JP2000241522A - Data-integrating and tracking apparatus between a plurality of sensors - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data-integrating and tracking apparatus, between a plurality of sensors, by which the position of a target can be tracked continuously and precisely. SOLUTION: A cross-fixing-line tracking and processing part 14 in which the position of a target is found by a cross fixing operation only between lines having the same frequency in different sensor arrays when received signals are integrated according to the target, in which the position of the target is tracked in terms of time and in which cross fixing lines at respective frequencies are created is provided. In addition, a discrimination processing part 16 in which whether the cross fixing lines at the respective frequencies are radiated from the same target or not is judged by using a distance to be standardized by a dispersion operation and in which the received signals are integrated according to the target is provided.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data integration / tracking apparatus for a plurality of sensors for integrating and tracking data for each target by using arrival direction and frequency information of signals received by a plurality of sensors.

[0002]

2. Description of the Related Art FIG. 3 is a block diagram showing a configuration of a conventional data integration / tracking device between a plurality of sensors, and shows a series of flows from signal detection to target tracking. In the figure, 1-1 and 1-2 are sensor arrays, 2-1 and 2-2 are fast Fourier transform (FFT) / phasing, 3-1 and 3-
2 is signal detection / tracking processing, 4-1 and 4-2 are integration / fusion processing, 5-1 and 5-2 are identification processing, 6 is cross fix processing, and 7 is target track tracking processing.

Here, as an example, an input signal is a sound wave,
A method of integrating data between a plurality of sensors using two sensor arrays arranged at different positions will be described. First,
FFT / phasing processing 2- is performed on a time-series signal input to a sensor array 1-1 in which a plurality of sensors are arranged on a straight line.
In step 1, a fast Fourier transform (FFT) and a phasing process are performed to convert a time-series signal into level data in an azimuth and frequency space.

The signal detection / tracking process 3-1 detects a local maximum point of the level in the azimuth and frequency space, and tracks it temporally, thereby forming a line (a temporal sequence of local maximum points). Perform detection. Then, the integration / fusion processing 4-
In step 1, an integrated process for generating a combination of lines determined to be radiated from the same sound source (hereinafter, referred to as a track) from the line direction information and a fusion direction calculated from the direction information of each line of the track To perform a fusion process.

In the identification process 5-1, the target of the track is identified by using the frequency information of the line included in each track integrated by the line integration / fusion processing unit. Also, sensor array 1-2, FFT / phasing processing 2
-2, automatic detection / tracking processing 3-2, data integration processing 4-
2. In the identification processing 5-2, the same processing as the processing of the signal input to the sensor array 1-1 is performed.

In the crossfix processing 6, the following formula (1) is obtained from the fusion direction of the tracks identified as the same target in the identification processing 5-1 and the identification processing 5-2 and the positions of the two sensor arrays. The target position is calculated according to the following. The symbols in the equation (1) are as shown in FIG.

[0007]

(Equation 1)

In the target track tracking process 7, the target position calculated at each time is tracked temporally. Also,
Tracks that cannot be cross-fixed in the target track tracking processing 7 (tracks that are not determined to be the same target in each of the identification processes 5-1 and 5-2, or tracks that cannot be calculated for the cross-fix by equation (1) even if they are identified as the same target) ) Temporally tracks the fusion direction in the integration / fusion processing 4-1 and 4-2.

[0009]

In the conventional data integration / tracking apparatus between a plurality of sensors as described above, the sensor array 1-1 and the sensor array 1-1 and the sensor array 1-1 and the target sound source levels decrease due to an increase in background noise and a decrease in a target sound source level.
When the detected SN ratio of the line detected in 1-2 decreases, the azimuth error increases, and each sensor array 1-1, 1-2
However, there is a problem that if the integration is performed only by the direction of arrival, the lines radiating from different targets cannot be correctly integrated.

For example, as shown in FIG. 5, there are actually targets at the positions of target 1 and target 2 and the frequency f
Despite radiating 1, f2, cross-fix processing cannot be performed with the direction (f1 + f2) obtained by integrating the directions in the sensor array 1-1 and the direction information (f1, f2) in the sensor array 1-2. In other words, it is impossible to correctly integrate azimuth information for each target. Further, in the cross fix processing 6, when the azimuth error increases, the fusion azimuth in each sensor array becomes parallel or more than parallel, and there is a problem that the target position cannot be calculated.

Further, when three or more sensor arrays are arranged on a straight line as shown in FIG. 6, a state in which two sensor arrays are simultaneously detected and a state in which only one sensor array is detected, Then, the state is shifted to a state detected by a sensor array different from the sensor array detected first, and the tracking from the target position to the tracking only in the azimuth is performed, so that the target position is continuously and accurately tracked. There was a problem that it was not possible.

[0012]

According to the present invention, there is provided an apparatus for integrating and tracking data between a plurality of sensors, wherein signals radiated from a plurality of targets are simultaneously received by sensor arrays arranged at different positions, and based on the received signals. In a multi-sensor data integration / tracking device that integrates received signals by target and tracks a target based on the integrated received signal, when integrating received signals by target, a line between lines having the same frequency in different sensor arrays is used. Only the position of the target by cross-fix is determined, the position of the target is tracked in time, the cross-fix line of each frequency is created, and the cross-fix line of each frequency is Means for judging whether or not the light is emitted from the same target, and performing integration for each target.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIG. 1 is a block diagram showing a configuration of an apparatus for integrating and tracking data between a plurality of sensors according to an embodiment of the present invention. In the figure, reference numeral 10-1 denotes a sensor array, which is connected to an FFT / phasing process 11-1, and an FFT / phasing process 11-1 connected to the sensor array 10-1 is a signal detection / tracking process 12-1. Is connected to the FFT / phasing process 11-1.
The tracking processing 12-1 is connected to the same frequency cross-fix processing 13.

A sensor array 10-2, an FFT / phasing process 11-2, and a signal detection / tracking process 12-2, which are arranged at positions different from the sensor array 10-1, also perform signal detection from the sensor array 10-1. The signal detection / tracking process 12-2 is also connected to the same frequency cross-fixing process 13, and the signal detection / tracking process 12-2 is connected in the same manner as the connection leading to the / tracking process 12-1.

The same-frequency crossfix processing 13 connected to the signal detection / tracking processing 12-1 and 12-2 is connected to the crossfix line tracking processing 14 and the azimuth line integration / fusion processing 17, and The cross fix line tracking processing 14 connected to the cross fix processing 13
It is connected to the fusion processing 15. The crossfix line integration / fusion processing 15 connected to the crossfix line tracking processing 14 is the azimuth line integration / fusion processing 17 connected to the same frequency crossfix processing 13.
Together with the identification processing 16.

Next, the operation of this embodiment will be described. First, the sensor arrays 10 arranged at different positions
-1 and 10-2 are respectively subjected to FFT / phasing 11-1 and FFT / phasing 11
At -2, fast Fourier transform and phasing are performed to convert the time-series signal into azimuth and level data in the frequency space.

In signal detection / tracking processing 12-1 and 12-2, a line maximum is detected by detecting a local maximum point in the azimuth and frequency space, and tracking it temporally. Cross-fix processing between same frequencies 1
3, the signal detection / tracking processing 12-1 and 1
2-2 A line set having the same frequency is selected from the created lines, and an intersection (hereinafter, referred to as a cross-fix point) of each line is determined from the positions of the two sensor arrays and the azimuth information of the selected line. calculate.

In the cross-fix line tracking process 14, the cross-fix points created in the same-frequency cross-fix process 13 are tracked temporally, and the temporal sequence of cross-fix points (hereinafter referred to as a cross-fix line). ) To create. Then, in the crossfix line integration / fusion processing 15, the integration / separation determination between the crossfix lines having different frequency information is determined by the distance normalized by the variance calculated by the following equation (2) (hereinafter, the statistical distance). ) And a preset threshold value, and a set of a plurality of cross-fix lines (hereinafter, referred to as a cross-fix track) is created.

[0019]

(Equation 2)

In the azimuth line integration / fusion processing 17, the lines detected by the signal detection / tracking processing 12-1 and 12-2 to the same frequency crossfix processing 13 are not determined to have the same frequency. Using a line (hereinafter referred to as an azimuth line), integration / separation determination is performed. Here, the integration / separation determination is performed by comparing the azimuth difference normalized by the variance calculated by the following equation (3) (hereinafter, referred to as statistical azimuth difference) with a preset threshold value. A set of azimuth lines (hereinafter referred to as azimuth track) is created.

[0021]

(Equation 3)

Then, in the identification processing 16, the frequency information of the cross-fix track created in the cross-fix line integration / fusion processing 15 and the azimuth track created in the azimuth line integration / fusion processing 17 are combined with the preset target-specific information. Compare sets of frequency information and identify which target it is.

In this embodiment, even when the azimuth error of the sensor array is large, the cross-fix track formed only between the same frequency lines detected by each sensor array and the line detected only by each sensor array are used. Since a separate track is created, the azimuth track to be created, it is possible to reduce mismerging, which mistakenly merges lines radiating from different targets.

Embodiment 2 FIG. FIG. 2 is a block diagram showing a configuration of a device for integrating and tracking data between a plurality of sensors according to another embodiment of the present invention. In the figure, reference numeral 20-1 denotes a sensor array, which is connected to an FFT / phasing process 21-1;
FFT / phasing process 2 connected to sensor array 20-1
1-1 is connected to the signal detection / tracking process 22-1, and the FF
The signal detection / tracking processing 22-1 connected to the T / phasing processing 21-1 is connected to the same frequency cross-fix line tracking processing 23.

Further, a sensor array 20-2, an FFT / phasing process 21-2, and a signal detection / tracking process 22-2 which are arranged at positions different from the sensor array 20-1 also perform signal detection / tracking from the sensor array 20-1. The connection is made in the same manner as the connection leading to the tracking processing 22-1, and the signal detection / tracking processing 22-2 is also connected to the same frequency cross-fix line tracking processing 23.

The same frequency crossfix line tracking processing 23 connected to the signal detection / tracking processing 22-1 and 22-2 is connected to the crossfix line integration / fusion processing 24 and the azimuth line integration / fusion processing 26. , Same frequency cross-fix line tracking processing 23
Cross-fix line integration and fusion processing 2 connected to
4 is a cross-fix line tracking process 23 between the same frequencies.
Are connected to the identification processing 25 together with the azimuth line integration / fusion processing 26 connected to the processing.

Next, the operation of this embodiment will be described. First, the sensor arrays 20 arranged at different positions
-1 and 20-2 are respectively subjected to FFT / phasing 21-1 and FFT / phasing 21
At -2, fast Fourier transform and phasing are performed to convert the time-series signal into azimuth and level data in the frequency space.

Then, in the signal detection / tracking processing 22-1 and 22-2, a line maximum is detected by detecting a local maximum point in the azimuth and frequency space, and tracking it temporally. In the inter-frequency cross-fix line tracking processing 23, the signal detection / tracking processing 22-1 is performed, respectively.
And a line set of the same frequency is selected from the lines created in 22-2, and the azimuth and frequency at each sensor array are used as observation values, and the cross-fix point, speed,
Using an extended Kalman filter for estimating a radiation frequency, tracking of a crossfix point is performed, and a crossfix line is created for each frequency.

In the crossfix line integration / fusion processing 24, the integration / separation determination between the crossfix lines having different frequency information is determined by comparing the statistical distance calculated by the above equation (2) with a preset threshold value. Make multiple cross-fix tracks by comparison. In the azimuth line integration / fusion processing 26, among the lines input to the inter-frequency crossfix line tracking processing 23 from the signal detection / tracking processing 22-1 and 22-2, the azimuth not determined to be the same frequency is used. The integration / separation determination is performed using the line. Here, the integration / separation determination is performed by comparing the statistical azimuth difference calculated by the above equation (3) with a preset threshold value to create a plurality of azimuth tracks.

In the identification process 25, the frequency information of the cross-fix track created in the cross-fix line integration / fusion process 24 and the azimuth track created in the azimuth line integration / fusion process 26 are combined with the preset target-specific information. Compare sets of frequency information and identify which target it is.

In this embodiment, the azimuth error of the sensor array is large by introducing an extended Kalman filter in the tracking processing in the same frequency cross-fix line tracking processing. Even when a point is not obtained, the cross-fix point can be estimated from the azimuth and azimuth variance of each sensor array that is continuously detected. Further, even in the case of being simultaneously detected by two or more sensor arrays once, and later being detected by only one sensor array, a cross-fix point is obtained based on the information estimated up to then,
The tracking of the cross fix line is performed, and the cross fic point can be continuously estimated. Therefore, stable tracking of the cross-fix line becomes possible from the above.

[0032]

As described above, according to the present invention, when the received signals are integrated for each target, the target position by the cross-fix is obtained only between the lines having the same frequency in the different sensor arrays, and the target position is determined by time. Tracking, create cross-fix lines for each frequency, use the distance standardized by variance to determine whether the cross-fix lines for each frequency are emitted from the same target, and integrate each target. Since it is performed, there is an effect that erroneous integration of erroneously integrating lines radiated by different targets can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an apparatus for integrating and tracking data between a plurality of sensors according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of an apparatus for integrating and tracking data between a plurality of sensors according to another embodiment of the present invention.

FIG. 3 is a block diagram showing a configuration of a conventional data integration / tracking device between multiple sensors.

FIG. 4 is an explanatory diagram for explaining reference numerals in Expression (1).

FIG. 5 is an explanatory diagram for explaining a conventional problem.

FIG. 6 is an explanatory diagram for explaining a conventional problem.

[Explanation of symbols]

 10-1, 10-2 Sensor array 11-1, 11-2 FFT / phasing processing 12-1, 12-2 Signal detection / tracking processing 13 Cross-fix processing between same frequencies 14 Cross-fix line tracking processing 15 Cross-fix line Integration / fusion processing 16 Identification processing 17 Azimuth line integration / fusion processing 20-1, 20-2 Sensor array 21-1, 21-2 FFT / phasing processing 22-1, 22-2 Signal detection / tracking processing 23 Same frequency Cross-fix line tracking processing 24 Cross-fix line integration / fusion processing 25 Identification processing 26 Azimuth line integration / fusion processing

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Mitsuru Yamada 1-7-12 Toranomon, Minato-ku, Tokyo Oki Electric Industry Co., Ltd. (72) Inventor Satoshi Tanaka 1-7-112 Toranomon, Minato-ku, Tokyo F term (reference) in Electric Industry Co., Ltd. 5J083 AA05 AB09 AD02 AD17 BC13 BE43 BE56 CA07 CA12

Claims (4)

[Claims] 1. A signal emitted by a plurality of targets is simultaneously received by sensor arrays arranged at different positions, the received signals are integrated for each target based on the received signals, and the targets are tracked based on the integrated received signals. When integrating received signals by target, the target position is determined by cross-fix only between lines with the same frequency in different sensor arrays, and the target position is tracked in time. Then, using a means for creating a cross-fix line of each frequency and a distance standardized by dispersion, determine whether the cross-fix line of each frequency is radiated from the same target, and perform integration for each target. Means for performing data integration and tracking between a plurality of sensors. 2. The method according to claim 1, wherein an extended Kalman filter is used in a tracking process when a target position between the lines having the same frequency is temporally tracked to form a crossfix line. Data integration and tracking device between multiple sensors. 3. The signals emitted by a plurality of targets are simultaneously received by at least two sensor arrays arranged at different positions, and the received signals are integrated for each target based on the received signals, and based on the integrated received signals. A data integration / tracking device between multiple sensors for tracking a target, wherein at least two fast Fourier transform / phasing processing means for performing fast Fourier transform and phasing processing on the signals received by the sensor array, respectively, the fast Fourier transform At least two signal detections for detecting a line by detecting a local maximum point in the azimuth and frequency space from each of the signals processed by the phasing processing means and tracking the detected point in time. Tracking processing means, and selecting a set of lines of the same frequency from the respective lines detected by the signal detection / tracking processing means, Cross-fix processing means for calculating the respective cross-fix points from the position of the sensor array and the azimuth information of the selected line; and tracking the cross-fix points created by the same-frequency cross-fix processing means in time. And a cross-fix line tracking processing means for creating a cross-fix line, and integration / separation determination between cross-fix lines having different frequency information created by the cross-fix line tracking processing means, a distance normalized by dispersion. Cross-fix line integration / fusion processing means for making a cross-fix track by comparing the magnitude with a predetermined threshold value, and a line input from the signal detection / tracking processing means to the same frequency cross-fix processing means. Did not determine the same frequency Azimuth line integration / fusion processing means for performing integration / separation determination using a position line; a cross fix track created by the cross fix line integration / fusion processing means; and an azimuth created by the azimuth line integration / fusion processing means. An apparatus for integrating and tracking data between a plurality of sensors, comprising: comparing frequency information of a track with a preset set of frequency information for each target, and identifying which target the target is. 4. A signal emitted by a plurality of targets is simultaneously received by at least two sensor arrays arranged at different positions, and the received signals are integrated for each target based on the received signals, and based on the integrated received signals. A data integration / tracking device between multiple sensors for tracking a target, wherein at least two fast Fourier transform / phasing processing means for performing fast Fourier transform and phasing processing on the signals received by the sensor array, respectively, the fast Fourier transform At least two signal detections for detecting a line by detecting a local maximum point in the azimuth and frequency space from each of the signals processed by the phasing processing means and tracking the detected point in time. Tracking processing means, and selecting a set of lines of the same frequency from the respective lines created by the signal detection / tracking processing means, Using the extended Kalman filter that sequentially estimates the cross-fix point, velocity, and emission frequency using the azimuth and frequency of each sensor array as observation values, track the cross-fix points, and create a cross-fix line for each frequency The same frequency crossfix line tracking processing means, and the integration / separation determination between the crossfix lines having different frequency information created by the same frequency crossfix line tracking processing means, A cross-fix line integration / fusion processing means for making a plurality of cross-fix tracks by comparing the magnitude with a set threshold value; and a line input to the same frequency cross-fix processing means from the signal detection / tracking processing means. The direction line that was not determined to be the same frequency Azimuth line integration / fusion processing means for performing integration / separation determination; frequency of a cross-fix track created by the cross-fix line integration / fusion processing means and azimuth track created by the azimuth line integration / fusion processing means A data integration / tracking device between multiple sensors, comprising: comparing information with a preset set of frequency information for each target, and identifying which target is the target.