JP2002196070A - Tracking device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a computing load required for eliminating an unnecessary signal. SOLUTION: An S/N ratio minimum value residual covariance matrix computing means 2 uses an S/N ratio minimum value to compute a residual covariance matrix independent of a detection data, while a rough gate inside/outside determining means 3 computes a rough gate according to the residual covariance matrix to determine the detection data and the S/N ratio inside the rough gate. Each S/N ratio residual covariance matrix computing means 4 computes a sampling time, an amount of the detection data, and a residual covariance matrix depending on the number of tracks, while a fine gate inside/outside determining means 5 computes a fine gate to determine the detection data and the S/N ratio within the fine gate.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a tracking apparatus for reducing a calculation load in a high-density target tracking environment.

[0002]

2. Description of the Related Art FIG. 11 shows, for example, "Samuel
S. Blackman, Multiple-Target
t Tracking with Radar App
ligations, Artech House, De
dham, 1986 "p83-p92, especially p88,
It is a block diagram which shows the structure of the conventional tracking device which has the gate determination method shown by the formula 4.2.2 (4.6).

In FIG. 11, 1 is an observation means, and 4 is each S
/ N ratio residual covariance matrix calculating means, 5 is a fine gate inside / outside determining means, 6 is a data updating means, 7 is a predicting means, and 8 is a delay means.

FIG. 12 is a view for explaining a coordinate system relating to the observation means 1 in FIG. In FIG. 12, O denotes a sensor provided in the observation means 1, T denotes a tracking target, R denotes a distance between the tracking target T and the sensor O, and E denotes a sensor O and the tracking target T.
Is an elevation angle formed by a line segment OT connecting the sensor O and the tracking target T, and B is an azimuth angle formed by an orthogonal projection vector on the XY plane of the line segment OT connecting the sensor O and the tracking target T to the X axis. Further, [R,
[E, B] indicates polar coordinates, and [X, Y, Z] indicates north reference rectangular coordinates. Hereinafter, unless otherwise specified, the coordinates indicate the north reference rectangular coordinates.

Next, the operation will be described. Observation means 1
Represents detection data and a signal accompanying the detection data (sign)
al) An S / N ratio, which is a noise ratio (noise), is detected from an internal sensor using a S / N ratio minimum value that is a threshold value for detecting a target, and detection data and The S / N ratio accompanying the detection data is output. Here, the types of the detection data include a target signal of the tracking target and other unnecessary signals, and the unnecessary signals may be referred to as clutter in the future.

Here, the detection data is represented by D_kjAnd the detection day
TA D_kjThe S / N ratio associated with _kjAnd The subscript k is
Indicates the sampling time, and the subscript j is the sampling time k
Shows the number of pieces of detection data in. Detection data D_kjIs
It is expressed by the following equation (1).

(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 detected data. That is, the detection data D _kj indicates the j-th detection data at the sampling time k. Further, x _okj indicates an X coordinate component of the detection data D _kj , and similarly, y _okj indicates a Y coordinate component of the detection data D _kj , z
_okj are respectively the Z coordinate component of the detection data D _kj. Further, the symbol “′” indicates transposition of a vector and a matrix.

Each S / N ratio residual covariance matrix calculation means 4
Detection data D obtained from observation means 1_kjAnd detection data D
_kjS / N ratio SNR accompanying_kjAnd enter the remaining S / N ratio
Difference covariance matrix S_kj(T) is calculated and output. here,
Let each S / N ratio residual covariance matrix be S_kjThe subscript k of (t) is sun.
The subscript j indicates the j-th detection data D _kj
(T) indicates the track number
ing.

Each S / N ratio residual covariance matrix calculation means 4
Each S / N ratio residual covariance matrix S _kj (t) is calculated from the following equation (2) based on the Kalman filter theory.

(Equation 2) In the above formula (2), H is an observation matrix, P _pk (t) is a prediction error covariance matrix of the track number (t) obtained from the prediction means 7 to be described later via the delay means 8, and R _kj (t) is The observation error covariance matrix of the track number (t) in the north reference rectangular coordinates is shown.

The observation matrix H in the above equation (2) is represented by the following equation (3).

(Equation 3)

Further, the observation error covariance matrix R _kj (t) in the north reference rectangular coordinates in the above equation (2) is expressed by the following equation (4).

(Equation 4)In the above equation (4), Λ_kjIs the observation error in polar coordinates
The covariance matrix, Γ _kj(T) is the Λ of the track number (t)
_kjIs a transformation matrix of.

The observation error covariance matrix Λ _kj in the polar coordinates shown in the above equation (4) is shown in the following equation (5).

(Equation 5)

In the above equation (5), σr _kj is the standard deviation of the distance observation noise, σe _kj is the standard deviation of the elevation angle observation noise, and σb _kj is the standard deviation of the azimuth angle observation noise. Range observation noise follows a normal distribution with mean 0 and variance σr ² _kj . Similarly, elevation angle observation noise follows a normal distribution with mean 0 and variance σe ² _kj , and azimuth angle observation noise follows a normal distribution with mean 0 and variance σb ² _kj .

The transformation matrix Γ _kj (t) of Λ _kj of the track number (t) in the above equation (4) is represented by the following equation (6).

(Equation 6) R _pk (t), e _pk (t), b in the above equation (6)
_pk (t) is the position component x _pk (t), y _pk (t), and z _pk (t) of the position vector x of the prediction vector VX _pk (t) obtained via the delay means 8 from the prediction means 7 described later. It shows a polar coordinate component.

Also, r_pk(T), e_pk(T), b
_pk(T) and x_pk(T), y_pk(T), z _pk(T) Seki
The relation is expressed by the following equation (7).

(Equation 7)

[0016] (5) Distance measurement noise standard deviation .sigma.r _kj, elevation of observation noise standard deviation Sigma] e _kj, standard deviation .sigma.b _kj of observation noise in the azimuth angle in the equation, by using the S / N ratio SNR _kj , (9), and (10).

(Equation 8) Α _r and α in the above equations (8), (9) and (10)
_e and α _b are positive constants determined from the frequency of the radar transmission signal, the bandwidth of the radar transmission signal, the effective aperture of the radar antenna, and the speed of light.

The fine gate used in the fine gate inside / outside determination means 5 is an area represented by the following equation (11).

(Equation 9) In the above equation (11), VZ _pk (t) indicates the position component of the prediction vector VX _pk (t), and the superscript “−1” in the equation (11) is a symbol indicating the inverse matrix of the matrix.

Further, dε in the equation (11) is a gate size parameter for determining the size of the fine gate. The gate size parameter dε is a constant, and the subscript ε of the gate size parameter dε indicates the probability that the target exists in the fine gate. The probability that the target exists in the fine gate is referred to as the gate probability or the target target existence expectation probability. Here, the gate size parameter dε indicates the size of the gate when the probability ε within the gate is used, and the degree of freedom and the probability ε within the gate that one wants to set are determined in advance by the chi-square distribution in statistics. Squared density function, or chi2
Calculate from the power distribution table.

The prediction vector VX _pk (t) is expressed by the following equation (12).

(Equation 10) In equation (12), x _pk (t), y _pk (t), z
_pk (t) is the predicted position of X, Y, Z, dx _pk (t), d
y _pk (t) and dz _pk (t) are predicted speeds at X, Y, and Z predicted positions.

The prediction vector VX in the above equation (11)
_pk (t) and position component VZ _pk of (t), equation (12) the predicted vector VX _pk (t), using the observation matrix H of the equation (3), the following equation (13) Relationship There is.

[Equation 11]

The size of the fine gate represented by the above equation (11) is determined by the gate size parameter dε and each S / N ratio residual covariance matrix S _kj (t). Next, why the gate size parameter dε and the residual covariance matrix S _kj (t)
Whether or not the size of the fine gate is determined by will be described using an example. Hereinafter, in order to simplify the description, the following equation (14):
Assume a matrix A and a vector x as in equation (15). Then, the above equation (11) becomes the equation (14), (1)
From the expression (5), it is determined as the following expression (16).

[0022]

(Equation 12)

Here, since the matrix A in the above equation (14) is a positive-value symmetric matrix having 3 rows and 3 columns, the following equation (17):
Equation 8) holds.

(Equation 13)

Here, it is assumed that the matrix A can be made diagonal by a diagonal matrix B and an orthogonal matrix C by algebra as shown in the following equation (19).

[Equation 14]

The diagonal matrix B in the above equation (19) is obtained by calculating the eigenvalues λ ₁ , λ ₂ , λ of the matrix A as shown in the following equation (20).
Suppose that it is indicated by ₃ . Here, from the relationship of the above equation (18), the eigenvalues λ ₁ , λ ₂ , λ ₃ are λ ₁ > 0, λ ₂ , respectively.
> 0, λ ₃ > 0.

(Equation 15)

The orthogonal matrix B in the above equation (19)
Eigenvector v of column A₁, V_Two, V _ThreeUsing the following
It is shown as in equation (21). Where v₁, V_Two, V
_ThreeAre vectors of 3 rows and 1 column, respectively.

(Equation 16)

The eigenvectors v ₁ , v ₂ , and v ₃ of the matrix A in the above equation (21) have the relationship of the following equation (22).

[Equation 17]

The above equation (20) is expressed by the following equation (23) using the above equation (19).

(Equation 18)

Here, when the component of Cx in the above equation (23) is represented by the following equation (24),
Equation 3) is obtained by using the above equation (20) and the following equation (24).
It is shown as the following equation (25).

[Equation 19]

Here, from the above equations (14) and (20),
And the eigenvalue of matrix A is λ₁, Λ_Two, Λ _ThreeSo each S /
N-ratio residual covariance matrix S_kjThe eigenvalue of (t) is 1 / λ₁, 1
/ Λ_Two, 1 / λ_ThreeIt is. Therefore, the above equation (25) becomes
The radius of the error ellipse is the S / N ratio residual covariance matrix S
_kjEigenvalue 1 / λ of (t)₁, 1 / λ_Two, 1 / λ_Three,as well as
Show that it is determined by the gate size parameter dε
I have. In other words, the size of the fine gate is the same for each S / N ratio residual.
Covariance matrix S_kj(T) and gate size parameter dε
Is determined. Then, each S / N ratio residual covariance matrix S
_kjThe larger the (t) is, the more the S / N ratio residual covariance row
Column S_kjThe smaller the inverse matrix of (t), the larger the size of the fine gate
Becomes larger. Further, the gate size parameter dε is
The larger the size, the larger the size of the fine gate.

Further, the prediction vector VX of the above equation (11)
_pk (t) position component VZ _pk (t) of indicates centers rectification gate (11) is the right side, the position component of which is the center of the detection data D _kj and fine gate prediction vector VX _pk (t) VZ _pk
The distance normalized by the S / N ratio residual covariance matrix S _kj (t) in (t) is shown.

Therefore, in the fine gate inside / outside determination means 5, the S / N ratio residual covariance matrix of the detection data D _kj and the position component VZ _pk (t) of the predicted vector VX _pk (t), which is the center of the fine gate, If the distance normalized by S _kj (t) is smaller than the gate size parameter dε, that is, (1)
The detection data D _kj that satisfies the expression 1) is converted into a target signal candidate aD _kj.
Output as Further, the candidate of the target signal is defined as aD _kj
Are output as aS _kj (t). Furthermore, it outputs the prediction vector VX _pk (t) position component Le VZ _pk (t).

Here, the detection data D _kj is obtained from the observation means 1 and the prediction vector VX which is the center of the fine gate is obtained.
position component VZ _pk of _pk (t) (t) is below the prediction unit 7
The predicted vector VX _{pk + 1} (t) after one sampling with respect to the current sampling time k, which is the output of (1), is set to VX _pk (t) at the current time via delay means 8 described later, and
3) using the equation to calculate the position component VZ _pk predictive vector _{VX pk (t) (t)} . Each S / N ratio residual covariance matrix S
_kj (t) is obtained from each S / N ratio residual covariance matrix calculation means 4.

The data updating means 6 is provided with a Kalman filter.
Based on the theory, the gain matrix K_k(T) is calculated by the following equation (26).
Is calculated from Also, the smooth vector VX_sk(T) to (2
6) Gain matrix K of equation_kUsing (t), the following (27)
It is calculated by the formula. Further, a smooth error covariance matrix P
_sk(T) is calculated by the gain matrix K of the equation (26)_kUsing (t)
Then, it is calculated by the following equation (28). Then, update the data
The new means 6 is calculated by the equations (27) and (28).
Smooth vector VX_sk(T) and the smooth error covariance matrix P _sk
(T) is output.

[0035]

(Equation 20)

The prediction means 7 calculates the current sampling time k by the following equation (29) based on the Kalman filter theory.
The prediction vector VX _{pk + 1} (t) at time k + 1 after one sampling is calculated and output. Further, the prediction means 7 calculates and outputs the prediction error covariance matrix P _{pk + 1} (t) at the time k + 1 one sampling after the current sampling time k by the following equation (30).

(Equation 21)

In the above equation (29), the state transition matrix Φ _k for performing one sampling extrapolation is expressed by the following equation (31).

(Equation 22) Δt in the above equation (31) indicates a current time k and a sampling interval from the current time to a time k + 1 after one sampling.

The covariance matrix Q _k of the driving noise in the above equation (30) is expressed by the following equation (32).

(Equation 23) Σ _a in the above equation (32) indicates the standard deviation of the driving noise, and is a constant indicating the ambiguity of the target motion model of the Kalman filter.

The delay unit 8 delays the prediction vector VX _{pk + 1} (t) and the prediction error covariance matrix P _{pk + 1} (t) input from the prediction unit 7 by one sampling, and predicts the current time. The vector VX _pk (t) and the prediction error covariance matrix P _pk (t) at the current time are output to each S / N ratio residual covariance matrix means 4.

In the tracking apparatus as described above, the inverse matrix calculation of each S / N ratio residual covariance matrix S _kj (t) in the above equation (11) is performed for all detection data D _kj . For example, if the number of tracks one sampling before the current time is L and the number of _pieces of detection data D _kj at the current time is M, the number of tracks becomes L when the track number (t) is t =
, L, and the number of _pieces of detection data D _kj (the number j of _pieces of detection data) means that the subscript j indicating the number of _pieces of detection data D _kj in the detection data D _kj is j. = 1,..., M.

Then, using the above equation (11), when the processing of the fine gate is performed in all of these combinations, L
The calculation of M inverse matrices is required, and the calculation load on the computer becomes enormous. The meaning of the fine gate in the fine gate inside / outside determination means 5 means that the fine gate processing is performed in all the combinations described above.

Here, the wake means at most one search from each time.
Knowledge data D_kjIs time-series data configured. Sailing
There are two types of tracks, existing tracks and new tracks.
Existing tracks are tracks created up to the previous time, and new tracks
Is the detection data D with the current time _kjMore newly launched voyages
It is a trace.

[0043]

Since the conventional tracking apparatus is configured as described above, the S / N ratio residual covariance matrix S in the above equation (11) is applied to all the detection data D _{kj. When} the inverse matrix calculation of _kj (t) is performed, if the number of wakes one sampling before the current time is L and the number of detection data _Dkj at the current time is M, these are calculated using Expression (11). In order to perform fine gate processing in all combinations, LM inverse matrix calculations are required, and there has been a problem that the calculation load on the computer becomes enormous.

The present invention has been made to solve the above-mentioned problems. In the operation of a fine gate for removing unnecessary signals from a target signal, the target signal is processed by the fine gate without applying a calculation load. The aim is to obtain a tracking device that can be captured inside.

[0045]

A tracking device according to the present invention uses detection data detected from an internal sensor using a minimum value of an S / N ratio, which is a threshold value for detecting a target. Observation means for outputting an S / N ratio associated with the detection data, and an observation error covariance at the current time which does not depend on the number of detection data calculated by using the output of the observation means and calculated using the minimum value of the S / N ratio From the matrix R _kmin (t), a residual covariance matrix S _kmin (t) independent of the number of detection data is calculated, and the calculated residual covariance matrix S _kmin (t), the input detection data and S / N S / N ratio minimum residual covariance matrix calculating means for outputting a ratio, and the residual covariance matrix S input from the S / N ratio minimum residual covariance matrix calculating means.
_{Based on kmin} (t), a coarse gate which is a target existence expected area is calculated, and for the input detection data and S / N ratio, the detection data and S / N ratio included in the coarse gate are determined, and the determined coarse is determined. Detection data and S in the gate
Using the coarse gate inside / outside determination means for outputting the / N ratio and the output of the coarse gate inside / outside determination means, each S / S by the coarse gate depending on the sampling time, the number of detected data, and the number of tracks
An N-ratio residual covariance matrix S _kj (t) is calculated, each calculated S / N ratio residual covariance matrix S _kj (t), and the detection data and S / Means for calculating the S / N ratio residual covariance matrix for outputting the N ratio;
Using the output of the ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, and the detection data and the S / N ratio contained in the coarse gate input are fine gates. Detection data and S / N contained in
The ratio is determined, and the S / N ratio residual covariance matrix S _kj (t) by the fine gate, which depends on the sampling time, the number of detected data, and the number of wakes, is calculated. It includes a fine gate inside / outside determination means for outputting the data, the S / N ratio, and the calculated S / N ratio residual covariance matrix S _kj (t).

The tracking device according to the present invention uses the S / N ratio minimum value, which is a threshold value for detecting a target, to detect detection data detected from an internal sensor and to detect S associated with the detection data. / N ratio output, an output of the observation unit, a prediction error covariance matrix of the current time and a prediction vector of the current time output by the delay unit described later,
An observation error covariance matrix R _kmin (t) at the current time which does not depend on the number of detection data calculated using the minimum value of the S / N ratio;
Using the input prediction error covariance matrix at the current time, a residual covariance matrix S _kmin (t) independent of the number of detection data is calculated, and the calculated residual covariance matrix S _kmin (t) and the input S / N ratio minimum residual covariance matrix calculating means for outputting the detected detection data, S / N ratio, current time prediction error covariance matrix and current time prediction vector, and the S / N ratio minimum value residual The residual covariance matrix S _kmin input from the covariance matrix calculation means
Based on (t) and the prediction vector at the current time, a coarse gate, which is a target existence expectation area, is calculated, and for the input detection data and S / N ratio, the detection data and S / N ratio contained in the coarse gate are determined. A coarse gate inside / outside determining means for outputting the detected data and the S / N ratio contained in the determined coarse gate, and the inputted prediction error covariance matrix of the current time and the prediction vector of the current time; Using the output of the determination means, each S / N ratio residual covariance matrix S _kj (t) based on the coarse gate, which depends on the sampling time, the number of detection data, and the number of wakes, is calculated. Ratio residual covariance matrix S _kj (t), input detection data in the coarse gate, S / N ratio in the coarse gate, prediction error covariance matrix of the current time, and prediction of the current time Each S / that outputs a vector Using the output of the ratio residual covariance matrix calculating means and the output of each of the S / N ratio residual covariance matrix calculating means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, and the Detection data and S
For the / N ratio, the detection data and the S / N ratio contained in the fine gate are determined, and the S / N ratio residual covariance matrix S by the fine gate depends on the sampling time, the number of detection data, and the number of tracks. _kj (t) is calculated, the detection data and the S / N ratio contained in the determined fine gate, and the calculated S / N
Using the output of the fine gate inside / outside determination means for outputting the ratio residual covariance matrix S _kj (t), the input prediction error covariance matrix of the current time and the prediction vector of the current time, and the output of the fine gate inside / outside determination means A data updating means for calculating and outputting the current time smoothed error covariance matrix and the current time smoothed vector,
A prediction means for calculating and outputting a prediction error covariance matrix one sampling after the current time and a prediction vector one sampling after the current time, using an output of the data updating means, and an output of the prediction means. , A delay means for outputting a prediction error covariance matrix at the current time and a prediction vector at the current time.

In the tracking apparatus according to the present invention, each S / N ratio residual covariance matrix S _kj output from the fine gate inside / outside determination means is provided.
The value of the S / N ratio minimum value used by the observation unit and the S / N ratio minimum residual covariance matrix calculation unit is changed according to the number of (t) and the number of detection data output by the observation unit. And an S / N ratio minimum value control means for notifying the observation means and the S / N ratio minimum value residual covariance matrix calculation means, wherein the observation means and the S / N ratio minimum value residual covariance matrix are provided. The calculating means performs processing using the changed S / N ratio minimum value.

The tracking device according to the present invention uses the S / N ratio minimum value, which is a threshold value for detecting a target, to detect detection data detected from a sensor provided therein, and to detect S associated with the detection data. / N ratio is selected, and the maximum value component of the observation error covariance matrix components is selected based on the maximum detection distance of the radar, and the sampling time is calculated using the maximum value component of the selected observation error covariance matrix. , Number of detected data,
Calculate the observation error covariance matrix R _min independent of the number of wakes, and use the output from the observation means and the calculated observation error covariance matrix R _min to obtain a simple residual covariance matrix independent of the number of detected data. S ′ _kmin (t) is calculated and the calculated simple residual covariance matrix S ′ _kmin (t), and the input detection data and the S / N ratio minimum value residual covariance matrix that outputs the S / N ratio Calculating means and a simple residual covariance matrix S ′ input from the S / N ratio minimum residual covariance matrix calculating means.
_{Based on kmin} (t), a coarse gate which is a target existence expected area is calculated, and for the input detection data and S / N ratio, the detection data and S / N ratio included in the coarse gate are determined, and the determined coarse is determined. Detection data and S in the gate
Using the coarse gate inside / outside determination means for outputting the / N ratio and the output of the coarse gate inside / outside determination means, each S / S by the coarse gate depending on the sampling time, the number of detected data, and the number of wakes is used.
An N-ratio residual covariance matrix S _kj (t) is calculated, each calculated S / N ratio residual covariance matrix S _kj (t), and the detection data and S / Means for calculating the S / N ratio residual covariance matrix for outputting the N ratio;
Using the output of the ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, and the detection data and the S / N ratio contained in the coarse gate input are fine gates. Detection data and S / N contained in
The ratio is determined, and the S / N ratio residual covariance matrix S _kj (t) by the fine gate, which depends on the sampling time, the number of detected data, and the number of wakes, is calculated. It includes a fine gate inside / outside determination means for outputting the data, the S / N ratio, and the calculated S / N ratio residual covariance matrix S _kj (t).

The tracking device according to the present invention uses the minimum value of the S / N ratio, which is the threshold value for detecting the target, to detect the detection data detected from the sensor provided therein and the S data accompanying the detection data. / N ratio is selected, and the maximum value component of the observation error covariance matrix components is selected based on the maximum detection distance of the radar, and the sampling time is calculated using the maximum value component of the selected observation error covariance matrix. , Number of detected data,
An observation error covariance matrix R _min independent of the number of wakes is calculated, an output from the observation means, a current time prediction error covariance matrix from a delay means described later, and a current time prediction vector,
Using the calculated observation error covariance matrix R _min and a simple residual covariance matrix S ′ independent of the number of detected data.
_kmin (t) is calculated, the calculated simple residual covariance matrix S ′ _kmin (t), and the input detection data, S / N
S / N ratio minimum residual covariance matrix calculation means for outputting the ratio, the current time prediction error covariance matrix and the current time prediction vector, and input from the S / N ratio minimum residual covariance matrix calculation means. From the simplified residual covariance matrix S ′ _kmin (t) and the prediction vector at the current time, a coarse gate that is a target existence expected area is calculated, and the input detection data and the S / N ratio _fall within the coarse gate. The detected detection data and the S / N ratio are determined, and the detection data and the S / N ratio contained in the determined coarse gate, the input prediction error covariance matrix of the current time and the prediction vector of the current time are output. Using the output of the coarse gate inside / outside determination means and the coarse gate inside / outside determination means,
Depends on sampling time, number of detected data, number of tracks,
Each S / N ratio residual covariance matrix S _kj (t) by the coarse gate is calculated, and each calculated S / N ratio residual covariance matrix S
_kj (t), the input detection data contained in the coarse gate, the S / N ratio contained in the coarse gate, the prediction error covariance matrix of the current time, and each S / P which outputs the prediction vector of the current time. Using the output of the N ratio residual covariance matrix calculation means and the output of each of the S / N ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, and the input coarse gate is input. The detection data and the S / N ratio contained in the fine gate are determined with respect to the detection data and the S / N ratio contained therein, and each S by the fine gate depends on the sampling time, the number of detected data, and the number of wakes. / N ratio residual covariance matrix S
_kj (t) is calculated, and the detection data and S / N ratio contained in the determined fine gate, each calculated S / N ratio residual covariance matrix S _kj (t), and prediction of the input current time Using the output of the fine gate inside / outside determination means for outputting the error covariance matrix and the prediction vector of the current time, and calculating the smooth error covariance matrix of the current time and the smooth vector of the current time using the output of the fine gate inside / outside determination means A data updating means for outputting, a prediction error covariance matrix one sampling after the current time and a prediction vector one sampling after the current time using the output of the data updating means, and A delay means for outputting a prediction error covariance matrix at the current time and a prediction vector at the current time by using an output of the prediction means.

The tracking device according to the present invention uses the minimum value of the S / N ratio, which is a threshold value for detecting a target, to detect detection data detected from a sensor provided therein and to detect an S value accompanying the detection data. / N ratio is selected, and the maximum value component of the observation error covariance matrix components is selected based on the maximum detection distance of the radar, and the sampling time is calculated using the maximum value component of the selected observation error covariance matrix. , Number of detected data,
The observation error covariance matrix R _min independent of the number of wakes is calculated, and the output from the observation means and the calculated observation error covariance matrix R _min are used to calculate the residual covariance independent of the number of detection data and the number of wakes. calculating a matrix E _k, calculated residual covariance matrix E _k, and the S / N ratio minimum residual covariance matrix calculation means for outputting the inputted detection data and the S / N ratio, the S / N ratio Based on the residual covariance matrix E _k input from the minimum residual covariance matrix calculation means, a coarse gate which is a target existence expected area is calculated, and the input detection data and S / N ratio are entered into the coarse gate. Using the output of the coarse gate inside / outside determination means for determining the detection data and the S / N ratio within the coarse gate, and outputting the detection data and the S / N ratio contained in the determined coarse gate; Sampling time, number of detected data, wake It depends on, to calculate the respective S / N ratio residual covariance matrix S _kj the crude gate (t), calculated the S / N ratio residual covariance matrix S _kj (t), as well as input crude gate Using the output of each of the S / N ratio residual covariance matrix calculating means for outputting the detection data and the S / N ratio contained therein, and the output of each of the S / N ratio residual covariance matrix calculating means, A fine gate, which is a narrower target existence expectation area, is calculated, and the detection data and S
For the / N ratio, the detection data and the S / N ratio contained in the fine gate are determined, and the S / N ratio residual covariance matrix S by the fine gate depends on the sampling time, the number of detection data, and the number of tracks. _kj (t) is calculated, and the detection data and the S / N ratio contained in the determined fine gate and the calculated S / N ratio residual covariance matrix S _kj (t) are output from the fine gate. Means.

The tracking device according to the present invention uses the minimum value of the S / N ratio, which is the threshold value for detecting the target, to detect the detection data detected from the sensor provided therein and the S accompanying the detection data. / N ratio is selected, and the maximum value component of the observation error covariance matrix components is selected based on the maximum detection distance of the radar, and the sampling time is calculated using the maximum value component of the selected observation error covariance matrix. , Number of detected data,
An observation error covariance matrix R _min independent of the number of wakes is calculated, an output from the observation means, a current time prediction error covariance matrix from a delay means described later, and a current time prediction vector,
Using the calculated observation error covariance matrix R _min , the number of detection data, a residual covariance matrix E _k independent of the number of wakes is calculated, the calculated residual covariance matrix E _k , and the input detection data, S / N ratio minimum residual covariance matrix calculation means for outputting the S / N ratio, the current time prediction error covariance matrix and the current time prediction vector, and the S / N ratio minimum residual covariance matrix calculation Based on the residual covariance matrix E _k input from the means and the prediction vector at the current time, a coarse gate which is a target existence expected area is calculated, and the input detection data and S / N ratio are detected in the coarse gate. The data and the S / N ratio are determined, and the detection data and the S / N ratio contained in the determined coarse gate, and the input / output of the current time prediction error covariance matrix and the current time prediction vector are output inside and outside the coarse gate. Determining means; Using the output of bets outside judgment means,
Depends on sampling time, number of detected data, number of tracks,
Each S / N ratio residual covariance matrix S _kj (t) by the coarse gate is calculated, and each calculated S / N ratio residual covariance matrix S
_kj (t), the input detection data contained in the coarse gate, the S / N ratio contained in the coarse gate, the prediction error covariance matrix of the current time, and each S / P which outputs the prediction vector of the current time. Using the output of the N ratio residual covariance matrix calculation means and the output of each of the S / N ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, and the input coarse gate is input. The detection data and the S / N ratio contained in the fine gate are determined with respect to the detection data and the S / N ratio contained therein, and each S by the fine gate depends on the sampling time, the number of detected data, and the number of wakes. / N ratio residual covariance matrix S
_kj (t) is calculated, and the detection data and S / N ratio contained in the determined fine gate, each calculated S / N ratio residual covariance matrix S _kj (t), and prediction of the input current time Using the output of the fine gate inside / outside determination means for outputting the error covariance matrix and the prediction vector of the current time, and calculating the smooth error covariance matrix of the current time and the smooth vector of the current time using the output of the fine gate inside / outside determination means A data updating means for outputting, a prediction error covariance matrix one sampling after the current time and a prediction vector one sampling after the current time using the output of the data updating means, and A delay means for outputting a prediction error covariance matrix at the current time and a prediction vector at the current time by using an output of the prediction means.

The tracking device according to the present invention detects a target.
Using the minimum value of the S / N ratio
Detection data and detection data detected from sensors
Observation means for outputting the S / N ratio attached to the data, and radar
Of the observation error covariance matrix components based on the maximum detection distance of
Select the maximum value component and calculate the maximum value of the selected observation error covariance matrix.
Large-value components and detection data and
For selecting the maximum value of the observation error covariance matrix that outputs the S / N ratio
And the output of the observation error covariance matrix maximum value selecting means.
Using a fixed sampling interval T_intBy sump
Radius r independent of ring time, number of detected data, number of tracks
Calculate the coarse gate of adius0 and input the detection data
And the S / N ratio detected in the coarse gate
Data and S / N ratio and enter the determined coarse gate
Inside and outside the coarse gate that outputs the detected detection data and S / N ratio
Using the output of the determination means and the rough gate inside / outside determination means.
Depends on the sampling time, the number of detected data, and the number of wakes.
, Each S / N ratio residual covariance matrix S by the coarse gate
_kj(T) is calculated, and the calculated S / N ratio residual covariance row is calculated.
Column S _kj(T), as well as within the input coarse gate
S / N ratio residual co-division outputting detection data and S / N ratio
Scatter matrix calculation means, and calculation of each of the above S / N ratio residual covariance matrices
Using the output of the means, the target existence expectation area narrower than the coarse gate
Calculate the fine gate, which is the area, and enter the coarse gate
The detected data and the S / N ratio
Judgment of the detected detection data and S / N ratio, and sampling
Precise gate depending on the scanning time, the number of detected data, and the number of tracks
S / N ratio residual covariance matrix S_kj(T) is calculated,
Detection data and S / N contained in the determined fine gate
Ratio and each calculated S / N ratio residual covariance matrix S
_kj(G) output means for determining whether the gate is inside or outside
It is.

The tracking device according to the present invention uses the minimum value of the S / N ratio, which is a threshold value for detecting a target, to detect detection data detected from a sensor provided therein and the S data accompanying the detection data. A maximum value component of the observation error covariance matrix component based on the maximum detection distance of the radar, and the maximum value component of the selected observation error covariance matrix, which is input from the observation unit. Detection data and S
/ N ratio, an observation error covariance matrix maximum value selecting means for outputting a prediction error covariance matrix at the current time and a prediction vector at the current time from a delay means described later, and an observation error covariance matrix maximum value selecting means. using the output, the fixed sampling interval T _int, the sampling time, the detection data number,
A coarse gate having a radius of radius 0 that does not depend on the number of wakes is calculated, and for the input detection data and S / N ratio, the detection data and the S / N ratio contained in the coarse gate are determined.
Detection data and S / N contained in the determined coarse gate
Using the output of the coarse gate inside / outside determination means for outputting the ratio, and the input prediction error covariance matrix of the current time and the prediction vector of the current time, the sampling time, the number of detected data, and the number of wakes , The respective S / N ratio residual covariance matrices S _kj (t) by the coarse gate are calculated, and the calculated S / N ratio residual covariance matrices S _kj (t) and the input coarse gate S / N ratio residual covariance matrix calculating means for outputting the detection data contained in the S / N ratio, the S / N ratio contained in the coarse gate, the prediction error covariance matrix at the current time, and the prediction vector at the current time. Using the output of each S / N ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, and the detection data and S / About N ratio, it is in the fine gate Determines knowledge data and the S / N ratio, the sampling time, the detection data number depends on the number of track, calculates the S / N ratio residual covariance matrix by rectification gate S _kj (t), the determined fine gate Detection data and S / N ratios contained in each of the calculated S / N ratio residual covariance matrices S
_kj (t), the input prediction error covariance matrix of the current time and the input current time prediction vector, and the output of the fine gate internal / external determination means are used to calculate the smoothed error of the current time. A data updating unit that calculates and outputs a variance matrix and a smooth vector at the current time; and a prediction error covariance matrix one sampling after the current time and a prediction one sampling after the current time using the output of the data updating unit. A prediction means for calculating and outputting a vector, and a delay means for outputting a prediction error covariance matrix at the current time and a prediction vector at the current time using the output of the prediction means.

[0054]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below. Embodiment 1 FIG. FIG. 1 is a block diagram showing a configuration of a tracking device according to Embodiment 1 of the present invention.
Detection data D _kj and detection data D output by the observation means 1
_The target is detected by inputting the S / N ratio SNR _kj attached to _kj , the current time prediction error covariance matrix P _pk (t) and the current time prediction vector VX _pk (t) output by the delay means 8. Time k, detection data number j, and S / N ratio minimum value SNR independent of the number of wakes, which are threshold values for
_The detection is performed using the observation error covariance matrix R _kmin (t) at the current time which does not depend on the detection data number j and the input prediction error covariance matrix P _pk (t) at the current time. and not calculated a residual covariance matrix S _{km in} (t) depending on the data number j, calculated residual covariance matrix S _kmin (t), and inputs the detection data D _kj, accompanying the detection data D _kj S
/ N ratio SNR _kj , prediction error covariance matrix P at the current time
S / N ratio minimum residual covariance matrix calculation means for outputting _pk (t) and the current time prediction vector VX _pk (t).

In FIG. 1, reference numeral 3 denotes a residual covariance matrix S from the S / N ratio minimum residual covariance matrix calculating means 2.
_kmin (t), detection data D _kj , S / N ratio SNR _kj attached to detection data D _kj , prediction error covariance matrix P _{pk at the} current time
(T) and the predicted vector VX _pk (t) at the current time are input, and the input residual covariance matrix S _kmin (t) and the predicted vector VX _pk (t) at the current time are used to calculate the rough calculating a gate, the S / N ratio SNR _kj accompanying detection data D _kj and the detection data D _kj inputted, the detection data D _kj and detection data D is within the coarse gate
determining the S / N ratio SNR _kj accompanying _kj, data detection has entered the determined coarse gate in D _kj and detection data D _kj
And a coarse gate inside / outside determination means for outputting an input S / N ratio SNR _kj , an input prediction error covariance matrix P _pk (t) at the current time, and a prediction vector VX _pk (t) at the current time. The configuration is the same as that shown in FIG.

Next, the S / N ratio minimum residual covariance matrix calculation means 2 will be described in detail. In the signal detection in the radar signal processing of the observation means 1, it is determined that the target has been detected with a signal having a power equal to or more than a certain value with respect to the noise power δ. The S / N ratio is a threshold for determining that the target is detected to a certain S / N minimum SNR _{mi n.} That is, it is assumed that a signal having a power _{equal to} or more than δ · SNR _min is detected. At this time, the S / N ratio SN of the detection data D _kj
R _kj satisfies the following equation (33).

(Equation 24)

Further, the observation noise of the distance, the elevation angle, and the azimuth angle calculated from the S / N minimum value SNR _min are represented by σr _min ,
σe _min , σb _min , σr _min , σe _min , σ
b _{mi n,} the above equation (8), (9), from (10),
The following equations (34), (35), and (36) are obtained.

(Equation 25)

The above equations (33), (34) and (35)
From equation (36), the following equation (37), equation (38):
The relationship of equation (39) holds.

(Equation 26)

Further, the above equations (34), (35) and (3)
From equation (6), the S / N ratio minimum value SNR _minCalculated from
Observation error covariance matrix in polar coordinates Λ_minIs the above (5)
From the equation, it is expressed by the following equation (40).

[Equation 27]

Here, when the above equations (5) and (40) are compared, the above equations (37), (38) and (39)
From the equation, the following (41) is established.

[Equation 28] The above equation (41) indicates that Λ _min −Λ _kj expressed in polar coordinates is
It shows that it is a positive value symmetric matrix by algebra.

Further, from the result of the equation (41) calculated from the S / N ratio minimum value SNR _min, the observation error covariance matrix R in the north reference Cartesian coordinates independent of the number of detection data j is obtained.
_kmin (t) becomes the following equation (42) from the above equation (4).

(Equation 29)

From the above equations (41), (4) and (42), the following equation (43) is obtained.

[Equation 30]

Further, S / S which does not depend on the detection data number j
The N-ratio minimum residual covariance matrix S _kmin (t) is given by the following (4)
4) Calculated by equation.

(Equation 31)

Since the above equations (2) and (44) are positive value symmetric matrices by algebra, when comparing equations (2) and (44), the following equation is obtained from the relationship of the above equation (43). Equation (45) is obtained.

(Equation 32) The result of the above equation (45) is the S / N ratio S of the detection data D _kj.
The detection data D _kj in the fine gate calculated from NR _kj is
This indicates that the signal always exists in the coarse gate calculated from the S / N ratio minimum value SNR _min .

Therefore, the S / N ratio minimum residual covariance matrix calculation means 2 calculates the above equations (34), (35), (3)
Using the expressions 6), (40), (42), and (44), the S / N ratio minimum value residual independent of the number of detected data j at the S / N ratio minimum value SNR _min Covariance matrix S
Calculate and output _kmin (t).

Next, the rough gate inside / outside determination means 3 will be described in detail. In the coarse gate inside / outside determination means 3, the S / N ratio minimum independent of the number j of detected data when the S / N ratio minimum value SNR _min calculated by the S / N ratio minimum value residual covariance matrix calculation means 2 is obtained. Using the value residual covariance matrix S _kmin (t),
Rough gate determination is made by the following equation (46), and (46)
And outputs the S / N ratio SNR _kj accompanying detection data D _kj and the detection data D _kj satisfies the expression.

[Equation 33]

Therefore, in the conventional tracking device, the above (1)
Using equation (1), all detection data D _kjWith respect to (1
Each S / N ratio residual covariance matrix S in equation (1)_kj(T)
Because the inverse matrix was to be calculated, the calculation of the computer
The load is enormous. But as mentioned above,
L is the number of tracks one sampling before the hour, and
Assuming that the number of detected data to be detected is M,
/ N ratio minimum residual covariance matrix S_kmin(T) is a sampler
Although it depends on the scanning time k and the number of wakes,
Are independent values, and only L inverse matrices need to be calculated. Ma
The number of detected data after the coarse gate determination by equation (46)
Not only can j be narrowed down, but also the number of wakes decreases,
The inside / outside determination means 5, the data updating means 6, and the prediction means 7
The calculation load of calculation in each process is reduced.

Next, the operation will be described. FIG. 2 is a flowchart showing a processing flow of the tracking device according to the first embodiment of the present invention. In step ST1, the observation means 1 does not depend on the sampling time k, the number of detected data j, and the number of wakes, which are threshold values for detecting a target.
Using the / N ratio minimum value SNR _min , the detection data D _kj detected from the sensor provided therein and the S / N ratio SNR _kj accompanying the detection data D _kj are converted into the S / N ratio minimum value residual covariance. Output to the matrix calculation means 2.

In step ST2, the S / N ratio minimum value
The residual covariance matrix calculation means 2 calculates the search result output by the observation means 1.
Knowledge data D_kjAnd detection data D_kjS / N ratio S associated with
NR _kj, And the prediction error of the current time output by the delay means 8
Covariance matrix P_pk(T) and prediction vector VX at the current time_pk
(T) and input the S / N ratio minimum value SNR_minUsing
Observation error at the current time independent of the number of detected data j
Covariance matrix R_kmin(T) and the prediction error of the input current time
Covariance matrix P_pk(T) and the number of detected data j
Nonexistent residual covariance matrix S_kminCalculate (t) and calculate
The residual covariance matrix S_kmin(T) and the input detection data
Data D_kj, Detection data D_kjS / N ratio SN associated with
R_kj, The prediction error covariance matrix P at the current time_pk(T) and current time
Hourly predicted vector VX_pk(T) is a rough gate inside / outside determination means
Output to 3.

In step ST3, the coarse gate inside / outside
The determination means 3 receives the input residual covariance matrix S_kmin(T) and
Predicted vector VX at current time_pkAccording to (t), target existence period
A rough gate which is a waiting area is calculated, and the input detection data D
_kjAnd detection data D_kjS / N ratio SNR accompanying_kjNitsu
And the detection data D contained in the coarse gate_kjAnd detection
Data D_kjS / N ratio SNR accompanying_kjJudge and judge
Detection data D contained in the rough gate_kjAnd detection
Data D_kjS / N ratio SNR accompanying_kj, And entered
Current time prediction error covariance matrix P_pk(T) and the current time
Measurement vector VX _pk(T) is calculated for each S / N ratio residual covariance matrix
Output to the output means 4.

In step ST4, each S / N ratio residual covariance matrix calculation means 4 calculates using the S / N ratio SNR _kj accompanying the detection data D _kj contained in the input coarse gate. The observation error covariance matrix R _kj (t) at the current time depending on the sampling time k, the number of detection data j, and the number of wakes
And the input prediction error covariance matrix P _pk (t) at the current time, each S / N ratio residual covariance matrix by a coarse gate, which depends on the sampling time k, the number of detection data j, and the number of tracks S _kj (t) is calculated, and each calculated S / N ratio residual covariance matrix S _kj (t), the input detection data D _kj contained in the coarse gate, and the detection data D _kj contained in the coarse gate The S / N ratio SNR _kj associated with the detection data D _kj , the prediction error covariance matrix P _pk (t) at the current time, and the prediction vector VX _{pk at the} current time
(T) is output to the fine gate inside / outside determination means 5.

In step ST5, the fine gate inside / outside determination means 5 determines whether the input S / N ratio residual covariance matrix S
_{Based on kj} (t) and the prediction vector VX _pk (t) at the current time, a fine gate which is a target existence expected area narrower than the coarse gate is calculated, and the detection data D _kj and the detection data D contained in the input coarse gate are calculated. for S / N ratio SNR _kj accompanying _kj, determines the S / N ratio SNR _kj accompanying detection data D _kj and the detection data D _kj is within seminal gate, detection data contained in the fine gate S / N accompanying D _kj
The current-time observation error covariance matrix R _kj (t) that depends on the sampling time k, the number of detection data j, and the number of wakes calculated using the ratio SNR _kj , and the input prediction error covariance matrix P _pk at the current time Using (t), each S / S by the fine gate depends on the sampling time k, the number of detected data j, and the number of wakes.
N ratio calculating a residual covariance matrix S _kj (t), detection is within the determined fine gate data D _kj and detection data D _kj
Accompanying S / N ratio SNR _kj, each S / N ratio residual covariance matrix S _kj calculated (t), as well as the prediction vector of the prediction error covariance matrix P _pk (t) and the current time of the current time entered VX
_pk (t) is output to the data updating means 6.

At step ST6, data updating means
6 is the detection data D contained in the fine gate input_kj
S / N ratio SNR accompanying_kj, Each S / N ratio residual covariance row
Column S _kj(T) and the prediction error covariance matrix P at the current time
_pkAccording to (t), the smoothing error covariance at the current time, which is the smoothing information.
Matrix P_sk(T) is calculated, and within the fine gate input
Detection data D contained in_kjAnd the prediction vector of the current time
VX_pkThe smooth vector of the current time which is the smooth information by (t)
Le VX_sk(T) is calculated, and the calculated current time smoothing error is calculated.
Covariance matrix P_sk(T), current time smoothed vector VX
_sk(T) is output to the prediction means 7.

In step ST7, the prediction means 7
Input smoothed error covariance matrix P at the current time_skAccording to (t)
Prediction after one sampling from the current time
Error covariance matrix P_{pk + 1}(T) is calculated and input.
The current time smoothed vector VX_skAccording to (t), the prediction information
Is the predicted vector VX one sampling after the current time
_{pk + 1}(T) is calculated.

In step ST8, if the processing is to be continued according to the instruction of the operator or automatic determination, the prediction error covariance matrix P _{pk + 1} (t) after one sampling from the current time calculated by the prediction means 7, the current time The prediction vector VX _{pk + 1} (t) after one sampling is output to the delay unit 8.

In step ST9, the delay means 8
The prediction error covariance matrix P _{pk + 1} (t) after one sampling from the input current time is delayed by one sampling, and the prediction error covariance matrix P _pk (t) at the current time is reduced to the S / N ratio minimum residual. The prediction vector VX _{pk + 1} (t) that is output to the covariance matrix calculation means 2 and one sampling after the input current time.
Is delayed by one sampling, and the prediction vector V at the current time is
X _pk (t) is calculated as S / N ratio minimum value residual covariance matrix calculation means 2
And the process from step ST1 is repeated.

As described above, according to the first embodiment, the coarse gate inside / outside determination means 5 determines whether the residual value does not depend on the number j of detection data calculated by the S / N ratio minimum residual covariance matrix calculation means 2. A coarse gate, which is a target existence expected area, is calculated from the difference covariance matrix S _kmin (t), and the input detection data D _kj
And detecting the S / N ratio SNR _kj accompanying data D _kj, the S / N ratio SNR _kj accompanying detection data D _kj and the detection data D _kj is within the coarse gate determines, seminal gate outside judgment means 5 calculates a fine gate that is a target existence expected area smaller than the coarse gate, and detects the detection data D _kj and the detection data D _kj contained in the input coarse gate.
/ N ratio SNR _kj , the detection data D _kj contained in the fine gate and the S / N ratio SN accompanying the detection data D _kj
By determining R _kj , the calculation load of the fine gate for eliminating unnecessary signals for the target signal in the fine gate inside / outside determination means 5 is reduced, and the data updating means 6 and the prediction means 7 perform the processing. The effect of reducing the calculation load of the calculation is obtained.

Embodiment 2 FIG. 3 is a block diagram showing a configuration of a tracking apparatus according to Embodiment 2 of the present invention. In the figure, reference numeral 12 denotes detection data D _kj output from the observation means 1 and an SNR ratio SNR accompanying the detection data D _{kj. kj} and the prediction error covariance matrix P _pk at the current time output by the delay means 8
(T) and the prediction vector VX _pk (t) at the current time are input, the maximum value component of the observation error covariance matrix R _kj component is selected by the maximum detection range RANGE _max of the radar, and the selected observation error covariance matrix R By the maximum value component of the _kj component, the observation error covariance matrix R _min in the north reference Cartesian coordinates independent of the sampling time k, the number of detection data j, and the number of wakes.
Is calculated, and using the calculated observation error covariance matrix R _min and the input prediction error covariance matrix P _pk (t) at the current time, a simple residual covariance matrix S independent of the number j of detection data. 'calculates _kmin (t), calculated simple residual covariance matrix S' _kmin (t), and inputs the detection data D _kj, S / N ratio SNR _kj accompanying detection data D _kj, the current time S / N ratio minimum residual covariance matrix calculation means for outputting the prediction error covariance matrix P _pk (t) and the prediction vector VX _pk (t) at the current time.

In FIG. 3, reference numeral 13 denotes a simple residual covariance matrix S ′ _kmin (t) from the S / N ratio minimum residual covariance matrix calculating means 12, detection data D _kj , and detection data D
S / N ratio SNR _kj accompanying _kj, the prediction vector VX of the prediction error covariance matrix P _pk (t) and the current time of the current time
_pk (t), and the input simple residual covariance matrix S ′
_{Based on kmin} (t) and the prediction vector VX _pk (t) at the current time, a coarse gate that is a target presence expectation area is calculated, and the input detection data D _kj and S / S attached to the detection data D _kj are calculated.
For the N ratio SNR _kj , the detection data D _kj contained in the coarse gate and the S / N ratio SNR accompanying the detection data D _kj
determines _kj, accompanying the detection data D _kj and the detection data D _kj is within determines crude gate S / N ratio SN
R _kj and the input prediction error covariance matrix P _pk at the current time.
(T) and a rough gate inside / outside determination means for outputting the prediction vector VX _pk (t) at the current time. Other configurations are the same as those in FIG. 1 of the first embodiment.

Next, the S / N ratio minimum residual covariance matrix calculation means 12 will be described in detail. Where the radar
Since the value of the maximum detection distance RANGE _max to be processed is known, as shown in the following equation (47), the observation error covariance matrix R
The maximum value component of the _kj component can be defined.

(Equation 34) In the above expression (47), σcr _min, σce _min, σ
cb _min is a positive constant. Also, max
select (A, B, C) is A, B, C,. . . Is a function that selects one maximum value from

Here, as shown in the following equation (48), the sampling time k, the number of detection data j, and the number of wakes do not depend.
When the observation error covariance matrix R _min in the north reference rectangular coordinates is defined, the number of detection data j in the above equation (44)
And the observation error covariance matrix R _kmin (t) that does not depend on the following equation (49).

(Equation 35)

Therefore, R in the above equation (44)
_{When kmin} (t) is replaced by R _min in the above equation (48),
The following equation (50) is obtained.

[Equation 36] The prediction error covariance matrix P _pk (t) in the above equation (50)
Has a different value for each track number (t), so (5
HP _pk (t) H ′ in equation (0) and residual covariance matrix S
_min (t) also has a different value for each track number (t).

From the observation error covariance matrix R _min,
0) A simple residual covariance matrix S ′ calculated using the equation
and _kmin (t), using the S / N ratio minimum value above the SNR _{mi n} (44) equation for each sampling time k, the relationship between successively calculated by the residual covariance matrix S _kmin (t), the following (51)
It becomes an expression.

(37)

The above equation (51) indicates that the observation error covariance matrix R
_The coarse gate calculated from the simple residual covariance matrix S ′ _kmin (t) calculated from Eq. (50) from _min is calculated from the SNR minimum value SNR _{min at} each sampling time k from (44) The equation shows that the value is the same as or wider than the coarse gate calculated using the S / N ratio minimum residual covariance matrix S _kmin (t) that is sequentially calculated.

Therefore, the S / N ratio minimum residual covariance matrix calculation means 12 uses the observation error covariance matrix R _min calculated using the above equation (48). That is, (44)
Since the observation error covariance matrix R _kmin (t) in the equation depends on the number j of detected data and the number of wakes, it is necessary to perform sequential calculation, whereas the observation error covariance matrix R _min in the above equation (50) is required. Is a constant that does not depend on the sampling time k, the detection data D _kj , and the wake (t).
The calculation load of the computer when using the expression (0) is further reduced as compared with the calculation load of the computer when using the expression (44).

Next, the operation will be described. FIG. 4 is a flowchart showing a processing flow of the tracking device according to the second embodiment of the present invention. Step ST1 is the same as FIG. 2 of the first embodiment.

In step ST11, the S / N ratio minimum residual covariance matrix calculation means 12 calculates the radar maximum detection distance RANGE _max by using the above equation (47).
Error variance σc in the distance direction of the observation error covariance matrix R _kj component
r ² _min , error variance in elevation direction RANGE ² _max σc
e _min , error variance in the azimuthal direction RANGE ² _max σcb
_{Among min} , the maximum component of the observation error covariance matrix R _kj component is selected.

In step ST12, the S / N ratio minimum value residual covariance matrix calculation means 12 uses the selected maximum value component of the observation error covariance matrix R _kj component to select the sampling time k, the number of detected data j, The observation error covariance matrix R _min in the north reference Cartesian coordinates independent of the number of wakes is given by (48)
It is calculated using the formula.

In step ST13, the S / N ratio is
The value residual covariance matrix calculation means 12 outputs from the observation means 1
Detection data D_kjAnd detection data D_kjS / N attached to
Ratio SNR_kj, And the prediction of the current time output by the delay means 8
Error covariance matrix P_pk(T) and the prediction vector V at the current time
X_pk(T) is input and the calculated observation error covariance matrix R
_minAnd the input prediction error covariance matrix P at the current time
_pkUsing (t), a simple method that does not depend on the number of detection data j
Simple residual covariance matrix S '_kmin(T) is calculated, and the calculated simple
Easy residual covariance matrix S '_kmin(T) and the input search
Knowledge data D_kj, Detection data D_kjS / N ratio SN associated with
R_kj, The prediction error covariance matrix P at the current time_pk(T) and current time
Hourly predicted vector VX_pk(T) is a rough gate inside / outside determination means
13 is output.

In step ST14, the coarse gate inside / outside determination means 13 receives the input simple residual covariance matrix S '.
_{Based on kmin} (t) and the predicted vector VX _pk (t) at the current time, a coarse gate that is a target existence expected area is calculated, and the input detection data D _kj and S / N accompanying the detection data D _kj are calculated.
For the ratio SNR _kj, S / N ratio SNR _kj accompanying detection data D _kj and the detection data D _kj is within the coarse gate
The determined, associated with the detection data D _kj is within determines coarse gate and the detection data D _kj S / N ratio SNR _kj,
And the input prediction error covariance matrix P _pk (t) at the current time
And the prediction vector VX _pk (t) at the current time are output to each S / N ratio residual covariance matrix calculation means 4.

The processing in steps ST4 to ST9 is the same as in FIG. 2 of the first embodiment. Therefore, by the above steps, the computational load of the computer when using the above equation (50) is larger than the computational load of the computer when using the above equation (44), the sampling time k, the number of detected data j,
Since the measurement error covariance matrix R _min that does not depend on the number of tracks is used, the weight is further reduced.

As described above, according to the second embodiment, the S / N ratio minimum residual covariance matrix calculating means 12 calculates the sampling error k independent of the sampling time k, the number of detected data j, and the number of wakes. Since a variance matrix R _min is calculated and a simple residual covariance matrix S ′ _kmin (t) independent of the number j of detection data is calculated using the calculated observation error covariance matrix R _min , 1 compared with the coarse gate inside / outside determination means
3. The effect of reducing the calculation load in each processing of each S / N ratio residual covariance matrix calculation means 4, fine gate inside / outside determination means 5, data update means 6, and prediction means 7 can be obtained.

Embodiment 3 FIG. 5 is a block diagram showing a configuration of a tracking device according to Embodiment 3 of the present invention. In the drawing, reference numeral 22 denotes detection data D _kj output from the observation means 1 and an S / N ratio SNR accompanying the detection data D _{kj. kj} and the prediction error covariance matrix P _pk at the current time output by the delay means 8
(T) and the prediction vector VX _pk (t) at the current time are input, the maximum value component of the observation error covariance matrix R _kj component is selected by the maximum detection range RANGE _max of the radar, and the selected observation error covariance matrix R By the maximum value component of the _kj component, the observation error covariance matrix R _min in the north reference Cartesian coordinates independent of the sampling time k, the number of detection data j, and the number of wakes.
Is calculated, and using the calculated observation error covariance matrix R _min and the input prediction error covariance matrix P _pk (t) at the current time, a residual covariance matrix independent of the number of detection data j and the number of wakes E _k is calculated, and the calculated residual covariance matrix E _k , input detection data D _kj , and S accompanying the detection data D _kj are calculated.
/ N ratio SNR _kj , prediction error covariance matrix P at the current time
S / N ratio minimum residual covariance matrix calculation means for outputting _pk (t) and the current time prediction vector VX _pk (t).

In FIG. 5, reference numeral 23 denotes a residual covariance matrix E _k , detection data D _kj from the S / N ratio minimum residual covariance matrix calculation means 22, detection data D _kj , and S / N values accompanying the detection data D _kj.
N ratio SNR _kj , prediction error covariance matrix P _pk (t) at the current time
And the prediction vector VX _pk (t) at the current time, and the input residual covariance matrix E _k and the prediction vector VX at the current time.
_pk (t) is used to calculate a coarse gate, which is a target existence expected area, and for the input detection data D _kj and the S / N ratio SNR _kj accompanying the detection data D _kj , the detection data D contained in the coarse gate is calculated. _kj and S / accompanying the detection data D _kj
Determines N ratio SNR _kj, accompanying the detection data D _kj is within determines coarse gate and the detection data D _kj S / N
Embodiment 1 is a coarse gate inside / outside determination means for outputting the ratio SNR _kj , the input prediction error covariance matrix P _pk (t) at the current time, and the prediction vector VX _pk (t) at the current time. Is equivalent to FIG.

Next, the S / N ratio minimum residual covariance matrix calculation means 22 will be described in detail. First, according to the definition of the Kalman filter, the smoothed error covariance matrix P _sk (t) is expressed by the following equation (52) using the state vector VX _k (t) and the smooth vector VX _sk (t) representing the target true value. Is shown as

(38) In the above formula (52), E [] is a symbol representing an average.

Here, as a simple estimate of the smooth vector VX _sk (t) and the smooth error covariance matrix P _sk (t), the smooth vector VVX _skj directly calculated from the detection data D _kj
(T) and the smooth error covariance matrix PP _sk (t) directly calculated from the detection data D _kj are given by the following equations (53) and (5), respectively.
4) Define as in equation.

[Equation 39] In the above equation (53), VZ _kj indicates target detection data excluding detection data of clutter in the coarse gate. T _k indicates the sampling time k.

Here, according to the Kalman filter theory,
The motion model is shown as in equation (55).

(Equation 40) In the above equation (55), Φ _k-1 is changed from time t _k-1 to t _k
Is the known transition matrix of the state vector to
It is shown as in equation 6).

[Equation 41] In the following equation (56), I is a 3 × 3 unit matrix.

In the above equation (55), W _k is a driving noise vector according to the trivariate normal distribution of the mean zero vector and the error covariance matrix Q _k , and corresponds to an acceleration vector. Further, Δ _k−1 is a known transformation matrix of the driving noise vector, and is represented as in the following equation (57).

(Equation 42)

The state vector VX in the above equation (55)
_k (t) is a vector composed of a target position vector and a velocity vector, and in the case of a three-dimensional space, is represented by the following equation (58).

[Equation 43] In the above equation (58), x _k , y _k , and z _k indicate an arbitrary position in the north reference rectangular coordinates, and the symbol “•” above x _k , y _k , and z _k indicates a first-order time derivative. , X _k , y _k , and z _k to which the symbol “·” above is given indicate the velocity at that position.

According to the Kalman filter theory, the target detection data VZ _kj excluding the clutter detection data in the coarse gate is given by the following equation (59).

[Equation 44] In the above equation (59), H is a known observation matrix, which is represented by the following equation (60).

[Equation 45]

In the above equation (59), U _kj is a mean zero vector, an observation noise vector at time t _k according to a trivariate white normal distribution of a covariance matrix Λ _kj , and Γ _kj is an observation noise vector U _{kj is} a known transformation matrix. Here, the relationship with Γ _kj (t) in the above equation (6) is used. When calculating the known transformation matrix Γ _kj of the observation noise vector U _kj, a true value is used. Since it is not known, a predicted value is inserted instead of the true value as in equation (6). Therefore, Γ _kj and Γ _kj (t) are treated as the same.

Therefore, from the above equation (54), the above (53)
By applying the equations, (55), (59), and (4),
The following equation (61) is obtained.

[Equation 46] Q _k in the above equation (61) is a covariance matrix of the driving noise W _k in the above equation (55).

Here, according to the theory of the Kalman filter, the smooth vector VX _sk (t) is expressed by the smooth error covariance matrix P
the sum of the diagonal elements of _sk (t) whereas the optimum value in the sense of minimizing, detection data D _kj smooth vector VVX _sk directly calculated from (t) is calculated directly from the detection data D _kj This is not an optimal value in the sense of minimizing the sum of the diagonal components of the smoothing error covariance matrix P _sk (t). Therefore, the following equation (62) holds.

[Equation 47]

From the relationship of the above equation (62), the following (63)
The relationship of the expression holds.

[Equation 48]

Here, from the Kalman filter theory, the prediction error covariance matrix P _{pk + 1} (t) is calculated by the following equation (64).

[Equation 49]

[0106] Detector error covariance matrix is calculated directly from the data D _kj using PP _sk (t), the prediction error covariance matrix directly calculated from the detection data D _kj after one sampling PP _{pk + 1} ( t) is calculated as in the following equation (65).

[Equation 50]

The above equations (65), (61) and (56)
Using the expression (57) and the definition (60) of the observation matrix H, the following expression (66) is obtained.

(Equation 51)

Here, the following equation (67) is established from the relationship between the above equations (43) and (49).

(Equation 52)

Therefore, when the above equation (66) is substituted into the equation (67), the following equation (68) is obtained.

(Equation 53)

Further, the above equations (66), (67) and (6)
Using the expression (8), the following expression (69) is obtained.

(Equation 54)

The left side of the above equation (69) is
Prediction error covariance matrix P on the right side _pk(T)
Data D_kjPrediction error covariance matrix PP directly calculated from_pk
(T). here
And the prediction error covariance matrix PP_pkRemaining calculated from (t)
Difference covariance matrix SS_k(T) is calculated by the following equation (70).
Determine. Also, the right side of the above equation (69) is expressed by the following equation (71).
Determined as follows.

[Equation 55] In the above equation (71), E _k represents a residual covariance matrix that does not depend on the number of detection data j and the number of wakes.

Then, the above equation (69), (70)
From the equation (71), the following equation (72) is calculated.

[Equation 56]

Further, the above equations (2), (63) and (6)
From the expressions 4) and (65), the following expression (73) is established.

[Equation 57]

Therefore, from the above equations (72) and (73), the following equation (74) is established.

[Equation 58]

The above equation (74) shows that the coarse gate defined by the following equation (75) is wider than the fine gate defined by the above equation (11). That is, the detection data D _kj in the fine gate defined by the equation (11) always exists in the coarse gate defined by the equation (75).

[Equation 59]

Therefore, in the S / N ratio minimum value residual covariance matrix calculation means 22, when the user determines that the calculation load is heavy, the S / N ratio residual covariance matrix calculation means 22 does not depend on the number of detection data j and the number of wakes using the above equation (71). Calculate the residual covariance matrix E _k .

Here, Q _k in the above equation (71) is
Since it is a fixed value determined before operation, the residual covariance matrix E _k in the equation (71) is obtained by a fixed value independent of the number of detection data j and the number of wakes. That is, the calculation load of the computer when the coarse gate of the above equation (75) is used is obtained by _replacing S _kmin (t) of the above equation (46) with S ′ _{km in}
The calculation load is further reduced as compared with the calculation load of the computer when the coarse gate replaced with (t) is used.

Next, the operation will be described. FIG. 6 is a flowchart showing a processing flow of the tracking device according to the third embodiment of the present invention. Step ST1 is the same as FIG. 2 of the first embodiment.

In step ST21, the S / N ratio minimum residual covariance matrix calculation means 22 calculates the radar maximum detection distance RANGE _max by using the above equation (47).
Error variance σc in the distance direction of the observation error covariance matrix R _kj component
r ² _min , error variance in elevation direction RANGE ² _max σc
e _min , error variance in the azimuthal direction RANGE ² _max σcb
_{Among min} , the maximum component of the observation error covariance matrix R _kj component is selected.

In step ST22, the S / N ratio minimum value residual covariance matrix calculating means 22 uses the selected maximum value component of the observation error covariance matrix R _kj component to perform sampling time k, detection data number j, The observation error covariance matrix R _min in the north reference Cartesian coordinates, which does not depend on the number of wakes,
8) Calculate using equation.

In step ST23, the S / N ratio minimum residual covariance matrix calculation means 22 outputs the detection data D _kj output from the observation means 1 and the S / N attached to the detection data D _kj.
The ratio SNR _kj , the prediction error covariance matrix P _pk (t) at the current time output by the delay means 8 and the prediction vector VX at the current time
_pk (t) is input and the calculated observation error covariance matrix R _min
And a prediction error covariance matrix PP _pk (t) directly calculated from the input detection data D _kj, to calculate a residual covariance matrix E _k independent of the number of detection data j and the number of wakes,
Calculated residual covariance matrix E _k, and inputs the detection data D _kj, accompanying the detection data D _{k j} S / N ratio SNR _kj,
The prediction error covariance matrix P _pk (t) at the current time and the prediction vector VX _pk (t) at the current time are output to the coarse gate inside / outside determination means 23.

In step ST24, the coarse gate inside / outside determination means 23 calculates a coarse gate, which is a target existence expected area, from the input residual covariance matrix E _k and the current time prediction vector VX _pk (t). Input detection data D _kj
And detecting the S / N ratio SNR _kj accompanying data D _kj, the S / N ratio SNR _kj accompanying detection data D _kj and the detection data D _kj is within the coarse gate determines, determined crude gate in detection data D are in the _kj and S / N ratio SNR _kj associated with detection data D _kj, and the prediction error covariance matrix at the current time input P _pk (t) and the current time of the prediction vector VX _pk (t) Is output to each S / N ratio residual covariance matrix calculation means 4.

The processing in steps ST4 to ST9 is the same as in FIG. 2 of the first embodiment. Therefore, according to the above steps, the calculation load of the computer when using the above equation (71) does not depend on the number j of detected data and the number of wakes as compared with the calculation load of the computer when using the above equation (50). Since the residual covariance matrix E _k is used, the weight is further reduced.

As described above, according to the third embodiment, the coarse gate inside / outside determination is performed by the residual covariance matrix E _k that does not depend on the number of detection data j and the number of wakes. The rough gate inside / outside determination means 23, each S /
An effect is obtained that the calculation load of calculation in each process of the N ratio residual covariance matrix calculation means 4, the fine gate inside / outside determination means 5, the data update means 6, and the prediction means 7 is reduced.

Embodiment 4 FIG. 7 is a block diagram showing the configuration of a tracking device according to Embodiment 4 of the present invention. In the drawing, reference numeral 32 denotes detection data D _kj output by the observation means 1 and the S / N ratio SNR accompanying the detection data D _{kj. kj} and the prediction error covariance matrix P _pk at the current time output by the delay means 8
(T) and the prediction vector VX _pk (t) at the current time are input, the maximum value component of the observation error covariance matrix R _kj component is selected by the maximum detection range RANGE _max of the radar, and the selected observation error covariance matrix R The maximum value component of the _kj component, the input detection data D _kj , and the S / N accompanying the detection data D _kj
It is an observation error covariance matrix maximum value selecting unit that outputs the ratio SNR _kj , the current time prediction error covariance matrix P _pk (t), and the current time prediction vector VX _pk (t).

In FIG. 7, reference numeral 33 denotes the maximum value component of the observation error covariance matrix R _kj component from the observation error covariance matrix maximum value selecting means 32, the detection data D _kj , and the detection data D
S / N ratio SNR _kj accompanying _kj, the prediction vector VX of the prediction error covariance matrix P _pk (t) and the current time of the current time
_pk (t) is input, a coarse gate having a radius of radius 0 independent of the sampling time k, the number of detection data j, and the number of wakes is calculated at a fixed sampling interval T _int , and the input detection data D _kj and detection data D _kj are calculated. S / attached to
For the N ratio SNR _kj , the detection data D _kj contained in the coarse gate and the S / N ratio SNR accompanying the detection data D _kj
determines _kj, accompanying the detection data D _kj and the detection data D _kj is within determines crude gate S / N ratio SN
R _kj and the input prediction error covariance matrix P _pk at the current time.
(T) and a rough gate inside / outside determination means for outputting the prediction vector VX _pk (t) at the current time. Other configurations are the same as those in FIG. 1 of the first embodiment.

Here, a rough gate having a radius of radius 0 will be described. Assuming that the sampling interval is a constant value T _int as in the following equation (76) and the driving noise covariance matrix Q _k is a diagonal matrix irrelevant to the coordinate axis and time, the following equation (77)
It is shown like an equation.

[Equation 60] Q in the above equation (77) is a driving noise constant.

Then, the above equation (48), (71)
From the equations (76), the rough gate in the above equation (75) can be rewritten as the following equation (78).

[Equation 61]

In the above equation (78), the radius radiu
s0 is represented by the following equation (79).

(Equation 62)

Here, with respect to the coarse gate defined by the above equation (75), the radius radi defined by the above equation (78)
Since the sampling interval of the coarse gate of us0 is fixed, it does not depend on the sampling interval. Therefore,
The calculation load of the computer when using the coarse gate having the radius of radius 0 in Expression (78) is further reduced as compared with the computer load when using the coarse gate in Expression (75).

Next, the operation will be described. FIG. 8 is a flowchart showing a processing flow of the tracking device according to the fourth embodiment of the present invention. Step ST1 is the same as FIG. 1 of the first embodiment.

In step ST31, the observation error
The scatter matrix maximum value selecting means 32 outputs the search
Knowledge data D_kjAnd detection data D_kjS / N ratio S associated with
NR _kj, And the prediction error of the current time output by the delay means 8
Covariance matrix P_pk(T) and prediction vector VX at the current time_pk
Input (t) and the maximum radar detection range RANGE_{ma x}
By the observation error covariance matrix R_kjSelect the maximum component of the component
And the selected observation error covariance matrix R_kjMaximum component value
Minutes and input detection data D_kj, Detection data D_kjTo
Accompanying S / N ratio SNR_kj, Current time prediction error covariance row
Row P_pk(T) and prediction vector VX at the current time_pk(T)
Output to the coarse gate inside / outside determination means 33.

In step ST 32, the coarse gate inside / outside determination means 33 determines whether the observation error covariance matrix maximum value selection means 3
Maximum value component of the observation error covariance matrix R _kj components from 2, detection data D _kj, accompanying the detection data D _kj S / N ratio SN
R _kj , the prediction error covariance matrix P _pk (t) of the current time and the prediction vector VX _pk (t) of the current time are input, and the sampling time k is calculated using the above equation (79) at a fixed sampling interval T _int. , A rough gate having a radius of radius 0 independent of the number of detection data j and the number of wakes is calculated.

In step ST33, the coarse gate inside / outside determination means 33 determines (78) the input detection data D _kj and the S / N ratio SNR _kj accompanying the detection data D _kj.
Using the equation, the detection data D _kj and the S / N ratio SNR _kj accompanying the detection data D _kj contained in the calculated coarse gate having the radius radius 0 are determined, and the detection data contained in the determined coarse gate are determined. The S / N ratio SNR _kj associated with D _kj and the detection data D _kj , the input current time prediction error covariance matrix P _pk (t), and the current time prediction vector VX _pk
(T) is output to each S / N ratio residual covariance matrix calculation means 4.

The processing in steps ST4 to ST9 is the same as in FIG. 2 of the first embodiment. Therefore, by the above steps, the sampling time k, the number of detected data j,
When a coarse gate having a radius of radius 0 that does not depend on the number of tracks is used to determine the inside / outside of the coarse gate using equation (78), the computational load on the computer is the number of detection data j and the residual covariance matrix E _k that does not depend on the number of tracks. Accordingly, the load on the computer can be further reduced as compared with the calculation load of the computer when performing the rough gate inside / outside determination using the above equation (71).

As described above, according to the fourth embodiment, the coarse gate inside / outside determination is performed by the coarse gate having a radius of radius 0 which does not depend on the sampling time k, the number of detection data j, and the number of wakes. The computational load of calculation in each process of the coarse gate inside / outside determination means 33, each S / N ratio residual covariance matrix calculation means 4, the fine gate inside / outside determination means 5, the data update means 6, and the prediction means 7 is reduced. Is obtained.

Embodiment 5 FIG. FIG. 9 is a block diagram showing a configuration of a tracking apparatus according to Embodiment 5 of the present invention. In the figure, reference numeral 41 denotes each S / N ratio residual covariance matrix S _kj (output from the fine gate inside / outside determination means 5). The number of t), that is, the number of wakes, is compared with the number j of detected data at the current sampling time detected by the observation means 1, and the observation means 1 and the S / N ratio minimum residual covariance matrix calculation means 2 use the data. S / N ratio minimum value control means for changing the value of the present S / N ratio minimum value SNR _min and notifying it to the observation means 1 and the S / N ratio minimum value residual covariance matrix calculation means 2. Fig. 1 of the first embodiment
Is equivalent to

In the signal detection in the observation means 1, it is determined that a target has been detected by a signal having power equal to or higher than the noise power δ, and a threshold value S for determining that the target has been detected is determined. / N ratio is the minimum S / N ratio S
When NR _min is set, the observation unit 1 converts the signal having the power _{equal to} or more than δ · SNR _min into the detected data D _kj ,
It is output as the S / N ratio SNR _kj accompanying the detection data D _kj . Here, the S / N ratio minimum value SNR _min is also a parameter for determining the false alarm probability.

Therefore, if the value of the S / N ratio minimum value SNR _min is increased and the false alarm probability is reduced, the total number of obtained detection data D _kj can be reduced. Therefore, each S / S output from the fine gate inside / outside determination means 5 before one sampling is output.
If the number of detection data j at the current sampling time is significantly larger than the number of N-ratio residual covariance matrix S _kj (t), that is, the number of wakes, the observation means 1 uses S / N larger than the S / N ratio minimum value SNR _min
If the N ratio minimum value SSNR _min is used, the number of detected data j
Can be narrowed down.

[Equation 63]

That is, S / S attached to the detection data D _kj
Only when the N ratio SNR _kj satisfies the following equation (81), S
The / N ratio minimum value SSNR _min is used for tracking, and the others are not used for tracking. By this processing, the number j of detection data used for tracking can be reduced, and the calculation load on the computer is reduced.

[Equation 64]

As described above, the observation means 1 has a predetermined S / N ratio.
Minimum SNR_minDetection data D using_kjHave gained
Is determined by the determination result of the inside / outside determination means 5 of the fine gate.
Detection data D_kjClutter detection data D_kjMany
Then, the S / N ratio minimum value SNR _minNeed to be bigger
is there. Conversely, target detection data D_kjIs small, S
/ N ratio minimum SNR_minNeeds to be smaller.

Next, the operation will be described. FIG. 10 is a flowchart showing a processing flow of the tracking device according to the fifth embodiment of the present invention. Steps ST1 to ST9 are the same as those in FIG. 1 of the first embodiment.

In step ST41, the S / N ratio minimum value control means 41 determines the number of each S / N ratio residual covariance matrix S _kj (t) output from the fine gate inside / outside determination means 5, ie, the number of wakes and Then, the number j of detected data at the current sampling time detected by the observation means 1 is compared.

In step ST42, the S / N ratio minimum value control means 41 determines the number of the compared S / N ratio residual covariance matrices S _kj (t), ie, the number of wakes, and the current number detected by the observation means 1. If the number j of detected data at the sampling time is significantly different, the S / N ratio minimum value SN used by the observation unit 1 and the S / N ratio minimum residual covariance matrix calculation unit 2 is used.
The value of R _min is changed and notified to the observation means 1 and the S / N ratio minimum value residual covariance matrix calculation means 2.

For example, for each S / N ratio residual covariance matrix S
If the number j of detected data at the current sampling time detected by the observation means 1 is significantly larger than the number of _kj (t), that is, the number of wakes, the observation means 1 and the S / N ratio minimum residual covariance matrix calculation The S / N ratio minimum value SSNR _{min that is} larger than the value of the S / N ratio minimum value SNR _min used by the means 2
To the observation means 1 and the S / N ratio minimum residual covariance matrix calculation means 2.

After changing the S / N ratio minimum value in step ST42, the process returns to step ST1, and the observation means 1
Using the modified S / N ratio minimum SNR _min, detection data D _kj in the same manner as in the first embodiment, and outputs the S / N ratio SNR _kj accompanying detection data D _kj, in step ST2, S / N ratio minimum residual covariance matrix calculation means 2
Using the changed S / N ratio minimum value SNR _min , an observation error covariance matrix R _kmin (t) is calculated as in the first embodiment. Subsequent processing is the same as in FIG. 2 of the first embodiment.

As described above, according to the fifth embodiment.
For example, each S / N ratio output from the fine gate inside / outside determination means 5
Residual covariance matrix S_kjThe number of (t), that is, the number of tracks,
Number of detected data at the current sampling time detected by the measuring means 1
j is significantly different, the S / N ratio minimum value control means 41
Is the observation means 1 and the S / N ratio minimum value residual covariance matrix calculation
SNR minimum value SNR used by means 2_minThe value of
Change the observation means 1 and the S / N ratio minimum residual covariance matrix
The calculating means 2 calculates the changed S / N ratio minimum value SNR _minTo
By performing processing using, for example, the number of detected data
If j is extremely large, the number of detection data j in the coarse gate
Can be reduced.
Calculation operation of each process of the updating means 6 and the prediction means 7
The effect that the load is reduced is obtained.

[0148]

As described above, according to the present invention, S /
A residual covariance matrix S _kmin (t) that does not depend on the number of detection data is calculated from the observation error covariance matrix R _kmin (t) at the current time that does not depend on the number of detection data calculated using the N-ratio minimum value. A coarse gate, which is a target existence expectation area, is calculated by using the / N ratio minimum residual covariance matrix calculation means and the residual covariance matrix S _kmin (t), and the coarse data is calculated for the input detection data and S / N ratio. Detection data and S /
A coarse gate inside / outside determining means for determining an N ratio and outputting detection data and an S / N ratio contained in the determined coarse gate;
Using the output from the coarse gate inside / outside determination means, each S / N for calculating each S / N ratio residual covariance matrix S _kj (t) by the coarse gate, which depends on the sampling time, the number of detected data, and the number of wakes Using the output from the ratio residual covariance matrix calculation means and the output from each S / N ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, For the detection data and S / N ratio contained in the fine gate, the detection data and the S / N ratio contained in the fine gate are determined, and each S / N by the fine gate depends on the sampling time, the number of detected data, and the number of tracks. N-ratio residual covariance matrix S _kj
The provision of the fine gate inside / outside determination means for calculating (t) has the effect of reducing the calculation load of the fine gate for removing unnecessary signals for the target signal.

According to the present invention, the observing means and the S / N ratio residual covariance matrix S _kj (t) outputted from the fine gate inside / outside judging means and the number of detection data outputted from the observing means are used for the observing means and the observing means. The S / N ratio minimum value used by the S / N ratio minimum residual covariance matrix calculation means is changed, and the observation means and the S / N ratio are changed.
S / N ratio minimum value control means for notifying the N ratio minimum value residual covariance matrix calculation means, and the observation means and the S / N ratio minimum value residual covariance matrix calculation means are provided with a modified S / N ratio. By performing processing using the minimum ratio value, for example, when the number of pieces of detection data is extremely large, the number of pieces of detection data in the coarse gate can be reduced, and there is an effect that the calculation load is reduced.

According to the present invention, the maximum value component of the observation error covariance matrix components is selected based on the maximum detection distance of the radar, and the sampling time, detection data number, calculates the observation error covariance matrix R _{mi n} that does not depend on the number of track, the output from the observation means, and using the observation error covariance matrix R _min calculated, simple residue that does not depend on the detection number of data 'calculates _kmin (t), calculated simple residual covariance matrix S' difference covariance matrix S and S / N ratio minimum residual covariance matrix calculation means for outputting _kmin (t), a simple residual Difference covariance matrix S ' _kmin (t)
Calculates the coarse gate which is the target existence expected area, determines the detection data and S / N ratio included in the coarse gate with respect to the input detection data and S / N ratio, and enters the determined coarse gate. Using the coarse gate inside / outside determining means for outputting the detected data and the S / N ratio, and using the output from the coarse gate inside / outside determining means, the sampling time, the number of detected data,
Each S / N ratio residual covariance matrix S _kj (t) by a coarse gate, which depends on the number of tracks, is calculated, and each calculated S / N ratio residual covariance matrix S _kj (t) is output. Using the output from the S / N ratio residual covariance matrix calculation means and the output from each S / N ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated and input. With respect to the detection data and S / N ratio contained in the coarse gate, the detection data and S / N ratio contained in the fine gate are determined, and the detection gate and the S / N ratio depend on the sampling time, the number of detection data, and the number of wakes. By providing fine gate inside / outside determination means for calculating each S / N ratio residual covariance matrix S _kj (t), coarse gate inside / outside determination means, each S / N ratio residual covariance matrix calculation means, The calculation load of calculation in each process of the gate inside / outside determination means is reduced. There is a result.

According to the present invention, the maximum value component of the observation error covariance matrix components is selected based on the maximum detection distance of the radar, and the sampling time, detection data number, calculates the observation error covariance matrix R _{mi n} that does not depend on the number of track, the output from the observation means, and using the observation error covariance matrix R _min calculated, detecting the number of data does not depend on the number of track and S / N ratio minimum residual covariance matrix calculation means for calculating a residual covariance matrix E _k, the residual covariance matrix E _k, the detection data to calculate a coarse gate that is a target presence expected areas, entered And the S / N ratio, the detection data and S / N
A coarse gate inside / outside determining means for determining an N ratio and outputting detection data and an S / N ratio contained in the determined coarse gate;
Using the output from the coarse gate inside / outside determination means, each S / N for calculating each S / N ratio residual covariance matrix S _kj (t) by the coarse gate, which depends on the sampling time, the number of detected data, and the number of wakes Using the output from the ratio residual covariance matrix calculation means and the output from each S / N ratio residual covariance matrix calculation means, a fine gate that is a target existence expected area narrower than the coarse gate is calculated, For the detection data and S / N ratio contained in the fine gate, the detection data and the S / N ratio contained in the fine gate are determined, and each S / N by the fine gate depends on the sampling time, the number of detected data, and the number of tracks. N-ratio residual covariance matrix S _kj
The provision of the fine gate inside / outside determination means for calculating (t) makes it possible to calculate the calculation load in each process of the coarse gate inside / outside determination means, each S / N ratio residual covariance matrix calculation means, and the fine gate inside / outside determination means. This has the effect of being reduced.

According to the present invention, the observation error covariance matrix maximum value selecting means for selecting the maximum value component of the observation error covariance matrix components based on the radar maximum detection distance, and the observation error covariance matrix maximum value selection. Using the output from the means, a coarse gate having a radius of radius 0 independent of the sampling time, the number of detection data, and the number of wakes is calculated at a fixed sampling interval T _int , and the coarse gate is calculated for the input detection data and S / N ratio. Detection data and S / N contained in
The S / N ratio residual covariance matrix S by the coarse gate, which depends on the sampling time, the number of detected data, and the number of tracks, using the coarse gate inside / outside determination means for determining the ratio and the output from the coarse gate inside / outside determination means. _The target existence expected area narrower than the coarse gate is obtained by using the output from each S / N ratio residual covariance matrix calculating means for calculating _kj (t) and the output from each S / N ratio residual covariance matrix calculating means. The fine gate is calculated, the detection data and the S / N ratio contained in the input coarse gate are determined, and the sampling time, the number of detected data, and the wake are determined. Means for calculating the S / N ratio residual covariance matrix S _kj (t) by fine gates, which depend on the number of fine gates. Difference covariance matrix calculation means,
This has the effect of reducing the calculation load of calculation in each process of the fine gate inside / outside determination means.

[Brief description of the 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 a flowchart showing a processing flow of the tracking device according to the first embodiment of the present invention.

FIG. 3 is a block diagram showing a configuration of a tracking device according to Embodiment 2 of the present invention.

FIG. 4 is a flowchart showing a processing flow of a tracking device according to Embodiment 2 of the present invention.

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

FIG. 6 is a flowchart showing a processing flow of a tracking device according to Embodiment 3 of the present invention.

FIG. 7 is a block diagram showing a configuration of a tracking device according to Embodiment 4 of the present invention.

FIG. 8 is a flowchart showing a processing flow of a tracking device according to Embodiment 4 of the present invention.

FIG. 9 is a block diagram showing a configuration of a tracking device according to a fifth embodiment of the present invention.

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

FIG. 11 is a block diagram illustrating a configuration of a conventional tracking device.

FIG. 12 is a diagram illustrating a coordinate system according to a conventional observation means.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Observation means, 2 S / N ratio minimum residual covariance matrix calculation means, 3 coarse gate inside / outside determination means, 4 S / N ratio residual covariance matrix calculation means, 5 fine gate inside / outside determination means, 6
Data updating means, 7 prediction means, 8 delay means, 12
S / N ratio minimum residual covariance matrix calculation means, 13 coarse gate inside / outside determination means, 22 S / N ratio minimum residual covariance matrix calculation means, 23 coarse gate inside / outside determination means, 32 observation error covariance matrix maximum Value selection means, 33 coarse gate inside / outside determination means, 41 S / N ratio minimum value control means.

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Yoshio Kosuge 2-3-2 Marunouchi, Chiyoda-ku, Tokyo F-term in Mitsubishi Electric Corporation (reference) 5J070 AC02 AC06 AE04 AH14 AH19 AH21 AK16 AK22 AK28 BB05 BB06

Claims (9)

[Claims] 1. A threshold value S for detecting a target.
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data by using the minimum value of the / N ratio; A residual covariance matrix S _kmin (t) that does not depend on the number of detection data is calculated from the observation error covariance matrix R _kmin (t) at the current time that does not depend on the number of detection data calculated using the ratio minimum value. S / N ratio minimum value residual covariance matrix calculating means for outputting the detected covariance matrix S _kmin (t) and the input detection data and S / N ratio, and the S / N ratio minimum value residual value Based on the residual covariance matrix S _{km in} (t) input from the variance matrix calculation means, a coarse gate which is a target existence expected area is calculated, and the input detection data and S
A coarse gate inside / outside determining means for determining the detection data and S / N ratio contained in the coarse gate for the / N ratio, and outputting the detected data and S / N ratio contained in the determined coarse gate; Using the output of the coarse gate inside / outside determination means, each S / N ratio residual covariance matrix S _kj (t) by the coarse gate, which depends on the sampling time, the number of detected data, and the number of wakes, is calculated,
Each calculated S / N ratio residual covariance matrix S _kj (t), and the detection data and S / N contained in the input coarse gate
Using the output of each S / N ratio residual covariance matrix calculating means for outputting a ratio, and the output of each S / N ratio residual covariance matrix calculating means, a fine gate that is a target existence expected area narrower than the coarse gate is determined. With respect to the calculated and detected detection data and S / N ratio in the coarse gate, the detection data and S / N ratio in the fine gate are determined and depend on the sampling time, the number of detection data, and the number of tracks. Calculate the S / N ratio residual covariance matrix S _kj (t) by the fine gate, find the detection data and the S / N ratio contained in the determined fine gate, and calculate the calculated S / N ratio residual. A tracking device comprising: a fine gate inside / outside determination means for outputting a difference covariance matrix S _kj (t). 2. A threshold value S for detecting a target.
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data using the minimum value of the / N ratio, an output of the observation means, and an output of a delay means to be described later. The current time prediction error covariance matrix and the current time prediction vector are input, and the current time observation error covariance matrix R independent of the number of detection data calculated using the S / N ratio minimum value.
_{Using the kmin} (t) and the input prediction error covariance matrix at the current time, the residual covariance matrix S independent of the number of detection data
_kmin (t), and the calculated residual covariance matrix S
_kmin (t), S / N ratio minimum residual covariance matrix calculating means for outputting the input detection data, S / N ratio, current time prediction error covariance matrix and current time prediction vector, A coarse gate, which is a target existence expected area, is calculated from the residual covariance matrix S _{km in} (t) input from the / N ratio minimum residual covariance matrix calculation means and the prediction vector at the current time,
With respect to the input detection data and S / N ratio, the detection data and S / N ratio contained in the coarse gate are determined, and the detection data and S / N ratio contained in the determined coarse gate and the input current value are determined. Coarse gate inside / outside determination means for outputting a prediction error covariance matrix of time and a prediction vector of the current time, and using the output of the rough gate inside / outside determination means, sampling time, number of detected data, number of wakes, coarse gate Calculate each S / N ratio residual covariance matrix S _kj (t) by
The calculated S / N ratio residual covariance matrix S _kj (t), the input detection data in the coarse gate, the S / N ratio in the coarse gate, and the prediction error covariance at the current time A target existence expectation narrower than a coarse gate, using each S / N ratio residual covariance matrix calculating means for outputting a matrix and a prediction vector at the current time; A fine gate, which is an area, is calculated, and for the detection data and S / N ratio contained in the input coarse gate, the detection data and S / N ratio contained in the fine gate are determined. Calculates the S / N ratio residual covariance matrix S _kj (t) by the fine gate depending on the number and the number of wakes, and determines the detection data and the S / N ratio contained in the determined fine gate. S
/ N ratio residual covariance matrix S _kj (t), and the input of the current time prediction error covariance matrix and the current time prediction vector. A data updating means for calculating and outputting a smoothed error covariance matrix at the current time and a smoothed vector at the current time, and a prediction error covariance matrix one sampling after the current time using the output of the data updating means. And a prediction unit that calculates and outputs a prediction vector one sample after the current time, and a delay unit that outputs a prediction error covariance matrix of the current time and a prediction vector of the current time using the output of the prediction unit. A tracking device, comprising: 3. The observing means and the S / N based on the number of each S / N ratio residual covariance matrix S _kj (t) outputted from the fine gate inside / outside judging means and the number of detection data outputted by the observing means.
S / N used by ratio minimum residual covariance matrix calculation means
An S / N ratio minimum value control unit that changes the value of the ratio minimum value and notifies the observation unit and the S / N ratio minimum value residual covariance matrix calculation unit, The tracking device according to claim 1, wherein the ratio minimum value residual covariance matrix calculation unit performs the process using the changed S / N ratio minimum value. 4. A threshold value S for detecting a target.
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data using the minimum value of the / N ratio, and an observation error covariance matrix component based on the maximum detection distance of the radar. Is selected, and using the maximum value component of the selected observation error covariance matrix, an observation error covariance matrix R _min independent of the sampling time, the number of detection data, and the number of wakes is calculated. Using the output from the means and the calculated observation error covariance matrix R _min , a simple residual covariance matrix S ′ _kmin (t) independent of the number of detection data is calculated, and the calculated simple residual covariance is calculated. S / N ratio minimum residual covariance matrix calculating means for outputting a matrix S ' _kmin (t) and input detection data and S / N ratio; and S / N ratio minimum residual covariance matrix calculating means Simple residual covariance matrix S 'input from _{Based on kmin} (t), a coarse gate which is a target existence expected area is calculated, and for the input detection data and S / N ratio, the detection data and S / N ratio included in the coarse gate are determined, and the determined coarse is determined. A coarse gate inside / outside determining means for outputting detection data and an S / N ratio contained in the gate; and a coarse gate depending on a sampling time, the number of detection data, and the number of wakes using an output of the coarse gate inside / outside determining means. Calculate each S / N ratio residual covariance matrix S _kj (t) by
Each calculated S / N ratio residual covariance matrix S _kj (t), and the detection data and S / N contained in the input coarse gate
Using the output of each S / N ratio residual covariance matrix calculating means for outputting a ratio, and the output of each S / N ratio residual covariance matrix calculating means, a fine gate that is a target existence expected area narrower than the coarse gate is determined. With respect to the calculated and detected detection data and S / N ratio in the coarse gate, the detection data and S / N ratio in the fine gate are determined and depend on the sampling time, the number of detection data, and the number of tracks. Calculate the S / N ratio residual covariance matrix S _kj (t) by the fine gate, find the detection data and the S / N ratio contained in the determined fine gate, and calculate the calculated S / N ratio residual. A tracking device comprising: a fine gate inside / outside determination means for outputting a difference covariance matrix S _kj (t). 5. A threshold value S for detecting a target.
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data using the minimum value of the / N ratio, and an observation error covariance matrix component based on the maximum detection distance of the radar. Is selected, and using the maximum value component of the selected observation error covariance matrix, an observation error covariance matrix R _min independent of the sampling time, the number of detection data, and the number of wakes is calculated. Using the output from the means, the prediction error covariance matrix of the current time and the prediction vector of the current time from the delay means described later, and the calculated observation error covariance matrix R _min ,
Simple residual covariance matrix S 'independent of the number of detection data
_kmin (t) is calculated, the calculated simple residual covariance matrix S ′ _kmin (t), and the input detection data, S / N
S / N ratio minimum residual covariance matrix calculation means for outputting the ratio, the current time prediction error covariance matrix and the current time prediction vector, and input from the S / N ratio minimum residual covariance matrix calculation means. From the simplified residual covariance matrix S ′ _kmin (t) and the prediction vector at the current time, a coarse gate that is a target existence expected area is calculated, and the input detection data and the S / N ratio _fall within the coarse gate. The detected detection data and the S / N ratio are determined, and the detection data and the S included in the determined coarse gate are determined.
And a coarse gate inside / outside determination means for outputting the prediction error covariance matrix of the current time and the input prediction vector of the current time, and using the output of the coarse gate inside / outside determination means, the sampling time, the number of detected data, Calculate each S / N ratio residual covariance matrix S _kj (t) by the coarse gate depending on the number of wakes,
The calculated S / N ratio residual covariance matrix S _kj (t), the input detection data in the coarse gate, the S / N ratio in the coarse gate, and the prediction error covariance at the current time A target existence expectation narrower than a coarse gate, using each S / N ratio residual covariance matrix calculating means for outputting a matrix and a prediction vector at the current time; A fine gate, which is an area, is calculated, and for the detection data and S / N ratio contained in the input coarse gate, the detection data and S / N ratio contained in the fine gate are determined. Calculates the S / N ratio residual covariance matrix S _kj (t) by the fine gate depending on the number and the number of wakes, and determines the detection data and the S / N ratio contained in the determined fine gate. S
/ N ratio residual covariance matrix S _kj (t), and the input of the current time prediction error covariance matrix and the current time prediction vector. A data updating means for calculating and outputting a smoothed error covariance matrix at the current time and a smoothed vector at the current time, and a prediction error covariance matrix one sampling after the current time using the output of the data updating means. And a prediction unit that calculates and outputs a prediction vector one sample after the current time, and a delay unit that outputs a prediction error covariance matrix of the current time and a prediction vector of the current time using the output of the prediction unit. A tracking device, comprising: 6. A threshold value S for detecting a target.
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data using the minimum value of the / N ratio, and an observation error covariance matrix component based on the maximum detection distance of the radar. Is selected, and using the maximum value component of the selected observation error covariance matrix, an observation error covariance matrix R _min independent of the sampling time, the number of detection data, and the number of wakes is calculated. Using the output from the means and the calculated observation error covariance matrix R _min , the number of detection data, a residual covariance matrix E _k independent of the number of wakes is calculated, and the calculated residual covariance matrix E _k , and Input detection data and S / N
The S / N ratio minimum residual covariance matrix calculation means for outputting the ratio, and the residual covariance matrix E _k input from the S / N ratio minimum residual covariance matrix calculation means, provide a target existence expectation area. A certain coarse gate is calculated, and the input detection data and S / N ratio are compared with the detection data and S / N contained in the coarse gate.
A coarse gate inside / outside determining means for determining a ratio and outputting detection data and an S / N ratio contained in the determined coarse gate; and using the output of the coarse gate inside / outside determining means, a sampling time, a number of detected data, Calculate each S / N ratio residual covariance matrix S _kj (t) by the coarse gate depending on the number of wakes,
Each calculated S / N ratio residual covariance matrix S _kj (t), and the detection data and S / N contained in the input coarse gate
Using the output of each S / N ratio residual covariance matrix calculating means for outputting a ratio, and the output of each S / N ratio residual covariance matrix calculating means, a fine gate that is a target existence expected area narrower than the coarse gate is determined. With respect to the calculated and detected detection data and S / N ratio in the coarse gate, the detection data and S / N ratio in the fine gate are determined and depend on the sampling time, the number of detection data, and the number of tracks. Calculate the S / N ratio residual covariance matrix S _kj (t) by the fine gate, find the detection data and the S / N ratio contained in the determined fine gate, and calculate the calculated S / N ratio residual. A tracking device comprising: a fine gate inside / outside determination means for outputting a difference covariance matrix S _kj (t). 7. A threshold value S for detecting a target.
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data using the minimum value of the / N ratio, and an observation error covariance matrix component based on the maximum detection distance of the radar. Is selected, and using the maximum value component of the selected observation error covariance matrix, an observation error covariance matrix R _min independent of the sampling time, the number of detection data, and the number of wakes is calculated. Using the output from the means, the prediction error covariance matrix of the current time and the prediction vector of the current time from the delay means described later, and the calculated observation error covariance matrix _Rmin , the residual independent of the number of detection data and the number of wakes is used. A difference covariance matrix E _k is calculated, and the calculated residual covariance matrix E _k , the input detection data, the S / N ratio, the current time prediction error covariance matrix, and the current time prediction vector S / S N ratio minimum A residual covariance matrix calculation means, the prediction vector of the S / N ratio minimum residual covariance matrix residual covariance matrix input from the calculating means E _k and the current time, the crude gate that is a target presence expected area With respect to the calculated and input detection data and S / N ratio, the detection data and S / N ratio in the coarse gate are determined, and the detection data and S / N ratio in the determined coarse gate, and Coarse gate inside / outside determination means for outputting the input current time prediction error covariance matrix and current time prediction vector, and using the output of the coarse gate inside / outside determination means, depends on the sampling time, the number of detected data, and the number of wakes. , Calculate each S / N ratio residual covariance matrix S _kj (t) by the coarse gate,
The calculated S / N ratio residual covariance matrix S _kj (t), the input detection data in the coarse gate, the S / N ratio in the coarse gate, and the prediction error covariance at the current time A target existence expectation narrower than a coarse gate, using each S / N ratio residual covariance matrix calculating means for outputting a matrix and a prediction vector at the current time; A fine gate, which is an area, is calculated, and for the detection data and S / N ratio contained in the input coarse gate, the detection data and S / N ratio contained in the fine gate are determined. Calculates the S / N ratio residual covariance matrix S _kj (t) by the fine gate depending on the number and the number of wakes, and determines the detection data and the S / N ratio contained in the determined fine gate. S
/ N ratio residual covariance matrix S _kj (t), and the input of the current time prediction error covariance matrix and the current time prediction vector. A data updating means for calculating and outputting a smoothed error covariance matrix at the current time and a smoothed vector at the current time, and a prediction error covariance matrix one sampling after the current time using the output of the data updating means. And a prediction unit that calculates and outputs a prediction vector one sample after the current time, and a delay unit that outputs a prediction error covariance matrix of the current time and a prediction vector of the current time using the output of the prediction unit. A tracking device, comprising: 8. A threshold value S for detecting a target.
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data using the minimum value of the / N ratio, and an observation error covariance matrix component based on the maximum detection distance of the radar. And a maximum value component of the selected observation error covariance matrix, and an observation error covariance matrix maximum value selection unit that outputs the detection data and the S / N ratio input from the observation unit, Using the output of the above-mentioned observation error covariance matrix maximum value selection means, a fixed sampling interval T _int and a radius rad independent of the sampling time, the number of detected data, and the number of wakes.
The rough gate of ius0 is calculated, the detection data and the S / N ratio in the coarse gate are determined for the input detection data and S / N ratio, and the detection data and S in the determined coarse gate are determined. Using the coarse gate inside / outside determination means for outputting the / N ratio, and using the output of the rough gate inside / outside determination means, the S / N ratio residual covariance of the coarse gate depending on the sampling time, the number of detected data, and the number of tracks. Calculate the matrix S _kj (t),
Each calculated S / N ratio residual covariance matrix S _kj (t), and the detection data and S / N contained in the input coarse gate
Using the output of each S / N ratio residual covariance matrix calculating means for outputting a ratio, and the output of each S / N ratio residual covariance matrix calculating means, a fine gate that is a target existence expected area narrower than the coarse gate is determined. With respect to the calculated and detected detection data and S / N ratio in the coarse gate, the detection data and S / N ratio in the fine gate are determined and depend on the sampling time, the number of detection data, and the number of tracks. Calculate the S / N ratio residual covariance matrix S _kj (t) by the fine gate, find the detection data and the S / N ratio contained in the determined fine gate, and calculate the calculated S / N ratio residual. A tracking device comprising: a fine gate inside / outside determination means for outputting a difference covariance matrix S _kj (t). 9. A threshold value S for detecting a target
Observation means for outputting detection data detected from an internal sensor and an S / N ratio accompanying the detection data using the minimum value of the / N ratio, and an observation error covariance matrix component based on the maximum detection distance of the radar. Is selected, the maximum value component of the selected observation error covariance matrix, the detection data and the S / N ratio input from the observation means, and the current time prediction error covariance from the delay means described later. Using the output of the observation error covariance matrix maximum value selecting means, which outputs a matrix and a prediction vector at the current time, using the output of the observation error covariance matrix maximum value selecting means, at a fixed sampling interval T _int , sampling time, detection data Radius rad independent of the number and number of tracks
The coarse gate of ius0 is calculated, the detection data and the S / N ratio in the coarse gate are determined for the input detection data and S / N ratio, and the detection data and S in the determined coarse gate are determined. And a coarse gate inside / outside determining means for outputting the / N ratio and the input prediction error covariance matrix of the current time and the current time predicted vector. Calculate each S / N ratio residual covariance matrix S _kj (t) by a coarse gate depending on the number of wakes,
Calculated S / N ratio residual covariance matrix S _kj (t), input detection data in coarse gate, S / N ratio in coarse gate, prediction error covariance at current time A target existence expectation narrower than the coarse gate, using each S / N ratio residual covariance matrix calculation means for outputting a matrix and a prediction vector at the current time, and using the output of each S / N ratio residual covariance matrix calculation means. A fine gate, which is an area, is calculated, and for the detection data and S / N ratio contained in the input coarse gate, the detection data and S / N ratio contained in the fine gate are determined. Calculates the S / N ratio residual covariance matrix S _kj (t) by the fine gate depending on the number and the number of wakes, and determines the detected data and the S / N ratio contained in the determined fine gate. S
/ N ratio residual covariance matrix S _kj (t), and the input of the current time prediction error covariance matrix and the current time prediction vector. Data updating means for calculating and outputting a smoothed error covariance matrix at the current time and a smoothed vector at the current time, and a prediction error covariance matrix one sampling after the current time using the output of the data updating means. And a prediction unit that calculates and outputs a prediction vector one sample after the current time, and a delay unit that outputs a prediction error covariance matrix of the current time and a prediction vector of the current time using the output of the prediction unit. A tracking device, comprising: