JP2012242181A - Detection device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a detection device and a detection method which can effectively obtain a positional relation with an object without requiring a communication infrastructure.SOLUTION: A detection device for detecting the position of an emission source of a radio wave comprises: a plurality of antennas (101 to 104); a plurality of receiving units (201 to 204) which respectively receive the radio wave from the emission source through the plurality of antennas; a calculation unit (300) which calculates relative values of respective reception intensities of the radio wave respectively received by the receiving units, normalizes the relative values and outputs the same as observation values; and a determination unit (400) which determines the position of the emission source on the basis of the observation values.

Description

  The present invention relates to a detection apparatus and a detection method for detecting the position of a radio wave transmission source.

  Currently, about 40% of traffic accidents occur in the vicinity of intersections in Japan, which is a serious social problem. Most of the accidents at intersections are entanglement accidents and face-to-face collisions at the time of turning left and right, mainly due to delayed recognition of drivers and jumping out of the blind spot (NLOS: non-line of sight). Therefore, research for assisting the driver in driving the vehicle by detecting an object such as a pedestrian in the blind spot and issuing an alarm to the driver is underway.

  For example, J. Ibanez-Guzman et al. Conducted a study for exchanging the positions of own vehicles obtained by GPS with each other by inter-vehicle communication. However, it is known that the GPS position accuracy in urban areas decreases due to radio wave irregular reflection and shielding. In the case of an automobile, the position accuracy can be improved because the position correction can be performed by referring to the road information of the gyro sensor or the car navigation system. However, in the case of a pedestrian or a bicycle, such a position correction is possible. It is difficult to use the function, and the use of GPS is not suitable for an application for preventing an accident at an intersection.

  At present, the position accuracy obtained by GPS is only 10 to several tens of meters, and the position accuracy required for this type of application cannot be obtained. The quasi-zenith satellite is attracting attention as a means to improve the position accuracy of GPS. However, in urban areas, the time zone in which four or more satellite signals can be received simultaneously is limited. In fact, it is not possible to expect an improvement in position accuracy by the quasi-zenith satellite.

  As an alternative to GPS, there is an in-vehicle system that detects an object using an in-vehicle camera, radar, and lidar (LIDAR). For example, P. Geismann et al. And Y. Huang et al. Proposed a system that supports driving by detecting pedestrians based on depth information measured using a stereo camera. The inventors of the present application have proposed a pedestrian detection technique using a vehicle-mounted monocular camera. To date, this type of in-vehicle system has been put into practical use as a forward collision prevention system using a stereo camera or a millimeter wave sensor.

  However, according to the above-described in-vehicle system that replaces the GPS, the infrastructure is not required and the system can be completed in the vehicle, but it exists in a blind spot region that is caused by a line of sight being blocked by another vehicle or a building. The object cannot be recognized. In addition, the situation that this type of in-vehicle system can handle is limited, and it is difficult to effectively cope with the situation at the intersection where there are multiple factors by vehicles, bicycles, and pedestrians.

  A road-vehicle cooperation system is mentioned as a system for compensating for the drawbacks of the in-vehicle system related to such blind spots. In this road-vehicle cooperation system, the roadside sensor installed at a height of about 5-10 meters measures the position and speed of the vehicle approaching the intersection, detects the position of the pedestrian near the pedestrian crossing, etc. Information obtained by the roadside sensor is provided to each vehicle by road-to-vehicle communication. As a result, the driver of each vehicle can recognize an object existing in a blind spot from the own vehicle, and can prevent an accident at an intersection such as a collision accident between a right turn vehicle and a straight-ahead vehicle or a pedestrian accident. Contributing.

  However, according to this road-vehicle cooperation system, in order to obtain a sufficient effect, it is necessary to install a number of road-side sensors. Due to cost constraints, the introduction is limited to a few large intersections.

  As a technique for solving the drawbacks of the various conventional techniques described above, there is a warning target detection apparatus disclosed in, for example, Japanese Patent Application Laid-Open No. 2008-186274 (Patent Document 1). According to this apparatus, the radio wave transmitted from the transmission source provided for the alert target such as a pedestrian is received by the antennas respectively disposed at the front part and the rear part of the vehicle. Then, by comparing the reception intensity at the front antenna and the reception intensity at the rear antenna, it is determined whether the alarm target exists in front of or behind the vehicle.

JP 2008-186274 A

  However, according to the technique disclosed in the above-mentioned publication, although the direction in which the alert target exists can be roughly determined, the position cannot be specified. Moreover, even if the distance to the alert target can be estimated based on the received signal strength of the radio wave from the transmission source, the environmental change of the radio wave propagation path, the decrease in the transmission output of the transmitter as the transmission source, etc. If the received signal strength fluctuates, the distance to the alert target cannot be estimated correctly. Therefore, it becomes difficult for the driver of the vehicle to grasp the position of the alert target and cannot pay attention to the alert target effectively.

  The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a detection apparatus and method that can effectively grasp the positional relationship with an object without requiring a communication infrastructure. .

  In order to solve the above problems, a detection apparatus according to the present invention is a detection apparatus for detecting a position of a radio wave transmission source, and includes a plurality of antennas and a plurality of antennas from the transmission source via the plurality of antennas. A detecting device comprising: a plurality of receiving units that respectively receive radio waves; and a determination unit that determines a position of the transmission source based on a combination of reception strengths of the radio waves respectively received by the plurality of receiving units. It has a configuration.

  In the detection device, the determination unit calculates a relative value of each reception intensity of the radio waves respectively received by the plurality of reception units, sets a combination of the relative values as an observation value, and based on the observation value, The position of the transmission source is determined.

  In the above detection device, the relative value is a normalized value.

  In the detection apparatus, the determination unit includes a reference table in which a coordinate value corresponding to a position of the transmission source and a reference value corresponding to the observation value are described, and the determination value is described in the observation value and the reference table. The correlation with the reference value is calculated, and the position of the transmission source is specified from the coordinate value corresponding to the reference value showing the highest correlation.

  In the detection apparatus, the reference value corresponds to a value to be obtained as the observation value when the transmission source is at the position indicated by the coordinate value.

  In the detection apparatus, the determination unit further tracks a change with time of the position of the transmission source, and determines a movement state of the transmission source based on the change with time.

  In the above detection device, the detection device is mounted on a vehicle, and means for detecting the movement state of the vehicle or a driving operation by the driver, and means for issuing an alarm based on the position of the transmission source. Further, the present invention is characterized in that the movement state of the vehicle or the driving operation by the driver is reflected in the alarm.

  The detection method according to the present invention is a detection method for detecting the position of a radio wave transmission source, comprising: receiving each radio wave from the transmission source via a plurality of antennas; and Determining a position of the transmission source based on a combination of received strengths of the received radio waves.

  The detection method according to the present invention is a detection method for detecting a position of a radio wave transmission source, wherein a plurality of receiving units respectively receive radio waves from the transmission source via a plurality of antennas. And calculating means for calculating a combination of relative values of the respective received intensities of the radio waves received by the plurality of receiving means, and outputting the combination of the relative values as an observed value; And determining the position of the transmission source based on the value.

According to the present invention, it is possible to detect the position of the transmission source with a simple configuration while suppressing the influence of fluctuations in the received signal strength of radio waves, and the object provided with the transmission source is located in the blind spot area. Even if it exists, its presence can be detected.
Therefore, for example, at an intersection where there are many blind spots, the driver can grasp the positions of other vehicles, pedestrians, and the like from the own vehicle, and can support driving of the vehicle.

It is a figure for demonstrating the example of the utilization environment of the detection apparatus by embodiment of this invention. It is a block diagram for showing the composition of the detecting device by the embodiment of the present invention. It is a figure for demonstrating arrangement | positioning on the vehicle of the some antenna with which the detection apparatus by embodiment of this invention is provided. It is explanatory drawing for demonstrating the method to produce the reference table by embodiment of this invention. It is a figure which shows the image which expressed the received signal strength obtained in the process of producing the reference table by embodiment of this invention with the gradation. It is a figure which shows an example of the reference table by embodiment of this invention. It is a figure which shows the example of a display of the position of the target object obtained by the detection apparatus by embodiment of this invention. It is a figure for demonstrating the trial production apparatus by this invention. It is a figure for demonstrating the experimental method of the prototype device by this invention. It is a figure for demonstrating the experimental result of the prototype device by this invention. It is a figure for demonstrating the experimental result of the prototype device by this invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
The present invention can be used to detect the position of a radio wave transmission source (or the distance to the transmission source and its direction) in a wide range of fields. In the present embodiment, the vehicle is operated at an intersection, in particular. A case where the position of another vehicle or a pedestrian with respect to the own vehicle is detected for assistance will be described as an example.
In the present embodiment, other vehicles, pedestrians, and the like equipped with a transmission source of radio waves received by the detection device are appropriately referred to as “objects”.

(Type of accident at the intersection assumed in this embodiment)
First, a typical traffic accident mode at an intersection assumed in the present embodiment will be described with reference to FIG.
Typical accidents at intersections are roughly divided into cases 1 to 3 shown in FIG.
Of these cases, Case 1 shows a case where the host vehicle collides with an oncoming vehicle when making a right turn. In this case 1, a blind spot area is generated by a large vehicle that is temporarily stopped on the opposite lane, and it is difficult to directly check other vehicles that approach the intersection from a distance on the opposite lane from the own vehicle. is there. For this reason, an accident may occur in which the host vehicle that is turning right collides with an oncoming vehicle that has emerged from the blind spot area. In order to prevent the occurrence of such an accident, it is required to accurately detect the oncoming vehicle existing in the blind spot area, and it is desirable to detect the oncoming direction including the traveling direction.

  Case 2 shows a case where the host vehicle collides with a pedestrian who is crossing when making a right turn (or a left turn). In Case 2, pedestrians around the pedestrian crossing may approach the vehicle that is turning right from various directions. Therefore, it is required to accurately detect pedestrians from various directions existing in the vicinity of the host vehicle. In addition, it is desirable to individually detect the position of each pedestrian so as not to issue an alarm for a pedestrian who is unlikely to collide with the host vehicle.

  Furthermore, case 3 shows a case where a subsequent bicycle or motorcycle that has been running on the left rear of the host vehicle is involved when making a left turn. In this case 3, the left rear region of the host vehicle is a region that is difficult for the driver to visually confirm, so the driver may overlook other vehicles or the like existing in such a region. Therefore, it is required to accurately detect other vehicles located on the left rear side of the host vehicle.

(Outline of this embodiment)
Next, an outline of the present embodiment will be described.
As described above, general in-vehicle systems and lidar (LIDAR) systems are based on detection of reflected waves, etc., so in principle, other vehicles, pedestrians, etc. existing in blind spots caused by obstacles It cannot be detected. In addition, the introduction of the road-vehicle cooperation system is restricted by the installation location and cost of the sensor.

  In this embodiment, in order to eliminate these inconveniences, a vehicle and a pedestrian that are detection objects include a transmitter that is a source of radio waves, and by receiving radio waves transmitted from this transmitter, The position of the object is detected in the vehicle on the receiving side. In this case, it is possible to grasp the object existing in the blind spot area by receiving the radio wave from the blind spot area using the diffraction phenomenon of the radio wave.

  In the present embodiment, antennas are installed at the four corners of the vehicle on the receiving side, and the positions of the objects are detected based on the relative values of the received signal strengths of the radio waves received by the total of four antennas. Alternatively, the direction of the object relative to the vehicle on the receiving side, the moving direction of the object (approach / separate), and the distance to the object (far / near) are estimated.

  In this case, the position of the object is detected based on the normalized value of the relative value in order to prevent erroneous detection due to fluctuations in radio field intensity caused by fluctuations in the transmission output of the transmitter and fluctuations in the surrounding environment. Then, by issuing an alarm to the driver based on the detection result, it is possible to effectively support the driving of the vehicle at the intersection where the occurrence of the accident as described above is concerned.

(Configuration of the detection apparatus according to the present embodiment)
Next, the configuration of the detection apparatus according to the present embodiment will be described in detail with reference to FIGS. 2 to 5.
As shown in FIG. 2, the detection apparatus includes a plurality of antennas 101 to 104, a plurality of receiving units (“receiving means”) 201 to 204, a calculation unit 300, a determination unit 400, and a storage unit 500. In the present embodiment, the calculation unit 300 and the determination unit 400 constitute “determination means” recited in the claims.
In the present embodiment, the storage unit 500 is a component of the determination unit 400. However, such a definition is for convenience of explanation, and the storage unit 500 may be handled as an independent component.

  Here, the antennas 101 to 104 are arranged at predetermined positions on the vehicle. In this embodiment, the antennas 101 to 104 are arranged at the four corners of the vehicle 600 as illustrated in FIGS. That is, the antenna 101 is disposed on the left front portion of the vehicle 600, the antenna 102 is disposed on the right front portion of the vehicle 600, the antenna 103 is disposed on the right rear portion of the vehicle 600, and the antenna 104 is disposed on the left rear portion of the vehicle 600. .

  In the present embodiment, the antennas 101 to 104 have directivity, and the reception directions of the antennas 101 and 102 arranged at the front are set to the left front and right front of the vehicle 600, respectively, and are arranged at the rear. The receiving directions of the antennas 103 and 104 are set to the right rear and left rear of the vehicle 600, respectively. The arrangement relationship of these antennas 101 to 104 is basically set to be the same as the arrangement relationship of the antennas when generating a “reference table” described later.

  In this embodiment, four antennas 101 to 104 are used. However, the present invention is not limited to this, and an arbitrary number of antennas can be used. Preferably, three or more antennas are used. Can be used. In this embodiment, antennas are arranged at the four corners of the vehicle. However, the present invention is not limited to this, and a “reference table” described later used for determining the position of the object can be generated significantly. In the limit, it may be arranged in any part on the vehicle, and the receiving direction may be set in any way. For example, the antenna may be arranged in any part of the vehicle such as the front center part, the rear center part, the left fender part, and the right fender part of the vehicle.

  In the present embodiment, the antennas 101 to 104 have directivity. However, the present invention is not limited to this, and an omnidirectional antenna may be employed. For example, the vehicle acts as a radio wave shield. Thus, the omnidirectional antenna may acquire directivity. In any case, the directivity of the antenna itself does not matter as long as a “reference table” described later can be generated significantly.

  Returning to FIG. 2, the four antennas 101 to 104 are connected to the receiving units 201 to 204, respectively. The receiving units 201 to 204 are for receiving radio waves from a transmission source provided in an object via the antennas 101 to 104, respectively. In this embodiment, the wireless communication device (Zigbee -Chip). Therefore, the radio wave transmission source provided for objects such as other vehicles and pedestrians also includes a transmitter conforming to the Zigbee standard.

  Here, the Zigbee standard is one of short-range wireless communication standards and is based on IEEE 802.15.4-2003. According to this standard, a mesh type network can be configured, and it is suitable for ad hoc communication because of low cost and low power consumption. The frequency band used in the Zigbee standard is the 2.4 GHz band, and is a frequency band that can be used freely on public roads in Japan. In this frequency band, radio wave diffraction phenomenon occurs relatively remarkably, so that radio waves can be received even from a blind spot region caused by an obstacle at an intersection or the like.

  In addition, because the Zigbee standard uses CSMA / CA (Carrier Sense Multiple Access / Collosion Avoidance) method, it automatically detects packet collisions and retransmits them. Can be avoided. In addition, in the Zigbee standard, a unique ID of 16 bits is assigned to each device, and the transmission packet includes the information. Therefore, using this information, packets from a plurality of transmitters can be identified, and it is possible to determine whether the radio wave is transmitted from a pedestrian, a bicycle, or a vehicle. .

The receiving units 201 to 204 measure received signal strength indications (RSSI) of radio waves respectively received via the antennas 101 to 104 installed at the four corners of the vehicle. This received signal strength is expressed as an observed value z (t) = [z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)] ^{T at time} t. Here, z ₁ (t), z ₂ (t), z ₃ (t), and z ₄ (t) represent received signal strengths received by the receiving units 201, 202, 203, and 204 at time t, respectively. . This observation value z (t) = [z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)] ^T is supplied to the arithmetic unit 300.

The calculation unit 300 calculates a relative value z _{n, m} (t) of the observed value z (t), normalizes the relative value, and outputs the normalized value as a value z ′ _{n, m} (t). The purpose of such calculation is to eliminate the influence of fluctuations in received signal strength caused by fluctuations in the transmission output of radio waves and fluctuations in the surrounding environment, as described above.

Here, for example, even if the intensity of the radio wave output from the transmission source fluctuates due to the fluctuation of the transmission output due to the decrease in the remaining battery level of the transmitter or the fluctuation of the surrounding environment, the degree of the influence of all the antennas 101 to 104 Therefore, the relative relationship between the received signal strengths obtained through these antennas 101 to 104 is maintained. The relative relationship between the received signal strengths is determined according to the position of the transmission source. Therefore, by normalizing the relative value z _{n, m} (t) based on such a relative relationship, it is possible to correctly detect the position of the target object by eliminating the influence of fluctuations in the transmission output of the transmitter, etc. become.

For convenience of explanation, the relative value z _n, a value obtained by normalizing the _m (t) z _{'n, m} (t) to "normalized relative observations z' referred to _{n, m} (t)".

Here, in the calculation unit 300, the relative value z _{n, m} (t) is calculated by the following equation (1).
z _{n, m} (t) = z _n -z _m ; 1 ≦ n <m ≦ 4 (1)

  In Expression (1), n and m represent two of the four antennas 101 to 104 (or the receiving units 201 to 204), and (n, m) = (1,2), (1,3 ), (1,4), (2,3), (2,4), (3,4), there are six combinations.

The computing unit 300 computes a difference (z _n −z _m ) between received signal strengths received by any two of the four antennas 101 to 104 at time t according to the equation (1), and calculates a relative value. Find z _{n, m} (t). In this case, the following six values are obtained as relative values z _{n, m} (t) according to six combinations of _{n and m} .

z _1,2 (t) = z ₁ -z ₂
z _1,3 (t) = z ₁ -z ₃
z _1,4 (t) = z ₁ -z ₄
z _2,3 (t) = z ₂ -z ₃
z _2,4 (t) = z ₂ -z ₄
z _3,4 (t) = z ₃ -z ₄

The relative value z _{n, m} (t) obtained in this way is normalized by the following equation (2), whereby a normalized relative observed value z ′ _{n, m} (t) is obtained.
z ' _{n, m} (t) = | z _{n, m} (t) / max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} |… (2)

In equation (2), max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} are four values z ₁ (t), z ₂ (t), Of z ₃ (t) and z ₄ (t), it represents the largest value.
Regarding the normalized relative observed values z ′ _{n, m} (t) obtained by the equation (2), the following six corresponding to the above-mentioned relative values z _{n, m} (t) are also obtained depending on the combination of _{n and m.} The normalized relative observation value z ′ _{n, m} (t) composed of these six values is supplied to the determination unit 400.

z ' _1,2 (t) = | z _1,2 (t) / max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} |
z ' _1,3 (t) = | z _1,3 (t) / max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} |
z ' _1,4 (t) = | z _1,4 (t) / max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} |
z ' _2,3 (t) = | z _2,3 (t) / max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} |
z ' _2,4 (t) = | z _2,4 (t) / max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} |
z ' _3,4 (t) = | z _3,4 (t) / max {z ₁ (t), z ₂ (t), z ₃ (t), z ₄ (t)} |

The determination unit 400 determines the position of the transmission source by referring to a later-described reference table based on the normalized relative observation value z ′ _{n, m} (t). It will be described in the explanation of the operation.

A reference table is stored in the storage unit 500 included in the determination unit 400. In this reference table, values corresponding to the above-mentioned normalized relative observation values z ′ _{n, m} (t) are described as reference values together with the coordinate values of the transmission source, and these reference values and coordinate values are obtained experimentally in advance. Is done.

Here, the reference table will be described in detail with reference to FIGS. 4 and 5.
As illustrated in FIG. 4, this reference table is obtained via the antennas 101 to 104 disposed in the vehicle 600 with a radio wave source (not shown) positioned on a predetermined grid on the xy plane. It is created by measuring the received signal strength of radio waves. In this example, the center position of the vehicle 600 is arranged at the origin, and the grids are arranged in a matrix at intervals of 2 m in a range of 40 m × 58 m.

  In the example of FIG. 4, the center position of the vehicle in the y-axis direction is not on the grid and exists between the grids, but such a definition of the grid is arbitrary and is not limited to this example.

The received signal strength c _n (x, y) measured via the antenna on the vehicle 600 when the radio wave source is located on the grid on the xy plane shown in FIG.
c _n (x, y); n = 1,2,3,4 -20 ≦ x ≦ 20, -29 ≦ y ≦ 29 (3)

In Expression (3), n represents any one of the antennas 101 to 104 (or the reception units 201 to 204). That is, c ₁ (x, y) represents the received signal strength received by the receiving unit 201 via the antenna 101. Similarly, c ₂ (x, y) to c ₄ (x, y) are respectively the antennas. The received signal strength received by the receiving units 202 to 204 via 102 to 104 is represented. Further, (x, y) represents the coordinate value of the transmission source.

For all grids, the respective values of c ₁ (x, y) to c ₄ (x, y) are measured, which is the observation value z (t) = [z ₁ (t), z ₂ ( t), z ₃ (t), z ₄ (t)] is a quantity corresponding to ^T.

FIG. 5 shows, as an example, an image in which the value of c ₂ (x, y) measured through the antenna 102 installed at the right front portion of the vehicle 600 is expressed by gradation. In the figure, the larger the value of c ₂ (x, y) (that is, the received signal strength), the higher the gray level.

  In the example shown in FIG. 5, the gray levels in the front and right direction areas are higher than in the left rear area of the vehicle 600, and the received signal intensity of radio waves coming from this area is high. In particular, the received signal strength of radio waves coming from the right front that matches the receiving direction of the antenna 102 is increased. On the other hand, it is understood from the antenna 102 that the gray level of the left rear area behind the vehicle 600 is low, and the received signal intensity of radio waves coming from this area is low. Further, it is understood that the closer to the antenna 102, the higher the received signal strength, and the farther away from the antenna 102, the lower the received signal strength.

Although not particularly illustrated, an image obtained from the value of c ₁ (x, y) measured through the antenna 101 installed at the left front portion of the vehicle 600 is basically the image of FIG. Line contrast with respect to the axis. In addition, the image obtained from the value of c ₃ (x, y) measured via the antenna 103 installed at the right rear portion of the vehicle 600 is basically a line contrast with the image of FIG. 5 and the x axis. is there. Furthermore, the image obtained from the value of c ₄ (x, y) measured via the antenna 104 installed in the left rear part of the vehicle 600 is basically point-contrast with respect to the image of FIG. 5 and the origin. .

As described above, the values of c ₁ (x, y) to c ₄ (x, y) obtained by receiving radio waves from the same transmission source via the four antennas 101 to 104 are different from each other. The combination of each value of c ₁ (x, y) to c ₄ (x, y) obtained for the coordinate value (x, y) is unique for each coordinate value (x, y). Therefore, conversely, the coordinate value of the transmission source can be specified from the combination of the values of c ₁ (x, y) to c ₄ (x, y).

From the received signal strength c _n (x, y) at each grid measured as described above, the relative value c _{n, m} (t) is calculated by the following equation (4).
c _{n, m} (x, y) = c _n (x, y) -c _m (x, y); 1 ≦ n <m ≦ 4… (4)

In Expression (4), n and m represent two of the four antennas 101 to 104 (or the reception units 201 to 204) installed in the vehicle 600 located at the origin on the xy plane in FIG. , (N, m) = (1,2), (1,3), (1,4), (2,3), (2,4), (3,4) . In accordance with these six combinations of _{n and m} , the following six values are obtained as relative values c _{n, m} (x, y).

c _1,2 (x, y) = c ₁ (x, y) -c ₂ (x, y)
c _1,3 (x, y) = c ₁ (x, y) -c ₃ (x, y)
c _1,4 (x, y) = c ₁ (x, y) -c ₄ (x, y)
c _2,3 (x, y) = c ₂ (x, y) -c ₃ (x, y)
c _2,4 (x, y) = c ₂ (x, y) -c ₄ (x, y)
c _3,4 (x, y) = c ₃ (x, y) -c ₄ (x, y)

The relative value c _{n, m} (x, y) obtained in this way is an amount corresponding to the relative value z _{n, m} (t) obtained from the observed value z (t).

The relative value c _{n, m} (x, y) is normalized by the following equation (5), and the value c ′ _{n, m} (x, y) is calculated. Hereinafter, for convenience of explanation, a value c ′ _{n, m} (x, y) obtained by normalizing the relative value c _{n, m} (x, y) is referred to as a “normalized relative reference value c ′ _{n, m} (x, y y) ".
c ' _{n, m} (x, y) = | c _{n, m} (x, y) / max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} |… (5)

In Expression (5), max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} is c ₁ (x, y), It represents the largest value among the values c ₂ (x, y), c ₃ (x, y), and c ₄ (x, y).
For the normalized relative reference value c ′ _{n, m} (x, y) obtained by the equation (5), the normalized relative observation value c ′ _{n, m} (t) is changed according to the combination of _{n and m.} The next corresponding 6 values are obtained.

c ' _1,2 (x, y) = | c _1,2 (x, y) / max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} |
c ' _1,3 (x, y) = | c _1,3 (x, y) / max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} |
c ' _1,4 (x, y) = | c _1,4 (x, y) / max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} |
c ' _2,3 (x, y) = | c _2,3 (x, y) / max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} |
c ' _2,4 (x, y) = | c _2,4 (x, y) / max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} |
c ' _3,4 (x, y) = | c _3,4 (x, y) / max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} |

The normalized relative reference value c ′ _{n, m} (x, y) obtained in this way is obtained when the source is at the position indicated by the coordinate value (x, y). ) To the normalized relative observed value z ′ _{n, m} (t) to be obtained by the calculation unit 300.
As shown in FIG. 6, the normalized relative reference value c ′ _{n, m} (x, y) is described in the reference table together with the coordinate value (x, y) when the value is measured, and is shown in FIG. It is stored in the storage device 500.

In this way, the reference table includes the normalized relative observation value z ′ _{n, m} (which will be obtained when the radio wave is received from the actual transmission source located within the range of 40 m × 58 m shown in FIG. t) is described as a normalized relative reference value c ′ _{n, m} (x, y). Therefore, from the normalized relative reference values c ' _{n, m} (x, y) described in the reference table, the value closest to the normalized relative observed value z' _{n, m} (t) actually obtained And the actual position of the object can be determined from the coordinate value (x, y) when the value is measured.

  In the present embodiment, the present detection device shown in FIG. 2 is configured as an in-vehicle system mounted on a vehicle, but how each component is realized is not particularly limited. For example, the calculation unit 300 and the determination unit 400 may be realized by the function of the navigation system. Moreover, you may comprise this detection apparatus as not a vehicle-mounted system but a fixed or portable terminal.

(Operation of the detection apparatus according to the present embodiment)
Next, the operation (detection method) of the detection apparatus will be described.
As described above, when the receiving units 201 to 204 receive radio waves from the transmission source via the antennas 101 to 104 shown in FIG. 2, the receiving units 201 to 204 obtain the observation value z (t) that is the received signal strength. Measure and give to the arithmetic unit 300. The calculation unit 300 calculates a normalized relative observation value z ′ _{n, m} (t) from the observation value z (t) and gives the result to the determination unit 400.

The determination unit 400 determines the position of the transmission source by referring to the reference table shown in FIG. 6 stored in the storage unit 500 based on the normalized relative observation value z ′ _{n, m} (t). In the present embodiment, the determination unit 400 includes the actually observed normalized relative observed value z ′ _{n, m} (t) and each normalized relative reference value c ′ _{n, m} (x, y), and among the normalized relative reference values c ' _{n, m} (x, y) specified in the reference table, the actually observed normalized relative observed values z' _{n, m} (t) Find the most correlated value. Then, the position of the object is determined from the coordinate value (x, y) corresponding to the retrieved normalized relative reference value.

Specifically, the determination unit 400 includes the actually observed normalized relative observed value z ′ _{n, m} (t) and each normalized relative reference value c ′ _{n, m} (x, Calculate the degree of correlation with y). Then, as described in the reference table normalization relative reference values c _{'n, m (x, y} ) of the highest normalized relative reference value c give _{correlation' n, m (x, y} ) is measured The coordinate value (x, y) is determined as the position of the transmission source, that is, the position of the object.

In the present embodiment, as the correlation, the Euclidean distance between the normalized relative observed value z ′ _{n, m} (t) and each normalized relative reference value c ′ _{n, m} (x, y) in the reference table. D is calculated. And among the normalized relative reference values c ′ _{n, m} (x, y) described in the reference table, the coordinate value (x, y) when the value giving the smallest Euclidean distance D is measured, that is, From the coordinate value (x, y) when the normalized relative reference value c ' _{n, m} (x, y) that is closest to the normalized relative observed value z' _{n, m} (t) is measured, Determine the position.

The Euclidean distance D is given by the following equation (6).
D = [{z ' _1,2 (t) -c' _1,2 (x, y)} ² + {z ' _1,3 (t) -c' _1,3 (x, y)} ² + { z ' _1,4 (t) -c' _1,4 (x, y)} ² + {z ' _2,3 (t) -c' _2,3 (x, y)} ² + {z ' _{2, 4} (t) -c ' _2,4 (x, y)} ² + {(z' _3,4 (t) -c ' _3,4 (x, y)} ² ] ^1/2 … (6)

In the present embodiment, the Euclidean distance D is evaluated as the correlation degree. However, the correlation degree is not limited to the Euclidean distance D, and the correlation degree may be calculated using any method. For example, one value z ′ out of six values z ′ _1,2 (t),..., Z ′ _3,4 (t) of normalized relative observation values z ′ _{n, m} (t) due to noise. _1,2 (t) is significantly different from the value c ' _1,2 (x, y) of the corresponding normalized relative reference value c' _{n, m} (x, y), so the Euclidean distance D is increased. Even so, if the other five values excluding this value are very close, it can be considered that the degree of correlation is high, and such a method may be adopted. Therefore, as long as the position of the object can be determined significantly, the meaning of the term “correlation” may be defined in any way, for example, by a similar concept.

  In addition, the coordinate value (x, y) described in the reference table is not limited to the xy component of the specific distance from the vehicle to the transmission source, but can be any limit as long as the position of the transmission source can be specified. It may be in the form of information. For example, a number may be given to the grid on the xy plane of FIG. 4 and the number may be used as a coordinate value. Further, a coordinate value may be defined as an address of a memory constituting the storage unit 500 for storing the reference table, and for example, a CAM (Content Addressable Memory) may be used as this memory.

The determination unit 400 determines the position of the object as described above, and outputs the position information.
If it is a general application that requires only position information, this position information can be used as the final output. In this embodiment, however, the determination is made for the purpose of supporting driving of the vehicle at an intersection or the like. Unit 400 includes alarm means for issuing an alarm based on the position information.
In the present embodiment, the warning means is formally included in the configuration of the determination unit 400, but the warning means may be an independent component without being limited thereto.

  The warning means included in the determination unit 400 issues a warning to the driver by, for example, voice information, image information, mechanical vibration, or the like based on the position information of the object (transmitting source). For example, as shown in FIG. 7, the position information is displayed on the screen of a display device (not shown). The three screens shown in the figure show how the position of the transmission source that moves in front of the host vehicle from right to left changes over time. That is, in the figure, the left screen displays the position of the transmission source located in the right front of the host vehicle, the central screen displays the position when the transmission source further moves leftward, and the right screen Further, the position when the transmission source moves to the left front of the host vehicle is displayed.

  Further, the determination unit 400 may track such a change in the position of the transmission source with time, determine the exercise status of the transmission source based on the change with time, and issue an alarm according to the exercise status. For example, in the example of FIG. 7, it is predicted that the transmission source continues to exercise so as to cross the traveling direction of the host vehicle from the temporal change in the position of the transmission source grasped from the left screen and the center screen. Since the possibility of collision between the host vehicle and the object having the transmission source is predicted from the situation of the movement, the determination unit 400 issues a warning to the driver and pays attention to the object having the transmission source. Prompt.

  In addition to the display example illustrated in FIG. 7, for example, the periphery of the vehicle may be divided into the first quadrant to the fourth quadrant on the xy plane, and the position of the other vehicle may be displayed in this quadrant. In this case, for example, the entire first quadrant of the xy plane is displayed as the position of the transmission source instead of the left screen and the central screen of FIG. 7, and the entire second quadrant is displayed as the position of the transmission source instead of the right screen of FIG. Is displayed. According to such a display form, the driver can intuitively grasp the direction in which attention should be paid.

  Moreover, it is good also as what issues a warning by audio | voice information with image information, or it is good also as what issues a warning only by audio | voice information.

  The detection apparatus further includes means for detecting an exercise situation of the own vehicle 600 or a driving operation by the driver, and the determination unit 400 detects the exercise situation of the own vehicle 600 or the driving operation by the driver as described above. You may reflect in the warning by a means. For example, when it is detected that the host vehicle 600 is turning in the right direction, or when it is detected that the driver has operated the indicator in the right direction, the determination unit 400, for example, The driver may be warned only about the transmission sources in the front and right directions of the host vehicle, excluding the rear and left directions.

  In this case, after determining the position with respect to the transmission sources in all directions, only the transmission sources in the front and right directions of the host vehicle may be selected as the alarm target. As a result, the processing load required for the warning can be reduced, and warnings unnecessary for the driver can be suppressed. Alternatively, the calculation unit 300 masks a part of the above-described reference table, for example, thereby omitting the calculation process required for determining the position of the object behind and in the left direction of the host vehicle. Only the position of the target object may be selected as a target for determination and warning, thereby reducing the processing load required for position determination.

As described above, according to the detection apparatus according to the present embodiment, it is possible to detect the presence of an object provided with a transmission source in a blind spot area with a simple configuration.
Therefore, for example, at an intersection with many blind spots, the driver can effectively grasp the positions of other vehicles, pedestrians, and the like from the own vehicle, and can support driving of the vehicle.

(Prototype device)
Next, with reference to FIGS. 8 to 11, an outline of a device (hereinafter referred to as “prototype device”) that has been prototyped for verifying the function of the detection device according to the present embodiment will be described.

  The prototype apparatus includes, for example, the antenna 101 and the receiving unit 201 shown in FIG. 2 described above as one receiving unit as shown in FIG. 8C, and has a total of four receiving units. Installed in the four corners. Based on the received radio wave intensity at each receiving unit, which varies depending on the position of the transmitter, the prototype device, as shown in FIG. Rough estimation of approach / separation) and distance (far / near) provides driving support at intersections.

  The four receiving units are configured using Xbee chips that conform to the Zigbee standard as described above. As shown in FIG. 8B, the receiving unit is accommodated in a shielding box covered with a radio wave absorber, and an opening is formed in a part of the shielding box. The antenna of the internal receiving unit receives radio waves through this opening. Therefore, due to the effect of the shielding box, the received radio wave intensity of the antenna facing the transmitter is high, the reception intensity on the opposite side is low, and the difference varies depending on the distance. The position of the source can be estimated using the intensity. In order to verify the function of the detection device according to the above-described embodiment, the four receiving units of the prototype device were temporarily installed on a tripod around the vehicle as shown in FIG.

In this prototype device, a particle filter was applied to suppress the influence of noise caused by reflection on the road surface and stabilize the position estimation. In the filter of this method, the state vector is defined as “x (t) = [x (t) y (t) v _x (t) v _y (t)] ^T ”. The transmitter position “(x, y)” is included here. In the particle filter, the probability density function “P (x (t) | z (t))” is a set of sample particles “S (t) = {s ⁽¹⁾ (t), s ⁽²⁾ (t), …} ”. Each particle has the same dimensions and elements as the state vector.

In the update phase of the particle filter, the new sample set “S (t + 1)” is derived from the previous sample “S (t)” and the corresponding likelihood “π (t) = {π ⁽¹⁾ (t ), π ⁽²⁾ (t),…} ”. Each likelihood “π ^(k) (t)” corresponds to each of the samples “s ^(k) (t)”, and its value is the likelihood function “L (z (t, n, m) | s ^{( k)} (t)) ”. The final filter performance is determined by this likelihood function. In this prototype, this function is defined as the following equation (7). However, in the following equation, the application range of “Σ” is 1 ≦ n <m ≦ 4.
L (z (t, n, m) | s ^(k) (t)) = exp [Σa (n, m) {z (t, n, m) -C (x, y, n, m)} ² / σ ² ] + O… (7)

Here, “O” is an offset for stabilizing the estimation, and “σ ² ” is a dispersion value of white noise assumed for convenience. “A (n, m)” is a weight of the normalized value for each likelihood. These values were determined empirically. “(X, y)” of “C (x, y, n, m)” is an element held by each “s ^(k) (t)”, and the function is at that position. Returns each possible normalized value. Therefore, the likelihood function “L (z | s)” is a function that compares an assumed value with an actual value, and gives a larger likelihood as it gets closer.

Next, state estimation of a transmitter used as a transmission source provided in the object will be described.
In this prototype device, as a result of the final estimation, three states are output: the approximate direction of the transmitter, whether it is approaching or moving away, and whether it is close or far away, and aiming at collision avoidance based on them. It was. Since all state estimation uses the above-described filtering results, the accuracy of collision avoidance depends on the accuracy of position estimation. The normalized value adopted in the prototype device is less affected by changes in radio waves due to the state of the transmitter and the surrounding environment, and the result estimation can be stabilized as shown in FIG. FIG. 10 includes two graphs representing received signal strengths obtained in the high-power transmission mode and the low-power transmission mode, respectively. In these graphs, near and far are estimated based on the above-described equation (7). doing. In each figure, the vertical axis represents the observed received signal strength (RSSI) value, the horizontal axis represents time, and a series of received signal strength until the object passes close by from far away This represents the state of change. As shown in FIG. 10, it can be seen that, regardless of whether the radio field intensity is high or low, it can be determined that the sensor value difference is “near” and the small value is “distant”.

Next, experimental results will be described.
In this prototype device, two types of experiments were conducted as shown in FIG. 9, assuming an accident near the intersection. One is a face-to-face collision with an oncoming vehicle that is turning right as shown in case 1 in FIG. 1, and the other is a collision with a nearby pedestrian or bicycle as shown in cases 2 and 3. It is. These accident modes are merely examples, and the present apparatus can be applied to other examples and accident prevention other than at intersections. In this experiment, received signal strength (RSSI) data was collected in time series, and estimation was performed based on the normalized relative reference value of the reference table described above.

Since the direction of the approaching vehicle is based on the intersection shown in FIG. 9A, it is necessary to convert the estimated position obtained by filtering. In this prototype device, a transmitter was attached to the bicycle, and the data was collected by moving on the dotted line shown in FIG. Numbers (1)-(4) are the scenario numbers of the experiments that were performed. They approach the intersection from different directions. The starting point was a point 20 m from the center of the intersection, and data traveled about 30 m from there was recorded.
In this experiment, the observation vehicle stopped waiting to turn right. Since the oncoming vehicle approaches from a distance, the vehicle approaches at first and leaves from the middle. For the direction, most were judged well.

Next, avoidance of collision with a pedestrian will be described.
Measurement was performed by tracing the line shown in FIG. Scenarios (5) to (12) are for testing the position and movement in the vicinity, and scenarios (13) and (14) are assumed to be a side-by-side bicycle with relative positions around. . From this experiment, it was possible to detect the state where the side-by-side bicycle was in the vicinity.

  FIG. 11 shows a comparison between the position obtained from the depth information of the stereo camera and the position detected by the prototype device for the same object. In this example, the object is moving in the left-right direction in front of the host vehicle. In FIG. 11A, the white circle plot represents the position estimated by the prototype device. FIG. 11B shows the error distance between the position estimated by the prototype device and the correct position, and the horizontal axis indicates the distance from the center of the vehicle to the correct position, and the vertical axis. Indicates the error distance. As understood from FIG. 11, the position detected by the prototype device was approximated to the position detected by the stereo camera, and it was confirmed that the position was sufficiently accurate.

Next, the effect of stabilization by normalization will be described.
In this prototype, in order to obtain robustness against changes in radio field strength due to transmitter conditions and the surrounding environment, the received signal strength (RSSI) was normalized and estimated. In order to verify this, an experiment was conducted by changing the transmitted radio wave intensity. From this experiment, it was confirmed that even if the transmitted radio wave intensity was changed, the result did not change, and the effect of normalization could be demonstrated.

  As described above, the embodiment according to the present invention and the verification by the prototype device have been described. However, the present invention is not limited to these examples and can be modified and modified in any manner without departing from the gist of the present invention. Good.

For example, in the above-described embodiment, the relative value z _{n, m} (t) and the relative value c _{n, m} (x, y) are normalized. If there is no need to consider the influence of fluctuations in the transmission output of the source, the surrounding environment, etc., each value z ₁ (t) of the received signal strength z _n (t) supplied from the plurality of receiving units 201 to 204, It is also possible to detect the position of the object (transmission source) based on the combination of z ₂ (t), z ₃ (t), and z ₄ (t). In this case, the reference table shown in FIG. 6 includes the values c ₁ (x, y), c ₂ (x, y), c ₃ (x, y) of the received signal strength c _n (x, y) as reference values. ), c ₄ (x, y). According to this example, the calculation unit 300 can be omitted, and the determination unit 400 can refer to the reference table to determine the object (transmission source) based on the combination of the observed values z _n (t). The position of is determined.

For example, if there is no need to consider the influence of fluctuations in the transmission output of the radio wave source or the surrounding environment, the relative value z _{n, m} (t) is not normalized and the relative value z _{n, m} ( t) values z _1,2 (t), z _1,3 (t), z _1,4 (t), z _2,3 (t), z _2,4 (t), z _3,4 ( It is also possible to detect the position of the object (transmission source) based on the combination of t). In this case, the reference table shown in FIG. 6 includes each value c _1,2 (t), c _1,3 (t), c _1,4 (t) of the relative value c _{n, m} (t) as a reference value. , c _2,3 (t), c _2,4 (t), c _3,4 (t) are described. According to this example, the calculation unit 300 calculates the relative value z _{n, m} (t) of the received signal strength supplied from the plurality of reception units 201 to 204, and calculates the relative value z _{n, m} (t). Each value is output as an observed value, and the determination unit 400 determines the position of the object (transmission source) based on the combination of the relative values z _{n, m} (t).

In the above-described embodiment, the relative value z _{n, m} (t) and the relative value c _{n, m} (x, y) are set to the maximum values max {z ₁ (t), t ₂ (t), z, respectively. ₃ (t), z ₄ (t)} and maximum value max {c ₁ (x, y), c ₂ (x, y), c ₃ (x, y), c ₄ (x, y)} The normalization method is not limited to this example, but other amounts can be used if the influence of fluctuations in the transmission output of the radio wave source and the surrounding environment can be suppressed. It is also possible to normalize using.

  101-104 ... Antenna (sensor), 201-204 ... Receiving unit, 300 ... Calculation unit, 400 ... Determination unit, 500 ... Storage unit, 600 ... Vehicle.

Claims (8)

A detection device for detecting the position of a radio wave transmission source,
Multiple antennas,
A plurality of receiving means for respectively receiving radio waves from the transmission source via the plurality of antennas;
Determination means for determining a position of the transmission source based on a combination of reception strengths of the radio waves respectively received by the plurality of reception means;
A detection device comprising: The determination means includes
Calculating a relative value of each received intensity of the radio waves respectively received by the plurality of receiving means, using a combination of the relative values as an observed value, and determining a position of the transmission source based on the observed value; The detection device according to claim 1.   The detection apparatus according to claim 2, wherein the relative value is a normalized value. The determination means includes
A reference table in which a coordinate value corresponding to the position of the transmission source and a reference value corresponding to the observed value are described;
The correlation between the observed value and the reference value described in the reference table is calculated, and the position of the transmission source is specified from the coordinate value corresponding to the reference value showing the highest correlation. Item 4. The detection device according to any one of Items 1 to 3.   5. The detection apparatus according to claim 4, wherein the reference value corresponds to a value to be obtained as the observation value when the transmission source is at a position indicated by the coordinate value.   6. The determination unit according to claim 1, further comprising: tracking a change with time of the position of the transmission source, and determining a movement state of the transmission source based on the change with time. Detection device. The detection device is mounted on the vehicle, and further includes means for detecting the movement state of the vehicle or a driving operation by the driver, and means for issuing an alarm based on the position of the transmission source,
The detection apparatus according to claim 1, wherein a motion state of the vehicle or a driving operation by the driver is reflected in the alarm. A detection method for detecting the position of a radio wave source,
Receiving each radio wave from the source via a plurality of antennas;
Determining the position of the source based on a combination of the received strengths of the radio waves received by the plurality of receiving means;
A detection method comprising: