JP2003149330A - Tracking device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a tracking device easily receiving a target signal and preventing an unnecessary signal from intruding. SOLUTION: This device is provided with an item calculating means 2 calculating a characteristic value and a characteristic vector in directions of distance, elevation angle and azimuth using detection data from an observation means 1 and an observation error covariance matrix from a single sampling delay means 8, a gate deciding means 3 carrying out a one-dimensional gate deciding in a range direction for the detection data, a gate deciding means 4 carrying out the one-dimensional gate deciding in a cross range direction, a gate deciding means 5 outputting a residual covariance matrix attached to the detection data corresponding to candidates of target signals from the gate deciding result, a first data updating means 6 updating the data based on the theorem of Kalman filter, a prediction means 7 calculating a predicted error covariance matrix a single sampling after the current time and a predictive vector, the single sampling delay means 8 delaying them by a single sampling, and a displaying means 9 displaying a target track and a target speed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tracking device,
In particular, the present invention relates to a target tracking device used for target tracking in a high density environment.

[0002]

2. Description of the Related Art A conventional tracking device will be described below with reference to the drawings. FIG. 45 shows, for example, "Samuel S. Blackm
an, Multiple-Target Tracking with Radar Applicatio
ns, Artech House, Dedham, 1986 ”p. 83-p. 9
2, especially p. 88, 4.2.2 is a block diagram showing a configuration of a conventional target tracking device having a gate determination method, which is shown in an equation (4.6).

In FIG. 45, 101 is observation means and 10
6 is a data updating means, 107 is a predicting means, 108 is a 1-sampling delaying means, 109 is a displaying means, 123 is a gate determination specification calculating means, and 124 is a gate determining means.

FIG. 2 is a diagram for explaining the coordinate system relating to the observing means 101 in FIG. In FIG.
O is a sensor, T is a tracking target, R is a tracking target T and sensor O
, E is a line segment O connecting the sensor O and the tracking target T.
The elevation angle T forms with the XY plane, B is the sensor O and the tracking target T
The orthogonal projection vector of the line segment OT connecting to and to the XY plane is X.
The azimuth angle with the axis. Further, [R, E, B] represent polar coordinates, and [X, Y, Z] represent north reference Cartesian coordinates. From now on, unless otherwise specified, the coordinates will be the north reference Cartesian coordinates.

In the observing means 101, the detection data and the S / N ratio which is the signal-to-noise ratio (noise) are detected.
The ratio is obtained from the sensor, and the S / N ratio is converted into the observation noise standard deviation of the distance, the observation noise standard deviation of the elevation angle, and the observation noise standard deviation of the azimuth angle. Input the observation noise standard deviation, elevation angle observation noise standard deviation, and azimuth angle observation noise standard deviation.

The types of detection data generally include a target signal of a tracking target and an unnecessary signal other than the target signal.

The detection data is represented by a symbol like D _kj , and the S / N ratio is represented by a symbol like SNR _kj . Here, the subscript k represents the sampling time, and the subscript j represents the number of detection data at the sampling time k.

The detection data D _kj is expressed by the following equation (1).

[0009]

[Equation 1]

In the above equation (1), o is a subscript indicating an observed value, k is a subscript indicating a sampling time, and j
Is a subscript indicating the number of detection data. That is, the detection data D _kj represents the j-th detection data at the sampling time k. Furthermore, x _Okj represents the X coordinate component of the detection data D _kj, likewise y _Okj is, Y coordinate components of the detection data D _kj, z _okj represent respectively the Z coordinate component of the detection data D _kj. The symbol "'" represents the transposition of vectors and matrices.

The observed noise standard deviation is obtained from the following equations (2), (3) and (4) using the S / N ratio SNR _kj .

[0012]

[Equation 2]

[0013]

[Equation 3]

[0014]

[Equation 4]

Σ in the above equations (2), (3) and (4)
r _kj , σ e _kj , and σ b _kj represent the observation noise standard deviation of the distance, the observation noise standard deviation of the elevation angle, and the observation noise standard deviation of the azimuth angle, respectively. Further, α _r , α _e , and α _b in the above equations (2), (3), and (4) are the frequencies of the radar transmission signal,
It is a positive constant determined from the bandwidth of the radar transmission signal, the effective aperture diameter of the radar antenna, and the speed of light.

In the gate determination parameter calculation means 123, the detection data, the observation noise standard deviation of the distance, the observation noise standard deviation of the elevation angle and the observation noise standard deviation of the azimuth angle are input from the observation means 101, and the prediction described later is made. The prediction vector and the prediction error covariance matrix are input from the means 107 via the one-sampling delay means 108. Furthermore, the gate determination specification calculation means 123 calculates the residual and the residual covariance matrix.

The residual and residual covariance matrix will be described below.

Hereinafter, the residual covariance matrix is represented as S _kj (t). The subscript k of S _kj (t) represents the sampling time, and j
From the th detection data, (t)
Represents the target number of the tracking target. The residual covariance matrix is calculated by the following equation (5) based on the Kalman filter theory.

[0019]

[Equation 5]

In equation (5), H is the observation matrix and P is
_pk (t) is the prediction error covariance matrix R _{kj at the} tracking target t obtained from the prediction means 107 described later represents the observation error covariance matrix at the north reference Cartesian coordinates.

The observation matrix H in the equation (5) is expressed by the following equation (6).
Can be expressed as

[0022]

[Equation 6]

The observation error covariance matrix R _kj in the north reference Cartesian coordinate in the equation (6) can be expressed by the following equation (7).

[0024]

[Equation 7]

In equation (7), Λ _kj is the observation error covariance matrix in polar coordinates, and Γ _kj (t) is the transformation matrix of Λ _kj for each tracking target.

In equation (7), the observation error covariance matrix Λ _kj in polar coordinates can be expressed by equation (8).

[0027]

[Equation 8]

The symbol " ² (or ^ in the equation (8)
2) ”represents the square.

The transformation matrix Γ _kj (t) of Λ _kj for each tracking target in the equation (7) can be expressed by the following equation (9).

[0030]

[Equation 9]

R _pk (t), e in equation (9)
_pk (t) and b _pk (t) are the prediction vector VX, respectively.
position component x _pk of _{pk (t) (t),} y pk (t), z
It is represented by the polar coordinate component of _pk (t).

R _pk (t), e _pk (t), b _pk (t) and x
The relationship between _pk (t), y _pk (t), and z _pk (t) is expressed by the following equation (1)
It is represented by 0).

[0033]

[Equation 10]

The residual is represented by the difference D _kj -VZ _pk (t) between the detection data and the position component of the prediction vector.

Therefore, the gate judgment specification calculation means 12
In 3, the residual, the residual covariance matrix, the detection data, the prediction vector, and the prediction error covariance matrix are input to the gate determination means 124 described later.

To explain the gate determination means 124, the gate will be described. The gate is the area of the error ellipse expressed by the following equation (11).

[0037]

[Equation 11]

In equation (11), VZ _pk (t) represents the position component of the prediction vector. In equation (11), d _ε
Is a gate size parameter that determines the size of the gate. Here, the gate size parameter is a constant, and the subscript ε of the gate size parameter represents the probability that the target exists in the gate. The probability that the target exists in the gate will be hereinafter referred to as the in-gate probability or the in-gate target existence expected probability. Here, the gate size parameter d _ε is
It represents the size of the gate when the probability is ε within the gate. The gate size parameter d _ε is obtained from the chi-square density function or the chi-square distribution table by predetermining the degree of freedom and the in-gate probability ε that one wants to set by the chi-square distribution in statistics. Further, the upper subscript “−1” in the equation (11) is a symbol representing the inverse matrix of the matrix.

The prediction vector VX _pk (t) is calculated by the following equation (1)
It is expressed as in 2).

[0040]

[Equation 12]

In equation (12), x _pk (t), y
_pk (t), z _pk (t) is the predicted position of X, Y, Z, dx
_pk (t), dypk (t), dzpk (t) are X,
It is the predicted speed of Y and Z.

Predicted vector position component VZ _pk (t) in equation (11) and predicted vector VX in equation (12)
_pk (t) is calculated by the following equation (1) using the observation matrix H of equation (6).
There is a relationship of 3).

[0043]

[Equation 13]

The size of the gate expressed by the equation (11) is
The gate size parameter d _ε and the residual covariance matrix S _kj
Determined by (t). Next, the reason why the gate size is determined by the gate size parameter d _ε and the residual covariance matrix S _kj (t) will be described below with an example. In order to simplify the description below, matrices and vectors are assumed as in the following equations (14) and (15). Then, (1
The equation (1) is determined as the following equation (16) from the following equations (14) and (15).

[0045]

[Equation 14]

[0046]

[Equation 15]

[0047]

[Equation 16]

Since the matrix A in the equation (14) is a positive-value symmetric matrix of 3 rows and 3 columns, the following equations (17) and (18)
Is established.

[0049]

[Equation 17]

[0050]

[Equation 18]

Here, the matrix A is expressed by the following equation (19).
It is assumed that the diagonal matrix B and the orthogonal matrix C can be diagonalized by algebra.

[0052]

[Formula 19]

The diagonal matrix B in equation (19) is given by the following equation (20): eigenvalues 1 / λ ₁ , 1 / λ ₂ ,
It is assumed to be represented by 1 / λ ₃ . Here, from the relationship of equation (18), the eigenvalues 1 / λ ₁ , 1 / λ ₂ , and 1 / λ ₃ are 1
/ Λ ₁ > 0, 1 / λ ₂ > 0, and 1 / λ ₃ > 0. Here, “/” is a symbol indicating division.

[0054]

[Equation 20]

The orthogonal matrix B in the equation (19) is obtained by using the eigenvectors v1 underbar, v2 underbar and v3 underbar of the matrix A as shown in the following equation (21).
It is expressed as in equation (21). Here, v1 underbar, v2 underbar and v3 underbar are each 3 lines 1
It is a vector of columns.

[0056]

[Equation 21]

The eigenvectors v1 underbar, v2 underbar and v3 underbar of the matrix A in the equation (21) have the relationship of the equation (22).

[0058]

[Equation 22]

Expression (20) is obtained by using Expression (19) and
It is expressed as in equation 3).

[0060]

[Equation 23]

Here, when the component of Cx in the equation (23) is set as the following equation (24), the equation (23) becomes the equation (2)
0) and equation (24) are used to express as equation (25).

[0062]

[Equation 24]

[0063]

[Equation 25]

From equations (14) and (20), the line
Column A eigenvalue is 1 / λ₁, 1 / λ₂, 1 / λ₃So
Mathematically, the residual covariance matrix S_kjThe eigenvalue of (t) is λ
₁, Λ ₂, Λ₃Is.

Therefore, the equation (25) represents that the radius of the error ellipse is determined by the eigenvalues λ ₁ , λ ₂ , λ _{3 of} the residual covariance matrix and the gate size parameter d _ε . That is, the size of the gate is determined by the residual covariance matrix and the gate size parameter d _ε .

Therefore, the larger the residual covariance matrix S _kj (t) and the smaller the inverse of the residual covariance matrix, the larger the gate size. Also,
The larger the gate size parameter, the larger the size of the gate.

Position component VZ of the prediction vector of equation (11)
_pk (t) represents the center of the gate. The left side of equation (11) is
The residual of the detection data D _kj and the position component VZ _pk (t) of the prediction vector that is the center of the gate is calculated as the residual covariance matrix S.
_It represents the distance normalized by _kj (t).

Therefore, in the gate determination means 124,
Detection data D_kjAnd the position of the prediction vector at the center of the gate
Table component VZ_pkResidual D of (t)_kj-VZ_pk(T) and the balance
Difference covariance matrix S_kj(T) is the gate determination specification calculation means 12
Does the input from 3 satisfy the formula (11) that is the gate?
The detection data D that satisfies the formula (11)
Let kj be a candidate for the target signal. Also, satisfy (11)
Detection data D_kjIf there are multiple
One piece of close detection data is the candidate aD of the target signal _kjAs
Output. Further, the target signal candidate is aD_kjCorresponding to
The residual covariance matrix_kj(T). That is,
The detection data aD in the gate is sent to the data updating means 106._kj
And aD_kjResidual covariance matrix aS corresponding to_kj(T), pre
Measurement error covariance matrix, prediction vector, position of prediction vector
Ingredient VZ_pkEnter (t). If the detection device
Data cannot be entered, that is, if it is a memory track
Is the in-gate detection data aD_kjIs the position of the prediction vector
As an ingredient, aD_kjResidual covariance matrix aS corresponding to
_kj(T) is calculated using the S / N set in advance.
Nau.

The memory track is a process of updating data without detecting data in the gate.

In the data updating means 106, the Kalman
Gain matrix K based on Ruta's theory _k(T) is given by the following equation (2
Calculated from 6). Also, the smooth vector VX_sk(T)
Gain matrix K of the following equation (26)_kUsing (t), (2
It is calculated by the equation 7). Furthermore, the smooth error covariance matrix P
_sk(T) is the gain matrix K of equation (26)_kUsing (t)
And calculated by the equation (28). In addition, data update
In the step 106, it is calculated by the equations (27) and (28).
Smooth vector VX_sk(T) and smooth error covariance matrix
P_sk(T) is output.

[0071]

[Equation 26]

[0072]

[Equation 27]

[0073]

[Equation 28]

In the predicting means 107, the prediction vector VXp at time k + 1, which is one sampling after the current time k, is calculated by the following equation (29) based on the Kalman filter theory.
k + 1 (t) is calculated and output. Also, the following equation (30)
Thus, the prediction error covariance matrix P _pk (t) at time k + 1 after one sampling from the current time k is calculated and output.

[0075]

[Equation 29]

[0076]

[Equation 30]

In the equation (29), the state transition matrix Φk for which one sampling extrapolation is performed is represented by the equation (31).

[0078]

[Equation 31]

Δt in the equation (31) represents the current time k and the sampling interval from the current time to the time k + 1 after one sampling.

The driving noise covariance matrix Qk in the equation (30) is expressed by the equation (32).

[0081]

[Equation 32]

Σ _{a in} equation (32) represents the standard deviation of drive noise. The standard deviation of drive noise is a constant that represents the ambiguity of the Kalman filter target motion model.

The one-sampling delay means 108 delays the prediction vector VX _pk (t) at the sampling time k and the prediction error covariance matrix P _pk (t) input from the prediction means 107 by one sampling, and the display means 109. To enter.

In the display means 109, the data updating means 10
Using the smoothing vector input from 6, the position components of the smoothing vectors for several past samplings are connected by a line and displayed as a track. Further, using the velocity component of the smooth vector, the velocity is displayed such that the position component of the smooth vector is the starting point and the end point is the target predicted position for the next sampling.

FIG. 4 is a schematic diagram showing the configuration of a conventional tracking device.
6 shows.

In the tracking device as described above, it is difficult to adjust the size of the gate. Hereinafter, in order to show the reason, FIG.
This will be explained using 7.

FIG. 47 is a diagram showing the relationship between the gate and the target predicted position, the target observation position, the target range direction and the target cross range direction.

In FIG. 47, G1 represents the gate at time t1. Further, G1 ′ represents a gate for which a gate size parameter larger than that of G1 at time t1 is set. Here, the gates G1 and G1 ′ have the same gate shape. Further, O1 is the target observation position at time t1, P1 is the target predicted position at time t1,
TGT1 is the target true position at time t1, C1, 1-C
Reference numerals 1 and 9 represent unnecessary signals. Further, O2 is the target observation position at time t2, P2 is the target predicted position at time t2, TGT2 is the target true position at time t2, C1, 1
C1 and C1 represent unnecessary signals. The target is to make a straight turn from TGT1 to TGT2.

In FIG. 47, the direction connecting the sensors S and R1 or the direction connecting the sensors S and R2 is called the target range direction. The axis orthogonal to the target range direction is called the target cross range direction.

For example, in FIG. 47, first, FIG.
When the gate G2 at the target t2 is opened like the ellipse of the solid line, the target observation position D2 is out of the range direction. Therefore, a gate size parameter that is larger than the gate size parameter that configures the gate G2 is set, and the gate G2 ′ is configured like a dashed ellipse so that the target observation position D2 is entered. Then, although the target observation position D2 which is out of the range direction can be supplemented by the gate G2 ', the gate G2' becomes large while keeping the same shape as the gate G2, so that the gate radius in the cross range direction becomes large. I will end up.

Here, generally, as shown in FIG. 48, the cross range error caused by the angle error becomes larger as the distance between the sensor and the target becomes larger than the range error. For example, if the target distance is on the order of hundreds of kilometers, the ratio of range error to cross range error is 1: 1000 to 1
The ratio is about 0000. Therefore, when the gate size parameter is increased, the gate radius in the cross range direction becomes considerably larger than the gate radius in the range direction.

Therefore, as in the example shown in FIG. 47, in the case of the gate G2 having a small gate size (solid line in the figure), even if no unnecessary signal is input, the gate G2 'having a large gate size (see FIG. There is a problem that many unnecessary signals are included in the broken line).

[0093]

The conventional tracking device is configured as described above, and the cross-range error caused by the angle error becomes larger as the sensor-target distance becomes longer than the range error. Therefore, when the gate size parameter is increased, the gate radius in the cross-range direction becomes much larger than the gate radius in the range direction, and even in the case where unnecessary signals do not enter in the case of a small-sized gate. , Gate G with large gate size
There is a problem that many unnecessary signals are included in 2 '.

The present invention has been made in order to solve such a problem, and it is an object of the present invention to obtain a tracking device capable of easily obtaining a target signal and performing gate determination for making it difficult for an unwanted signal to enter. To aim.

[0095]

The present invention obtains an S / N ratio which is a signal-to-noise ratio on the basis of received power and converts it into an observation noise standard deviation of distance, elevation angle and azimuth angle, Observation means for outputting the observation noise standard deviation and the detection data representing the position associated with them, the prediction error covariance matrix and the detection data is input, from the residual covariance matrix calculated from them, in the distance direction First gate determination parameter calculation means for calculating an eigenvalue, an eigenvalue in the elevation direction, an eigenvalue in the azimuth direction, an eigenvector in the distance direction, an eigenvector in the elevation direction and an eigenvector in the azimuth direction, and the detection data is 1 in the distance direction. It is determined whether or not the dimension gate is present, and a range direction gate determination means for outputting a signal indicating the determination result, the observing means, and the first gate determination parameter calculator. Using the data output from the cross range direction, the cross range direction gate determination means for performing a gate determination of a predetermined dimension and the determination result of whether or not the gate of the predetermined dimension in the distance direction is entered are shown. From the signal, the signal indicating the result of determination as to whether or not the gate is in the predetermined dimension in the elevation direction, and the signal indicating the result of determination as to whether or not the gate is in the predetermined dimension in the azimuth direction. , When the detection data is contained in all the gates of the predetermined dimension in the elevation direction and the azimuth direction, the detection data and the residual covariance matrix associated with the detection data are output as candidates of the target signal. Based on the theory of Kalman filter, the first gate determination means, the detection data and the residual covariance matrix output from the first gate determination means are input. Based on the theory of the Kalman filter and a first data updating unit that updates the data and outputs a smoothing error covariance matrix and a smoothing vector, and outputs a prediction error covariance matrix and a prediction vector one sampling after the current time. Predicting means for calculating, a prediction error covariance matrix and a prediction vector input from the predicting means are delayed by one sampling, and a prediction error covariance matrix is output to the first gate determination parameter calculating means 1 The tracking device comprises a sampling delay means and a display means for displaying a target track and a target speed from a smooth vector input from the first data updating means.

Further, the gate determining means in the cross range direction determines whether or not the detection data is in the one-dimensional gate in the elevation direction, outputs a signal indicating the determination result, and the detection data indicates the azimuth angle. It is comprised of a first cross-range direction gate determining means for determining whether or not the gate is in the one-dimensional gate in the direction and outputting a signal indicating the determination result.

Further, the gate determining means in the cross-range direction has an eigenvector in the elevation angle direction, an eigenvalue in the elevation angle direction,
It is composed of a second cross-range direction gate determination means for performing two-dimensional gate determination in the cross-range direction using the azimuth-direction eigenvector and the azimuth-direction eigenvalue.

Further, there is further provided first driving noise control means for setting a large standard deviation of the driving noise vector when the velocity component of the smoothing vector exceeds a certain threshold value.

Further, based on the logistic curve with the magnitude of the standard deviation of the driving noise vector as the vertical axis and the velocity component of the smooth vector as the horizontal axis, the velocity component of the smooth vector becomes larger than a certain threshold value. In this case, it further comprises second drive noise control means for setting a large standard deviation of the drive noise vector.

Further, according to the present invention, an S / N ratio which is a signal-to-noise ratio is obtained on the basis of the received power, and the S / N ratio is converted into an observation noise standard deviation of a distance, an elevation angle and an azimuth angle to obtain the observation noise standard. The eigenvalues in the distance direction are obtained from the observation means that outputs the deviation and the detection data representing the position associated with them, the prediction error covariance matrix and the observation error covariance matrix, and the residual covariance matrix calculated from them. , An eigenvalue in the elevation direction, an eigenvalue in the azimuth direction, an eigenvector in the distance direction, an eigenvector in the elevation direction and an eigenvector in the azimuth direction, and the detection data is one-dimensional in the distance direction. It is determined whether or not the gate is in the gate, and the range direction gate determination means for outputting a signal indicating the determination result, and the detection data are in the elevation direction one-dimensional gate. It is determined whether or not it is determined, a signal indicating the determination result is output, it is determined whether the detection data is in a one-dimensional gate in the azimuth direction, and a signal indicating the determination result is output. The gate determination means in the cross range direction, the gate determination means in the range direction, and the gate determination means in the first cross range direction, as a result of the gate determination of the detection data, the detection data in the gate are detected as the gate center and Based on the theory of the Kalman filter, the second data updating means for converting the data distance into the likelihood of a normal distribution, and using the likelihood, and the prediction error after one sampling from the current time The prediction means for calculating the variance matrix and the prediction vector and the prediction error covariance matrix and the prediction vector input from the prediction means are delayed by one sampling. And a display for displaying the target track and the target speed from the 1-sampling delay means output to the first gate determination specification calculation means and the smooth vector input from the first data update means. And a tracking device.

Further, there is further provided first gate size parameter setting means for setting the gate size parameter according to the sampling interval.

Further, there is further provided an observation error fixing means for setting the observation noise standard deviation of distance, the observation noise standard deviation of elevation angle and the observation noise standard deviation of azimuth angle to fixed values.

Further, it further comprises sampling interval fixing means for inputting a control signal for fixing the sampling interval to the predicting means and fixing the sampling interval when calculating the prediction error covariance matrix and the prediction vector.

Further, gate center fixing means for fixing the center of the gate for several samplings from the initial stage of tracking is further provided.

Further, there is further provided a gate size parameter varying means for changing the magnitude of the gate size parameter by using the magnitude of the velocity component of the smooth vector obtained from the data updating means.

Further, a first display smoothing means for calculating the display smoothing vector by inputting the smoothing vector obtained from the data updating means to the display filter having a fixed time constant is further provided.

Further, a second display smoothing means for calculating the display smoothing vector by inputting the smoothing vector obtained from the data updating means to a display filter having a variable time constant is further provided.

Further, in the observing means, when the S / N ratio which is the signal-to-noise ratio is obtained and the S / N ratio is larger than a certain threshold value, it is judged as multipath, and when it is judged as multipath, The control signal for controlling the residual covariance matrix to be calculated by using a fixed value that is preset for the observation noise standard deviation of the elevation angle used when calculating the residual covariance matrix is the first gate judgment. The multipath determination means which outputs to a specification calculation means is further provided.

Further, using the condition number that is the ratio of the maximum value of the eigenvalue and the minimum value of the eigenvalue among the eigenvalues input from the first gate judgment parameter calculation means, and if the condition number is below a certain threshold value, Further, there is further provided condition number determination means for inputting to the display means a danger signal indicating that the tracking calculation has become unstable.

[0110]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiment 1. A tracking device according to Embodiment 1 of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a tracking device according to an embodiment of the present invention. In each of the following embodiments, the same reference numerals in the drawings represent the same or corresponding parts.

In FIG. 1, 1 is an S / N ratio which is a signal-to-noise ratio based on the received power obtained from the sensor,
It is an observing means that converts it into observation noise standard deviations of distance, elevation angle, and azimuth angle, and outputs the observation noise standard deviation and detection data representing the position associated with them. 2, the prediction error covariance matrix is input from the 1-sampling delay means 8 and the observation error covariance matrix is input from the observing means 1 and the residual covariance matrix calculated from them is used to calculate the eigenvalues in the distance direction and the elevation angle direction. Eigenvalue, eigenvalue in azimuth direction,
It is a first gate determination specification calculation means for calculating an eigenvector in the distance direction, an eigenvector in the elevation angle direction, and an eigenvector in the azimuth angle direction. Reference numeral 3 is a range-direction gate determination means that determines whether or not the detection data is in a one-dimensional gate in the distance direction and outputs a signal indicating the determination result. Reference numeral 4 determines whether or not the detection data is in a one-dimensional gate in the elevation direction, outputs a signal indicating the determination result, and determines whether or not the detection data is in a one-dimensional gate in the azimuth direction. It is a first cross-range direction gate determination means that performs determination and outputs a signal indicating the determination result. Reference numeral 5 denotes a signal indicating whether or not the vehicle has entered the one-dimensional gate in the distance direction, a signal indicating whether or not the vehicle has entered the one-dimensional gate in the elevation direction, and 1 for the azimuth direction. Based on the signal indicating the result of determination as to whether or not the object has entered the three-dimensional gate, if the detection data is included in all one-dimensional gates in the distance direction, the elevation angle direction, and the azimuth angle direction, it is determined as a candidate for the target signal. Outputting the detection data and residual covariance matrix associated with the detection data,
Is a gate determination means. Reference numeral 6 is a first data updating means for updating the data based on the Kalman filter theory and outputting the smoothing error covariance matrix and the smoothing vector. Reference numeral 7 is a prediction means for calculating a prediction error covariance matrix and a prediction vector after one sampling from the current time, based on the theory of Kalman filter. Reference numeral 8 is a one-sampling delay means that delays the prediction error covariance matrix and the prediction vector input from the prediction means 7 by one sampling and outputs the delayed one to the first gate determination parameter calculation means 2. Reference numeral 9 is a display means for displaying the target track and the target speed from the smooth vector input from the first data updating means. Note that 6, 7, 8, and 9 are 106, 107, 108, and 109 of FIG. 45 described above, respectively.
The detailed description is omitted here.

The first gate judgment specification calculation means 2 will be described below.

In the first gate judgment parameter calculation means 2, the detection data, the observation noise standard deviation of the distance,
The elevation observation noise standard deviation and the azimuth observation noise standard deviation are input, and the prediction vector and the prediction error covariance matrix are input from the prediction unit 7 via the one sampling delay unit 8.

Further, the first gate judgment specification calculation means 2
Then, the eigenvector in the range direction and the eigenvector in the cross range direction are calculated, and the calculation method will be described below.

FIG. 49 shows the relationship between the eigenvectors in the range direction and the eigenvectors in the cross range direction.

The three-dimensional elliptical region in FIG. 49 is a gate.
is there. In FIG. 49, A is the predicted position and B is in the gate
Represents observation positions and their position vector is VZ
_pk(T), D _kjAnd In addition, in the distance direction (OA direction)
The unit vector is Ur, the unit vector in the z direction is Uel,
Cross-range lateral direction (HI direction) orthogonal to the distance in the horizontal plane
The unit vector of (direction) is Uby. Furthermore, inside the gate
Let OG, CD, and EF orthogonal to each other be x1, x2, and x3 axes
It x1, x2, and x3 are x in the equation (25).₁,
x₂, X₃It corresponds to. x1, x2, x3 axes
The direction is the eigenvector v in equation (21).₁Underscore
ー 、 v₂Underbar, v₃Each underbar
Stipulated.

Hereinafter, the eigenvector v ₁ underbar, v ₂
Underbar, v ₃ Underbar to distance direction eigenvector v _r Underbar, elevation direction eigenvector v _el ,
A method of assigning an eigenvector v _{by in the} azimuth direction will be described. Here, the distance direction has the same meaning as the range direction, the elevation direction has the same meaning as the cross range vertical direction, and the azimuth direction has the same meaning as the cross range horizontal direction.

Of the eigenvectors v ₁ underbar, v ₂ underbar, and v ₃ underbar, the eigenvector having the largest inner product with the unit U _r in the distance direction is defined as the eigenvector in the distance direction. The eigenvalue corresponding to the eigenvector is λ _r . That is, it is determined by the following equation (33).

[0119]

[Expression 33]

In equation (33), max is the inner product U _r
-It works to select the largest one among v ₁ , U _r · v ₂ and U _r · v ₃ .

However, U _r is defined by the following equation (34).

[0122]

[Equation 34]

Here, the symbol "||" in the equation (34) is used.
Represents the Euclidean norm in algebra.

Next, the eigenvectors v ₁ underbars excluding those selected as the eigenvectors in the distance direction,
The remaining two eigenvectors of v ₂ underbar and v ₃ underbar are v ₁ ′ and v ₂ ′, respectively. v1'v
Among 2 ', the eigenvector having the largest inner product with the unit vector U _el in the elevation direction is referred to as the eigenvector in the elevation direction and is referred to as v _el . Also, its eigenvalue is λ _el . Here, v _el is determined by the equation (35). The unit vector U _el in the elevation direction is defined by the equation (36).

[0125]

[Equation 35]

[0126]

[Equation 36]

Of the eigenvectors v ₁ underbar, v ₂ underbar, and v ₃ underbar, the remaining one eigenvector excluding the two selected as the eigenvector in the distance direction and the eigenvector in the elevation direction is called the eigenvector in the azimuth direction. And v _by . Also, its eigenvalue is λ _by
And

By the above processing, the eigenvector in the distance direction, the eigenvector in the elevation direction, and the eigenvector in the azimuth angle direction are determined. That is, the eigenvector in the range direction is
The eigenvectors in the distance direction and the cross range direction are the eigenvectors in the elevation angle direction and the eigenvectors in the azimuth angle direction.
Therefore, the residuals of the eigenvectors in the distance direction, the eigenvectors in the azimuth direction, and the eigenvectors in the elevation direction in the respective directions are represented _by v _r · x, v _{e l} · x, and v _by · x, respectively.

Therefore, the first gate determination parameter calculation means 2 uses the detection data obtained from the observation means 1 and the prediction vector obtained from the prediction means 7 via the 1-sampling delay means 8 to calculate the residual ( The residual covariance matrix is calculated by the equation (16). Furthermore, the eigenvalue λ _{r in} the distance direction and the residual v _r of the eigenvector in the distance direction
Input x, the prediction vector, the prediction error covariance matrix, the detection data, the residual and the residual covariance matrix to the gate determination means 3 in the range direction. Similarly, the eigenvalue λ in the elevation direction
el, eigenvector residual v _el x, azimuth eigenvalue λ _by , azimuth eigenvector residual v
_Byx , the prediction vector, the prediction error covariance matrix, the detection data, the residual and the residual covariance matrix are input to the first cross-range direction gate determination means 4.

Next, the range direction gate determination means 3 will be described.

In the range direction gate determination means 3, (3
Using expression 7), it is determined whether or not the detection data is included in the one-dimensional gate in the distance direction. Where (37)
D _r on the right side of the equation is the gate size parameter to be set in advance by a distance direction.

[0132]

[Equation 37]

Therefore, in the range direction gate determination means 3, the determination signal indicating whether or not the distance gate of the equation (37) is entered, the prediction vector, the prediction error covariance matrix, the detection data, the residual error and the residual error The variance matrix is input to the first gate determination means 5.

Next, the gate determining means 4 in the first cross range direction will be described.

The first cross-range direction gate determination means 4 uses the following equation (38) to determine whether or not the detection data is in the one-dimensional gate in the elevation direction. Here, (38) d _el of the right side of the equation is the gate size parameter to be set in advance by the elevation direction.

[0136]

[Equation 38]

Further, the first cross-range direction gate determination means 4 determines whether or not the detection data is in the one-dimensional gate in the azimuth direction by using the equation (39). Here, d _{by on} the right side of the equation (39) is a gate size parameter preset in the azimuth direction.

[0138]

[Formula 39]

Therefore, in the first cross range direction gate determination means 4, a determination signal indicating whether or not the elevation angle gate of formula (38) is entered, and whether or not the azimuth angle gate of formula (39) is entered. The decision signal, the prediction vector, the prediction error covariance matrix, the detection data, the residual and the residual covariance matrix indicating

In the first gate determination means 5, the determination signal, the prediction vector, the prediction error covariance matrix and the detection data indicating whether or not the distance gate of the formula (37) is entered from the range direction gate determination means 3. Is entered. Further, in the first cross-range direction gate determination means 4, a determination signal from the cross-range direction gate determination means 4 indicating whether or not the vehicle is in the elevation angle gate of the formula (38), and the azimuth angle of the formula (39). A determination signal indicating whether or not the gate is entered, a prediction vector, a prediction error covariance matrix, and detection data are input.

In the first gate judgment means 5, a judgment signal indicating whether or not the distance gate of the formula (37) is entered, (3
In the case where all of the determination signal indicating whether the gate is in the elevation angle gate of the formula 8) and the determination signal indicating whether the gate is in the azimuth angle gate of the formula (39) are signals input in the gate, Detection data D _kj corresponding to these determination signals
Is a candidate for the target signal. Further, a determination signal indicating whether or not the distance gate of the formula (37) is entered, a determination signal indicating whether or not the elevation gate of the formula (38) is entered, (39)
Judgment signal indicating whether or not the azimuth gate of the formula is entered,
In the case where there are a plurality of signals that all are in the gate, one piece of detection data closest to the prediction vector is output as the target signal candidate aD _kj . It also outputs the residual covariance matrix aS _kj (t) corresponding to the detection data aD _kj . Further, the position component of the prediction vector is calculated by the equation (13).

Therefore, the first gate determination means 5
To the first data updating means 6, the target signal candidates aD _kj in the detection data and the residual covariance matrix aS corresponding to aD _kj are detected.
_{Input kj} (t), the position component of the prediction vector, the prediction error covariance matrix, the prediction vector, the residual and the residual covariance matrix.

Therefore, in the conventional tracking device, (1
Since the gate determination process is performed using the equation (1), as shown in FIG. 47, if the gate size parameter is increased in order to supplement the detection data in the range direction, the entire gate becomes large, and in particular, the cross In the range direction,
Since the gate is expanded, there is a problem that many unwanted signals are picked up. However, equation (37), equation (38),
By separately performing the gate determination in the range direction and the cross range direction as in the expression (39), the gate size can be set independently in the range direction and the cross range direction. Therefore, even if you want to expand the gate only in the range direction, it does not affect the cross range direction, so it is possible to set the gate optimally in the sense that the target signal is captured and unnecessary signals are not captured as much as possible. is there.

Next, FIG. 3 and FIG. 4 show the first embodiment.
3 is a flowchart illustrating the operation of the above.

First, in step ST1, S / N and detection data obtained from the sensor are input.

Next, in step ST2, a residual and a residual covariance matrix are calculated.

Next, in step ST3, eigenvalues and eigenvectors are calculated from the residual covariance matrix.

Next, in step ST4, the inner product of the three eigenvectors calculated from the residual covariance matrix and the unit vector in the range direction is calculated.

Next, in step ST5, the one having the largest inner product of the unit vector in the range direction and the eigenvector is determined as the eigenvector in the range direction. The eigenvector in the range direction is synonymous with the eigenvector in the distance direction. Of the three eigenvectors, the two eigenvectors in the cross-range direction excluding the eigenvectors in the range direction are input to step ST6.

Next, in step ST6, the inner products of the two eigenvectors in the cross-range direction and the unit vector in the z-axis direction are calculated and input to step ST7.

Next, in step ST7, of the two eigenvectors in the cross-range direction, the one having the largest inner product of the unit vector and the eigenvector in the z-axis direction is determined as the eigenvector in the cross-range vertical direction, and the remaining one is determined. The eigenvector is determined as the eigenvector in the cross-range lateral direction. The eigenvector in the vertical direction of the cross range is synonymous with the eigenvector in the elevation direction. Similarly, the eigenvector in the cross-range lateral direction is synonymous with the eigenvector in the azimuth direction.

Next, in step ST8, residuals in the range direction, the cross range vertical direction, and the cross range horizontal direction are calculated from the eigenvectors in the range direction and the cross range direction.

Next, in step ST9, the gate size parameters in the range direction, the cross range vertical direction, and the cross range horizontal direction are set.

Next, in step ST10, (3
When Expressions 7), 38, and 39 are satisfied, that is, when all the gates in the range direction, the cross range vertical direction, and the cross range horizontal direction are satisfied, step ST11.
At, the detection data is determined to be in-gate detection data. Further, if the expressions (37), (38), and (39) are not satisfied, that is, if even one of the range direction, the cross range vertical direction, and the cross range horizontal direction does not satisfy the gate, step ST12. At, the detection data is determined to be out-of-gate detection data.

Next, in step ST13, the number of all detection data Nos times at the sampling time k,
The counter n counting the processing from ST1 to ST12 is n =
If Nobs, the process proceeds to step ST14. if,
If n = Nobs is not satisfied, the processes of ST1 to ST12 are continued.

Next, in step ST14, one of the in-gate detection data closest to the center of the gate is selected, and the selected in-gate detection data and the residual covariance matrix associated therewith are calculated as follows. Input in step ST15.

Next, in step ST15, the data is updated based on the theory of the Kalman filter, and the smooth vector and the smooth error covariance matrix are calculated. Further, using the smoothing vector and the smoothing error covariance matrix at the current time, one sampling extrapolation is performed, and the prediction vector and the prediction error covariance matrix after one sampling from the current time are calculated.

Next, in ST16, the smooth position and the smooth velocity are displayed using the smooth vector at the current time.

Next, in step ST17, when the sampling time k becomes larger than the sampling end time kend, the processing is interrupted. If the sampling time k is less than or equal to the sampling end time kend, the process continues.

As described above, in the target tracking device according to the present embodiment, it is possible to perform the gate determination for making it easy for the target signal to enter and for preventing the unwanted signal from entering.

Embodiment 2. A tracking device according to Embodiment 2 of the present invention will be described with reference to the drawings. Figure 5
FIG. 6 is a block diagram showing a configuration of a tracking device according to Embodiment 2 of the present invention. In FIG. 5, reference numeral 10 is a gate determination unit in the second cross range direction. The same reference numerals in the drawings represent the same or corresponding parts.

The gate determining means 10 in the second cross range direction will be described below. In the gate determination means 10 in the second cross range direction, the first gate determination specification calculation means 2
To elevation eigenvalue λ _el , elevation eigenvector residual v _el · x, azimuth eigenvalue λ _by , azimuth eigenvector residual v _by · x, prediction vector, prediction error covariance matrix, detection data , Residuals and residual covariance matrix are input.
Furthermore, two-dimensional gate determination is performed using the following equation (40).

[0163]

[Formula 40]

In the equation (40), the gate size parameter d _elby is set in advance.

Therefore, in the gate determining means 10 in the second cross range direction, the determination signal indicating whether or not the gate is in the gate of the equation (40), the prediction vector, the prediction error covariance matrix, the detection data, the residual and First set the residual covariance matrix
It is input to the gate determination means 5.

In the first gate judgment means 5, the judgment signal indicating whether or not the distance gate of the formula (37) is entered, the prediction vector, the prediction error and the covariance matrix are input from the range direction gate judgment means 3. To be done. Further, from the gate determining means 10 in the second cross range direction, a determination signal indicating whether or not the distance gate of the formula (40) is entered, a prediction vector, a prediction error, and a covariance matrix are input. (3
A determination signal indicating whether or not the gate of the equation (7) is entered,
In the case where all of the judgment signals indicating whether the gate is in the equation (40) are in the gate, the detection data Dkj corresponding to these judgment signals are set as the target signal candidates. Further, in the case where there are a plurality of signals indicating that all of the determination signals indicating whether the distance gate of the formula (37) has entered the gate and the determination signal indicating whether the gate of the formula (40) have entered the gate With respect to the plurality of detection data of, the one detection data closest to the prediction vector is output as the candidate aD _kj of the target signal. It also outputs the residual covariance matrix aS _kj (t) corresponding to the detection data aD _kj . Further, the position component of the prediction vector is calculated by the equation (13).

By performing the gate determination according to the equations (37) and (40), it is possible to perform the gate determination independently in the range direction and the cross range direction in which the ratio of the gate radii greatly differs, and as shown in FIG. It is possible to prevent the cross range direction from becoming too large when the gate in the direction is enlarged. Further, by performing the gate determination according to the equation (40), the gate size parameter in the vertical direction in the cross range direction, the gate size parameter in the horizontal direction in the cross range, that is, the gate size parameter in the elevation angle direction,
It is only necessary to set the gate size parameter once without setting two gate size parameters in the azimuth direction.

6 and 7 are flow charts for explaining the operation of the second embodiment. Here, steps with the same reference numerals are omitted. That is, here, step ST
Only 8b and step ST9b will be described.

In step ST8b, the sum of the residual in the cross-range vertical direction divided by the eigenvalue in the cross-range vertical direction and the sum of the residual in the cross-range lateral direction divided by the eigenvalue in the cross-range lateral direction is added. calculate.

In step ST9b, the gate size parameter d _r set in the range direction gate in the expression (37) and the gate size parameter d set in the two-dimensional cross range direction gate in the expression (40).
Set _elby .

Therefore, by the above steps, (3
By performing the gate determination according to the equations (7) and (40), the gate determination can be performed independently in the range direction and the cross range direction in which the ratio of the gate radii greatly differs.
As shown in Fig. 50, when the gate in the range direction is enlarged,
It is possible to prevent the cross range direction from becoming too large.
Further, by performing the gate determination according to the equation (40), the gate size parameter in the vertical direction of the cross range direction, the gate size parameter in the horizontal direction of the cross range, that is, the gate size parameter in the elevation angle direction, the gate size parameter in the azimuth angle direction. It is only necessary to set the gate size parameter once without setting two.

Third Embodiment A tracking device according to Embodiment 3 of the present invention will be described with reference to the drawings. Figure 8
FIG. 9 is a block diagram showing a configuration of a tracking device according to Embodiment 3 of the present invention. In FIG. 8, 11 is a first drive noise control means. The same reference numerals in the drawings represent the same or corresponding parts.

According to the Kalman filter theory, the prediction error covariance matrix P _pk (t) can be calculated by the equation (30).

In the first drive noise control means 11, the standard deviation σa of the drive noise vector in the equation (32) at the time of calculating the prediction error covariance matrix is set to the smooth vector obtained by the first data update means 6. When the magnitude of the velocity component is larger than a certain threshold, it is set to a large value, and the first data updating means 6
If the magnitude of the velocity component of the smoothing vector obtained in step 3 is smaller than a certain threshold, it is set to a small value.

Increasing the standard deviation σa of the driving noise vector in the equation (32) increases the prediction error covariance matrix by the equation (30), and further increases the residual covariance matrix by the equation (5). Therefore, the gate becomes large. Therefore, the target cannot be captured in the gate if the driving noise is set to be small when the vehicle turns when the target speed is high. However, in the first drive noise control means 11, when the magnitude of the velocity component of the smoothing vector becomes larger than a certain threshold value, the standard deviation of the drive noise vector is increased and the gate is widened, which makes it easier to capture the target. .

Next, FIGS. 9 and 10 show flow charts for explaining the operation of the third embodiment. Here, steps with the same reference numerals as those in FIGS. 3 and 4 are omitted. That is, step ST15b and step ST15c will be described.

In step ST15b, data is updated based on the theory of Kalman filter to calculate a smooth vector and a smooth error covariance matrix.

In step ST15c, if the magnitude of the velocity component of the smoothing vector becomes larger than a certain threshold value before performing the prediction process, the standard deviation of the driving noise vector is increased. Then, a prediction process is performed based on the Kalman filter theory, and a prediction vector and a prediction error covariance matrix are calculated.

Therefore, by performing the above steps, when the magnitude of the velocity component of the smoothing vector becomes larger than a certain threshold value, the standard deviation of the driving noise vector is increased and the gate is widened to supplement the target. It will be easier.

Fourth Embodiment A tracking device according to Embodiment 4 of the present invention will be described with reference to the drawings. Figure 11
FIG. 9 is a block diagram showing a configuration of a tracking device according to Embodiment 4 of the present invention. In FIG. 8, reference numeral 12 is a second drive noise control means. The same reference numerals in the drawings represent the same or corresponding parts.

In the second drive noise control means 12, the drive noise standard deviation σa is calculated using the equation (41) according to the magnitude of the velocity component of the smooth vector obtained by the first smoothing means 6.
Switch.

[0182]

[Formula 41]

Equation (41) is a logistic curve.
A conceptual diagram of the logistic curve is shown in FIG.

The horizontal axis of FIG. 12 is the magnitude V of the velocity component of the smooth vector, and the vertical axis is the standard deviation σa of drive noise. Here, V0 is a threshold for switching the magnitude V of the position component of the smoothed vector, and σa0 is the standard deviation of the drive noise corresponding to V0.

In the equation (41), a, b and c are parameters for determining the shape of the curve. Here, as for the scaling parameters a, b, and c, when a is increased, the vertical width of the curve is increased as shown in FIG. 12, and when b is increased, the vertical width is decreased and the gradient is smoothed. When is large, the height is the same and the gradient becomes steep.

Therefore, when the standard deviation σa of the drive noise is suddenly increased, the gain changes abruptly, so that the calculated values of the smooth vector and the smooth error covariance matrix become strange. However, by switching the standard deviation σa of the driving noise as in the equation (41), the transition of the gain becomes smooth before and after the switching, so that the switching of the velocity vector can be stably performed. Since the standard deviation of the driving noise vector is gradually increased and the gate is widened, it becomes easier to supplement the target.

Next, FIGS. 13 and 14 show flowcharts for explaining the operation of the fourth embodiment. Here, steps with the same reference numerals are omitted. That is, here, only step ST15d will be described.

In step ST15d, after determining the scaling parameter of the logistic curve in advance,
The standard deviation of drive noise is switched according to the magnitude V of the velocity component of the target smooth vector.

Therefore, by changing the standard deviation σa of the drive noise by performing the above steps, the gain transition becomes smooth before and after the change,
In addition, it becomes possible to stably switch the velocity vector. Further, the standard deviation of the driving noise vector is gradually increased and the gate is widened, so that the target can be easily captured.

Embodiment 5. A tracking device according to Embodiment 5 of the present invention will be described with reference to the drawings. Figure 15
FIG. 9 is a block diagram showing the configuration of a tracking device according to Embodiment 5 of the present invention. In FIG. 15, 13 is the second
Data updating means. The same reference numerals in the drawings represent the same or corresponding parts.

The second data updating means 13 will be described with reference to FIGS. 46 and 50. In FIG. 46, the detection data D _k1 , in the area of the gate centered on the position component of the prediction vector VX _pk (t) surrounded by the broken line at the current time tk,
D _k2 and D _k3 exist. In the gate determination and data updating means 13 considering the weighting, the determination signal input from the range direction gate determination 3 means indicating whether or not the distance gate of the formula (37) is entered, the first cross range direction A determination signal input from the gate determination means 4 and indicating whether or not the vehicle is in the elevation angle gate of the formula (38), (39)
In the case where all of the determination signals indicating whether or not the azimuth gate of the expression is in the gate are signals, the detection data Dkj corresponding to these determination signals are set as the target signal candidates. Here, a judgment signal indicating whether the distance gate of the formula (37) is entered, a judgment signal indicating whether the elevation gate of the formula (38) is entered, and an azimuth gate of the formula (39) is entered. Consider a case where there are a plurality of signals that all of the determination signals indicating whether or not there are gates. In this case, in the second data updating means 13, the smooth vector and the smooth error covariance in which the distance between the gate centers is taken into consideration by the likelihood of converting the distance from the gate center by the normal distribution for the plurality of detection data. Calculate the matrix.

Approximate likelihood calculated by converting the distance from the gate center
A conceptual drawing is shown in FIG. D in FIG. 50 _k1, D_k2, D_k3, D
_k4, D_k5Is detection data, and D_k1, D_k2, D_k3Is a game
It is supposed to be inside the package. Where exists in the gate
Detection data D_k1, D_k2, D_k3The distance from the center of the gate
The converted likelihood is the likelihood of a normal distribution as shown in FIG.
γ _k1, Γ_k2, Γ_k3It is represented by.

Therefore, when the detection data closest to the gate center is an unnecessary signal, when the smooth vector and the smooth error covariance matrix are calculated, the center of the gate after the next sampling is extremely pulled to the unnecessary signal at the current time. , Tracking may fail. However, by calculating the smooth vector and the smooth error covariance matrix in consideration of the distance between the gate centers by the likelihood obtained by converting the distance from the gate center by the gate determination and data update means 13 considering the weighting, The center of the gate after the next sampling is not extremely pulled by the unnecessary signal at the current time, and stable tracking can be performed in the unnecessary signal.

16 and 17 are flow charts for explaining the operation of the fifth embodiment. Here, steps with the same reference numerals are omitted. That is, step ST14
Only b and step ST15e will be described.

In step ST14b, the likelihood of the detection data in the gate is calculated.

In step ST15e, a smooth vector and a smooth error covariance matrix are calculated using the likelihood of the detection data. Also, extrapolating the smoothing vector considering the distance between the gate centers at the current time k by one sampling,
Compute the prediction vector and the prediction error covariance matrix.

By the above steps, therefore, when the detection data closest to the center of the gate is an unnecessary signal, when the smooth vector and the smooth error covariance matrix are calculated, the center of the gate after the next sampling is the unnecessary signal at the current time. However, the smooth vector and smooth error covariance matrix considering the distance between the gate centers should be calculated according to the likelihood of the detection data converted from the distance from the gate center. Therefore, the center of the gate after the next sampling is not extremely pulled by the unnecessary signal at the current time, and stable tracking can be performed in the unnecessary signal.

Sixth Embodiment A tracking device according to Embodiment 6 of the present invention will be described with reference to the drawings. FIG.
FIG. 13 is a block diagram showing a configuration of a tracking device according to Embodiment 6 of the present invention. In FIG. 13, 14 is the first
Is a gate size parameter setting means. The same reference numerals in the drawings represent the same or corresponding parts.

First gate size parameter setting means 1
In FIG. 4, the time interval between the current sampling time and the previous sampling time obtained from the observing means is checked, and when the time interval is large, a signal for setting the gate size parameter to a small value is applied to the gate determining means 3 in the range direction To the gate determination means 4 in the cross range direction.

The gate radius that determines the size of the gate is
Like the denominator on the left side of the equation (25), 0.5 to the product of eigenvalues λ ₁ and d _ε , 0.5 to the product of λ ₂ and d _ε , and 0.5 to the product of λ ₃ and d _ε. The power of 0.5 is the product of the eigenvalues λ ₁ , λ ₂ , and λ ₃ and the gate size parameter d _ε .

Here, as the sampling interval increases,
Since the eigenvalues λ ₁ , λ ₂ , and λ ₃ are large, the gate radius is also large. However, when the sampling interval becomes extremely long, the eigenvalue also becomes extremely large, and the gate becomes very large. Therefore, not only the target signal, but also a large amount of unnecessary signals intrude into the gate, which makes the gate useless. Therefore, if the sampling interval is larger than a certain threshold, set the gate size parameter to a small value to maintain an appropriate gate radius to prevent the target signal from being captured and unwanted signals from entering. To

Next, FIGS. 19 and 20 show flow charts for explaining the operation of the sixth embodiment. Here, steps with the same reference numerals are omitted. That is, step ST9b
Only I will explain.

In step ST9b, when the sampling interval is larger than a certain threshold value, the gate size parameter values set in the range direction, the cross range horizontal direction and the cross range vertical direction are controlled to be small.

By the above steps, when the sampling interval is larger than a certain threshold value, the gate size parameter is set to a small value so as to keep the gate radius of an appropriate size, so that the target signal is not supplemented or unnecessary. Try to prevent signal intrusion.

Seventh Embodiment A tracking device according to Embodiment 7 of the present invention will be described with reference to the drawings. Figure 21
FIG. 13 is a block diagram showing the configuration of a tracking device according to Embodiment 7 of the present invention. In FIG. 21, 15 is an observation error fixing means. The same reference numerals in the drawings represent the same or corresponding parts.

The observation error fixing means 15 uses the distance observation noise standard deviation, the elevation angle observation noise standard deviation, and the azimuth angle observation noise standard deviation σr _kj , σe _kj , σb _{kj in the} equation (8).
Set a fixed value to.

R _kj in the residual covariance matrix S _kj (t) of equation (5) is calculated by equation (7), and R _{kj of} equation (7) is calculated.
The calculation of Λ _kj is calculated by the equation (8). Therefore, the observation noise standard deviation of the distance, the observation noise standard deviation of the elevation angle, and the observation noise standard deviation σ of the azimuth angle in Eq. (8)
r _kj, Sigma] e _kj, by setting the value of fixed .sigma.b _kj, every sampling time, since each detection data is calculated in equation (8) becomes unnecessary, the calculation load becomes lighter.

Next, FIGS. 22 and 23 show flowcharts for explaining the operation of the seventh embodiment. Here, steps with the same reference numerals are omitted. That is, step ST2b
Will be described.

At step ST2b, the residual is calculated by the equation (15). Further, with the standard deviation of observation noise fixed, Λ _kj is calculated from equation (8), Λ _kj is substituted into equation (7), R _kj is calculated, and then the residual covariance matrix is calculated. .

By the above steps, fixed values are set for the distance observation noise standard deviation, the elevation angle observation noise standard deviation, and the azimuth angle observation noise standard deviation σr _kj , σe _kj , and σb _kj in the equation (8). Then, the calculation load is lightened because the calculation of the formula (8) is not necessary for each sampling time and each detection data.

[Embodiment 8] A tracking device according to Embodiment 8 of the present invention will be described with reference to the drawings. Figure 24
FIG. 13 is a block diagram showing the configuration of a tracking device according to Embodiment 8 of the present invention. In FIG. 17, 16 is a sampling interval fixing means. The same reference numerals in the drawings represent the same or corresponding parts.

The sampling interval fixing means 16 inputs a control signal for controlling the sampling interval to the predicting means 7. Therefore, since the state transition matrix Φk in the equation (31) has a fixed value, the prediction vector calculation formula of the equation (29) including the state transition matrix Φk and the prediction error covariance of the equation (30) including the state transition matrix Φk are included. The calculation load on the matrix is reduced.

Next, FIGS. 25 and 26 show flowcharts for explaining the operation of the eighth embodiment. Here, steps with the same reference numerals are omitted. That is, step ST15
Only f will be described.

In step ST15f, using the state transition matrix Φk with a fixed sampling interval, (29)
The calculation formula of the prediction vector of the formula and the prediction error covariance matrix of the formula (30) are calculated.

By handling the state transition matrix Φk in equation (31) as a fixed value through the above steps, the calculation load of the prediction vector and the prediction error covariance matrix is reduced.

Embodiment 9. FIG. A tracking device according to Embodiment 9 of the present invention will be described with reference to the drawings. FIG. 27
FIG. 13 is a block diagram showing the configuration of a tracking device according to Embodiment 9 of the present invention. In FIG. 19, 17 is a gate center fixing means. The same reference numerals in each figure represent the same or corresponding parts.

The gate center fixing means 17 continues to use fixed detection data of the first sampling instead of the value of the position component VZ _pk (t) of the prediction vector on the right side of the equation (15) for several samplings from the initial tracking. .

The sampling number for fixing the gate is set in advance.

Therefore, in the initial tracking, when the prediction vector is calculated from the scattered detection data, the variation of the velocity vector is large. That is, when calculating the prediction vector,
Using a disparate velocity vector will likely not be able to capture the target signal because the center of the gate is not stable. However, the variation in the initial tracking is reduced by continuing to use the fixed detection data of the first sampling instead of the position component VZ _pk (t) of the prediction vector, which is the center of the gate, for a few samplings in the initial tracking. To be done.

Next, FIGS. 28 and 29 show flowcharts for explaining the operation of the ninth embodiment. Here, steps with the same reference numerals are omitted. That is, step ST2c
Will be described only.

At step ST2c, the gate center is fixed for the initial number of samplings.

By the above steps, the tracking is performed by using the fixed detection data of the first sampling instead of the position component VZ _pk (t) of the prediction vector, which is the center of the gate, for a few samplings in the initial tracking. Initial variations are reduced.

[Embodiment 10] Embodiment 1 of the present invention
A tracking device relating to 0 will be described with reference to the drawings. FIG. 30 is a block diagram showing the configuration of the tracking device according to the embodiment of the present invention. In FIG. 30, reference numeral 18 denotes a gate size parameter varying means according to the target speed.
The same reference numerals in the drawings represent the same or corresponding parts.

The gate speed parameter varying means 18 based on the target speed calculates the magnitude of the speed component of the smoothing vector from the smoothing vector obtained one sampling before and obtained from the first data updating means 6. When the magnitude of the velocity component of the smoothing vector one sampling before is larger than a certain threshold value, the magnitude of the gate size parameter at the current time in the gate determining means 3 in the range direction and the gate determining means 4 in the first cross range direction A control signal for increasing the height is input to each of the gate determining means 3 in the range direction and the gate determining means 4 in the first cross range direction.
The gate determining means 3 in the range direction and the gate determining means 4 in the first cross-range direction use the control signal for increasing the size of the gate size parameter input from the gate size parameter varying means 18 according to the target speed, Increase the size parameter.

Therefore, in the gate size parameter varying means 18 based on the target velocity, when the magnitude of the velocity component of the smoothing vector before one sampling is larger than a certain threshold value, the gate determining means 4 in the first cross range direction By increasing the size of the gate size parameter at the current time and the size of the gate size parameter at the current time in the range direction gate determination means 3, it becomes easier to supplement the target signal.

Next, FIG. 31 and FIG. 32 show the first embodiment.
9 shows a flowchart illustrating the operation of 0. Here, steps with the same reference numerals are omitted. That is, step ST
Only 9c will be described.

In step ST9c, if the magnitude of the velocity component of the smoothing vector one sampling before is larger than a certain threshold value, the gate size parameter at the current time in the range direction, the cross range horizontal direction and the cross range vertical direction is increased. Set.

By the above steps, when the magnitude of the velocity component of the smoothing vector one sampling before is larger than a certain threshold value, the target signal is supplemented by increasing the magnitude of the gate size parameter at the current time. Make it easier.

Eleventh Embodiment Embodiment 1 of the present invention
A tracking device according to No. 1 will be described with reference to the drawings. FIG. 33 is a block diagram showing the configuration of the tracking device according to Embodiment 11 of the present invention. In FIG. 33, 19
Is a first display smoothing means. The same reference numerals in the drawings represent the same or corresponding parts.

In the first display smoothing means 19, the smoothing vector VX input from the first data updating means 6 is inputted.
_The display smoothing vector VX _sk (t *) is calculated from _sk (t) using the following equation (42), and the display smoothing vector VX is obtained.
Input _sk (t *) to the display means 9. Further, the smoothing vector for display is stored for one sampling.

[0231]

[Equation 42]

In the equation (42), VX _sk (t) is the smooth vector input from the first data updating means 6 at time k, VX _sk (t *) is the display smooth vector at time k, ηk is a coefficient that determines the smoothing weight.

Ηk is calculated by the following equation (43).

[0234]

[Equation 43]

On the right side of the equation (43), Δt is a sampling interval and τ is a time constant.

As the time constant is larger, the display smoothing vector has a larger delay than the target detection data at the current time, but the smoothing effect is large, that is, it looks smooth in appearance.

Therefore, in the first display smoothing means 19, the time constant of the equation (43) is determined in advance, and the smoothing vector VX input from the first data updating means 6 is inputted.
_{By obtaining} the display smoothing vector VX _sk (t *) from _sk (t) using the equation (42), the display for the operator can be easily seen.

Next, FIG. 34 and FIG. 35 show the eleventh embodiment.
6 is a flowchart illustrating the operation of the above. Here, steps with the same reference numerals are omitted. That is, step ST1
Only 6b will be described.

In step ST16b, the display smoothing vector is calculated from the smoothing vector, and the smoothing position smoothing speed is displayed.

[0240] When the smooth vector obtained from the scattered detection data is displayed, the display track is rattled. However, by the above steps, the display smooth vector is calculated and the display smooth vector is displayed. When is displayed, the displayed track for the operator becomes smooth and easy to see.

[Embodiment 12] Embodiment 1 of the present invention
A tracking device according to No. 2 will be described with reference to the drawings. FIG. 36 is a block diagram showing the configuration of the tracking device according to the embodiment of the present invention. In FIG. 36, 20 is a second display smoothing means. The same reference numerals in each figure represent the same or corresponding parts.

In the second display smoothing means 20, the smoothing vector VX input from the first data updating means 6 is input.
_sk a (t), using the following equation (44) obtains a smooth vector VX _sk (t **) for display, display smoothing vector VX _sk
Input (t **) to the display means 9. Further, the smoothing vector for display is stored for one sampling.

[0243]

[Equation 44]

In the equation (44), VX _sk (t) is the smooth vector input from the first data updating means 6 at time k, and VX _sk (t **) is the display smooth vector at time k. , Θ _k are coefficients that determine the smoothing weight.

The calculation of θ _k is performed by the equation (45).

[0246]

[Equation 45]

On the right side of the equation (45), Δt is a sampling interval, and ρ _k is a time constant that varies _depending on the sampling time.

Ρ _k is calculated by the following equation (46).

[0249]

[Equation 46]

In (46), ρ_kWhen sampling
Time constant that changes with time, φ_kIs a factor that determines the smoothing weight
Is a number. Also, ρ_infIs ρ_kIs a steady-state fixed value of
It ρ _kIs set in advance. Φ in equation (46)_kIs
It is determined by the following equation (47).

[0251]

[Equation 47]

In the right side of the expression (45) in the expression (47), Δt is a sampling interval, and τc is a time constant of a time constant for controlling the time constant ρ _k which varies depending on the sampling time.

Expression (44), Expression (45), Expression (46),
By using the equation (47), in the initial state, the rattling of the smoothing vector is larger than the delay than the target detection data at the current time. Therefore, the time constant θk of the smoothing vector for display is set to be large, and The wake for display can be displayed smoothly. Further, when the initial state ends and the state approaches a steady state, the rattling of the smoothing vector becomes small, so that the delay with the detection data is made small, the time constant θ _k
Can be set small. Therefore, the display to the operator becomes easy to see.

Next, FIG. 37 and FIG. 38 show the twelfth embodiment.
6 is a flowchart illustrating the operation of the above. Here, steps with the same reference numerals are omitted. That is, step ST1
Only 6c will be described.

In step ST16c, a smoothing vector for display is calculated from the smoothing vector with a variable time constant, and the smoothed position smoothing speed is displayed.

By the above steps, the equations (44) and (4
By using equations (5), (46), and (47) to change the time constant and calculating the display smoothing vector,
It is possible to delay the smooth vector before obtaining the display smooth vector and improve the smoothness of the smooth vector.

Thirteenth Embodiment Embodiment 1 of the present invention
A tracking device according to No. 3 will be described with reference to the drawings. FIG. 39 is a block diagram showing the configuration of the tracking device according to the embodiment of the present invention. In FIG. 39, reference numeral 21 is a multipath determining means. The same reference numerals in the drawings represent the same or corresponding parts.

The multipath is a reflected radio wave in which the radio wave emitted from the radar hits the ground, the sea surface, or an obstacle and is scattered.

If the S / N ratio is larger than a certain threshold value obtained from the observing means 1, the multipath judging means 21
The control signal for setting the observation noise standard deviation of the elevation angle in the expression (8) to a fixed value given in advance and calculating the residual covariance matrix using the expression (13) is used to calculate the first gate determination specifications. Input to means 2. In addition, the multipath determination means 21
When the S / N ratio is larger than a certain threshold value obtained from the observing means 1, the gate size parameter in the cross-range vertical direction in the gate determining means 4 in the first cross-range direction indicates that the gate determination result is always within the gate. A control signal for setting a sufficiently large value for determination is input to the gate determination means 4 in the first cross range direction.

Therefore, in the multipath environment in which the accuracy in the altitude direction is not good, the multipath determination means 21 does not use the observation noise standard deviation of the elevation angle that is most affected by the multipath. The tracking performance of can be improved.

Next, FIG. 40 and FIG. 41 show the thirteenth embodiment.
6 is a flowchart illustrating the operation of the above. Here, steps with the same reference numerals are omitted. That is, step ST2
Only d and step ST9d will be described.

In step ST2d, the elevation observation noise due to the S / N threshold is fixed, that is, the observation noise standard deviation of the elevation angle in equation (8) is set to a fixed value in advance, and then the residual covariance matrix is set. Is calculated by the equation (13).

If multipath is determined in step ST9d, the gate size parameter is set assuming multipath, that is, in the gate determination in the crossrange vertical direction, all the detection data are determined to be within the gate. Set the gate size parameter for the direction.

By the above steps, in the multipath environment where the accuracy in the altitude direction is not good, the tracking noise performance in the multipath environment is improved by not using the observation noise standard deviation of the elevation angle that is most affected by the multipath. Can be improved.

Fourteenth Embodiment Embodiment 1 of the present invention
The tracking device according to No. 4 will be described with reference to the drawings. FIG. 42 is a block diagram showing the configuration of the tracking device according to Embodiment 14 of the present invention. In FIG. 42, 22
Is a condition number determination means. The same reference numerals in the figure
Represents the same or corresponding part.

Among the eigenvalues input from the first gate judgment parameter calculation means 2, the ratio of the maximum eigenvalue λkmax to the minimum eigenvalue λkmin is calculated to calculate the condition number COND _k in the following equation (48). .

[0267]

[Equation 48]

If COND _k is small, there is a problem that a digit loss occurs in the matrix operation, the inverse matrix operation cannot be performed, and the tracking operation is unstable.

In the condition number judging means 22, the condition number COND _k is set to a certain threshold value C which is set in advance as in the following equation (49).
When OND _th or less, the operator inputs a danger signal indicating that the tracking calculation is unstable to the display means 9 and outputs it to the operator's display screen.

[0270]

[Equation 49]

Therefore, by using the condition number judging means 22, the operator can detect a danger signal that the tracking calculation is unstable. For example, it is possible to retry the tracking calculation or to cancel the tracking calculation. You can

Next, the fourteenth embodiment will be described with reference to FIGS. 43 and 44.
6 is a flowchart illustrating the operation of the above. Here, steps with the same reference numerals are omitted. That is, step ST1
Only 6d will be described.

At step ST16d, the condition number is calculated by the expression (48), and when the expression (49) is satisfied, the operator is inputted a danger signal that the tracking calculation is unstable, Output to the operator's display screen.

Through the above steps, the operator can detect the danger signal that the tracking calculation is unstable.

[0275]

The present invention obtains an S / N ratio which is a signal-to-noise ratio on the basis of received power and converts it into an observation noise standard deviation of a distance, an elevation angle and an azimuth angle to obtain the observation noise standard. Observing means for outputting the deviation and the detection data representing the position associated with them, the prediction error covariance matrix and the detection data are input, from the residual covariance matrix calculated from them,
First gate determination specification calculation means for calculating an eigenvalue in the distance direction, an eigenvalue in the elevation direction, an eigenvalue in the azimuth direction, an eigenvector in the distance direction, an eigenvector in the elevation direction and an eigenvector in the azimuth direction, and the detection data is the distance. It is output from the range direction gate determination means that determines whether or not it is in a one-dimensional gate in the direction and outputs a signal indicating the determination result, and the observation means and the first gate determination specification calculation means. Using the data, in the cross-range direction, a cross-range direction gate determination means for performing gate determination in a predetermined dimension, and a signal indicating the determination result of whether or not the gate is in the predetermined dimension in the distance direction, A signal indicating the result of determination as to whether the gate is in the predetermined dimension in the elevation direction, and a signal indicating whether or not the gate is in the predetermined dimension in the azimuth direction. It results a signal indicating the distance direction,
When the detection data is included in all the gates of the predetermined dimension in the elevation direction and the azimuth direction, the detection data and the residual covariance matrix associated with the detection data are output as candidates for the target signal. 1 gate determination means, the detection data and the residual covariance matrix output from the first gate determination means are input, the data is updated based on the theory of Kalman filter, and the smooth error covariance matrix and First data updating means for outputting a smooth vector, prediction means for calculating a prediction error covariance matrix and a prediction vector after one sampling from the current time based on the theory of the Kalman filter, and input from the prediction means The prediction error covariance matrix and the prediction vector are delayed by one sampling to make a prediction with respect to the first gate determination parameter calculation means. Since the tracking device is provided with the one-sampling delay means for outputting the difference covariance matrix and the display means for displaying the target track and the target speed from the smoothing vector input from the first data updating means, The target signal can easily enter, and it is possible to prevent unwanted signals from entering.

Further, the gate determining means in the cross range direction determines whether or not the detection data is in the one-dimensional gate in the elevation direction, outputs a signal indicating the determination result, and the detection data indicates the azimuth angle. It is easy to perform gate determination in the cross range direction because it is configured by first gate determination means in the cross range direction that determines whether or not the gate is in the one-dimensional gate in the direction and outputs a signal indicating the determination result. Can be done.

Further, the gate determining means in the cross-range direction has an eigenvector in the elevation angle direction, an eigenvalue in the elevation angle direction,
The azimuth angle eigenvector and the azimuth angle eigenvalue are used to configure a second cross range direction gate determination means for performing two-dimensional gate determination in the cross range direction. Therefore, since the gate determination can be performed independently, it is possible to prevent the gate in the range direction from being increased together with the gate in the range direction. Also, the gate size parameter can be set once.

Further, when the velocity component of the smoothing vector becomes larger than a certain threshold value, a first drive noise control means for setting the standard deviation of the drive noise vector to be large is further provided. Widening the gate makes it easier to capture the goal.

Further, based on the logistic curve with the magnitude of the standard deviation of the driving noise vector as the ordinate and the velocity component of the smooth vector as the abscissa, the velocity component of the smooth vector becomes larger than a certain threshold value. In this case, since the second drive noise control means for setting the standard deviation of the drive noise vector to be large is further provided, the standard deviation of the drive noise vector is gradually increased and the gate is widened, so that the target can be easily captured.

Further, the present invention obtains an S / N ratio which is a signal-to-noise ratio on the basis of received power and converts it into an observation noise standard deviation of a distance, an elevation angle and an azimuth angle to obtain the observation noise standard. The eigenvalues in the distance direction are obtained from the observation means that outputs the deviation and the detection data representing the position associated with them, the prediction error covariance matrix and the observation error covariance matrix, and the residual covariance matrix calculated from them. , An eigenvalue in the elevation direction, an eigenvalue in the azimuth direction, an eigenvector in the distance direction, an eigenvector in the elevation direction and an eigenvector in the azimuth direction, and the detection data is one-dimensional in the distance direction. It is determined whether or not the gate is in the gate, and the range direction gate determination means for outputting a signal indicating the determination result, and the detection data are in the elevation direction one-dimensional gate. It is determined whether or not it is determined, a signal indicating the determination result is output, it is determined whether the detection data is in a one-dimensional gate in the azimuth direction, and a signal indicating the determination result is output. The gate determination means in the cross range direction, the gate determination means in the range direction, and the gate determination means in the first cross range direction, as a result of the gate determination of the detection data, the detection data in the gate are detected as the gate center and Based on the theory of the Kalman filter, the second data updating means for converting the data distance into the likelihood of a normal distribution, and using the likelihood, and the prediction error after one sampling from the current time The prediction means for calculating the variance matrix and the prediction vector and the prediction error covariance matrix and the prediction vector input from the prediction means are delayed by one sampling. And a display for displaying the target track and the target speed from the 1-sampling delay means output to the first gate determination specification calculation means and the smooth vector input from the first data update means. Since it is a tracking device provided with means, by calculating the smooth vector and the smooth error covariance matrix considering the distance between the gate centers by the likelihood of the detection data,
The center of the gate after the next sampling is not extremely pulled by the unnecessary signal at the current time, and stable tracking can be performed in the unnecessary signal.

Further, since the first gate size parameter setting means for setting the gate size parameter according to the sampling interval is further provided, when the sampling interval is larger than a predetermined threshold value, the gate size parameter should be set small. Thus, by keeping the gate radius of an appropriate size, the supplement of the target signal and the intrusion of the unnecessary signal are prevented.

Further, since the observation error fixing means for setting the observation noise standard deviation of the distance, the observation noise standard deviation of the elevation angle, and the observation noise standard deviation of the azimuth angle to fixed values is further provided, the fixed values are set to them. By setting it, it becomes unnecessary to calculate the observation error covariance matrix in polar coordinates for each sampling time and each detection data, so that the calculation load can be reduced.

Further, since the control signal for fixing the sampling interval is input to the predicting means and the sampling interval fixing means for fixing the sampling interval when the prediction error covariance matrix and the prediction vector are calculated is further provided, the state transition By handling the matrix as a fixed value, the calculation load of the prediction vector and the prediction error covariance matrix can be reduced.

Further, since the gate center fixing means for fixing the center of the gate for several samplings from the initial tracking is further provided, the position component of the prediction vector which is the center of the gate is replaced by only several samplings at the initial tracking. In addition, by continuing to use the fixed first sampling detection data, variations at the initial stage of tracking are reduced.

Further, since the gate size parameter varying means for changing the magnitude of the gate size parameter by using the magnitude of the velocity component of the smoothing vector obtained from the data updating means is further provided, the smoothing before one sampling is performed. When the magnitude of the velocity component of the vector is larger than the predetermined threshold value, it becomes easy to supplement the target signal by increasing the magnitude of the gate size parameter at the current time.

Since the smoothing vector obtained from the data updating means is inputted to the display filter having a fixed time constant, the first smoothing means for display is further provided. , When the smooth vector obtained from the scattered detection data is displayed, there is rattling of the display track, but if the display smooth vector is calculated and the display track is displayed from the display smooth vector,
The displayed track for the operator becomes smooth and easy to see.

Since the smoothing vector obtained from the data updating means is inputted to the display filter having a variable time constant, the second smoothing means for display is further provided. , Make the time constant variable,
By calculating the display smooth vector, the delay of the smooth vector before obtaining the display smooth vector and the smoothness of the smooth vector can be improved.

In the observing means, when the S / N ratio which is the signal-to-noise ratio is obtained, and the S / N ratio is larger than a certain threshold value, it is judged as multipath, and when it is judged as multipath, The control signal for controlling the residual covariance matrix to be calculated by using a fixed value that is preset for the observation noise standard deviation of the elevation angle used when calculating the residual covariance matrix is the first gate judgment. Since it is further equipped with a multipath judging means for outputting to the specification calculating means, in a multipath environment where accuracy in the altitude direction is not good, by not using the observation noise standard deviation of the elevation angle that is most affected by the multipath. , It is possible to improve the tracking performance in a multipath environment.

Further, using the condition number which is the ratio of the maximum value of the eigenvalue and the minimum value of the eigenvalue among the eigenvalues input from the first gate judgment parameter calculation means, and if the condition number is below a certain threshold value, Since the operator is further provided with a condition number determination means for inputting a danger signal indicating that the tracking calculation is unstable to the display means, the operator can output a danger signal indicating that the tracking calculation is unstable. Can detect.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a tracking device according to a first embodiment of the present invention.

FIG. 2 is an explanatory diagram showing a coordinate system related to an observation means provided in the tracking device.

FIG. 3 is a flowchart showing a processing flow of the tracking device according to the first embodiment of the present invention.

FIG. 4 is a flowchart showing a processing flow of the tracking device according to the first embodiment of the present invention.

FIG. 5 is a block diagram showing a configuration of a tracking device according to a second embodiment of the present invention.

FIG. 6 is a flowchart showing a processing flow of the tracking device according to the second embodiment of the present invention.

FIG. 7 is a flowchart showing a processing flow of the tracking device according to the second embodiment of the present invention.

FIG. 8 is a block diagram showing a configuration of a tracking device according to a third embodiment of the present invention.

FIG. 9 is a flowchart showing a processing flow of the tracking device according to the third embodiment of the present invention.

FIG. 10 is a flowchart showing a processing flow of the tracking device according to the third embodiment of the present invention.

FIG. 11 is a block diagram showing a configuration of a tracking device according to a fourth embodiment of the present invention.

FIG. 12 is an explanatory diagram showing a conceptual diagram of a logistic curve.

FIG. 13 is a flowchart showing a processing flow of the tracking device according to the fourth embodiment of the present invention.

FIG. 14 is a flowchart showing a processing flow of the tracking device according to the fourth embodiment of the present invention.

FIG. 15 is a block diagram showing a configuration of a tracking device according to a fifth embodiment of the present invention.

FIG. 16 is a flowchart showing a processing flow of the tracking device according to the fifth embodiment of the present invention.

FIG. 17 is a flowchart showing a processing flow of the tracking device according to the fifth embodiment of the present invention.

FIG. 18 is a block diagram showing a configuration of a tracking device according to a sixth embodiment of the present invention.

FIG. 19 is a flowchart showing a processing flow of the tracking device according to the sixth embodiment of the present invention.

FIG. 20 is a flowchart showing a processing flow of the tracking device according to the sixth embodiment of the present invention.

FIG. 21 is a block diagram showing a configuration of a tracking device according to a seventh embodiment of the present invention.

FIG. 22 is a flowchart showing a processing flow of the tracking device according to the seventh embodiment of the present invention.

FIG. 23 is a flowchart showing a processing flow of the tracking device according to the seventh embodiment of the present invention.

FIG. 24 is a block diagram showing the configuration of a tracking device according to an eighth embodiment of the present invention.

FIG. 25 is a flowchart showing a processing flow of the tracking device according to the eighth embodiment of the present invention.

FIG. 26 is a flowchart showing a processing flow of the tracking device according to the eighth embodiment of the present invention.

FIG. 27 is a block diagram showing the configuration of a tracking device according to a ninth embodiment of the present invention.

FIG. 28 is a flowchart showing a processing flow of the tracking device according to the ninth embodiment of the present invention.

FIG. 29 is a flowchart showing a processing flow of the tracking device according to the ninth embodiment of the present invention.

FIG. 30 is a block diagram showing the configuration of a tracking device according to a tenth embodiment of the present invention.

FIG. 31 is a flowchart showing a processing flow of the tracking device according to the tenth embodiment of the present invention.

FIG. 32 is a flowchart showing a processing flow of the tracking device according to the tenth embodiment of the present invention.

FIG. 33 is a block diagram showing the configuration of a tracking device according to an eleventh embodiment of the present invention.

FIG. 34 is a flowchart showing a processing flow of the tracking device according to the eleventh embodiment of the present invention.

FIG. 35 is a flowchart showing a processing flow of the tracking device according to the eleventh embodiment of the present invention.

FIG. 36 is a block diagram showing the configuration of a tracking device according to a twelfth embodiment of the present invention.

FIG. 37 is a flowchart showing a processing flow of the tracking device according to the twelfth embodiment of the present invention.

FIG. 38 is a flowchart showing a processing flow of the tracking device according to the twelfth embodiment of the present invention.

FIG. 39 is a block diagram showing the configuration of a tracking device according to a thirteenth embodiment of the present invention.

FIG. 40 is a flowchart showing a processing flow of the tracking device according to the thirteenth embodiment of the present invention.

FIG. 41 is a flowchart showing a processing flow of the tracking device according to the thirteenth embodiment of the present invention.

FIG. 42 is a block diagram showing the configuration of a tracking device according to a fourteenth embodiment of the present invention.

FIG. 43 is a flowchart showing a processing flow of the tracking device according to the fourteenth embodiment of the present invention.

FIG. 44 is a flowchart showing a processing flow of the tracking device according to the fourteenth embodiment of the present invention.

FIG. 45 is a block diagram showing a configuration of a conventional tracking device.

FIG. 46 is an explanatory diagram showing a configuration of a conventional tracking device.

FIG. 47 is a relationship diagram of a gate, a target predicted position, a target observation position, a target range direction, and a target cross range direction.

FIG. 48 is an explanatory diagram showing a relationship between a cross range error and a range error with respect to an angle error.

FIG. 49 is an explanatory diagram showing a relationship between an eigenvector in the range direction and an eigenvector in the cross range direction.

FIG. 50 is a conceptual diagram of likelihood converted from the distance from the gate center.

[Explanation of symbols]

1 Observing Means, 2 First Gate Judgment Specification Calculation Means, 3
Range direction gate determination means, 4 first cross range direction gate determination means, 5 first gate determination means,
6 First data updating means, 7 Predicting means, 8 1 Sampling delaying means, 9 Displaying means, 10 2nd cross-range direction gate determining means, 11 1st driving noise control means, 12 2nd driving noise control Means, 13 Second data updating means, 14 First gate size parameter setting means, 15 Observation error fixing means, 16 Sampling interval fixing means, 17 Gate center fixing means, 18 Gate size parameter varying means by target speed, 19
First smoothing means for display, 20 second smoothing means for display,
21 multipath judging means, 22 condition number judging means, 1
01 observation means, 106 data updating means, 107 prediction means, 108 1 sampling delay means, 109 display means, 123 gate determination specification calculation means, 124 gate determination means.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Yoshio Kosuge             2-3 2-3 Marunouchi, Chiyoda-ku, Tokyo             Inside Ryo Electric Co., Ltd. F-term (reference) 5J070 AC02 AC11 AD01 BB04

Claims (15)

[Claims] 1. A signal-to-noise ratio S based on received power.
/ N ratio is obtained, converted into observation noise standard deviations of distance, elevation angle and azimuth angle, and the observation noise standard deviation and detection data representing the position associated with them are output, The variance matrix and the detection data are input,
A first gate for calculating the eigenvalue in the distance direction, the eigenvalue in the elevation direction, the eigenvalue in the azimuth direction, the eigenvector in the distance direction, the eigenvector in the elevation direction, and the eigenvector in the azimuth direction from the residual covariance matrix calculated from them. Determination specification calculation means, range determination gate determination means for determining whether or not the detection data is in a one-dimensional gate in the distance direction, and outputting a signal indicating the determination result, the observation means and the Using the data output from the first gate determination specification calculation means, a gate determination means in the cross range direction for performing gate determination of a predetermined dimension in the cross range direction and a gate of the predetermined dimension in the distance direction are entered. Signal indicating the determination result of whether or not it is in the azimuth direction, From the signal indicating the determination result of whether or not the gate enters the predetermined dimension gate, the target signal when the detection data is contained in all the gates of the predetermined dimension in the distance direction, the elevation direction, and the azimuth direction. Of the detection data and the residual covariance matrix associated with the detection data, and the detection data and the residual covariance matrix output from the first gate determination means. Is input to update the data based on the theory of the Kalman filter, and outputs a smooth error covariance matrix and a smooth vector, and 1 based on the theory of the Kalman filter
Prediction means for calculating the prediction error covariance matrix and the prediction vector after sampling, and the prediction error covariance matrix and the prediction vector input from the prediction means are delayed by one sampling to obtain the first gate determination parameter calculation means. A sampling delay means for outputting the prediction error covariance matrix, and a display means for displaying the target track and the target speed from the smoothing vector input from the first data updating means. A tracking device characterized by the above. 2. The cross range direction gate determination means determines whether or not the detection data is included in a one-dimensional gate in the elevation angle direction, outputs a signal indicating the determination result, and the detection data indicates the azimuth angle. 2. The gate determining means for the first cross-range direction for determining whether or not the gate is in the one-dimensional gate in the direction, and outputting a signal indicating the determination result. Tracking device. 3. The cross-range direction gate determination means uses the elevation angle direction eigenvector, elevation angle direction eigenvalue, azimuth angle eigenvector, and azimuth angle eigenvalue,
The tracking device according to claim 1, wherein the tracking device comprises a second cross-range direction gate determination means for performing two-dimensional gate determination in the cross-range direction. 4. The first drive noise control means for setting a large standard deviation of the drive noise vector when the velocity component of the smoothing vector becomes larger than a certain threshold value. The tracking device according to any one of 1. 5. A logistic curve having the standard deviation of the driving noise vector as the ordinate and the velocity component of the smooth vector as the abscissa has a velocity component of the smooth vector larger than a certain threshold based on the logistic curve. In the case of setting a large standard deviation of the driving noise vector,
5. The tracking device according to claim 4, further comprising: drive noise control means. 6. The signal-to-noise ratio S based on the received power.
/ N ratio is obtained, converted into observation noise standard deviations of distance, elevation angle and azimuth angle, and the observation noise standard deviation and detection data representing the position associated with them are output, Covariance matrix and observation error covariance matrix are input, and from the residual covariance matrix calculated from them, the eigenvalue in the distance direction, the eigenvalue in the elevation direction, the eigenvalue in the azimuth direction,
First gate determination parameter calculation means for calculating an eigenvector in the distance direction, an eigenvector in the elevation direction, and an eigenvector in the azimuth direction; and determining whether or not the detection data is in a one-dimensional gate in the distance direction, Range direction gate determination means for outputting a signal indicating the determination result, and determination as to whether or not the detection data is in a one-dimensional gate in the elevation angle direction, and outputs a signal indicating the determination result to detect the detection data. A first cross-range direction gate determination means for determining whether or not the vehicle is in a one-dimensional gate in the azimuth direction and outputting a signal indicating the determination result; the range direction gate determination means; As a result of the gate determination of the detection data by the gate determination unit 1 in the cross range direction, the detection data in the gate,
Based on the theory of the Kalman filter, a second data updating unit that converts the distance between the gate center and the detection data into the likelihood of a normal distribution, and uses the likelihood to update the data by 1 from the current time.
Prediction means for calculating the prediction error covariance matrix and the prediction vector after sampling, and the prediction error covariance matrix and the prediction vector input from the prediction means are delayed by one sampling to obtain the first gate determination parameter calculation means. Tracking, characterized by comprising: 1-sampling delay means for outputting the target track and display means for displaying the target track and the target speed from the smoothing vector input from the first data updating means. apparatus. 7. The tracking device according to claim 1, further comprising first gate size parameter setting means for setting a gate size parameter according to a sampling interval. 8. The observation error fixing means for setting the observation noise standard deviation of the distance, the observation noise standard deviation of the elevation angle, and the observation noise standard deviation of the azimuth angle to fixed values, further comprising: The tracking device according to any one of 3 above. 9. A sampling interval fixing means for inputting a control signal for fixing a sampling interval to a predicting means and fixing a sampling interval when calculating a prediction error covariance matrix and a prediction vector. The tracking device according to claim 1. 10. The tracking device according to claim 1, further comprising gate center fixing means for fixing the center of the gate for several samplings from the initial stage of tracking. 11. The gate size parameter varying means for changing the magnitude of the gate size parameter by using the magnitude of the velocity component of the smoothing vector obtained from the data updating means. The tracking device according to any one of 1 to 3. 12. A first display smoothing means for calculating a display smoothing vector by inputting the smoothing vector obtained from the data updating means to a display filter having a fixed time constant. Claim 1 to 3 characterized by the above-mentioned.
The tracking device according to any one of 1. 13. A second display smoothing means for calculating a display smoothing vector by inputting the smoothing vector obtained from the data updating means to a display filter having a variable time constant. The tracking device according to claim 12, which is characterized in that. 14. The observation means obtains an S / N ratio which is a signal-to-noise ratio, and when the S / N ratio is larger than a certain threshold value, it is determined as multipath, and when it is determined as multipath, A control signal for controlling the residual covariance matrix to be calculated by using a fixed value that is preset for the observation noise standard deviation of the elevation angle used when calculating the residual covariance matrix is
The tracking device according to any one of claims 1 to 3, further comprising a multipath determining means for outputting to the gate determining specification calculation means. 15. Using the condition number which is the ratio of the maximum value of the eigenvalue and the minimum value of the eigenvalue among the eigenvalues input from the first gate judgment parameter calculation means, and if the condition number is below a certain threshold value, The tracking device according to any one of claims 1 to 3, further comprising condition number determination means for inputting, to the display means, a danger signal indicating that the tracking calculation is unstable.