close

Вход

Забыли?

вход по аккаунту

?

JP2006234523

код для вставкиСкачать
Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JP2006234523
In an object detection method using an ultrasonic sensor, the calculation accuracy of object
detection is shortened by maintaining the detection accuracy of the distance to the object and the
orientation of the object with required accuracy, and improvement of real time property of object
detection Plan. SOLUTION: A distance and an azimuth of an object are detected by using a wave
receiving element array in which ultrasonic wave receiving elements are arranged in an array.
First, transmission and reception of ultrasonic waves (# 1), digitization of reception signals (# 2),
and recording of digitized data (# 3) are performed. Thereafter, the three-dimensional space to be
subjected to object detection is divided into individual small spaces defined by distance and
orientation (# 4). A delay-and-add operation for object detection is performed for each of the
selected small spaces (# 5). The division and selection of the small space is dynamically
performed depending on conditions such as whether or not an object has been detected in
advance, which range the object is focused on, and at what resolution the object is to be detected.
In the case of recalculation or remeasurement, the number of divisions in the three-dimensional
space and the calculation conditions are changed (# 7). [Selected figure] Figure 1
Object detection method using ultrasonic sensor
[0001]
The present invention relates to an object detection method using an ultrasonic sensor.
[0002]
Conventionally, by measuring the distance to an ultrasonic reflection point and its direction using
04-05-2019
1
an ultrasonic wave transmitting element and a wave receiving element array in which a plurality
of ultrasonic wave receiving elements are arranged in one or two dimensions. Techniques for
detecting obstacles are known.
Various proposals have been made in connection with such technology. For example, there is
known a device in which wave receiving elements are arranged horizontally on the front and
back surfaces of a mobile body to detect an obstacle in a road width area (see, for example,
Patent Document 1). Furthermore, there is known a semiconductor device in which a plurality of
wave receiving elements and a circuit for delaying and adding output signals from the respective
wave receiving elements are formed into chips (see, for example, Patent Document 2). JP-A-2179491 JP-A-2002-156451
[0003]
However, in the ultrasonic technology as disclosed in Patent Document 1 described above, only
the presence of an obstacle can be detected, and a plurality of isolated obstacles can not be
individually recognized. In general, ultrasonic sensors for detecting an object including an
obstacle are wave receiving elements incorporated in an array to increase the number of
ultrasonic sensors in order to grasp surrounding objects in a wide range, or to grasp objects
accurately. There is a tendency to increase the number of These ultrasonic sensors, generally
array sensors in which receiving elements are arrayed, perform so-called electronic scanning by
an operation of delaying and adding output signals from the receiving elements for each
direction. Therefore, an increase in calculation time due to the improvement of the object
detection accuracy in the array sensor becomes a problem. In this regard, the technology
disclosed in Patent Document 4 merely indicates that a plurality of wave receiving elements are
integrated to form an ultrasonic sensor composed of the wave transmitting elements and the
wave receiving element array in a small size and at low cost. The problem of calculation time
arises, for example, when a mobile body installs an ultrasonic sensor and moves while detecting
an obstacle in the moving direction. In order for a mobile to move efficiently, obstacle detection
requires real-time capability, but an increase in computation time hinders it.
[0004]
The present invention solves the above-mentioned problems, and can maintain the accuracy of
detection of the distance to the object and the orientation of the object to the required accuracy
to shorten the calculation time of object detection, and improve the real time property of object
detection. An object of the present invention is to provide an object detection method using an
ultrasonic sensor that can be realized.
04-05-2019
2
[0005]
In order to achieve the above object, the invention according to claim 1 uses an ultrasonic sensor
comprising an ultrasonic wave source and a wave receiving element array in which a plurality of
wave receiving elements are arranged in an array. After the air pressure wave radiated from the
sound source is reflected on the object surface, it is received by each of the wave receiving
elements, and the time from the radiation of the air pressure wave to the wave receiving by each
wave receiving element is measured. In the object detection method for detecting the direction of
an object based on the difference in time of incidence of each air pressure wave on each
receiving element while detecting the distance to the object surface based on the determined
time, the object detection is an object It is performed for each small space defined by the distance
and direction formed by dividing the three-dimensional space to be detected, and the number of
divisions of the three-dimensional space is changed depending on the direction.
[0006]
According to the second aspect of the present invention, in the object detection method
according to the first aspect, first, the number of divisions in one of two directions which defines
the orientation of the three-dimensional space to be detected is set to one. The object detection is
performed with the number of divisions in the other direction being larger than one, and then the
number of space divisions in one direction is increased only in the small space in which the
object is detected among the small spaces in the other direction. Object detection is performed.
[0007]
According to the third aspect of the present invention, in the object detection method according
to the first aspect, first, an object is detected with a small number of space divisions using a part
of the wave receiving elements of the wave receiving elements, and then, The object detection is
performed with a large number of space divisions using all the wave receiving elements in the
direction in which the object is detected.
[0008]
According to the fourth aspect of the present invention, in the object detection method according
to the first aspect, when an object is detected, next, the number of space divisions in the direction
in which the object is detected is increased, and the space for other orientations is Object
detection is performed with the number of divisions reduced.
[0009]
04-05-2019
3
The invention according to claim 5 uses an ultrasonic sensor comprising an ultrasonic wave
source and a wave receiving element array in which a plurality of wave receiving elements are
arranged in an array, and the air pressure emitted from the ultrasonic wave source After the
wave is reflected on the object surface, the wave is received by each of the wave receiving
elements, and the time from the radiation of the air pressure wave to the wave reception by each
wave receiving element is measured, and the object surface is measured based on the measured
time. In an object detection method for detecting an azimuth of an object based on a difference
between incident times of the air pressure wave to each of the wave receiving elements while
detecting a distance up to the point, a three-dimensional space in which object detection is an
object detection object The division is performed for each small space defined by the distance
and the direction formed by dividing. The number of divisions of the three-dimensional space is
changed according to the distance.
[0010]
According to a sixth aspect of the present invention, in the object detection method according to
the fifth aspect, first, a distance to an object is detected using an output signal output from one of
the receiving elements, and then the receiving is performed. Object detection is performed in the
space before and after including the distance to the detected object using output signals output
from other wave receiving elements in the wave element array.
[0011]
According to the invention of claim 7, in the object detection method according to claim 5, next,
when an object is detected, the object detection is performed in a space closer than the distance
at which the object is detected.
[0012]
The invention according to claim 8 is the object detection method according to any one of claims
1 to 7, wherein the output signals from the respective wave receiving elements are recorded as
digital data, and based on the difference in the incident time. In the calculation for detecting the
direction of the object, the recorded digital data is used by thinning out.
[0013]
The invention according to claim 9 is the object detection method according to claim 8, wherein
the recorded digital data is selected by skipping one and used for the calculation.
[0014]
04-05-2019
4
The invention according to claim 10 is the object detection method according to any one of
claims 1 to 9, wherein object detection is performed only in a three-dimensional space consisting
of an effective field of view determined by the attachment position of the ultrasonic sensor. .
[0015]
The invention according to claim 11 relates to the object detection method according to any one
of claims 1 to 10, wherein when the ultrasonic wave source and the ultrasonic wave sensor are
attached to a moving body, the traveling speed of the moving body Accordingly, the field of view
of the ultrasonic sensor is changed, and object detection is performed in a three-dimensional
space in the field of view.
[0016]
The invention according to claim 12 is the object detection method according to any one of
claims 1 to 10, wherein the traveling direction of the moving body is the ultrasonic source and
the ultrasonic sensor are attached to the moving body. Accordingly, the field of view of the
ultrasonic sensor is changed, and object detection is performed in a three-dimensional space in
the field of view.
[0017]
According to the invention of claim 1, when forming a small space for detecting an object, the
number of divisions of the space, and hence the size of each small space, is changed according to
the azimuth, so the size of the small space of interest in the azimuth of interest The number of
small spaces can be reduced by decreasing the size of the small space and increasing the size of
the small space in the non-focused space area.
Therefore, it is possible to shorten the calculation time of object detection by maintaining the
detection accuracy of the distance to the object and the orientation of the object to the required
accuracy.
[0018]
According to the second aspect of the present invention, the space division stage relating to the
orientation is performed as two stages of coarse division and fine division, and object detection is
performed in each stage, and in the second stage, only the attention orientation narrowed down
Since the object detection is performed, the required detection accuracy can be maintained to
04-05-2019
5
shorten the calculation time of the object detection.
[0019]
According to the third aspect of the present invention, there is a small number of wave receiving
elements used, and after covering the entire space of interest with a coarse spatial accuracy,
object detection is performed with a large number of wave receiving elements and fine spatial
accuracy for important orientations. The calculation accuracy of object detection can be
shortened by maintaining the detection accuracy.
[0020]
According to the invention of claim 4, the accuracy of object detection, that is, the number of
space divisions, is changed depending on whether the object is detected or not, so that the
required detection accuracy is maintained to shorten the operation time of object detection. Can.
[0021]
According to the invention of claim 5, when forming a small space for detecting an object, the
number of divisions of the space, and hence the size of each small space, is changed according to
the distance, so the size of the small space of interest in the direction of interest The number of
small spaces can be reduced by decreasing the size of the small space and increasing the size of
the small space in the non-focused space area.
Therefore, it is possible to shorten the calculation time of object detection by maintaining the
detection accuracy of the distance to the object and the orientation of the object to the required
accuracy.
[0022]
According to the invention of claim 6, since the object detection is performed focusing on the
distance and thereafter the object detection is performed including the measurement of the
azimuth, the required detection accuracy is maintained to shorten the calculation time of the
object detection. Can.
04-05-2019
6
For example, by distance detection using one wave receiving element, highly accurate azimuth
detection is performed only when an object is detected at a distance position important for a
moving object, for example, a short distance within a predetermined distance, and the object
When the object is detected at a distance, the calculation time of the object detection can be
shortened by performing the object detection with the accuracy of the azimuth detection
lowered.
[0023]
According to the invention of claim 7, since the space area for object detection is limited to a
close space, for example, object detection in the vicinity of a moving object important for the
moving object can be efficiently performed by shortening the calculation time. It is possible to
improve the real time property of object detection.
For example, only with respect to an object moving closer to the front side, the computation time
of object detection can be shortened by detecting the direction and position of the object with
higher accuracy.
[0024]
According to the invention of claim 8, the digital data can be thinned out and used by changing
the thinning degree of digital data as necessary, so that the object detection can be performed
efficiently by shortening the calculation time appropriately. it can.
[0025]
According to the invention of claim 9, the object detection can be performed in approximately
half of the calculation time.
[0026]
According to the tenth aspect of the present invention, it is possible to perform efficient object
detection without using unnecessary calculation time.
[0027]
According to the invention of claim 11, for example, when moving the moving object at high
speed, the field of view is concentrated in the traveling direction to perform object detection
04-05-2019
7
satisfying the real time property at high speed in a narrow space. Object detection in line with
reality.
[0028]
According to the invention of claim 12, safe and efficient movement by the moving object can be
performed by changing the field of view in accordance with the change of course of the moving
object.
[0029]
Hereinafter, an object detection method using an ultrasonic sensor according to an embodiment
of the present invention will be described with reference to the drawings.
FIG. 1 shows the procedure of object detection in the object detection method according to the
present invention.
In this object detection method, a so-called electronic scan method is used in which a distance
and an orientation are detected using a wave receiving element array in which a plurality of wave
receiving elements are arranged in an array.
In this object detection method, the time from radiation of an air pressure wave (hereinafter
referred to as ultrasonic wave) to reception by each wave receiving element is measured, and the
distance to the object surface is detected based on the measured time. A delay and add operation
is performed to detect the direction of the object based on the difference in the time of incidence
of the sound wave on each wave receiving element.
The electronic scan and the delay addition operation will be described later.
[0030]
In the present object detection method, first, transmission of ultrasonic waves by the ultrasonic
sensor and reception of reflected ultrasonic waves are performed (# 1).
04-05-2019
8
The wave reception is performed by each wave receiving element in the wave receiving element
array.
Next, digitization of the wave receiving signal (output signal) from each wave receiving element is
performed (# 2), and the digitized digital data is recorded in the memory (# 3).
Thereafter, the three-dimensional space to be subjected to object detection is divided into
individual small spaces defined by distance and orientation (# 4).
Object detection is performed for each small space selected from the divided individual small
spaces.
The division regarding the orientation is performed by, for example, angle division in which a
predetermined range of the viewing angle of the ultrasonic sensor is divided in the horizontal
direction and the vertical direction.
The division of the distance is performed for a predetermined distance range in front of the
ultrasonic sensor.
The division of the three-dimensional space into small spaces and control of each procedure of
object detection are performed by an object detection control unit 34 in FIG. 2 described later.
[0031]
After the above-described small space is determined, an object detection operation is performed
for each selected small space (# 5).
The object detection operation is a so-called delayed addition operation in which each digital data
of each wave receiving element stored in the memory is subjected to a predetermined delay and
added to each other.
04-05-2019
9
The division and selection of the small space is dynamically performed depending on conditions
such as whether or not an object has been detected in advance, which range the object is focused
on, and at what resolution the object is to be detected.
When the calculation in step # 5 is completed, a series of electronic scans are completed.
Therefore, when the object detection is ended (YES in # 6), the object detection is ended.
If the object detection is not finished (NO in # 6), the process proceeds to step # 7.
[0032]
In step # 7, the number of divisions in the three-dimensional space and the calculation conditions
are changed.
The spatial resolution can be changed by changing the number of divisions. The calculation
conditions include, for example, conditions such as which small space to select. Next, using the
digital data of the received wave signal already recorded, recalculation is performed as
recalculation under the changed calculation condition (YES in # 8) or remeasurement is
performed. Determine if there is any (NO at # 8). This determination is determined depending on
the result of the object detection and the state in which the object detection is performed. The
above steps are repeated from step # 4 in the case of recalculation and from step # 1 in the case
of remeasurement. In step # 7, the remeasurement is repeated with the coarse division number
(low resolution) without changing the condition, and when the object is detected, the condition is
changed and recalculation or remeasurement is performed. Good.
[0033]
Next, an ultrasonic sensor used in the present invention and an object detection apparatus
incorporating the same will be described with reference to FIGS. 2, 3 and 4. FIG. FIG. 2 shows a
block configuration of the object detection apparatus. This apparatus comprises an ultrasonic
04-05-2019
10
sensor 2 and an object detection control means 34, a memory 35, and an arithmetic means 36.
The ultrasonic sensor 2 has a transmission element 21 which is a transmission source for
transmitting ultrasonic waves, and a plurality of reception elements 31. The transmission
element 21 is driven by the drive circuit 20. The wave receiving element 31 receives the
ultrasonic wave reflection wave WR in which a part of the transmission wave WS transmitted
from the wave transmission element 21 is reflected by the ultrasonic wave reflection object 22.
Alternatively, they are arrayed in two dimensions, that is, arrayed, and arranged and integrated
on a substrate to constitute the receiving element array 3. An analog type output signal received
and output by each wave receiving element 31 is amplified by the amplifier 32, converted into
digital data by the A / D conversion means 33, and directly via the subsequent object detection
control means 34 or , Is stored in the memory 35.
[0034]
The operations of the ultrasonic sensor 2 and the object detection control means 34 described
above will be described. The object detection control means 34 sends a drive signal to the drive
circuit 20 of the ultrasonic sensor 2 to cause the transmission element 21 to transmit an
ultrasonic wave. A part of the transmission wave WS transmitted from the transmission element
21 is reflected by the ultrasonic reflection object 22 in front, and the reflected wave WR is
received by the reception element 31. The object detection control means 34 causes the
calculation means 36 to perform delay addition operation on the data stored in the memory 35,
and detects the distance to the object surface based on the time from transmission to reception,
and The azimuth of the object is detected based on the difference in incident time to each wave
receiving element (described in detail later).
[0035]
The transmission element 21 will be described. FIG. 3 shows the structure of the transmission
element, and FIG. 4 shows the ultrasonic waveform transmitted by the transmission element 21.
As shown in FIG. The transmission element 21 is a thermally excited ultrasonic element that
generates an ultrasonic wave by heat generation. As shown in FIG. 3, the wave transmitting
element 21 includes a substrate 21b made of single crystal silicon or the like, a heat insulating
layer 21c made of porous nanocrystalline silicon or the like provided on the substrate 21b, and
the insulating layer 21c. And 21c having a heat generating member 21d formed of a metal thin
film such as tungsten formed on the surface 21c. The substrate 21b is mounted on the mounting
substrate 21a. By providing an energization electrode terminal 21e to the heating element 21d
and applying an alternating voltage thereto, compressional compression wave USW can be
04-05-2019
11
generated in the air according to the frequency.
[0036]
The thermally excited ultrasound source as described above emits a sound wave of a frequency
according to the frequency of the supplied power. Its frequency characteristics are flat and have
no resonance. Thereby, as shown in FIG. 4, the ultrasonic wave WS of a sinusoidal pulse
waveform can be generated. (Refer Unexamined-Japanese-Patent No. 2002-156451.). As the
wave receiving element 31, a capacitive microphone having a gap on a silicon diaphragm
(membrane part) can be used (not shown). When a pressure wave (reflected ultrasonic wave) is
received at the membrane portion, the gap interval of the gap fluctuates, and the capacitance
based on this space fluctuates. The pressure wave can be detected by taking out the fluctuation
of the capacitance as voltage fluctuation through the electrode to the outside.
[0037]
Next, with reference to FIG. 5 to FIG. 8A and FIG. 8B, the electronic scan and the delay addition
operation in the object detection apparatus incorporating the above-mentioned ultrasonic sensor
will be described. Here, first, reflection ultrasonic wave reception by a single wave receiving
element 31 will be described with reference to FIG. 2 described above. A transmission start signal
is sent from the object detection control means 34 to the drive circuit 20, and a burst sound
wave or a sound wave with a small wave number is transmitted from the transmission element
21 driven by the drive circuit 20. The object detection control means 34 starts the reception of
the reflected ultrasonic wave at the same time as the transmission start.
[0038]
The transmitted wave WS transmitted by the wave transmitting element 21 is attenuated by
distance and attenuated by absorption into air before reaching the ultrasonic reflecting object 22
and reflected, and gradually attenuates with the propagation distance. Furthermore, even when
reflecting, there is reflection attenuation. Such a reflected wave (echo wave) WR receives distance
attenuation and absorption attenuation also in the return path, and reaches the wave receiving
element 31.
04-05-2019
12
[0039]
The reflected wave WR that has reached the wave receiving element 31 has a sound pressure
attenuated by about 45 dB as compared to the sound pressure when transmitting, for example,
when the distance to the ultrasonic wave reflecting object is 1 m (reciprocal 2 m). The time t for
the sound wave to travel at the sound velocity V = 340 m / s at a distance D = 1 m is t = 2 × D /
V = 6 ms. Therefore, when detecting an object up to a distance of 1 m, the echo wave may be
sampled for 6 ms from the generation of the transmission start signal. The sound pressure
fluctuation due to the received ultrasonic wave is converted into an electric signal by the wave
receiving element 31 and amplified by the amplifier 32.
[0040]
The amplification factor of the amplifier 32 at this time is, for example, 40 dB to 60 dB. Also, the
S / N ratio is about 60 dB. The analog signal amplified by the amplifier 32 is converted into a
digital signal by, for example, 12-bit or 16-bit A / D conversion means 33, and stored in the
memory 35 via the object detection control means 34 or directly. When the sampling frequency
is 1 MHz and A / D conversion is 12 bits, the distance resolution is 1 bit = 2.4 mm, and when the
A / D conversion is 16 bits, the distance resolution is 1 bit = 0.15 mm. Processing of each signal
and data from each wave receiving element 31 to the memory 35, that is, ultrasonic wave
reception and electric signal conversion, electric signal amplification, A / D conversion, and
memory storage are independently performed for each wave receiving element 31. To be done.
[0041]
Here, as shown in FIG. 5, an example will be described in which the horizontal distances and
directions of the ultrasonic wave reflective objects (hereinafter referred to as objects) 22a and
22b distributed in the horizontal plane are recognized. The objects 22 a and 22 b are present in
the direction of the front right angle θ 1 and the left angle θ 2 of the ultrasonic sensor 2. The
reflected waves from the objects 22a and 22b are, as shown in FIG. 6, a direction n1 in which the
angle θ1 swings to the right with respect to the normal direction N of the wave receiving
elements 31 arranged in a line in the horizontal direction The light receiving element 31 is
incident on the light receiving element 31 from the direction n2 where the angle .theta. At this
time, it can be assumed that the incident reflected wave is a plane wave.
04-05-2019
13
[0042]
That is, the wave front of the group A of the reflected wave incident on each of the wave
receiving elements 31 from the object 22a is a plane, and the wave front of the group B is also a
plane. The group A will be described below. At the time when the same wave front, ie, the same
phase surface, is incident on each wave receiving element 31, a time delay depending on the
angle θ1 occurs. In other words, by performing delay processing DL on the signals output from
the respective wave receiving elements 31 and performing addition processing ADD, it is possible
to extract an ultrasonic signal incident from a specific angular direction. That is, when the wave
receiving element array 3 in which the plurality of wave receiving elements 31 are arranged is
used, the angular scan can be performed by electronic processing (electronic scan) without using
mechanical scan. The calculation means 36 performs the delay process DL and the addition
process ADD shown in FIG. 6 under the control of the object detection control means 34.
[0043]
The result of obtaining the ultrasonic signal corresponding to the object 22a by the abovedescribed electronic scan (delayed addition processing) is composed of a concentrated waveform
by group A and a dispersed waveform by group B, as shown in the upper right part of FIG. . The
waveforms concentrated as a result of the delay addition are as shown in FIGS. 8 (a) and 8 (b).
Assuming that the transmission time is the time measurement origin point t = 0, the waveform
appearance times t1 and t2 are times when ultrasonic waves reciprocate between the
transmission element and the objects 22a and 22b. That is, the distances of the objects 22a and
22b can be obtained from t1 and t2, and the directions (angles θ1 and θ2) of the objects 22a
and 22b can be obtained from the delay amount used for the delay addition process. If such a
delay-and-add operation is performed not only in the horizontal direction but also in the vertical
direction, object detection can be performed for each orientation in the field of view in front of
the ultrasonic sensor.
[0044]
In the following, in the object detection method using the ultrasonic sensor as described above
and the object detection device incorporating the same, the object detection method is performed
by maintaining the detection accuracy of the distance to the object and the orientation of the
object to the required accuracy. A method for shortening calculation time and improving the real
time property of object detection will be described. FIG. 9 shows how a three-dimensional space
04-05-2019
14
(hereinafter referred to as detection space 5) to be subjected to object detection is divided into
individual small spaces defined by distance and orientation. The Z direction indicates the forward
direction of the ultrasonic sensor, the θ direction indicates the horizontal direction, and the φ
direction indicates the vertical direction (vertical direction). The detection space 5 has, for
example, a part of a spherical shell centering on the ultrasonic sensor 2 by a viewing angle range
J in the horizontal direction θ, a viewing angle range M in the vertical direction φ, and a
predetermined distance range K in the forward direction Z It can be defined as a restricted
hexahedron space.
[0045]
The detection space 5 is a set of hexahedron small spaces consisting of parts of smaller spherical
shells by dividing the distance range K into k, dividing the angle range J into j, and dividing the
angle range M into m. In the object detection method according to the present invention, object
detection is performed for each small space thus divided. As an example of division, j = 18
division at an equal interval of 5 ° to the angular range J = 90 ° in the horizontal direction, and
m = 18 division at an equal interval of 5 ° to an angular range M = 90 ° in the vertical
direction Furthermore, when division of k = 50 division is performed at equal intervals of 100
mm with respect to the distance range K = 5 m in the forward direction, the number of small
spaces is 16,200.
[0046]
When the number of small spaces mentioned above increases, spatial resolution (angular
resolution, distance resolution) improves, but when the operation of object detection is
performed for all small spaces, the amount of delay addition operation increases and real time
property in object detection Will be lost. Such a problem is eliminated by not performing
unnecessary calculations. For example, when forming a small space for detecting an object, the
size of the small space of interest in the direction of interest or distance is reduced by changing
the number of divisions of the space according to the azimuth or distance, and hence the size of
each small space The resolution is increased, and the small space size is increased to reduce the
resolution in the unfocused space region. This makes it possible to reduce the total number of
small spaces. Moreover, the distance to the object and the distance to the object can be obtained
by changing the resolution as appropriate by performing arithmetic processing or measurement
through a plurality of steps, or limiting the small space to be calculated to a specifically selected
small space. The calculation time of the object detection can be shortened by maintaining the
detection accuracy of the azimuth to the required accuracy.
04-05-2019
15
[0047]
FIG. 10 shows an example in which the number of divisions in the detection space 5 is changed
according to the direction, and FIG. 11 shows an example in which the number of divisions in the
detection space is changed according to the distance. These detection spaces 5 show an example
of reducing the amount of calculation by changing the division width in each direction depending
on the direction or distance. In the detection space 5 shown in FIG. 10, the angle division width is
5 ° in the angle ranges J1 and M1 which are angle ranges close to the front direction, and the
angle division width is 10 ° in the other angle ranges J2 and M2. There is. Further, in the
detection space 5 shown in FIG. 11, the division width is 50 mm in the short distance range K1,
and the division width is 100 mm in the distance range K2 farther therefrom.
[0048]
As another method, an example in which the amount of operation is reduced by performing
operation processing or measurement through multiple steps will be described. First, the number
of divisions in one of two directions defining the orientation of the detection space 5 to be
detected by an object, for example, the vertical direction φ, is 1, and the number of divisions in
the other direction, ie, the horizontal direction θ is 1 Make the object detection larger. Next,
object detection is performed by increasing the number of space divisions in the vertical
direction φ only in the small space in which the object is detected in the small space in the
horizontal direction θ. In the second object detection, object detection is performed by
performing a delay addition operation using data already stored in the memory as digital data. As
described above, calculation is not performed on a small space in the vertical direction in which
no object is detected in the horizontal direction, and in the next step, object detection is
performed only on the focused direction of interest. Time can be reduced. This method
corresponds to the case of the recalculation in FIG. 1 described above (YES in step # 8). In this
case, re-measurement (NO in step # 8) can also be performed.
[0049]
In addition, in the case of performing time-sequential object detection as in the case of detecting
an object while tracking a moving object, the following is performed after the object detection is
performed by emitting ultrasonic waves from an ultrasonic source. In the detection of the object
04-05-2019
16
detection, the object detection is performed as a large division number in a narrow angle range
including the direction in which the object was detected in the previous detection, a fine angle
division width, and a coarse division number in other angle ranges. As described above, it is
possible to maintain the required detection accuracy and shorten the calculation time of the
object detection by changing the accuracy of the object detection, that is, the number of space
divisions, depending on whether the object is the detected azimuth.
[0050]
Also, as shown in FIG. 12, first, object detection is performed with a small number of space
divisions using a part of the wave receiving elements 31 of the wave receiving elements 31 in the
wave receiving element array 3, and then the object The computation time can also be shortened
by performing object detection with a large number of space divisions, that is, increasing the
resolution, using all the wave receiving elements for the detected azimuth. When only a part of
the wave receiving elements 31 among the arrayed wave receiving elements 31 are used, since
the spatial resolution necessarily becomes low, object detection is performed under rough space
division. When an object is detected, the object detection with higher resolution may be
performed using all the wave receiving elements 31 for the detected azimuth and its vicinity, for
example, an angle range of ± 5 °.
[0051]
In the object detection method as described above, one of a method of repeating calculation (YES
in step # 8 in FIG. 1) and a method of repeating measurement (NO in step # 8 in FIG. 1) can be
used. In the former case, at the time of the first object detection, digital data of the output signal
from all the wave receiving elements 31 is recorded in the memory, and the operation for object
detection is only digital data of some of the wave receiving elements 31 You should use it. When
the processing speed of A / D conversion is fast due to parallel processing using a dedicated
electronic element or circuit for A / D conversion, the former method is preferable.
[0052]
Similar to the shortening of the calculation time focusing on the azimuth in the above, it is
possible to shorten the calculation time focusing on the distance, which will be described with
reference to FIG. 13, FIG. 14 (a) (b) and FIG. . As shown in FIG. 13, first, the distance to the object
04-05-2019
17
is detected using the output signal output from one of the wave receiving elements 31 in the
wave receiving element array 3, and then the distance to the detected object By performing
object detection in the space of the distance range before and after including, the calculation time
of object detection can be shortened.
[0053]
In the first object detection, as shown in FIG. 14A, only the distance Z = Z1 of the object is
detected, and the detection angle range in the horizontal and upper and lower directions is
divided into one or a small number of divisions. Do. In the next object detection, all the wave
receiving elements 31 are used, and as shown in FIG. 14 (b), the resolution regarding the azimuth
is increased, that is, the number of angle divisions in the horizontal direction and the number of
angle divisions in the vertical direction are increased. Perform object detection. As a result, the
orientation of the object is accurately determined in the azimuth space of the horizontal direction
θ and the vertical direction φ, for example, as in the area 51. Depending on where the distance
Z = Z1 at which the object is detected is located in the small space divided with respect to the
distance, the resolution with respect to the azimuth is increased or decreased. For example,
highly accurate azimuth detection is performed only when the distance Z1 is a short distance,
and when an object is detected at a distant position, the computation time of the object detection
is performed by performing object detection with lowered azimuth detection accuracy. It can be
shortened. According to such an object detection method, the object detection is performed
focusing on the distance, and thereafter the object detection is performed including the
measurement of the azimuth, so the required detection accuracy is maintained and the
computation time of the object detection is shortened. be able to.
[0054]
Further, by selecting and using only data corresponding to the predetermined distance range as
digital data used for the delay addition operation, the operation can be performed in a short time
as compared with the case of using a wide range of data. This will be explained. The distance Z =
Z1 at which the object is detected is determined by the position where the output waveform by
the output signal from the wave receiving element appears on the time axis, as shown in FIG.
Therefore, digital data within a predetermined time width including the time width Δt in which
the output waveform appears is selected from the digital data of each wave receiving element 31,
and the delay addition operation is performed using the data. That is, the delay addition
operation is not performed on data outside the focused time zone. As a result, calculation time
can be shortened based on the distance measured by one wave receiving element, and the
04-05-2019
18
accuracy of object detection in the distance before and after the distance can be increased.
[0055]
In addition, another example of shortening the calculation time focusing on the distance will be
described with reference to FIGS. 16 (a) and 16 (b) as described above. In this object detection
method, as shown in FIG. 16 (a), when an object is detected at a distance Z = Z1, in the next
object detection, an object is detected as shown in FIG. 16 (b). Object detection is performed in a
space 52 closer than the distance Z = Z1. For example, only with respect to an object moving
closer to the front side, the computation time of object detection can be shortened by detecting
the direction and position of the object with higher accuracy. In such an object detection method,
it is possible to determine whether the object detected in the previous time is approaching by
performing the second object detection by remeasurement (NO in step # 8 of FIG. 1). .
[0056]
Such a method of object detection is effective for performing obstacle detection and obstacle
avoidance of the moving object quickly when the ultrasonic sensor 2 is attached to the moving
object and used as an obstacle detection sensor. In addition, after confirming that the detected
object is not approaching, the object detection calculation is performed on all the distances in the
predetermined detection space. As described above, according to the present object detection
method, since the space area for object detection is limited to a close space, the object detection
in the vicinity of the moving object important for the moving object can be efficiently performed
by shortening the calculation time. , Real-time property detection can be improved.
[0057]
Next, shortening of the operation time by the change of the operation condition (the process of
step # 7 in FIG. 1) will be described with reference to FIGS. While recording the output signal
from the wave receiving element 31 as digital data, in the delay addition operation for detecting
the direction of the object based on the difference in incident time, thinning out and using the
recorded digital data for object detection Operation time can be shortened. For example, when
sampling data by A / D converting a 50 kHz signal, the original waveform can be reproduced if
sampling is performed at 500 kHz which is twice the frequency of 100 kHz, and 10 times the
frequency in practice, considering the so-called sampling theorem can do. Therefore, as using 50
04-05-2019
19
kHz ultrasonic waves, the output signal from each of the wave receiving elements 31 is sampled
at 1 MHz to store digital data in the memory. Then, at the time of the delay addition operation,
every other digital data is thinned and used as sampling data equivalent to 500 kHz. The
calculation time can be shortened by changing such calculation conditions.
[0058]
Here, the identification of the ultrasonic wave between each receiving element is considered. As
shown in FIG. 17, it is assumed that ultrasonic waves are incident at an incident angle α to a
wave receiving element array 3 in which wave receiving elements a, b, c, etc. in the wave
receiving element array 3 are arranged at a constant interval d. . The identification of the
ultrasonic waves is a concept related to the directivity or angular resolution of the ultrasonic
sensor. In order to obtain an ultrasonic wave incident on the wave receiving element a and an
output signal identifying an ultrasonic wave incident on the wave receiving element b, at a
sampling frequency of a predetermined frequency F or higher (or the same, a predetermined
sampling period The output signal must be changed A / D in a cycle of T = 1 / F or less. For
example, in the case of the interval d = 4 mm, the incident angle α = 5 °, and the sound velocity
c = 340 m / sec, the predetermined cycle T is T = d × sin (α) / c = 1 μsec. It becomes F = 1
MHz. The predetermined period T means that the optical path difference of the ultrasonic waves
incident on the wave receiving element a and the wave receiving element b when the incident
angle α = 5 ° is one wavelength λ. In the description of the equation of FIG. 17, the incident
angle α which is such an optical path difference is expressed as a detection resolution.
[0059]
The above-described delay and add operation when using the ultrasonic sensor of the wave
receiving element array 3 having the wave receiving element distance d = 4 mm will be
described. In the front direction (incident angle α = 0 °) of the ultrasonic sensor, the addition
operation of the output signal is performed with no delay in all of the wave receiving elements a,
b and c. Conversely, the signal added without delay becomes a signal for extracting the signal of
the ultrasonic wave incident from the front. When object detection is performed in the direction
of the incident angle α of ultrasonic waves α = 5 °, that is, the azimuth angle of 5 °, digital
data of the wave receiving element b is delayed by 1 μs with respect to the wave receiving
element c. Data is delayed by 2 microseconds to perform an addition operation. Similarly to the
above, the signal added after such delay becomes a signal for extracting the signal of the
ultrasonic wave incident from the direction of the incident angle α. When performing object
detection at an azimuth of 5 ° (incident angle α = 5 °), it is necessary to delay each one
04-05-2019
20
microsecond between adjacent elements, so a sampling period of 1 microsecond is also required.
[0060]
Therefore, the output signal of the wave receiving element 31 is acquired every T = 1 μs using
the wave receiving element array shown in FIG. 17 and an ultrasonic wave of 50 kHz, and in the
delay addition process, data every 2 μs are obtained. It can be said that using it is a reasonable
process. A state of such waveform processing is shown in FIG.
[0061]
Next, with reference to FIGS. 19 and 20 (a) and 20 (b), an example of applying the abovedescribed ultrasonic sensor and the method of object detection using the ultrasonic sensor to
obstacle detection in a moving object will be described. Do. FIG. 19 shows how the ultrasonic
sensor 2 is installed on the moving body 6. The ultrasonic sensor 2 is attached at a height H from
a reference surface (traveling surface) on the front surface of the moving body 6, and a space
within the upper and lower viewing angle range ± φ and the forward distance D is defined as a
detection space 53. ing. However, below the forward Z direction, the detection space is expanded,
and there is a detection space 54 where H <D × sin φ, and this detection space 54 is a space
below the traveling surface and in a space where no object is detected is there. Therefore, for
such unnecessary detection space 54, calculation of object detection is not performed. As
described above, by performing object detection only in a three-dimensional space having an
effective field of view determined by the installation position of the ultrasonic sensor,
unnecessary calculation time can be omitted, calculation time can be shortened, and efficient
object detection can be performed.
[0062]
20 (a) and 20 (b) show a moving body 6 which travels while detecting an object by an object
detection method using the ultrasonic sensor 2 with the ultrasonic sensor 2 installed. The
moving body 6 has a total of seven ultrasonic sensors 2 in the front in the traveling direction,
two each in the rear, and two each on the left and right sides, and the signal processing of these
ultrasonic sensors 2 , And travel control means 60 for traveling and moving autonomously. Each
ultrasonic sensor 2 detects an obstacle around the entire periphery of the moving body 6 by the
detection space 5 having a spread of about 90 ° in the horizontal direction.
04-05-2019
21
[0063]
In such a mobile unit 6, the visual field of the ultrasonic sensor 2, that is, the detection space 5 is
changed according to the traveling speed of the mobile unit 6. In this case, when the moving
speed is slow, the calculation time can be shortened by shortening the detection distance. In
addition, when moving a moving object at high speed, the field of view is concentrated in the
traveling direction to perform object detection in a narrow space at high speed to satisfy realtime characteristics. Object detection can be performed.
[0064]
Further, the detection space 5 is appropriately changed at the time of changing the field of view
of the ultrasonic sensor 2, that is, the traveling direction such as a left turn or a right turn, in
accordance with the traveling direction of the movable body 6. By changing the field of view in
accordance with the change of course of the mobile, safe and efficient movement by the mobile
can be achieved. In this case, selective object detection processing such as operating only the
sensor attached to the front of the moving object is performed when the moving object is moving
forward, and detection accuracy of the distance to the object and the orientation of the object is
required. To reduce the calculation time of object detection and improve the real-time property
of object detection. In addition, reducing the angular resolution and the distance resolution other
than the sensors attached to the front is also effective for shortening the calculation time.
[0065]
The present invention is not limited to the above-described configuration, and various
modifications are possible. For example, in the division of the three-dimensional space (detection
space) in step # 4 shown in FIG. 1, the user sets division patterns in advance and stores them in
the memory, and the object detection control unit 34 detects It may be selected appropriately
from the patterns dynamically.
[0066]
04-05-2019
22
The flowchart which shows the procedure of the object detection in the object detection method
concerning this invention. The block configuration figure of the ultrasonic sensor used for the
object detection method same as the above. Sectional drawing of the ultrasonic source of an
ultrasonic sensor same as the above. The ultrasonic waveform figure which the ultrasonic source
of an ultrasonic sensor same as the above emits. The figure of the example of object distribution
for demonstrating the object detection operation | movement of an ultrasonic sensor same as the
above. The figure which shows the relationship between the incident direction of the ultrasonic
wave which injects into an ultrasonic sensor same as the above, and a wave front. The figure
which demonstrates the delay addition process of the output signal from an ultrasonic sensor
same as the above. (A) (b) is a figure of the example of a waveform obtained by carrying out the
delay addition of the output signal from the ultrasonic sensor same as the above. The perspective
view of three-dimensional space which shows the example which divided | segmented threedimensional space used as the object which performs object detection by the object detection
method of this invention by distance and direction. The perspective view of the three-dimensional
space which shows the example which changed the division | segmentation number of threedimensional space same as the above by direction. The perspective view of the three-dimensional
space which shows the example which changed the division | segmentation number of threedimensional space same as the above with distance. The front view of the wave receiving element
array explaining a mode that the number of wave receiving elements is changed and the object
detection is performed in the object detection method of this invention. The front view of the
wave receiving element array explaining a mode that an object detection is performed using all
after object detection which used one wave receiving element in an object detection method
same as the above. (A) is a perspective view of a three-dimensional space to be an object
detection target showing a state in which the distance of the object is detected by one wave
receiving element in the same object detection method as above; A two-dimensional view. The
output signal waveform figure from one wave receiving element. (A) is a perspective view of a
three-dimensional space to be an object detection target showing a state in which the distance of
the object is detected by one wave receiving element in the above object detection method, (b) is
the distance of the object shown in (a) The perspective view of the same three-dimensional space
explaining that object detection is performed in the space more near. Sectional drawing of the
wave receiving element array for demonstrating the relationship between the space | interval of
the wave receiving element in the wave receiving element array used for the object detection
method same as the above, sampling period, and detection resolution. The output signal
waveform figure which digitized and displayed the output signal from three different wave
receiving elements used for the object detection method same as the above. The side view
explaining an unnecessary three-dimensional space portion in object detection by the same
above-mentioned object detection method performed using an ultrasonic sensor attached to a
mobile. (A) is a schematic plan view of a moving body traveling while installing an ultrasonic
sensor on the moving body and performing object detection using the ultrasonic sensor with the
ultrasonic sensor, (b) is a perspective view of the moving body Figure.
04-05-2019
23
Explanation of sign
[0067]
Reference Signs List 2 ultrasonic sensor 3 receiving element array 5 detection space (threedimensional space) 6 moving body 21 transmitting element (ultrasound source) 22, 22a, 22b
(Atmospheric pressure wave) USW Coarse wave (air pressure wave)
04-05-2019
24
Документ
Категория
Без категории
Просмотров
0
Размер файла
41 Кб
Теги
jp2006234523
1/--страниц
Пожаловаться на содержимое документа