vehicle, receiving an echo of a transmission wave, and detecting a distance and a speed of a reflection point of the received echo and estimating an axis deviation amount of the radar In
A stationary object extraction unit that selects a stationary object echo from the echo of the transmission wave based on the distance and speed;
An observation area dividing unit for dividing the radar coverage into a plurality of areas;
The reflection point of the echo of the stationary object selected by the stationary object extraction unit, and the distribution of the reflection point belonging to a part of the region divided by the observation area dividing unit is linearly approximated to the traveling track of the vehicle. A reference straight line calculation unit for obtaining a parallel reference straight line, and obtaining an axis deviation amount of the radar based on the reference straight line;
An axis deviation amount estimation device comprising: In the axial deviation amount estimation apparatus according to claim 5,
The observation area dividing unit divides the radar coverage into a plurality of areas by a boundary line where an angle formed with a traveling track of a vehicle on which the radar is mounted is approximately 0 °. In the axis deviation estimation device according to claim 6,
The observation area dividing unit divides the radar coverage into left and right with the radar axis as a boundary line. In the axial deviation amount estimation apparatus according to claim 5,
The reference straight line calculation unit is a first approximation by linearly approximating the distribution of reflection points of the echoes of the stationary object selected by the stationary object extraction unit and belonging to a part of the region divided by the observation region dividing unit. And calculating a second straight line approximating the distribution of reflection points of echoes of a stationary object whose distance from the first straight line is within a predetermined value as a reference straight line. apparatus. In the axial deviation amount estimation apparatus according to claim 5,
A reference straight line storage unit for storing the reference straight line;
The reference straight line calculation unit is a first approximation by linearly approximating the distribution of reflection points of the echoes of the stationary object selected by the stationary object extraction unit and belonging to a part of the region divided by the observation region dividing unit. In the case where the first straight line is determined as the reference straight line based on the distance between the first straight line and the reflection point, the first straight line is used as the reference straight line. If the first straight line is stored in the reference straight line storage unit and output as a reference straight line, and the first straight line is not used as the reference straight line, the reference straight line is calculated using the reference straight line stored in the reference straight line storage unit. ,
An apparatus for estimating an axis deviation amount. In the axial deviation amount estimation apparatus according to claim 5,
A trajectory curvature acquisition unit for acquiring the curvature of the trajectory of the vehicle;
A track curvature storage unit for storing the curvature of the track of the vehicle;
A reference straight line storage unit for storing the reference straight line;
The reference straight line calculation unit acquires the curvature of the track of the vehicle stored in the track curvature storage unit as the curvature of the track of the first vehicle, and acquires the curvature of the track of the second vehicle from the track curvature acquisition unit. The curvature of the trajectory of the second vehicle is stored in the trajectory curvature storage unit, and the curvature of the trajectory of the first vehicle and the curvature of the trajectory of the second vehicle are both predetermined. If the value is within the range, the reflection point of the echo of the stationary object selected by the stationary object extraction unit and belonging to a part of the area divided by the observation area dividing unit is linearly approximated and used as a reference A straight line is calculated, and the reference straight line is stored using the reference straight line stored in the reference straight line storage unit when either the curvature of the trajectory of the first vehicle or the curvature of the trajectory of the second vehicle exceeds a predetermined value. calculate,
An apparatus for estimating an axis deviation amount.

Images