JP5853489B2 - Target motion estimation system and method - Google Patents

Description

  The present invention relates to a target motion estimation system and method for tracking and tracking a target detected from received radio waves, sound waves, light waves, etc., and estimating a motion parameter of the target, and in particular, estimating a target speed and a temporal change in the speed. The present invention relates to a target motion estimation system and method.

  As a method for detecting and tracking a target using a radar (Radar: Radio detecting and ranging), a sonar (Sonar: Sound navigation and ranging), a rider (Lida: Light detecting and ranging), etc., for example, described in Patent Document 1 A known MHT (Multiple Hypothesis Tracking) method is known. In addition, when the frequency of false detection of a target or detection omission is low, the target can be detected and tracked by using a well-known Kalman filter or particle filter.

  In any of the above methods, it is possible to estimate the speed of the target and the temporal change (acceleration, etc.) of the target from the target position obtained by a plurality of searches and the temporal change of the position. Furthermore, by using the target Doppler velocity obtained from the wave information such as the received radio wave, sound wave, and light wave as well as the target position, the target velocity and the time change of the velocity can be estimated with higher accuracy.

JP 2009-192550 A

  As described above, since the conventional method uses a time series filter such as a Kalman filter or a particle filter, multiple searches are required to obtain the target speed and the time change of the speed. For this reason, if the target speed or acceleration is large, the estimation accuracy of the target speed or temporal change of the speed may be lowered depending on the search interval.

  In addition, when the target Doppler velocity is used in the Kalman filter or particle filter processing, the linear approximation is used in the nonlinear conversion from the Doppler velocity to the velocity vector, so that the target velocity estimation accuracy may be greatly reduced. . The same applies to the MHT method using the Kalman filter.

  The present invention has been made in order to solve the problems of the background art as described above, and provides a target motion estimation system and method capable of acquiring a target speed and a temporal change in the speed for each detection. The purpose is to do.

In order to achieve the above object, a target motion estimation system according to the present invention includes a plurality of sensors, and estimates a target motion parameter based on waveform information indicating time-series fluctuations of waves generated from the target obtained by the sensors. A target motion estimation system,
Detecting means for detecting the target from the waveform information;
Doppler speed calculation means for calculating the target Doppler speed detected by the detection means;
Azimuth estimation means for estimating the azimuth of the target detected by the detection means;
Transmission time estimation means for estimating a time at which a wave was transmitted from the target detected by the detection means;
A velocity parameter vector composed of the target velocity component which is an unknown parameter and its time-varying component is generated, and the unknown parameter is obtained using Doppler velocity obtained from observation data of the sensor equal to or more than the number of the unknown parameters. Generate a Doppler vector having the same number of components as the number of parameters, and use the azimuth of the target obtained from the observation data of the sensor equal to or more than the number of unknown parameters and the wave transmission time from the target. An equation building means for generating an orientation matrix having the same number of components as the number of parameters, and constructing simultaneous equations indicating the velocity parameter vector by a multiplication result of the Doppler vector and the orientation matrix;
A speed parameter estimating means for obtaining the speed parameter vector by solving the simultaneous equations, and calculating a target speed component and a time change component of the speed from each vector component of the speed parameter vector;
A statistical parameter calculating means for calculating a speed obtained for each combination of the sensors and a statistical value of a temporal change in the speed, and outputting the statistical value as a target true speed component and a true speed temporal change component;
Have

Alternatively, a target motion estimation system that includes a plurality of sensors and estimates motion parameters of the target based on waveform information indicating time-series fluctuations of waves emitted from the target obtained by the sensors,
Detecting means for detecting the target from the waveform information;
Doppler speed calculation means for calculating the target Doppler speed detected by the detection means;
Azimuth estimation means for estimating the azimuth of the target detected by the detection means;
Transmission time estimation means for estimating a time at which a wave was transmitted from the target detected by the detection means;
An optimal speed parameter vector composed of the target speed component which is an unknown parameter and its time-varying component is generated, and the unknown parameter is obtained using Doppler speed obtained from observation data of the sensor equal to or more than the number of the unknown parameters. Generating a Doppler moment vector having the same number of components as the number of parameters, and using the direction of the target obtained from the observation data of the sensor equal to or more than the number of unknown parameters and the wave transmission time from the target An equation construction means for generating an azimuth moment matrix having the same number of components as the number of unknown parameters, and constructing simultaneous equations indicating the optimum velocity parameter vector by the multiplication result of the Doppler moment vector and the azimuth moment matrix;
An optimum speed parameter estimating means for obtaining the optimum speed parameter vector by solving the simultaneous equations, calculating a target speed component and a time change component of the speed from each vector component of the optimum speed parameter vector; and
Have

On the other hand, the target motion estimation method of the present invention includes a plurality of sensors, and a target for estimating motion parameters of the target based on waveform information indicating time-series fluctuations of waves emitted from the target obtained by the sensors. A motion estimation method,
Computer
Detecting the target from the waveform information;
Calculating the detected Doppler velocity of the target;
Estimating the orientation of the detected target;
Estimating the time when a wave was transmitted from the detected target;
Generating a velocity parameter vector comprising the target velocity component which is an unknown parameter and its time-varying component;
Generate a Doppler vector having the same number of components as the number of unknown parameters using Doppler velocity obtained from observation data of the sensor equal to or more than the number of unknown parameters,
Generate an orientation matrix having the same number of components as the number of unknown parameters using the target orientation obtained from observation data of the sensor equal to or greater than the number of unknown parameters and the wave transmission time from the target,
Construct a simultaneous equation indicating the velocity parameter vector by the multiplication result of the Doppler vector and the orientation matrix,
The speed parameter vector is obtained by solving the simultaneous equations, a target speed component and a time change component of the speed are calculated from each vector component of the speed parameter vector,
This is a method of calculating a speed obtained for each combination of the sensors and a statistical value of the temporal change of the speed, and outputting the statistical value as a target true speed component and a true speed temporal change component.

Alternatively, a target motion estimation method comprising a plurality of sensors and for estimating a motion parameter of the target based on waveform information indicating time-series fluctuations of a wave emitted from the target obtained by the sensor,
Detecting the target from the waveform information;
Calculating the detected Doppler velocity of the target;
Estimating the orientation of the detected target;
Estimating the time when a wave was transmitted from the detected target;
Generating an optimal speed parameter vector composed of the target speed component which is an unknown parameter and its time-varying component;
Generate a Doppler moment vector having the same number of components as the number of unknown parameters using Doppler velocity obtained from observation data of the sensor equal to or more than the number of unknown parameters,
A azimuth moment matrix having the same number of components as the number of unknown parameters is generated using the azimuth of the target obtained from observation data of the sensor equal to or more than the number of unknown parameters and the time of wave transmission from the target. ,
Constructing simultaneous equations indicating the optimum velocity parameter vector by the multiplication result of the Doppler moment vector and the azimuth moment matrix;
In this method, the optimum speed parameter vector is obtained by solving the simultaneous equations, and a target speed component and a time change component of the speed are calculated and output from each vector component of the optimum speed parameter vector.

  According to the present invention, it is possible to acquire a target speed and a temporal change in the speed for each search.

It is a block diagram which shows the example of 1 structure of the target motion estimation system of 1st Embodiment. It is a block diagram which shows the example of 1 structure of the target motion estimation system of 2nd Embodiment. It is a block diagram which shows the example of 1 structure of the target motion estimation system of 3rd Embodiment. It is a schematic diagram which shows the example of a definition of the target direction vector in three-dimensional space. It is a schematic diagram which shows the example of a definition of the target velocity vector in three-dimensional space. It is a schematic diagram which shows the example of a definition of the target acceleration vector in three-dimensional space. It is a schematic diagram which shows the example of a definition of the target velocity vector and acceleration vector in a two-dimensional plane coordinate. It is a schematic diagram which shows the example of a definition of the target velocity vector and acceleration vector in a two-dimensional spherical coordinate.

Next, the present invention will be described with reference to the drawings.
(First embodiment)
FIG. 1 is a block diagram illustrating a configuration example of a target motion estimation system according to the first embodiment.

  As shown in FIG. 1, the target motion estimation system 1 according to the first embodiment includes a detection unit 11, a Doppler speed calculation unit 12, an azimuth estimation unit 13, a transmission time estimation unit 14, an equation construction unit 15, a speed parameter estimation. Means 16 and statistical parameter calculation means 17 are provided.

  The target motion estimation system 1 shown in FIG. 1 may be realized by a D / A converter, an A / D converter, an LSI including various logic circuits, or the like, and is provided with a CPU that executes processing according to a program. It can also be realized by a processing device (computer). Observation data such as radio waves, sound waves and light waves observed by the sensor data input means 4 is input to the target motion estimation system 1. The observation data is input to the detection means 11 provided in the target motion estimation system 1. Further, the target speed estimated by the target motion estimation system 1 and the time change of the speed are output to the recording display means 5 and displayed on the recording display means 5.

  The sensor data input means 4 is a radar, a sonar, a rider or the like provided with one or a plurality of sensors that receives incoming radio waves, sound waves, light waves, etc. and converts them into electrical signals. The observation data of the sensor data input means 4 includes waveform information that is data in which the wavelength, amplitude, phase, modulation method, etc. of radio waves, sound waves, light waves, etc. emitted from the target fluctuate in time series.

  In the sensor data input means 4 having a plurality of sensors, in general, the plurality of sensors are often arranged on a straight line at half-wavelength intervals (half-wavelength intervals of detection wavelengths). In the present embodiment, the sensor included in the sensor data input unit 4 may be arranged in any shape such as a ring shape, a spherical shape, or a three-dimensional lattice shape similar to the crystal lattice shape. Moreover, the sensitivity characteristic and wavelength characteristic of each sensor may be the same or different. In the present embodiment, a multi-static configuration is assumed in which there is one wave transmission source such as a radio wave, a sound wave, and a light wave and a plurality of sensors that receive the wave. As long as interference between the transmission waves can be avoided by using the above, a plurality of wave transmission sources may be used.

  The detection means 11 performs target detection processing for detecting a target to be tracked on the observation data input from the sensor data input means 4, and the detection results are used as Doppler velocity calculation means 12 and direction estimation means 13. And the transmission time estimation means 14 respectively. The number of targets detected by the target detection process may be 0, 1 or a plurality of cases. The detection unit 11 may directly perform target detection processing on the observation data input from the sensor data input unit 4, and converts the observation data input from the sensor data input unit 4 with a predetermined orthogonal function (for example, Fourier After the conversion, the target detection process may be performed.

  A well-known target detection method may be used for the target detection process performed by the detection unit 11. As the simplest method, for example, there is a method in which a wave transmitted from a sensor is reflected by a target and the target is detected by receiving the reflected wave (echo). In that case, there is a method of using the information (frequency or wavelength) of the transmission wave as a template to obtain a correlation with the reception wave and determining that the target is the target when the correlation coefficient is larger than a preset threshold value. In addition, information on the wave transmitted from the target in various noise environments is registered in advance, and the registered wave information is used as a template to correlate with the received wave, or a known neural network, support Create a wave template transmitted from the target by learning using a machine learning algorithm such as Vector Machine or Adaboost, and determine whether the target exists by pattern recognition of the received wave based on the template You may use the method to do. The target detection process may be performed every time observation data is input from the sensor data input means 4, or may be periodically performed at a preset timing, or may be performed by an output signal of another external sensor or an operator. It may be implemented in response to an instruction from. For example, when the target detection process is performed from the observation data of the active sonar, the target detection process is not performed because the sound wave (noise) reflected from an object at a very short distance is large for a certain period from the time of transmission of the sound wave.

  The Doppler velocity calculation means 12 measures the frequency or wavelength of the wave transmitted from the target detected by the detection means 11 or the wave reflected from the target (hereinafter collectively referred to as a transmitted wave) and stored in advance. The reference signal is compared with the frequency or wavelength, and the target Doppler velocity is calculated from the displacement using the Doppler effect formula.

  As a method for obtaining the frequency of the target transmission wave, for example, there is a method in which the observation data is subjected to Fourier transform, and the frequency of the peak value larger than a preset threshold value is set as the frequency of the target transmission wave. The reference signal has a known frequency and wavelength when transmitting a radio wave, a sound wave, or the like from a sensor like an active radar or an active sonar. On the other hand, in the case of a passive radar, a passive sonar, or the like, the reference signal may be prepared in advance with a database of frequency components of the transmitted wave corresponding to the target characteristics, such as ship engine sound.

  The direction estimating means 13 estimates the target direction detected by the detecting means 11. As a method for estimating the target azimuth, for example, a reception beam is formed in various azimuths and waveform information for each azimuth is recorded, and an azimuth having a peak value larger than a preset threshold value is set as a true target. The direction may be used. The target orientation may be estimated using a known method such as the MUSIC (MUltiple SIgnal Classification) method.

  In the present embodiment, the vertical direction is referred to as the vertical direction or the vertical direction, and the horizontal direction is referred to as the horizontal direction or the horizontal direction. There are various other ways to call the vertical direction, such as angle of attack and depression. There are various other ways of calling the horizontal direction, such as an azimuth angle and a horizontal angle. In a radar or sonar, a required function may be sufficiently achieved by scanning radio waves or sound waves only in the vertical direction or the horizontal direction. In such radar and sonar, the direction of arrival of radio waves and sound waves is sometimes simply referred to as azimuth or target azimuth.

  The transmission time estimation means 14 estimates the time when the wave was transmitted from the target. In an active radar or active sonar, the distance between each sensor and the target can be determined from the positional relationship between the sensor and the target, the wave transmission time from the sensor, the wave reception time by the sensor, and the wave propagation speed. Can be estimated at the time of reflection at the target. On the other hand, in passive radar and passive sonar, for example, it is assumed that a true target exists at a position where the target orientations detected by a plurality of sensors intersect, and based on the distance to the true target and the wave propagation speed, The time when a wave is transmitted can be estimated.

  The equation constructing means 15 constructs simultaneous equations for obtaining a target velocity component and its time-varying component.

  First, the processing of the equation construction means 15 will be described by taking as an example the case of obtaining a target velocity component and its time change component in a three-dimensional space. In this case, the target velocity component and its time change component are each represented by three parameters. In other words, since the unknown parameter is 6, the equation construction means 15 selects the waveform information of at least 6 sensors out of N sensors (N is 6 or more) arranged in the three-dimensional space, and sets the target parameters. The velocity component and its time change component are obtained. Note that there is one target to be detected, and the acceleration of the target is constant. When a plurality of targets are detected, the same process may be executed for each detected target.

  In the following, a coordinate system is considered in which the x-axis and y-axis are set on the horizontal plane with the sensor position as the origin, and the direction perpendicular to the horizontal plane is the z-axis (upward is positive).

  At this time, the target azimuth vector viewed from the sensor n is defined as shown in FIG. 4a, the target velocity vector is defined as shown in FIG. 4b, and the target acceleration vector is defined as shown in FIG. 4c. That is, the target azimuth vector viewed from the sensor n is expressed by using the target angle φn with respect to the x-axis and the target angle θn with respect to the z-axis. The target velocity vector is expressed by using a target velocity magnitude v, an angle ξ with respect to the x axis, and an angle η with respect to the z axis. The target acceleration vector is expressed by using a target acceleration magnitude a, an angle σ with respect to the x-axis, and an angle ρ with respect to the z-axis.

  Here, when the target Doppler velocity obtained from the waveform information of the sensor n (n = 1 to 6) is Vn, the target Doppler vector G has Vn as a Doppler vector component.

で表される６次元ベクトルとなる。

This is a 6-dimensional vector represented by

  Further, a speed parameter vector X composed of a target speed component which is an unknown parameter and its time-varying component is given by

の６次元ベクトルで表される。なお、式（２）の中央のベクトルは、式の見通しをよくするために各ベクトル成分を簡略化して示したものである。

This is represented by a 6-dimensional vector. Note that the vector in the center of equation (2) is a simplified illustration of each vector component in order to improve the visibility of the equation.

Of the waves that reach the sensor n, the time of the wave transmitted from the target (transmission time) is t _n ,

と定義すると、未知のパラメータ数と同数の成分をから成る目標の方位行列Ｆは

The target orientation matrix F consisting of the same number of components as the number of unknown parameters is

の６×６行列で表される。

This is represented by a 6 × 6 matrix.

  Therefore, from the Doppler vector G, the velocity parameter vector X, and the orientation matrix F, the simultaneous equations to be solved are

となる。上述したように、ドップラーベクトルＧ及び方位行列Ｆは各センサの波形情報から得られる値であり、未知のパラメータはＸである。したがって、方位行列Ｆが逆行列を持つなら、速度パラメータベクトルＸは

It becomes. As described above, the Doppler vector G and the orientation matrix F are values obtained from the waveform information of each sensor, and the unknown parameter is X. Therefore, if the orientation matrix F has an inverse matrix, the velocity parameter vector X is

で求めることができる。なお、方程式構築手段１５は、式（５）で示した方程式を構築するまでの処理を実施し、式（６）で示した方程式を解くのは後段の速度パラメータ推定手段１６である。

Can be obtained. The equation construction means 15 performs processing until the equation shown in the equation (5) is constructed, and the latter-stage speed parameter estimation means 16 solves the equation shown in the equation (6).

  By the way, when obtaining a time differential component of a higher rank speed, such as a target speed time differential (acceleration) or acceleration time differential, it can be solved using the same method as described above. However, since the time derivative of each speed is represented by a three-dimensional vector, the number of unknown parameters increases by three each time the rank of the time derivative increases by one, and the number of simultaneous equations to be solved also increases by three. Become.

  For example, when obtaining a vector of m-th order time derivative with respect to speed, the Doppler vector G is

で表される３（ｍ＋１）次元ベクトルとなる。

Is a 3 (m + 1) -dimensional vector.

Assuming that the magnitude of the m-th time derivative of the velocity vector is a _m , the angle of the m-th derivative of the velocity vector with respect to the z-axis is η _m , and the angle of the m-th derivative of the velocity vector with respect to the x-axis is ξ _m. , The speed parameter vector X is

の３（ｍ＋１）次元ベクトルで表される。なお、式（８）の中央のベクトルは、式の見通しをよくするために各ベクトル成分を簡略化して示したものである。

Of 3 (m + 1) -dimensional vectors. Note that the vector in the center of equation (8) is a simplified illustration of each vector component in order to improve the visibility of the equation.

Of the waves that have reached sensor n, the time of the wave that is transmitted from the target (transmission time) is defined as t _n, and the mth-order time derivative of the velocity vector is

と定義すると、目標の方位行列Ｆは

The target orientation matrix F is

の３（ｍ＋１）×３（ｍ＋１）の行列で表される。

3 (m + 1) × 3 (m + 1) matrix.

  Therefore, from the Doppler vector G, the velocity parameter vector X, and the orientation matrix F, the simultaneous equations to be solved are

となる。ここで、ドップラーベクトルＧ及び方位行列Ｆは各センサの波形情報から得られる値であり、未知のパラメータはＸである。したがって、方位行列Ｆが逆行列を持つなら、速度パラメータベクトルＸは

It becomes. Here, the Doppler vector G and the orientation matrix F are values obtained from the waveform information of each sensor, and the unknown parameter is X. Therefore, if the orientation matrix F has an inverse matrix, the velocity parameter vector X is

で求めることができる。なお、方程式構築手段１５は、式（１１）で示した方程式を構築するまでの処理を実施し、式（１２）で示した方程式を解くのは後段の速度パラメータ推定手段１６である。

Can be obtained. The equation construction means 15 performs processing until the equation shown in Expression (11) is constructed, and the speed parameter estimation means 16 in the subsequent stage solves the equation shown in Expression (12).

  Next, an example in which the target speed and acceleration are obtained by scanning radio waves, sound waves, light waves, etc. in the horizontal direction, which is an important direction for many radars and sonars. Here, the case where there is one target to be detected and the acceleration of the target is constant will be described as an example.

  When scanning only in the horizontal direction, a target in a two-dimensional space is detected, so that the target velocity component and its time-varying component are each represented by two parameters. That is, since the unknown parameter is 4, the equation constructing unit 15 selects the waveform information of at least four sensors among N sensors (N is 4 or more) arranged in the two-dimensional space, and the simultaneous equations. Build up. Note that the positive direction of the x-axis is north, the positive direction of the y-axis is east, and the target velocity vector and acceleration vector viewed from the sensor n are defined as shown in FIG. That is, the target angle (azimuth) viewed from the sensor n with respect to the y-axis is φn, and the target velocity vector is expressed by using the target velocity magnitude v and the angle (azimuth) ξ with respect to the y-axis. The target acceleration vector is expressed by using a target acceleration magnitude a and an angle (azimuth) η with respect to the y-axis. The x-axis means a wave scanning direction (horizontal direction), that is, an axis in a predetermined direction, and the y-axis means a direction perpendicular to the designated axis.

At this time, the target Doppler velocity obtained from the sensor _n is defined as t _{n when} the wave transmission time from the target among the waves reaching the sensor _n is defined as t _n .

であり、ドップラーベクトルＧは

And the Doppler vector G is

で表される４次元ベクトルとなる。

The four-dimensional vector represented by

  The speed parameter vector X is

の４次元ベクトルで表される。なお、式（１５）の中央のベクトルは、式の見通しをよくするために各ベクトル成分を簡略化して示したものである。

It is expressed by a four-dimensional vector. Note that the vector in the center of equation (15) is a simplified illustration of each vector component in order to improve the visibility of the equation.

  Also,

と定義すると、目標の方位行列Ｆは

The target orientation matrix F is

の４×４行列で表される。

Of 4 × 4 matrix.

  From the above Doppler vector G, velocity parameter vector X and orientation matrix F, the simultaneous equations to be solved are

となる。ここで、ドップラーベクトルＧ及び方位行列Ｆは各センサの波形情報から得られる値であり、未知のパラメータはＸである。したがって、方位行列Ｆが逆行列を持つなら、速度パラメータベクトルＸは

It becomes. Here, the Doppler vector G and the orientation matrix F are values obtained from the waveform information of each sensor, and the unknown parameter is X. Therefore, if the orientation matrix F has an inverse matrix, the velocity parameter vector X is

で求めることができる。なお、方程式構築手段１５は、式（１８）で示した方程式を構築するまでの処理を実施し、式（１９）で示した方程式を解くのは後段の速度パラメータ推定手段１６である。

Can be obtained. The equation constructing means 15 performs processing until the equation shown in the equation (18) is constructed, and the latter-stage speed parameter estimating means 16 solves the equation shown in the equation (19).

  Thus, in the case of a two-dimensional space, attention must be paid to the definition of the angle. However, the target speed and the time change of the speed can be obtained by essentially the same method as in the case of the three-dimensional space. The same applies to the case where the target speed is time-differentiated or the time is differentiated with a higher rank.

  Next, an example will be shown in which radio waves, sound waves, light waves, etc. are scanned only in the horizontal direction in spherical coordinates, which is an important orientation used for target detection by many radars and sonars, to obtain the target speed and acceleration. Here, the case where there is one target to be detected and the acceleration of the target is constant will be described as an example.

  Since this case also becomes a problem in a two-dimensional space, in order to obtain a target speed and its time change, waveform information obtained from four sensors among N sensors (N is 4 or more) is selected and set simultaneously. An equation should be constructed. Note that the target velocity vector and acceleration vector viewed from the sensor n are defined as shown in FIG.

In FIG. 6, a is a meridian passing through the sensor n, and c is a meridian passing through the target. b is a great circle connecting the target and the sensor n, not a latitude line. φ _n is the orientation of the target viewed from the sensor, and φ _n ′ is the angle between the great circle connecting the sensor n and the target viewed from the target and the meridian.

The angle between the meridian passing through the target and the great circle connecting the target and the sensor n is θ _n , the angle between the meridian a and the meridian c is η, the angle between the target velocity vector and the meridian c is ξ, the acceleration vector Is the angle formed by the meridian c. θ _n can be easily obtained by using a trigonometric formula in a spherical triangle if the target coordinates and sensor coordinates are known.

In this case, if the target Doppler velocity V _n obtained from the sensor n (n = 1 to 4) is defined as t _n , the transmission time of the wave transmitted from the target among the waves reaching the sensor _n is defined as:

となる。

It becomes.

  The equation (20) has the same structure as the equation (13), and the target speed and the change with time of the target can be obtained by constructing simultaneous equations as in the equations (14) to (19).

  The speed parameter estimation means 16 obtains a speed parameter vector X by solving the simultaneous equations constructed by the equation construction means 15 and calculates a target speed vector and a time-varying component of the speed vector from each component of the speed parameter vector X. . The simultaneous equations may be solved using a well-known Gaussian elimination method, LU decomposition method, or the like.

  The statistical parameter calculation means 17 obtains a statistical value such as a target velocity vector obtained from all sensor combinations and an average value or median value of its time-varying components, and calculates the statistical value as a target true velocity vector and a true value. Let it be the time-varying component of the velocity vector. Then, the target true speed vector and the time change component of the true speed vector are output to the external recording display means 5.

  The recording display means 5 is realized by a computer provided with a display, for example.

According to this embodiment, a velocity parameter vector composed of a target velocity component that is an unknown parameter and its time-varying component is generated, and a Doppler velocity obtained from observation data of sensors equal to or more than the number of unknown parameters is used. Generating a Doppler vector having the same number of components as the number of unknown parameters, and further using the direction of the target obtained from the sensor observation data equal to or greater than the number of unknown parameters and the time of wave transmission from the target. An azimuth matrix having the same number of components as the number of parameters is generated, and simultaneous equations indicating the velocity parameter vector as a result of multiplication of the Doppler vector and the azimuth matrix are constructed. Then, by solving the simultaneous equations, it is possible to obtain a target speed component and its time change component which are unknown parameters. Therefore, the target speed and the time change of the speed can be acquired for each search.
(Second Embodiment)
FIG. 2 is a block diagram illustrating a configuration example of the target motion estimation system according to the second embodiment.

  As shown in FIG. 2, the target motion estimation system 2 according to the second embodiment includes a detection unit 11, a Doppler velocity calculation unit 12, an azimuth estimation unit 13, a transmission time estimation unit 14, an optimum equation construction unit 25, and an optimum velocity. Parameter estimation means 26 is provided. Observation data measured by the external sensor data input means 4 is input to the detection means 11. The parameter calculated by the optimum speed parameter estimation means 26 is output to the external recording display means 5.

  The sensor data input means 4, the detection means 11, the Doppler speed calculation means 12, the azimuth estimation means 13, the transmission time estimation means 14 and the recording display means 5 are the same as those in the first embodiment, and thus description thereof is omitted. .

  As in the first embodiment, the target motion estimation system 2 shown in FIG. 2 may be realized by a D / A converter, an A / D converter, an LSI including various logic circuits, etc. Therefore, it can be realized by an information processing apparatus (computer) including a CPU for executing processing.

  Hereinafter, the processes of the optimum equation constructing means 25 and the optimum speed parameter estimating means 26 shown in FIG. 2 will be described.

  In the present embodiment, an example is shown in which the speed for each target in the three-dimensional space with the highest applicability and the m-th time differential of the speed are obtained. Therefore, the required number N of sensor waveform information is 3 (m + 1) or more. The method for setting coordinates, velocity vectors, and the like is the same as in the first embodiment.

  Optimal equation construction means 25 generates a Doppler moment vector having a Doppler moment calculated based on the Doppler speed, transmission time, and target orientation obtained from the waveform information of each sensor as a vector component for each target.

  The target Doppler velocity obtained from the waveform information of sensor n is Vn, the wave transmission time from the target that has reached sensor n is tn, and the m-th time differential of the velocity vector is the same as in equation (9). In

と定義すると、ドップラーモーメントベクトルＧは

The Doppler moment vector G is

で表される３（ｍ＋１）＋１次元ベクトルとなる。

3 (m + 1) +1 dimensional vector expressed by

  The optimum equation constructing means 25 obtains the target speed magnitude, the target speed longitudinal and lateral directions, the target speed time-varying component magnitude, and the target obtained from the waveform information of each sensor. An optimum speed parameter vector having the vertical direction azimuth and the horizontal direction azimuth of the time-varying component of the speed as vector components is generated. Alternatively, the magnitude of the target speed, the direction of the speed with respect to the direction perpendicular to the pre-specified axis, the magnitude of the time-varying component of the speed, and the direction of the time-varying component of the speed with respect to the direction perpendicular to the pre-specified axis To generate an optimum speed parameter vector.

  The optimum speed parameter vector X has one term added to the above equation (8),

の３（ｍ＋１）＋１次元ベクトルで表される。なお、式（２３）の中央のベクトルは、式の見通しをよくするために各ベクトル成分を簡略化して示したものである。また、Ｄは計算の便宜上設けた変数である。

Of 3 (m + 1) +1 dimensional vectors. Note that the vector in the center of Expression (23) is a simplified illustration of each vector component in order to improve the visibility of the expression. D is a variable provided for convenience of calculation.

  The optimum equation constructing means 25 generates an azimuth moment matrix having the target azimuth as a component calculated from the wave transmission time from the target obtained from the waveform information for each sensor and the target azimuth. For the azimuth moment matrix F, using the notation shown in equation (21),

の３（ｍ＋１）＋１×３（ｍ＋１）＋１行列で表される。

3 (m + 1) + 1 × 3 (m + 1) +1 matrix.

  At this time, the simultaneous equations to be solved are

となる。ドップラーモーメントベクトルＧ及び方位モーメント行列Ｆは各センサの波形情報から得られる値であり、未知のパラメータはＸである。

It becomes. The Doppler moment vector G and the azimuth moment matrix F are values obtained from the waveform information of each sensor, and the unknown parameter is X.

  Note that the simultaneous equations shown in equation (25) are

で定義される尤度関数Ｌにおいて、

In the likelihood function L defined by

として得られる最適解である。式（２６）に示す尤度関数は一例であり、測定値と理論値との差が最小となるような尤度関数であれば他の式を用いてもよい。

Is the optimal solution obtained as The likelihood function shown in Expression (26) is an example, and other expressions may be used as long as the likelihood function is such that the difference between the measured value and the theoretical value is minimized.

  The optimum speed parameter estimation means 26 obtains an optimum speed parameter vector that is an unknown parameter from the Doppler moment vector and the azimuth moment matrix that can be calculated from the waveform information of the sensor data input means 4 in the simultaneous equations, and each optimum speed parameter vector A target speed vector and a time-varying component of the speed vector are calculated from the components.

  Similar to the speed parameter estimation means 16 of the first embodiment, the optimum speed parameter estimation means 26 uses a known method such as a Gaussian elimination method or an LU decomposition method to obtain a target speed vector or time of the speed vector. What is necessary is just to calculate a change component. The optimum speed parameter estimation means 26 outputs the obtained speed vector and the time change component of the speed vector to the external recording display means 5.

According to the present embodiment, the optimum equation construction means 25 generates the Doppler moment vector G and the azimuth moment matrix F, and the optimum speed parameter estimation means 26 obtains the optimum speed parameter vector from the Doppler moment vector G and the azimuth moment matrix F. Even if there is no statistical parameter calculation means 17 as in the first embodiment, a value corresponding to the target speed component obtained from the combination of the sensors, the average value of the time-varying component, and the like can be obtained. Therefore, with a simpler configuration than the target motion estimation system 1 of the first embodiment, it is possible to acquire the target speed and the time change of the speed for each search.
(Third embodiment)
FIG. 3 is a block diagram illustrating a configuration example of the target motion estimation system according to the third embodiment.

  As shown in FIG. 3, the target motion estimation system 3 of the third embodiment has a configuration in which a defective state detection means 38 is added to the target motion estimation system 1 of the first embodiment shown in FIG. The defective state detection means 38 shown in FIG. 3 may be added to the target motion estimation system 2 of the second embodiment shown in FIG.

  Similar to the first and second embodiments, the target motion estimation system 3 shown in FIG. 3 may be realized by a D / A converter, an A / D converter, an LSI including various logic circuits, or the like. Alternatively, it may be realized by an information processing apparatus (computer) including a CPU that executes processing according to a program.

  The defective state detection means 38 acquires the position of each sensor from the sensor data input means 4, and when the plurality of sensors and the target are aligned on the same straight line, or when the plurality of sensors and the target are in the same plane. Then, the equation construction means 15 or the optimum equation construction means 25 is made to construct simultaneous equations excluding the waveform information of the sensor that is in a defective state.

  For example, the defective state detection unit 38 may connect each sensor and the target with straight lines, and determine whether or not the defective state is based on whether these straight lines are parallel or not. Whether each straight line is parallel or not is determined in advance by determining an allowable range of angles between the straight lines, and when the angle is within the range, the corresponding straight lines may be regarded as parallel.

  According to the present embodiment, in addition to the same effects as those of the first or second embodiment, the sensors that are arranged on the same straight line as the target by the defective state detection unit 38 or in the same plane with respect to the target Excluding the waveform information of a certain sensor, the equation construction means 15 or the optimum equation construction means 25 constructs the simultaneous equations, so that the target speed and the estimation accuracy of the time change of the speed are improved.

1, 2, 3 Target motion estimation system 4 Sensor data input means 5 Record display means 11 Detection means 12 Doppler speed calculation means 13 Direction estimation means 14 Transmission time estimation means 15 Equation construction means 16 Speed parameter estimation means 17 Statistical parameter calculation means 25 Optimal equation construction means 26 Optimal speed parameter estimation means 38 Defective state detection means

Claims (6)

A target motion estimation system that includes a plurality of sensors and estimates a motion parameter of a target based on waveform information indicating time-series fluctuations of a wave emitted from a target obtained by the sensor,
Detecting means for detecting the target from the waveform information;
Doppler speed calculation means for calculating the target Doppler speed detected by the detection means;
Azimuth estimation means for estimating the azimuth of the target detected by the detection means;
Transmission time estimation means for estimating a time at which a wave was transmitted from the target detected by the detection means;
A velocity parameter vector composed of the target velocity component which is an unknown parameter and its time-varying component is generated, and the unknown parameter is obtained using Doppler velocity obtained from observation data of the sensor equal to or more than the number of the unknown parameters. Generate a Doppler vector having the same number of components as the number of parameters, and use the azimuth of the target obtained from the observation data of the sensor equal to or more than the number of unknown parameters and the wave transmission time from the target. An equation building means for generating an orientation matrix having the same number of components as the number of parameters, and constructing simultaneous equations indicating the velocity parameter vector by a multiplication result of the Doppler vector and the orientation matrix;
A speed parameter estimating means for obtaining the speed parameter vector by solving the simultaneous equations, and calculating a target speed component and a time change component of the speed from each vector component of the speed parameter vector;
A statistical parameter calculating means for calculating a speed obtained for each combination of the sensors and a statistical value of a temporal change in the speed, and outputting the statistical value as a target true speed component and a true speed temporal change component;
A target motion estimation system. A target motion estimation system that includes a plurality of sensors and estimates a motion parameter of a target based on waveform information indicating time-series fluctuations of a wave emitted from a target obtained by the sensor,
Detecting means for detecting the target from the waveform information;
Doppler speed calculation means for calculating the target Doppler speed detected by the detection means;
Azimuth estimation means for estimating the azimuth of the target detected by the detection means;
Transmission time estimation means for estimating a time at which a wave was transmitted from the target detected by the detection means;
An optimal speed parameter vector composed of the target speed component which is an unknown parameter and its time-varying component is generated, and the unknown parameter is obtained using Doppler speed obtained from observation data of the sensor equal to or more than the number of the unknown parameters. Generating a Doppler moment vector having the same number of components as the number of parameters, and using the direction of the target obtained from the observation data of the sensor equal to or more than the number of unknown parameters and the wave transmission time from the target An equation construction means for generating an azimuth moment matrix having the same number of components as the number of unknown parameters, and constructing simultaneous equations indicating the optimum velocity parameter vector by the multiplication result of the Doppler moment vector and the azimuth moment matrix;
An optimum speed parameter estimating means for obtaining the optimum speed parameter vector by solving the simultaneous equations, calculating a target speed component and a time change component of the speed from each vector component of the optimum speed parameter vector; and
A target motion estimation system.   A failure state detecting means for acquiring the position of the sensor and constructing the simultaneous equations except for waveform information obtained from a sensor arranged on the same straight line as the target or a sensor in the same plane with respect to the target; The target motion estimation system according to claim 1 or 2. A target motion estimation method for estimating a motion parameter of a target based on waveform information including a plurality of sensors and indicating time-series fluctuations of a wave emitted from the target obtained by the sensor,
Computer
Detecting the target from the waveform information;
Calculating the detected Doppler velocity of the target;
Estimating the orientation of the detected target;
Estimating the time when a wave was transmitted from the detected target;
Generating a velocity parameter vector comprising the target velocity component which is an unknown parameter and its time-varying component;
Generate a Doppler vector having the same number of components as the number of unknown parameters using Doppler velocity obtained from observation data of the sensor equal to or more than the number of unknown parameters,
Generate an orientation matrix having the same number of components as the number of unknown parameters using the target orientation obtained from observation data of the sensor equal to or greater than the number of unknown parameters and the wave transmission time from the target,
Construct a simultaneous equation indicating the velocity parameter vector by the multiplication result of the Doppler vector and the orientation matrix,
The speed parameter vector is obtained by solving the simultaneous equations, a target speed component and a time change component of the speed are calculated from each vector component of the speed parameter vector,
A target motion estimation method of calculating a speed obtained for each combination of the sensors and a statistical value of a temporal change in the speed, and outputting the statistical value as a true speed component of the target and a temporal change component of the true speed. A target motion estimation method for estimating a motion parameter of a target based on waveform information including a plurality of sensors and indicating time-series fluctuations of a wave emitted from the target obtained by the sensor,
Detecting the target from the waveform information;
Calculating the detected Doppler velocity of the target;
Estimating the orientation of the detected target;
Estimating the time when a wave was transmitted from the detected target;
Generating an optimal speed parameter vector composed of the target speed component which is an unknown parameter and its time-varying component;
Generate a Doppler moment vector having the same number of components as the number of unknown parameters using Doppler velocity obtained from observation data of the sensor equal to or more than the number of unknown parameters,
A azimuth moment matrix having the same number of components as the number of unknown parameters is generated using the azimuth of the target obtained from observation data of the sensor equal to or more than the number of unknown parameters and the time of wave transmission from the target. ,
Constructing simultaneous equations indicating the optimum velocity parameter vector by the multiplication result of the Doppler moment vector and the azimuth moment matrix;
A target motion estimation method for obtaining the optimum speed parameter vector by solving the simultaneous equations, calculating a target speed component and a time-varying component of the speed from each vector component of the optimum speed parameter vector, and outputting them.   The position of the sensor is acquired, and the simultaneous equations are constructed by excluding waveform information obtained from a sensor arranged on the same straight line as the target or a sensor in the same plane with respect to the target. The target motion estimation method described.

Images