JP6220717B2 - Apparatus and method for estimating moving direction and speed of object - Google Patents

Description

  The present invention relates to an apparatus and method for estimating a moving direction and speed of an object using a multistatic method for detecting an object that hardly reflects energy such as sound waves and electromagnetic waves.

  In a multi-static object detection system (see, for example, Patent Documents 1 to 3), a large number of transmission devices and reception devices are arranged in a search space, and observation is performed with all reception devices while switching the transmission devices. Usually, the velocity vector of the target object can be calculated from the Doppler shift amounts obtained from a plurality of receiving apparatuses.

JP 2002-168937 A JP 2007-304071 A JP 2011-058906 A

However, if the target object is a constituent material or structure that is difficult to reflect energy, the reflected wave sensed by the receiving device becomes very weak. Therefore, by picking up noise (white noise) as a result of lowering the threshold, a false candidate Since points occur, it is difficult to analytically calculate a velocity vector for the following reason.
(1) In some cases, the energy reaching the receiving apparatus does not reach the threshold value, and only a part of the velocity vector component of the Doppler shift amount necessary for the calculation is obtained.
(2) A large number of irrelevant velocity vector components (false candidate values) obtained from false candidate points caused by white noise are mixed in a form indistinguishable from true components.

  If the Doppler shift information obtained from the true target object can be reliably retained and the moving direction and speed can be estimated after suppressing the influence of the false candidate points, the movement of the target object can be calculated for each transmission. As a result of greatly increasing the amount of information, the tracking accuracy of the moving object is improved, and the search procedure is made more efficient by reducing the number of transmissions required until the target is specified. However, an apparatus and method for estimating the moving direction and speed of such an object have not been developed yet.

  An object of the present invention is to provide an object moving direction and speed estimation apparatus and method capable of estimating the moving direction and speed of a target object with higher tracking accuracy than in the prior art. It is in.

An object movement direction and speed estimation device according to a first invention includes a plurality of transmission devices and a plurality of reception devices, and the plurality of transmission devices transmit predetermined radio signals one by one and are reflected by the target object. A plurality of receiving devices receive wireless signals that are not reflected or reflected, convert the arrival time from the transmitting device to the receiving device into a distance from the speed of the wireless signal, and target object on an ellipse that focuses on the transmitting device and the receiving device A multi-static object detection system that determines the position of a target object by analytically calculating the position where multiple ellipses overlap by acquiring ellipses in a similar manner in a plurality of other receiving devices. A moving azimuth and speed estimation device for an object,
Based on the known position of the transmission device, the known position of the reception device, and the determined position of the target object, the movement direction of the target object is all set in a state where the posture and movement direction of the target object match. The estimated speed of the target object is calculated by calculating the Doppler shift amount, which is the frequency shift amount of the radio signal due to the Doppler phenomenon generated by the moving target object, discretely around the target object. characterized in that it comprises a maximum likelihood velocity vector calculation unit for estimating integrates speed estimation value calculated from the candidate values for all of the velocity vector components, including false candidate values of the vector components.

  In the moving azimuth and speed estimation device of the object, the maximum likelihood speed vector calculation unit votes a speed estimation value in each direction at a point or a region with respect to a voting space having a polar coordinate shape, and obtains a final vote distribution The moving direction and speed of the target object are estimated based on the above.

  Further, in the moving direction and speed estimation apparatus of the object, the maximum likelihood speed vector calculation unit performs voting with different weights so that votes from a plurality of different receiving apparatuses are highly evaluated at the time of voting, and the speed of the target object The estimation value is estimated.

  Furthermore, in the moving direction and speed estimation device for the object, the maximum likelihood speed vector calculation unit performs voting by changing the weight of the vote based on a predetermined physical constraint condition at the time of voting, and calculates the speed estimation value of the target object. It is characterized by estimating.

  Still further, in the moving azimuth and velocity estimation device of the object, the maximum likelihood velocity vector calculation unit is a region in which a voting range is widened based on an azimuth error of a receiving device or a measurement error of a frequency shift amount at the time of voting. And estimating the speed estimation value of the target object.

  Furthermore, in the moving direction and speed estimation device for the object, the maximum likelihood speed vector calculation unit limits the speed or direction that the target object can take from the vote distribution in the voted polar coordinate voting space, and A speed estimation value is estimated.

An object movement direction and speed estimation method according to a second invention includes a plurality of transmission devices and a plurality of reception devices, and each of the plurality of transmission devices transmits a predetermined radio signal one by one and is reflected by the target object. A plurality of receiving devices receive wireless signals that are not reflected or reflected, convert the arrival time from the transmitting device to the receiving device into a distance from the speed of the wireless signal, and target object on an ellipse that focuses on the transmitting device and the receiving device A multi-static object detection system that determines the position of a target object by analytically calculating the position where multiple ellipses overlap by acquiring ellipses in a similar manner in a plurality of other receiving devices. A method for estimating the moving direction and speed of an object,
Based on the known position of the transmitting device, the known position of the receiving device, and the determined position of the target object, the maximum likelihood velocity vector calculation unit Estimating the velocity of the target object by computing the Doppler shift amount, which is the frequency shift amount of the radio signal due to the Doppler phenomenon generated by the moving target object, assuming the moving direction of the target object discretely around the entire circumference The method includes a step of estimating a value by integrating velocity estimation values calculated from candidate values of all velocity vector components including false candidate values of velocity vector components of the target object .

  In the moving direction and speed estimation method of the object, the maximum likelihood speed vector calculation unit votes the speed estimation value in each direction at a point or a region with respect to a voting space having a polar coordinate shape, and obtains a final vote distribution The method further includes the step of estimating the moving direction and speed of the target object based on

  Further, in the method for estimating the moving direction and speed of the object, the maximum likelihood speed vector calculation unit performs voting by changing the weight so that votes from a plurality of different receiving devices are highly evaluated at the time of voting, and the speed of the target object The method further includes the step of estimating an estimated value.

  Further, in the moving direction and speed estimation method of the object, the maximum likelihood speed vector calculation unit votes while changing the weight of the vote based on a predetermined physical constraint condition at the time of voting, and calculates the speed estimation value of the target object. The method further includes the step of estimating.

  Still further, in the moving azimuth and velocity estimation method of the object, the maximum likelihood velocity vector calculation unit gives a range to the voting range based on the azimuth error of the receiving device or the measurement error of the frequency shift amount at the time of voting. And estimating the speed estimation value of the target object.

  Still further, in the moving direction and speed estimation method of the object, the maximum likelihood speed vector calculation unit limits the speed or direction that the target object can take from the vote distribution in the voted polar coordinate voting space, The method further includes the step of estimating a speed estimation value.

Therefore, according to the apparatus and method for estimating the moving direction and speed of an object according to the present invention, it is possible to estimate the moving direction and speed of the target object with higher tracking accuracy than in the prior art. Thereby, the present invention has the following specific effects.
(1) The velocity vector of the target object can be estimated from a large number of detection candidate points including false candidates.
(2) Two or more target objects having different moving directions that overlap at the same position can be identified.
(3) Since the speed assumption for all directions is always maintained, it is possible to flexibly respond to requests for correction and review in target object tracking and target object discrimination in the upper process.
(4) Feature quantities useful for verification and evaluation of target object movement between time series can be calculated.

It is a block diagram which shows the structure of the moving direction and speed estimation apparatus of the object which concerns on one Embodiment of this invention. FIG. 3 is a plan view showing a case where the presence of a target object 31 can be limited on an ellipse focusing on a transmission device 11 and a reception device 21 in the movement direction and speed estimation processing of the object movement direction and speed estimation device of FIG. 1. FIG. 3 is a plan view showing a case where a plurality of receiving devices 22 are further present in space in the case of FIG. 2. It is a top view which shows the case where the direction of a target object and a moving method are assumed the same as a moving azimuth | direction and speed estimation process of the object of FIG. FIG. 2 is a diagram illustrating a relationship among a transmission apparatus 11, a reception apparatus 21, and a target object 31 in a movement direction and speed estimation process of an object movement direction and speed estimation apparatus in FIG. It is a top view which shows the velocity vector which receives the influence of the Doppler effect, and the velocity vector which receives the influence of the Doppler effect at the time of reflection. FIG. 3 is a plan view showing a case where a velocity vector is voted in a voting space as a region or as a point in the movement azimuth and velocity estimation processing of the object movement azimuth and velocity estimation device of FIG. It is a figure which shows the movement azimuth | direction and speed estimation process of the object's movement azimuth | direction and speed estimation apparatus of FIG. 1, Comprising: (a) is a figure which shows the velocity voting space of a polar coordinate shape, (b) is vote distribution of voting space (C) is a bar graph showing a vote distribution in which the moving direction of the object is fixed, and (d) is a bar graph showing a vote distribution in which the moving speed of the object is fixed. 3 is a flowchart showing a voting process for a polar coordinate-shaped voting space executed by a maximum likelihood velocity vector calculation unit 101 in FIG. 1.

  Hereinafter, embodiments according to the present invention will be described with reference to the drawings. In addition, in each following embodiment, the same code | symbol is attached | subjected about the same component.

FIG. 1 is a block diagram showing a configuration of an object moving direction and speed estimation apparatus according to an embodiment of the present invention. In FIG. 1, the object movement direction and speed estimation apparatus according to the present embodiment is a predetermined one from each transmission apparatus 11 in an object detection system including a plurality of transmission apparatuses 11 and a plurality of reception apparatuses 21. A wireless signal (sound wave, ultrasonic wave, radio wave signal) is transmitted, and arrival times (received from the transmission time from the transmission device 11) when the reflected signals reflected by the plurality of target objects 31 are received by the plurality of reception devices 21. An object detection system using a multi-static method for detecting the position of each target object 31 by measuring the time until the reception time of the device 21, and estimating the moving direction and speed of each target object 31. It is a feature. Here, the moving direction and speed estimation device of the object are:
(1) Input including, for example, a keyboard, a mouse, and the like, which inputs position coordinates of a plurality of transmission apparatuses 11, position coordinates of a plurality of reception apparatuses 21, and arrival times measured by the reception apparatuses 21 as described above. Part 110;
(2) An arithmetic processing unit 100 including a maximum likelihood velocity vector calculation unit 101 that calculates a maximum likelihood velocity vector by executing a maximum likelihood velocity vector calculation process described later based on input data by the input unit 110;
(3) It includes a display unit such as a display and a printer, and includes an output unit 120 that outputs a calculation result of the maximum likelihood velocity vector calculated by the calculation processing unit 100.

  First, the outline of the maximum likelihood velocity vector calculation processing by the moving direction and velocity estimation device of the object according to the embodiment will be described below.

  In the present embodiment, if the moving direction of the target object 31 can be determined in advance, the moving speed of the target object 31 can be estimated from the frequency shift amount detected by the single receiving device 21. Accordingly, the moving azimuth of the target object 31 is assumed for all azimuths in advance, and the estimated speed value in each moving azimuth is calculated. By integrating the estimated speed values of each moving direction calculated from all detection candidate points including false candidate points, the influence of false candidate values due to false candidates is eliminated, and the maximum likelihood velocity vector that the target object can take is calculated. it can.

  Here, as a specific method for integrating speed estimation values, a polar coordinate-shaped speed voting space having a voting space in an azimuth circle is defined using each direction as a key. When voting the estimated speed value calculated in each assumed moving direction to the voting space of the corresponding direction, voting based on the physical grounds such as Doppler phenomenon overlaps in the voting space, and the voting of false candidates caused by white noise Are distributed over the voting space. For this reason, the overlap in the voting space indicates the likelihood of the velocity vector that the target object 31 can take. From the vote distribution in this voting space, it is possible to estimate the most likely velocity vector that the target object can take without being affected by false candidates by detecting a threshold value or peak in a highly overlapping part. .

  As a result, there is a limit to the speed component that can be reproduced from the amount of frequency shift obtained by the single receiver 21. Therefore, an estimation result based on velocity components from a plurality of directions can be obtained by giving a weight so that votes from different receiving devices 21 are highly evaluated.

  Further, by adding weights based on physical constraints (limit speed, distance between the transmission device 11 and the reception device 21, incidence angle and reflection angle, reflectance of surface material, etc.) of the target object 31, the vote distribution In contrast, an index indicating the presence rate of the target object 31 can be added. For example, when the physical constraint condition of the target object 31 is satisfied, voting by adding a weight of 1 increases the number of votes near the target object 31, and when the physical constraint condition such as white noise is not satisfied, 0 By voting with weights, the number of votes is reduced and rejected.

  Furthermore, in order to improve the accuracy of speed estimation, it is necessary to increase the density of the voting space. However, when voting is performed using the speed estimation value as a point, the voting space is sparsely distributed and a significant overlap of votes cannot be obtained. Therefore, voting that reproduces a certain overlap that does not depend on the density of the voting space by expanding the voting space to the region by using variations such as the measurement error of the frequency shift amount and the azimuth observation error of the receiving device. Is possible.

  Next, object detection by the multistatic method according to the present embodiment will be described below.

  A large number of transmitters 11 and receivers 21 are distributed in the space where the detection target object 31 exists. A predetermined radio signal is transmitted one by one from the plurality of transmission apparatuses 11 and received by a large number of reception apparatuses 21.

  FIG. 2 is a plan view showing a case where the presence of the target object 31 can be limited on an ellipse focusing on the transmission device 11 and the reception device 21 in the movement direction and speed estimation processing of the object of FIG. is there. As indicated by the solid line arrow in FIG. 2, when the energy of the radio signal transmitted by the transmission device 11 reflects off the target object 31 and is detected by a certain reception device 21, the arrival time from the transmission device 11 to the reception device 21. Can be converted into a distance from the velocity of the propagating wave, so that the presence of the target object can be limited on an ellipse with the transmitting device 11 and the receiving device 21 as the focus.

  3 is a plan view showing a case where a plurality of receiving devices 22 and 23 are further present in the space in the case of FIG. 2. In FIG. 3, a plurality of receiving devices 22 and 23 are further present in the space. Similarly, the other receiving apparatuses 22 and 23 can obtain an ellipse, and the position of the target object can be determined by analytically calculating the position where the ellipses overlap. Such an object detection method is called a multi-static method. Actually, not only the stationary target object 31 but also the moving target object 31 exists, and such a target object 31 generates a frequency shift (Doppler shift) due to the Doppler phenomenon. The expansion / contraction effect of the wave of the radio signal generated by the movement of the target object 31 is the cause of the Doppler shift, and the velocity vector of the target object 31 can be calculated backward from the deviation amount, but the velocity component causing the Doppler phenomenon is projected onto the propagation path. Since it is limited to the component, the velocity vector cannot be determined from the deviation amount obtained by the single receiving device 21. Only when detection is obtained by a plurality of receiving devices 21, a moving azimuth component indicating a different direction can be calculated from a deviation amount of each, and an analytical method of multivariable parameters such as a least square method is used. The moving direction and speed of the target object 31 can be determined.

However, when the target object 31 is a constituent material or structure that hardly reflects energy, the reflected wave that can be sensed by the receiving device 21 becomes very weak, and it is difficult to obtain detection simultaneously by a plurality of receiving devices 21. Therefore, it is necessary to lower the threshold value to improve the sensitivity of the receiving device 21, but as a result, a lot of noise is picked up and a large number of "false" candidate points are generated. The “false” candidate point includes information caused by white noise that is completely unrelated to the true movement of the target object, and cannot be distinguished from the true detection candidate point that captured the target object 31 at all. It is a hindrance to derive the speed analytically.

  Next, the assumption of the moving direction and the rejection of the speed performed by the maximum likelihood speed vector calculation unit 101 will be described below.

  FIG. 4 is a plan view showing a case where the direction of the object and the direction of movement are assumed to be the same in the movement direction and speed estimation processing of the object shown in FIG. . FIG. 5 is a diagram showing the relationship among the transmission apparatus 11, the reception apparatus 21, and the target object 31 in the movement direction and speed estimation process of the object in FIG. It is a top view which shows a vector, the velocity vector which receives the influence of the Doppler effect at the time of incidence, and the velocity vector which receives the influence of the Doppler effect at the time of reflection.

  The calculation formula of the Doppler shift amount obtained from the single receiving device 21 assumes that the transmitting device 11 and the receiving device 21 are stationary and the positions are known, and the posture of the target object and the moving direction are the same. The Doppler shift amount DS is expressed by the following equation.

  Here, each vector is normalized, and c is the velocity of the carrier wave. When the moving direction vector of the target object is Vp, the speed of the target object is V, the incident vector is PS, and the reflection vector is RP, the expressions in parentheses on the right side of Expression (1) are obtained by adding projection components to each. Become. This indicates that if the moving direction vector of the target object 31 can be determined in advance, the speed of the target object can be determined from the Doppler shift value.

  Here, if the moving direction vector of the target object 31 is appropriately placed in advance, an estimated speed when the target object 31 moves in that direction can be obtained. However, in reality, there are physical constraints on the speed that can be taken by the target object 31, and if the estimated speed of the target object deviates from the actual speed, an appropriate temporary placement is required. Can be rejected. If this assumption and the concept of rejection are implemented for the entire periphery, a plurality of velocity vectors that are likely to be taken by the target object 31 can be derived, and estimation for unrealistic directions can be rejected in advance.

  Furthermore, creation of a polar voting speed voting space performed by the maximum likelihood velocity vector calculation unit 101 will be described below.

  FIG. 6 is a plan view showing a case where a velocity vector is voted in a voting space as a region or as a point in the movement azimuth and velocity estimation processing of the object in FIG. 7 is a diagram showing the movement direction and speed estimation processing of the object movement direction and speed estimation apparatus of FIG. 1, and FIG. 7A is a diagram showing a polar coordinate-shaped speed voting space. (B) is a graph showing the vote distribution in the voting space, FIG. 7 (c) is a bar graph showing the vote distribution in which the moving direction of the object is fixed, and FIG. 7 (d) is a vote in which the moving speed of the object is fixed. It is a bar graph which shows distribution.

  In order to efficiently carry out the above assumptions and rejections, as shown in FIGS. 6 and 7A, a velocity voting of a polar coordinate shape constructed by a mesh of azimuth circles with a specified azimuth as a key. Think of space. In this voting space, the previous “temporary placement of the moving direction vector of the target object” is replaced with “selection of the voting angle”, and “physical constraints on the speed that the target object can take” are expressed as “weight at the time of voting” . When a plurality of detection candidate points including false candidates are voted, as shown in FIG. 7B, votes based on the physical basis such as Doppler phenomenon overlap, and the vote value becomes high. Votes that are not distributed will be dispersed, resulting in lower votes. In other words, the vote distribution obtained from this voting space can be regarded as indicating the high possibility of moving azimuth and speed that the target object can take, as shown in FIGS. 7C and 7D. . In addition, the velocity distribution can be acquired when the moving direction of the target object is obvious, and the direction distribution can be acquired when the velocity of the target object is obvious. It can be used as an effective feature amount.

  FIG. 8 is a flowchart showing a voting process in a polar coordinate-shaped voting space executed by the maximum likelihood velocity vector calculation unit 101 of FIG.

  In FIG. 8, first, in step S1, search candidate points scattered around the specified position (a predetermined range from the specified position) are searched. Next, in step S2, the moving direction of the target object 31 is assumed discretely with respect to all directions. In the moving direction assumed in step S3A, the speed of the target object 31 is estimated from the Doppler shift amount. In step S3B, a weight based on a predetermined physical constraint condition is added to the estimated speed. Further, in step S3C, a vote is given by a point or a region with respect to the estimated orientation of the voting space. In step S4, the processes of steps S3A to S3C are repeated in all directions from 0 degrees to 360 degrees. In step S5, the voting process is performed by repeating the processes in steps S2 to S4 for all the detection point candidates scattered.

  According to the moving direction and speed estimation device of the object according to the present embodiment configured as described above, if the moving direction of the target object can be determined in advance, the speed can be determined from the frequency shift amount detected by a single receiving device. It can be calculated. Therefore, assuming the moving direction of the target object with respect to the entire circumference in advance, the speed in each direction is calculated. By integrating the speed estimation values calculated from all the detection candidates including the false candidates, the influence of the false candidates can be eliminated. Here, as a specific method, a velocity voting space having a polar coordinate shape is defined. Vote the estimated speed value in the specified direction in the voting space with points or areas. Votes based on physical grounds such as Doppler phenomenon overlap in space, and votes from false candidates caused by white noise are dispersed in space. By detecting the peak on the vote distribution in the voting space, it is possible to estimate the moving direction and speed of the target object that is not affected by the false candidate. In addition, it is possible to force estimation based on a basis from a plurality of directions by giving a weight so that votes of different receiving apparatuses are highly evaluated at the time of voting.

  Therefore, according to the present embodiment, it is possible to estimate the moving direction and speed of the target object without being affected by the false candidate. In addition, the vote distribution in the velocity voting space can be treated as a probability distribution of the velocity that the target object can take in the nearby space, verification of the moving direction, determination of the stationary target object, tracking of moving objects in time series, and multiple moving objects It can be used as a feature amount for overlap detection and the like. Furthermore, the velocity vector of the target object can be estimated from a large number of detection candidate points including false candidates. It is also possible to identify two or more target objects having different moving directions that overlap at the same position. Furthermore, since the speed assumption for all directions is always maintained, it is possible to flexibly respond to requests for correction and review in target object tracking and target object discrimination in the upper process. Feature quantities useful for verification and evaluation of target object movement between time series can be calculated.

  In the above embodiment, when the target object 31 is an underwater object that moves in water, this embodiment is an underwater object detection system for detecting an underwater object. Here, the transmission device 11 is a transmission device for transmission, the reception device 21 is a reception device for reception, and the target object 31 is an underwater moving body that moves in water.

  Further, at the time of voting, voting may be performed by changing the weight of the vote based on the physical constraints of the target object such as the limit speed, the distance between the transmission device 11 and the reception device 21, the incident angle and the reflection angle, and the reflectance of the surface material. Good.

  Further, at the time of voting, the voting range may be voted as an area having a width based on the azimuth error of the receiving device and the measurement error of the frequency shift amount.

  Further, the most likely velocity vector that the target object can take may be estimated from the final vote distribution in the voted polar coordinate voting space. Here, threshold value processing, peak processing, or the like can be used to extract the velocity vector from the vote distribution.

  Further, a limited azimuth distribution or velocity distribution may be acquired by assuming a velocity or azimuth that can be taken by the target object from the vote distribution in the voted polar coordinate voting space.

As described above in detail, according to the object moving direction and speed estimation apparatus and method according to the present invention, the moving direction and speed of the target object can be estimated with higher tracking accuracy than in the prior art. Thereby, the present invention has the following specific effects.
(1) The velocity vector of the target object can be estimated from a large number of detection candidate points including false candidates.
(2) Two or more target objects having different moving directions that overlap at the same position can be identified.
(3) Since the speed assumption for all directions is always maintained, it is possible to flexibly respond to requests for correction and review in target object tracking and target object discrimination in the upper process.
(4) Feature quantities useful for verification and evaluation of target object movement between time series can be calculated.

11: Transmitter,
21, 22, 23 ... receiving device,
31, 32 ... target object,
100 ... arithmetic processing unit,
101 ... Maximum likelihood velocity vector calculation unit,
110 ... input section,
120: Output unit.

Claims (12)

A plurality of transmission devices and a plurality of reception devices are provided, each of the plurality of transmission devices transmits a predetermined radio signal one by one, and the plurality of reception devices receive and transmit a radio signal reflected or not reflected by the target object. The arrival time from the device to the receiving device is converted from the speed of the radio signal to the distance, and the target object exists on an ellipse that focuses on the transmitting device and the receiving device. An object movement azimuth and velocity estimation device for an object detection system of a multi-static method that analytically calculates a position where a plurality of ellipses overlap by acquiring a position of a target object,
Based on the known position of the transmission device, the known position of the reception device, and the determined position of the target object, the movement direction of the target object is all set in a state where the posture and movement direction of the target object match. The estimated speed of the target object is calculated by calculating the Doppler shift amount, which is the frequency shift amount of the radio signal due to the Doppler phenomenon generated by the moving target object, discretely around the target object. moving azimuth and velocity estimation apparatus of an object, characterized in that it comprises a maximum likelihood velocity vector calculation unit for estimating integrates speed estimation value calculated from the candidate values for all of the velocity vector components, including false candidate values of the vector components.   The maximum likelihood velocity vector calculation unit votes a velocity estimation value in each direction at a point or region with respect to a voting space having a polar coordinate shape, and moves and velocity of the target object based on a vote distribution of the final number of votes. The apparatus for estimating the moving direction and speed of an object according to claim 1.   The maximum likelihood velocity vector calculation unit performs voting by changing weights so as to highly evaluate votes from a plurality of different receiving devices at the time of voting, and estimates a velocity estimation value of the target object. The moving azimuth and speed estimation apparatus of the object described.   4. The maximum likelihood velocity vector calculation unit, based on a predetermined physical constraint condition at the time of voting, performs voting by changing the weight of a vote, and estimates a velocity estimation value of the target object. Moving direction and speed estimation device for an object.   The maximum likelihood velocity vector calculation unit votes as an area having a voting range having a width based on the azimuth error of the receiving device or the measurement error of the frequency shift amount at the time of voting, and estimates the velocity estimation value of the target object The apparatus for estimating the moving direction and speed of an object according to any one of claims 2 to 4.   3. The maximum likelihood velocity vector computing unit limits a velocity or direction that the target object can take from a vote distribution in a voted polar coordinate voting space, and estimates a velocity estimation value of the target object. The moving direction and speed estimation device for an object according to any one of? A plurality of transmission devices and a plurality of reception devices are provided, each of the plurality of transmission devices transmits a predetermined radio signal one by one, and the plurality of reception devices receive and transmit a radio signal reflected or not reflected by the target object. The arrival time from the device to the receiving device is converted from the speed of the radio signal to the distance, and the target object exists on an ellipse that focuses on the transmitting device and the receiving device. A method for estimating the moving direction and speed of an object for a multi-static object detection system that analytically calculates a position where a plurality of ellipses overlap by acquiring a position of a target object,
Based on the known position of the transmitting device, the known position of the receiving device, and the determined position of the target object, the maximum likelihood velocity vector calculation unit Estimating the velocity of the target object by computing the Doppler shift amount, which is the frequency shift amount of the radio signal due to the Doppler phenomenon generated by the moving target object, assuming the moving direction of the target object discretely around the entire circumference Object moving azimuth and speed, including a step of estimating a value by integrating speed estimation values calculated from candidate values of all speed vector components including false candidate values of speed vector components of the target object Estimation method.   The maximum likelihood velocity vector calculation unit votes the estimated velocity value in each direction at a point or area with respect to a voting space having a polar coordinate shape, and the moving direction and velocity of the target object based on the vote distribution of the final vote number The method of estimating the moving direction and speed of an object according to claim 7, further comprising the step of estimating   The maximum likelihood velocity vector calculation unit further includes a step of voting with varying weights so as to highly evaluate votes from a plurality of different receiving devices at the time of voting, and estimating a velocity estimation value of the target object, The method for estimating the moving direction and speed of an object according to claim 8.   The maximum likelihood velocity vector calculation unit further includes a step of voting by changing the weight of a vote based on a predetermined physical constraint condition at the time of voting, and estimating a velocity estimation value of the target object. The moving direction and speed estimation method of an object according to 8 or 9.   The maximum likelihood velocity vector calculation unit votes as a region in which the voting range has a width based on the azimuth error of the receiving device or the measurement error of the frequency shift amount at the time of voting, and estimates the velocity estimation value of the target object The method according to any one of claims 8 to 10, further comprising a step.   The maximum likelihood velocity vector calculation unit further includes a step of limiting a velocity or azimuth that the target object can take from a vote distribution in the voted polar coordinate voting space, and estimating a velocity estimation value of the target object. The moving direction and speed estimation method for an object according to any one of claims 8 to 11.

Images