JP2013228310A - Device and method for tracking target - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately estimate the position of a target by using plural sensors different in observation accuracy from each other.SOLUTION: On the basis of the correlation determination results of a tracking processing unit 4, the determination results of a continuity determination unit 7, and the determination results of an identity determination unit 8, a wake selection unit 9 selects an integrated wake calculated by a tracking processing unit 3, or a high accuracy sensor wake calculated by the tracking processing unit 4. Thus, a low accuracy sensor 1 and a high accuracy sensor 2 different in observation accuracy are used to estimate the position of a target with high accuracy.

Description

  The present invention uses a plurality of sensors with different observation accuracy such as radar and GPS (Global Positioning System), for example, to observe the position of a target such as an aircraft, a ship, or a vehicle and estimate the target position with high accuracy. The present invention relates to a target tracking device and a target tracking method.

Non-Patent Document 1 below discloses a target tracking device that reduces target tracking errors by performing tracking processing using a tracking filter using observation values of targets of a plurality of sensors. .
However, when the observation values of a plurality of sensors having different observation accuracy are used, the tracking position accuracy may be deteriorated by the observation values of the sensor having a low observation accuracy.

In the following Patent Document 1, tracking is performed using observation values of wide-area sensors (low accuracy, wide coverage, with communication delay) and onboard sensors (high accuracy, narrow coverage, no delay). A target tracking device that performs processing is disclosed.
In this target tracking device, the number of detection targets is identified from the observation values of the wide-area sensor, and the reliability of the hypothesis is increased using the detection target number, while the tracking processing is performed using the observation values of the on-board sensor. Yes.
That is, in this target tracking device, the target position is estimated by using only the observation value of the high-precision mounted sensor without using the observation value of the low-precision wide-area sensor in the target tracking process. Yes.

JP 2007-147306 A (paragraph numbers [0006] to [0007])

Science theory B, vol. J83-B, no. 12, pp. 1747-1759 (200.12), Kameda / Shinji / Kominato, “Turning multiple target tracking with wide area multiple radars”

  Since the conventional target tracking device is configured as described above, when tracking processing is performed using observation values of a plurality of sensors having different observation accuracy (Non-patent Document 1), observation of a sensor with low observation accuracy is performed. Depending on the value, the tracking position accuracy may deteriorate. Further, in the target tracking process, when the target position is estimated using only the observation value of the high-precision mounted sensor without using the observation value of the low-accuracy wide-area sensor (Patent Document 1), the target tracking process is highly accurate. If the observation value of the on-board sensor is missing, there is a problem that the tracking position accuracy deteriorates.

  The present invention has been made to solve the above-described problems, and provides a target tracking device and a target tracking method capable of estimating a target position with high accuracy using a plurality of sensors having different observation accuracy. For the purpose.

  A target tracking device according to the present invention includes a first sensor that observes a target position, a second sensor that observes the target position with higher observation accuracy than the first sensor, and an observation result of the first sensor. The first track calculation for determining the correlation between the target track calculated before the current calculation and the observation result of the first sensor while calculating the target track using the observation result of the second sensor. A second track for calculating a target track using the means and the observation result of the second sensor, and determining a correlation between the target track calculated before the current calculation and the observation result of the second sensor The wake calculation means and the continuity determination means for determining the continuity of the target wake calculated by the first wake calculation means from the determination results of the first wake calculation means and the first wake calculation means Target wake and second wake calculation means Identity determination means for determining the identity with the calculated target track, the wake selection means includes a determination result of the second wake calculation means, a determination result of the continuity determination means, and an identity determination means. Based on the determination result, the target track calculated by the first track calculation unit or the target track calculated by the second track calculation unit is selected.

  According to this invention, the wake selection means is calculated by the first wake calculation means based on the determination result of the second wake calculation means, the determination result of the continuity determination means, and the determination result of the identity determination means. Since the target track or the target track calculated by the second track calculating means is selected, the target position can be estimated with high accuracy using a plurality of sensors having different observation accuracy. There is an effect that can be done.

It is a block diagram which shows the target tracking apparatus by Embodiment 1 of this invention. It is a flowchart which shows the processing content of the wake selection output part 5 of the target tracking apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the characteristic of a low precision sensor and a high precision sensor. It is explanatory drawing which shows the scenario which a subject generate | occur | produces at the time of target constant-velocity straight advance. It is explanatory drawing which shows the scenario which a subject generate | occur | produces at the time of target turning. It is explanatory drawing which shows the case where a highly accurate sensor track and highly accurate sensor data are correlated. It is explanatory drawing which shows the case where there is no continuity of an integrated wake. It is explanatory drawing which shows the case where the identity of an integrated track and a highly accurate sensor track is not recognized. It is explanatory drawing which shows the case where the identity of an integrated track and a high precision sensor track is recognized. It is a block diagram which shows the target tracking apparatus by Embodiment 2 of this invention. It is a flowchart which shows the processing content of the wake selection output part 15 of the target tracking apparatus by Embodiment 2 of this invention. It is explanatory drawing which shows the characteristic of the sensor from which observation accuracy differs. It is explanatory drawing which shows the case where the 1st high precision sensor track and the 1st high precision sensor data are correlated. It is explanatory drawing which shows the case which can consider that a 2nd highly accurate sensor track is correct. It is explanatory drawing which shows the case where there is no continuity of an integrated wake. It is explanatory drawing which shows the case where the identity of an integrated track and the selection high precision sensor track is not recognized. It is a block diagram which shows the target tracking apparatus by Embodiment 3 of this invention.

Embodiment 1 FIG.
1 is a block diagram showing a target tracking apparatus according to Embodiment 1 of the present invention.
In FIG. 1, a low-precision sensor 1 is a first sensor that observes a target position and outputs low-precision sensor data that is the observed value.
In the first embodiment, as shown in FIG. 3, the low accuracy sensor 1 is characterized by low position observation accuracy, high detection rate at the time of target turning, and many unnecessary signals. .
The high-precision sensor 2 is a second sensor that observes the target position with higher observation accuracy than the low-precision sensor 1 and outputs high-precision sensor data that is the observed value.
In the first embodiment, as shown in FIG. 3, the high-precision sensor 2 is characterized in that the position observation accuracy is high, the detection rate during the target turn is low, and there are few unnecessary signals. .

  The tracking processing unit 3 is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like, and the low precision sensor data output from the low precision sensor 1 and the high precision sensor 2 output. A target track (hereinafter referred to as “integrated track”) is calculated by performing a target tracking process (for example, a tracking process using a tracking filter such as a known Kalman filter) using the accuracy sensor data. Then, a process of determining the correlation between the target integrated wake calculated before the current calculation and the low accuracy sensor data output from the low accuracy sensor 1 is performed. The tracking processing unit 3 constitutes a first wake calculation unit.

  The tracking processing unit 4 is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like, and uses the high-precision sensor data output from the high-precision sensor 2 to perform target tracking processing (for example, In addition to calculating a target track (hereinafter referred to as “high-precision sensor track”) by performing a tracking process using a tracking filter such as a known Kalman filter, the target calculated before the current calculation is calculated. The process of determining the correlation between the high-accuracy sensor track and the high-accuracy sensor data output from the high-accuracy sensor 2 is performed. The tracking processing unit 4 constitutes a second wake calculation unit.

The wake selection output unit 5 includes a correlation determination result input unit 6, a continuity determination unit 7, an identity determination unit 8, and a wake selection unit 9.
The correlation determination result input unit 6 is composed of, for example, a semiconductor integrated circuit mounted with a CPU or a one-chip microcomputer. The correlation determination result of the tracking processing unit 4 is input, and the correlation determination result is selected as a wake. Processing to be output to the unit 9 is performed.

The continuity determination unit 7 is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted or a one-chip microcomputer. The correlation determination result of the tracking processing unit 3 is input, and the continuation of the integrated wake is continued from the correlation determination result. A process for determining sex is performed. The continuity determination unit 7 constitutes continuity determination means.
The identity determination unit 8 is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like. The integrated track calculated by the tracking processing unit 3 and the high accuracy calculated by the tracking processing unit 4 A process for determining the identity with the sensor track is performed. The identity determination unit 8 constitutes identity determination means.

The wake selection unit 9 is composed of, for example, a semiconductor integrated circuit mounted with a CPU or a one-chip microcomputer, and the correlation determination result of the tracking processing unit 4, the determination result of the continuity determination unit 7, and the identity determination unit Based on the determination result of No. 8, the process of selecting the integrated track calculated by the tracking processing unit 3 or the high-accuracy sensor track calculated by the tracking processing unit 4 is performed.
That is, the wake selection unit 9 selects the high-precision sensor wake calculated by the tracking processing unit 4 if the correlation determination result of the tracking processing unit 4 indicates that they are correlated.
On the other hand, if the correlation determination result of the tracking processing unit 4 indicates that there is no correlation, the tracking processing unit 4 determines that the determination result of the continuity determination unit 7 indicates that there is no continuity of the integrated track. Select the calculated high accuracy sensor track.
If the determination result of the continuity determination unit 7 indicates that the integrated track has continuity, the tracking processing unit 4 determines that the determination result of the identity determination unit 8 indicates that there is identity. If the calculated high-precision sensor track is selected and the determination result of the identity determination unit 8 indicates that there is no identity, the integrated track calculated by the tracking processing unit 3 is selected.
The wake selection unit 9 constitutes wake selection means.

In the example of FIG. 1, the low-accuracy sensor 1, the high-accuracy sensor 2, the tracking processing unit 3, the tracking processing unit 4, the correlation determination result input unit 6, the continuity determination unit 7, and the identity determination, which are components of the target tracking device. Although it is assumed that each of the unit 8 and the track selection unit 9 is configured by dedicated hardware, all or a part of the target tracking device may be configured by a computer.
For example, when a part of the target tracking device (for example, a portion excluding the low accuracy sensor 1 and the high accuracy sensor 2) is configured by a computer, the tracking processing unit 3, the tracking processing unit 4, the correlation determination result input unit 6, A program describing the processing contents of the continuity determination unit 7, the identity determination unit 8, and the track selection unit 9 is stored in the memory of a computer, and the CPU of the computer executes the program stored in the memory You can do it.
FIG. 2 is a flowchart showing the processing contents of the wake selection output unit 5 of the target tracking device according to Embodiment 1 of the present invention.

Prior to describing the specific operation of the target tracking device, the premise of the invention will be described.
An example of a system for observing a target using observation results of a plurality of sensors is an air traffic control support system for observing an aircraft (target).
In aircraft observation, it is necessary to accurately grasp the current position of the aircraft in order to ensure the safety of the aircraft.
Examples of sensor devices for knowing the current position of an aircraft include an airport surveillance radar (ASR), an air route surveillance radar (ARSR), and a wide area multi-lateration (WAM). And an automatic dependent surveillance (ADS) receiver.

The airport monitoring radar is mainly a radar for detecting an aircraft or the like that is flying in the airspace around the airport. For example, a primary surveillance radar (PSR) and a secondary surveillance radar (SSR: Secondary Surveillance). Radar).
The primary surveillance radar radiates radio waves from a directional antenna that rotates once at a fixed period, and receives the reflected waves reflected by the radiated radio waves hitting an aircraft or the like, so that the distance between the directional antenna and the aircraft Detecting the direction of the aircraft. However, the primary monitoring radar has a feature that there are many unnecessary signals because it also receives reflected waves from other than aircraft (such as clouds and the sea surface) as unnecessary signals.

The secondary monitoring radar radiates a radio wave (question signal) of 1030 MHz from a directional antenna that rotates together with the antenna of the primary monitoring radar.
Thus, when a transponder mounted on an aircraft receives a 1030 MHz radio wave (question signal), a 1090 MHz radio wave (response signal) is transmitted as a response to the interrogation signal.
The secondary monitoring radar receives a 1090 MHz radio wave (response signal) to detect the distance between the directional antenna and the aircraft and the direction of the aircraft.
Since the response signal includes the identification code of the aircraft, unnecessary signals other than aircraft (such as clouds and sea surface) are not acquired. In order to prevent radio waves (response signals) transmitted from multiple aircrafts in the same direction from congesting and becoming unidentifiable, the secondary monitoring radar specifies the identification code of the aircraft that should respond. Some aircraft transponders having the identification code have a function of returning radio waves (response signals) (SSR mode S).

Wide area multilateration measures the arrival time by detecting the response of the capture squitter and secondary surveillance radar of the aircraft collision prevention device transmitted by the transponder mounted on the aircraft at a plurality of receiving stations.
Wide area multilateration calculates the difference in arrival time between the aircraft and the plurality of receiving stations by calculating the arrival time difference between the plurality of receiving stations from the measured arrival time. The position of the aircraft is calculated by obtaining the intersection of the elliptic hyperboloids composed of the distance difference.

  An automatic subordinate monitoring receiver uses an GPS (Global Positioning System) receiver, an inertial navigation system (INS), etc. to calculate its own position, speed, etc. When information is transmitted through a broadcast-type automatic dependent surveillance (ADS-B) device, the position of the aircraft is specified by receiving information such as position and speed.

Hereinafter, the subject which the present invention presupposes is demonstrated.
FIG. 3 is an explanatory view showing the characteristics of the low accuracy sensor 1 and the high accuracy sensor 2.
FIG. 4 is an explanatory diagram showing a scenario in which a problem occurs when the target goes straight at a constant speed, and FIG. 5 is an explanatory diagram showing a scenario in which a problem occurs when the target turns.
As shown in FIG. 3, the low-precision sensor 1 is characterized by low position observation accuracy, a high detection rate during target turning, and many unnecessary signals. It is characterized by high accuracy, low detection rate during target turning, and few unnecessary signals.

Therefore, for example, in the air traffic control support system, the low-precision sensor data output from the low-precision sensor 1 corresponds to PSR data, and the high-precision sensor data output from the high-precision sensor 2 corresponds to SSR data.
When the target turns, the SSR data has a high possibility that the detection rate will drop due to the fact that the transponder mounted on the aircraft is hidden behind the aircraft. Regarding the number of unnecessary signals detected, PSR data increases because it does not have an aircraft identification code.

As shown in FIG. 4, when the target performs a constant-velocity linear motion, the integrated track (the target track calculated by performing the tracking process using the low-precision sensor data and the high-precision sensor data) Compared with the accuracy sensor track (the target track calculated by performing the tracking process using only the high-precision sensor data), the integrated track uses the low-accuracy sensor data. There is a problem that the tracking accuracy deteriorates.
In addition, as shown in FIG. 5, when the target performs a turning motion, there is a high possibility that high-precision sensor data will be lost when the target turns, so there is a problem that tracking accuracy deteriorates only with the high-precision sensor track. is there.

In the present invention, deterioration in tracking accuracy during straight traveling or turning is prevented, and the target position can be estimated with high accuracy.
Hereinafter, the operation of the target tracking device of FIG. 1 will be specifically described.

The low accuracy sensor 1 observes the target position and outputs the low accuracy sensor data that is the observed value to the tracking processing unit 3.
The high-accuracy sensor 2 observes the target position with higher observation accuracy than the low-accuracy sensor 1 and outputs the high-accuracy sensor data as the observation values to the tracking processing units 3 and 4.

When the tracking processing unit 3 receives the low-precision sensor data output from the low-precision sensor 1 and the high-precision sensor data output from the high-precision sensor 2, the tracking processing unit 3 uses the low-precision sensor data and the high-precision sensor data to The target integrated wake is calculated by executing the tracking process, and the integrated wake is output to the wake selection output unit 5.
The process of calculating the wake by executing the target tracking process is a known technique, and thus detailed description thereof is omitted. For example, the target tracking process is performed by using a tracking filter such as a known Kalman filter. can do.

The target integrated track includes a target smooth value (a state vector smooth value indicating a target position / velocity and a smooth error covariance matrix) calculated by a tracking filter and a target predicted value (target position / speed). State vector prediction value and prediction error covariance matrix), or both.
Hereinafter, “estimated value” means a state vector smoothed value or state vector predicted value, and “estimated error covariance matrix” means a smoothed error covariance matrix or a predicted error covariance matrix.
Further, the tracking processing unit 3 determines the correlation between the target integrated wake calculated before the current calculation and the low accuracy sensor data output from the low accuracy sensor 1 this time, and selects the correlation determination result as the wake selection. Output to the output unit 5.
Note that the process itself for determining the correlation between the target integrated track and the low-precision sensor data is a known technique, and thus detailed description thereof is omitted.

Here, at the tracking update time t, the tracking processing unit 3 recognizes a correlation between the integrated integrated track calculated before the current calculation and the low accuracy sensor data output from the low accuracy sensor 1 this time. If, as an integrated track the target, the estimated value x _t hat (specification are, on the relationship between the electronic application, it is not possible to subject the symbol "^" over a x _t, as "x _t hat" outputting the representation to) the estimation error covariance matrix P _t to the track selection output section 5 in.
On the other hand, when the correlation is not recognized, the estimated value x _t hat extrapolated in time (formula (Eq. (2)) is calculated from the time difference between the time t ′ when the correlation with the latest low-precision sensor data is recognized and the tracking update time t. an estimate x _t hat) shown in 1), and outputs a time extrapolated estimate error covariance matrix P _t (equation (estimation error covariance matrix P _t shown in 2)) to track selection output unit 5.

In Equations (1) and (2), Φ (t−t ′) is defined as a state transition matrix of a constant velocity linear motion model, for example, as shown in Equation (3) below.
Q (t−t ′) is a driving noise covariance matrix, I _{n × n} is an n _{× n} unit matrix, 0 _{n × n} is an n _{× n} 0 matrix, and a superscript T is transposed. Means the symbol.

式（４）において、ｘ_t,obsは低精度センサデータが示す目標の観測位置、Ｓ_tは残差共分散行列、ε_gateは閾値であり、所定の有意水準に基づいてカイ二乗分布表などから求められる。 The tracking processing unit 3 uses the low-accuracy sensor data and the high-accuracy sensor data so that the time t when calculating the target integrated track is equal to or longer than a certain time from the observation time of the low-accuracy sensor data and the high-accuracy sensor data. If the time has elapsed, the prediction error covariance matrix included in the target prediction value becomes too large (the prediction error covariance matrix increases with the passage of time). A target integrated wake is calculated using a prediction error covariance matrix included in a prediction value calculated after a certain time from the observation time of the accuracy sensor data.
In addition, in the correlation processing with the low-precision sensor data, the tracking processing unit 3 removes the low-precision sensor data as an abnormal value if the value ε _t shown in the following equation (4) is equal to or greater than a predetermined threshold ε _gate. Perform gate inside / outside determination processing.

In the formula _(4), x t, _obs is the observed position of the target indicated by the low accuracy sensor data, S _t is the residual covariance matrix, epsilon _Gate is a threshold, such as chi-square distribution table, based on a predetermined significance level It is requested from.

When the high-precision sensor data output from the high-precision sensor 2 is input, the tracking processing unit 4 calculates a target high-precision sensor wake by performing a target tracking process using the high-precision sensor data. The highly accurate sensor track is output to the track selection output unit 5.
In addition, the tracking processing unit 4 determines the correlation between the target high-precision sensor track calculated before the current calculation and the high-precision sensor data output from the high-precision sensor 2, and selects the correlation determination result as a track selection. Output to the output unit 5.
The track calculation processing by the tracking processing unit 4 is basically the same as the track calculation processing by the tracking processing unit 3, and is different in that data used for calculation is limited to high-precision sensor data.
The tracking processing unit 4 also performs the gate inside / outside determination processing in the same manner as the tracking processing unit 3.

In the tracking processing unit 4, as described above, since the data used for calculation is limited to high-precision sensor data, the estimation error covariance matrix is divergent when the high-precision sensor data is lost when the target turns. There is a high possibility of doing.
Therefore, when the time difference (t−t ′) exceeds a predetermined threshold T _th , the estimated error covariance matrix P _t may be calculated as in the following equation (5). Further, when the lack of high-precision sensor data continues for a certain period, the wake may be deleted.

The correlation determination result input unit 6 of the wake selection output unit 5 inputs the correlation determination result of the tracking processing unit 4 (determination result indicating whether or not the high accuracy sensor track and the high accuracy sensor data are correlated), The correlation determination result is output to the wake selection unit 9.
The continuity determination unit 7 inputs the correlation determination result of the tracking processing unit 3 (determination result indicating whether or not the integrated track and the low-accuracy sensor data are correlated), and the continuity of the integrated track is determined from the correlation determination result. Determine.
That is, the continuity determination unit 7 assumes that an unnecessary signal is observed by the low accuracy sensor 1 and if the integrated track and the low accuracy sensor data are continuously correlated (for example, preset). If the number of times Th _TQ correlates sequentially), determines that there is continuity of integrated track, if integrated track and the low accuracy sensor data correlated continuously, if there is no continuity of integrated track judge.

The identity determination unit 8 determines the identity between the integrated track calculated by the tracking processing unit 3 and the high-accuracy sensor track calculated by the tracking processing unit 4.
This identity can be determined using a statistical method. For example, by performing the chi-square test, the identity of the integrated track and the high-precision sensor track can be determined.
Specifically, the integrated wake estimated value x _{t, FUSE} hat, the integrated wake estimated error covariance matrix P _{t, FUSE} , the high precision sensor wake estimated value x _{t, SNS} hat, and the high precision sensor wake estimated error The test value ε is calculated by substituting the variance matrix P _{t, SNS} into the following equation (6).

ここで、閾値ε_thは、所定の有意水準に基づいて、例えば、カイ二乗分布表から求めたものである。例えば、有意水準５％で検定すると、危険率５％で、本来同一である航跡が、同一の航跡でないとする誤判定がなされることを意味する。 After calculating the test value ε, the identity determination unit 8 compares the test value ε with a predetermined threshold value ε _th as shown in the following equation (7), and if the test value ε is equal to or greater than the threshold value ε _th. For example, the identity of the integrated track and high-precision sensor track is not recognized.
On the other hand, if the test value ε is less than the threshold value ε _th , the identity of the integrated track and the high-precision sensor track is recognized.

Here, the threshold value ε _th is obtained from, for example, a chi-square distribution table based on a predetermined significance level. For example, if a test is performed at a significance level of 5%, it means that a misjudgment is made that the originally identical track is not the same track with a risk rate of 5%.

The wake selection unit 9 is based on the correlation determination result of the tracking processing unit 4 output from the correlation determination result input unit 6, the determination result of the continuity determination unit 7, and the determination result of the identity determination unit 8. The integrated track calculated by the above or the high accuracy sensor track calculated by the tracking processing unit 4 is selected.
Hereinafter, the wake selection process by the wake selection output unit 5 will be described in detail.

The track selection unit 9 of the track selection output unit 5 indicates the correlation determination result of the tracking processing unit 4 output from the correlation determination result input unit 6 (shows whether or not the high-accuracy sensor track and the high-accuracy sensor data are correlated). If the determination result indicates that they are correlated (step ST1 in FIG. 2), the high-precision sensor track calculated by the tracking processing unit 4 is selected (step ST2).
When the high-accuracy sensor track and the high-accuracy sensor data are correlated, a case as shown in FIG. 6 can be considered.
When the wake selection unit 9 selects the high-precision sensor track calculated by the tracking processing unit 4, the continuity determination unit 7 of the wake selection output unit 5 uses a track quality value TQ (Track Quality) that becomes an index of continuity of the integrated wake. ) Is initialized to “0” (step ST3).

When the continuity determination unit 7 indicates that the correlation determination result of the tracking processing unit 4 is not correlated (step ST1), the correlation determination result of the tracking processing unit 3 (correlation between the integrated wake and the low-accuracy sensor data) If the determination result indicating whether or not the recording is correlated (step ST4), the track quality value TQ is incremented by one (step ST5).
On the other hand, if the correlation determination result of the tracking processing unit 3 indicates that there is no correlation (step ST4), the track quality value TQ is counted down by one (step ST6).
If the track quality value TQ is less than the predetermined threshold Th _TQ (step ST7), the continuity determination unit 7 determines that there is no continuity of the integrated wake, and the wake selection unit 9 is calculated by the tracking processing unit 4. A highly accurate sensor track is selected (step ST8).
When the track quality value TQ is less than the threshold Th _TQ (when there is no continuity of the integrated track), the low-accuracy sensor data may be an unnecessary signal as shown in FIG. As long as this is not secured, the possibility of deviating from the actual true trajectory remains, so a highly accurate sensor track is selected.

If the track quality value TQ is equal to or greater than a predetermined threshold Th _TQ (step ST7), the continuity determination unit 7 determines that there is continuity of the integrated track.
If the identity determination unit 8 determines that the continuity determination unit 7 has the continuity of the integrated wake, if the test value ε of the equation (6) is equal to or greater than the threshold ε _th (step ST9), the identity track and the high accuracy are obtained. The track selection unit 9 determines that the sensor tracks are not identical, and selects the high-accuracy sensor track calculated by the tracking processing unit 4 (step ST10).
When the identity of the integrated track and the high-precision sensor track is not recognized, as shown in FIG. 8, it is considered that the integrated track in which continuity is ensured is correct.
If the test value ε of equation (6) is less than the threshold value ε _th (step ST9), the identity determination unit 8 determines that the integrated track and the high-precision sensor track are identical, and the track selection unit 9 The highly accurate sensor track calculated by the tracking processing unit 4 is selected (step ST8).
When the identity of the integrated track and the high-precision sensor track is recognized, a case as shown in FIG. 9 can be considered.

  As is apparent from the above, according to the first embodiment, the wake selection unit 9 is based on the correlation determination result of the tracking processing unit 4, the determination result of the continuity determination unit 7, and the determination result of the identity determination unit 8. Since the integrated track calculated by the tracking processing unit 3 or the high-accuracy sensor track calculated by the tracking processing unit 4 is selected, the low-accuracy sensor 1 and the high-accuracy sensor 2 having different observation accuracy are selected. By using this, there is an effect that the target position can be estimated with high accuracy.

  In addition, although the example applied to the air traffic control support system has been described in the first embodiment, it suffices if the target position is estimated using a plurality of sensors having different observation accuracy, and other than the air traffic control support system. Needless to say, it may be applied to other systems.

Embodiment 2. FIG.
FIG. 10 is a block diagram showing a target tracking apparatus according to Embodiment 2 of the present invention.
In FIG. 10, a low-precision sensor 11 is a first sensor that observes a target position and outputs low-precision sensor data that is the observed value.
In the second embodiment, as shown in FIG. 12, the low-accuracy sensor 11 is characterized in that the position observation accuracy is low, the detection rate at the time of target turning is high, and there are many unnecessary signals. .
The high-precision sensor 12a is a second sensor that observes the target position with higher observation accuracy than the low-precision sensor 1 and outputs high-precision sensor data that is the observed value.
In the second embodiment, as shown in FIG. 12, the high-accuracy sensor 12a is characterized in that the position observation accuracy is high, the detection rate at the time of target turning is low, and there are few unnecessary signals. .

The high-precision sensor 12b is a second sensor that observes the target position with higher observation accuracy than the low-precision sensor 1 and outputs high-precision sensor data that is the observation value.
In the second embodiment, as shown in FIG. 12, the high accuracy sensor 12b has a feature that the position observation accuracy is high to medium, the detection rate at the time of target turning is low, and unnecessary signals are few. It shall be.
The high-precision sensor 12c is a second sensor that observes the target position with higher observation accuracy than the low-precision sensor 1 and outputs high-precision sensor data that is the observation value.
In the second embodiment, as shown in FIG. 12, the high-precision sensor 12c has the characteristics that the position observation accuracy is medium, the detection rate at the time of target turning is low, and there are few unnecessary signals. To do.

  Here, the observation accuracy is high accuracy sensor 12a> high accuracy sensor 12b> high accuracy sensor 12c. For example, in an air traffic control support system, the low accuracy sensor data output from the low accuracy sensor 11 corresponds to PSR data. The high-precision sensor data output from the high-precision sensor 12a corresponds to ADS-B data, and the high-precision sensor data output from the high-precision sensor 12b corresponds to WAM data and is output from the high-precision sensor 12c. High-precision sensor data corresponds to SSR data.

  The tracking processing unit 13 is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like. From the low-precision sensor data output from the low-precision sensor 11 and the high-precision sensors 12a, 12b, and 12c. Using the output high-precision sensor data, target tracking processing (for example, tracking processing using a tracking filter such as a known Kalman filter) is performed to calculate the integrated track of the target and before this time The process of determining the correlation between the integrated track of the target calculated in step S3 and the low accuracy sensor data output from the low accuracy sensor 11 is performed. The tracking processing unit 13 constitutes a first wake calculation unit.

  The tracking processing unit 14a is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like, and uses the high-precision sensor data output from the high-precision sensor 12a to perform target tracking processing (for example, The tracking process using a tracking filter such as a known Kalman filter) is performed to calculate the target high-precision sensor track and the target high-precision sensor track and the high-precision sensor calculated before the current calculation. Processing for determining the correlation with the high-precision sensor data output from 12a is performed.

  The tracking processing unit 14b is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like, and uses the high-precision sensor data output from the high-precision sensor 12b to perform target tracking processing (for example, The tracking process using a tracking filter such as a known Kalman filter) is performed to calculate the target high-precision sensor track and the target high-precision sensor track and the high-precision sensor calculated before the current calculation. Processing for determining the correlation with the high-precision sensor data output from 12b is performed.

The tracking processing unit 14c is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like, and uses the high-precision sensor data output from the high-precision sensor 12c to perform target tracking processing (for example, The tracking process using a tracking filter such as a known Kalman filter) is performed to calculate the target high-precision sensor track and the target high-precision sensor track and the high-precision sensor calculated before the current calculation. The process which determines the correlation with the high precision sensor data output from 12c is implemented.
The tracking processing units 14a, 14b, and 14c constitute second wake calculation means.

  The wake selection output unit 15 is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted or a one-chip microcomputer, and the tracking processing unit is based on the correlation determination results of the tracking processing units 13, 14a, 14b, and 14c. A process of selecting one wake from the integrated wake calculated by 13 and the high-accuracy sensor wake calculated by the tracking processing units 14a, 14b, and 14c is performed. The wake selection output unit 15 constitutes wake selection means.

In the example of FIG. 10, each of the low-precision sensor 11, the high-precision sensors 12 a, 12 b, 12 c, the tracking processing units 13, 14 a, 14 b, 14 c and the wake selection output unit 15 that are components of the target tracking device is dedicated hardware. Although what is comprised with the wear is assumed, all or one part of the target tracking apparatus may be comprised with the computer.
For example, when a part of the target tracking device (for example, a part excluding the low accuracy sensor 11 and the high accuracy sensors 12a, 12b, and 12c) is configured by a computer, the tracking processing units 13, 14a, 14b, and 14c and the wake selection A program describing the processing contents of the output unit 15 may be stored in a memory of a computer so that the CPU of the computer executes the program stored in the memory.
FIG. 11 is a flowchart showing the processing contents of the wake selection output unit 15 of the target tracking device according to Embodiment 2 of the present invention.

Next, the operation will be described.
The low accuracy sensor 11 observes the target position and outputs low accuracy sensor data, which is the observed value, to the tracking processing unit 13.
The high-precision sensor 12a observes the target position with higher observation accuracy than the low-precision sensor 11, and tracks high-precision sensor data (hereinafter referred to as “first high-precision sensor data”) as the observation value. To the units 13 and 14a.
The high-accuracy sensor 12b observes the target position with higher observation accuracy than the low-accuracy sensor 11, and tracks high-accuracy sensor data (hereinafter referred to as “second high-accuracy sensor data”) as the observation value. To the units 13 and 14b.
The high-precision sensor 12c observes the target position with higher observation accuracy than the low-precision sensor 11, and tracks high-precision sensor data (hereinafter referred to as “third high-precision sensor data”) as the observation value. To the units 13 and 14c.

When the tracking processing unit 13 receives the low-precision sensor data output from the low-precision sensor 11 and the high-precision sensor data output from the high-precision sensors 12a, 12b, and 12c, the low-precision sensor data and the three high-precision sensors are input. The target tracking process is performed using the data to calculate the target integrated track, and the integrated track is output to the track selection output unit 15.
Since the process of calculating the wake by performing the target tracking process is a known technique, detailed description thereof is omitted.
Further, the tracking processing unit 13 determines the correlation between the target integrated wake calculated before the current calculation and the low accuracy sensor data output from the low accuracy sensor 11, and the correlation determination result is used as the wake selection output unit. 15 is output.

When the tracking processing unit 14a receives the first high-accuracy sensor data output from the high-precision sensor 12a, the tracking processing unit 14a performs the target tracking process using the first high-accuracy sensor data as in the tracking processing unit 4 of FIG. By performing the calculation, a target high-precision sensor track (hereinafter referred to as “first high-precision sensor track”) is calculated, and the first high-precision sensor track is output to the track selection output unit 15.
Further, the tracking processing unit 14a determines the correlation between the target first high-precision sensor track calculated before the current calculation and the first high-precision sensor data output from the high-precision sensor 12a, and determines the correlation. The result is output to the wake selection output unit 15.

When the tracking processing unit 14b receives the second high-accuracy sensor data output from the high-precision sensor 12b, the tracking processing unit 14b performs the target tracking process using the second high-accuracy sensor data as in the tracking processing unit 4 of FIG. By performing the calculation, a target high-precision sensor track (hereinafter referred to as “second high-precision sensor track”) is calculated, and the second high-precision sensor track is output to the track selection output unit 15.
Further, the tracking processing unit 14b determines the correlation between the target second high-precision sensor track calculated before the current calculation and the second high-precision sensor data output from the high-precision sensor 12b, and determines the correlation. The result is output to the wake selection output unit 15.

When the tracking processing unit 14c receives the third high-accuracy sensor data output from the high-precision sensor 12c, the tracking processing unit 14c performs target tracking processing using the third high-accuracy sensor data as in the tracking processing unit 4 of FIG. By performing the calculation, a target high-accuracy sensor track (hereinafter referred to as “third high-accuracy sensor track”) is calculated, and the third high-accuracy sensor track is output to the wake selection output unit 15.
Further, the tracking processing unit 14c determines the correlation between the target third high accuracy sensor track calculated before the current calculation and the third high accuracy sensor data output from the high accuracy sensor 12c, and determines the correlation. The result is output to the wake selection output unit 15.

The wake selection output unit 15 is based on the correlation determination result of the tracking processing units 13, 14a, 14b, and 14c, and the high accuracy calculated by the integrated wake and tracking processing units 14a, 14b, and 14c calculated by the tracking processing unit 13. One track is selected from the sensor tracks.
Hereinafter, the wake selection process by the wake selection output unit 15 will be specifically described.

The wake selection output unit 15 uses the observation result of the high-precision sensor 12a having the highest observation accuracy, and the correlation determination result of the tracking processing unit 14a that calculates the target wake (the first high-precision sensor wake and the first high-precision sensor wake). If the determination result indicating whether or not the accuracy sensor data is correlated (step ST11 in FIG. 11), the first highly accurate sensor wake calculated by the tracking processing unit 14a is used. Select (step ST12).
When the first high-precision sensor track calculated by the tracking processing unit 14a and the first high-precision sensor data are correlated, a case as shown in FIG. 13 is conceivable.
When the wake selection output unit 15 selects the first high-precision sensor wake calculated by the tracking processing unit 14a, the wake selection output unit 15 initializes the track quality value TQ that is an index of the continuity of the integrated wake to “0” (step ST13). .

  The wake selection output unit 15 indicates that the correlation determination result of the tracking processing unit 14a does not correlate (step ST11), and uses the observation result of the high-precision sensor 12b having the second highest observation accuracy. The correlation determination result (determination result indicating whether or not the second high-accuracy sensor track and the second high-accuracy sensor data are correlated) of the tracking processing unit 14b that calculates the wake of the vehicle is correlated If so (step ST14), the identity of the first high accuracy sensor track calculated by the tracking processing unit 14a and the second high accuracy sensor track calculated by the tracking processing unit 14b is determined.

航跡選択出力部１５は、検定値ε１を算出すると、下記の式（９）に示すように、その検定値ε１を所定の閾値ε_thと比較し、その検定値ε１が閾値ε_th以上であれば、第１高精度センサ航跡と第２高精度センサ航跡の同一性を認定しない。
一方、その検定値ε１が閾値ε_th未満であれば、第１高精度センサ航跡と第２高精度センサ航跡の同一性を認定する。

That is, the wake selection output unit 15 includes the estimated value x _{t of} the first high accuracy sensor track, the _SNS1 hat, the estimated error covariance matrix P _{t, SNS1} of the first high accuracy sensor track, and the estimated value of the second high accuracy sensor track. The test value ε1 is calculated by substituting the estimated error covariance matrix P _{t, SNS2} of the x _{t, SNS2} hat and the second high-precision sensor track into the following equation (8) by chi-square test.

Track selection output unit 15, calculating the test values .epsilon.1, there in as shown in the following equation (9), comparing the test value .epsilon.1 with a predetermined threshold epsilon _th, the test value .epsilon.1 threshold epsilon _th or For example, the identity of the first high-precision sensor track and the second high-precision sensor track is not recognized.
On the other hand, the test value ε1 is less than the threshold epsilon _th, certified first and precision sensor track the identity of the second high-precision sensor track.

Track selection output unit 15, if the test value ε1 is smaller than the threshold epsilon _th (step ST15), since the identity of the first high-precision sensor track and the second high-precision sensor track is recognized by the tracking processing unit 14a The calculated first high-accuracy sensor track is selected (step ST16).
On the other hand, if the test value ε1 is equal to or greater than the threshold value ε _th (step ST15), since the identity of the first high-precision sensor track and the second high-precision sensor track is not recognized, the first calculated by the tracking processing unit 14b. 2. Select the high-accuracy sensor track (step ST17).
This is because the second high-precision sensor data has a feature that the observation accuracy is good and the number of unnecessary signals detected is small, and the abnormal value is removed by the gate inside / outside determination processing of the tracking processing unit 14b. This is because when the accuracy sensor data is input, the second high-accuracy sensor track can be regarded as correct as shown in FIG.
When selecting the first high-accuracy sensor track or the second high-accuracy sensor track, the wake selection output unit 15 initializes the track quality value TQ that is an index of the continuity of the integrated wake to “0” (step ST18).

  The wake selection output unit 15 indicates that the correlation determination result of the tracking processing unit 14b does not correlate (step ST14), and uses the observation result of the high-precision sensor 12c having the third highest observation accuracy. The correlation determination result (determination result indicating whether or not the third high-accuracy sensor track and the third high-accuracy sensor data are correlated) of the tracking processing unit 14c that calculates the wake of the vehicle is correlated If so (step ST19), the identity of the first high accuracy sensor track calculated by the tracking processing unit 14a and the third high accuracy sensor track calculated by the tracking processing unit 14c is determined.

航跡選択出力部１５は、検定値ε２を算出すると、下記の式（１１）に示すように、その検定値ε２を所定の閾値ε_thと比較し、その検定値ε２が閾値ε_th以上であれば、第１高精度センサ航跡と第３高精度センサ航跡の同一性を認定しない。
一方、その検定値ε２が閾値ε_th未満であれば、第１高精度センサ航跡と第３高精度センサ航跡の同一性を認定する。

That is, the wake selection output unit 15 includes the estimated value x _{t of} the first high accuracy sensor track, the _SNS1 hat, the estimated error covariance matrix P _{t, SNS1 of} the first high accuracy sensor track, and the estimated value of the third high accuracy sensor track. The test value ε2 is calculated by substituting the estimated error covariance matrix P _{t, SNS3} of the x _{t, SNS3} hat and the third high accuracy sensor track into the following equation (10) by chi-square test.

Track selection output unit 15, calculating the test values .epsilon.2, there in as shown in the following formula (11), comparing the test value .epsilon.2 with a predetermined threshold epsilon _th, the test value .epsilon.2 threshold epsilon _th or For example, the identity of the first high-precision sensor track and the third high-precision sensor track is not recognized.
On the other hand, the test value ε2 is less than the threshold epsilon _th, certified first and precision sensor track the identity of the third high-precision sensor track.

Track selection output unit 15, if the test value ε2 is smaller than the threshold epsilon _th (step ST20), since the identity of the first high-precision sensor track and the third high-precision sensor track is recognized by the tracking processing unit 14a The calculated first high-accuracy sensor track is selected (step ST21).
On the other hand, if the test value ε2 is equal to or greater than the threshold value ε _th (step ST20), since the identity of the first high-precision sensor track and the third high-precision sensor track is not recognized, the first calculated by the tracking processing unit 14b. The identity of the second high-accuracy sensor track and the third high-accuracy sensor track calculated by the tracking processing unit 14c is determined.

航跡選択出力部１５は、検定値ε３を算出すると、下記の式（１３）に示すように、その検定値ε３を所定の閾値ε_thと比較し、その検定値ε３が閾値ε_th以上であれば、第２高精度センサ航跡と第３高精度センサ航跡の同一性を認定しない。
一方、その検定値ε３が閾値ε_th未満であれば、第２高精度センサ航跡と第３高精度センサ航跡の同一性を認定する。

That is, the wake selection output unit 15 includes the estimated value x _{t of} the second high accuracy sensor track, the _SNS2 hat, the estimated error covariance matrix P _{t, SNS2 of} the second high accuracy sensor track, and the estimated value of the third high accuracy sensor track. The test value ε3 is calculated by substituting the estimated error covariance matrix P _{t, SNS3} of the x _{t, SNS3} hat and the third high-accuracy sensor track into equation (12) by the following chi-square test.

When calculating the test value ε3, the wake selection output unit 15 compares the test value ε3 with a predetermined threshold value ε _th as shown in the following equation (13), and if the test value ε3 is equal to or greater than the threshold value ε _th. For example, the identity of the second high accuracy sensor track and the third high accuracy sensor track is not recognized.
On the other hand, the test value ε3 is less than the threshold epsilon _th, certified a second precision sensor track the identity of the third high-precision sensor track.

If the test value ε3 is less than the threshold value ε _th (step ST22), the wake selection output unit 15 recognizes the identity of the second high-precision sensor track and the third high-precision sensor track, and therefore the tracking processing unit 14b The calculated second high accuracy sensor track is selected (step ST23).
On the other hand, if the test value ε3 is equal to or greater than the threshold value ε _th (step ST22), the second high-precision sensor track and the third high-precision sensor track are not identical, so the first calculated by the tracking processing unit 14c. 3. Select a high-accuracy sensor track (step ST24).
This is because the third high-accuracy sensor data has a feature that the observation accuracy is good and the number of unnecessary signals detected is small, and the abnormal value is removed by the gate inside / outside determination processing of the tracking processing unit 14c. This is because the third high-accuracy sensor track can be regarded as correct when the accuracy sensor data is input.
When the first high-precision sensor track, the second high-precision sensor track, or the third high-precision sensor track is selected, the wake selection output unit 15 initially sets the track quality value TQ that is an index of continuity of the integrated wake to “0”. (Step ST25).

When the track selection output unit 15 indicates that the correlation determination result of the tracking processing unit 14b is not correlated (step ST19), the correlation determination result of the tracking processing unit 13 (correlation between the integrated track and the low-accuracy sensor data is correlated) If the determination result indicating whether or not the recording is correlated (step ST26), the track quality value TQ is incremented by one (step ST27).
On the other hand, when the correlation determination result of the tracking processing unit 13 indicates that there is no correlation N times (step ST28), the track quality value TQ is counted down by one (step ST29).
Here, N is a parameter, and when the observation cycle between a plurality of sensors is different and the observation cycle of the low-precision sensor 11 is longer than the observation cycle of the high-precision sensors 12a, 12b, 12c, the high-precision sensors 12a, 12b, In order to prevent the track quality value TQ from falling at every observation period of 12c, it is confirmed that there is no correlation N times continuously.
For example, when the observation period of the high-precision sensors 12a, 12b, and 12c is n seconds and the observation period of the low-precision sensor 11 is m seconds, N = m / n (however, an integer) may be set.

航跡選択出力部１５は、第１〜第３高精度センサ航跡の中で、トレースＴＲＡＣＥ_iが最小の高精度センサ航跡を「選抜高精度センサ航跡」に決定する。 Next, the track selection output unit 15 compares the accuracy of the first to third high-precision sensor tracks, and determines the high-precision sensor track with the highest accuracy as the “selected high-precision sensor track” (step ST30).
That is, the wake selection output unit 15 calculates the trace TRACE _i that is the sum of the diagonal terms of the estimated error covariance matrix as shown in the following equation (14) for each of the first to third high-precision sensor wakes. To do. i = 1,2,3.

The wake selection output unit 15 determines the high-precision sensor track having the smallest trace TRACE _i as the “selected high-precision sensor wake” among the first to third high-precision sensor tracks.

Next, when the track quality value TQ is less than the predetermined threshold Th _TQ (step ST31), the wake selection output unit 15 selects the selected high-precision sensor wake (step ST32).
When the track quality value TQ is less than the threshold Th _TQ , as shown in FIG. 15, the low-accuracy sensor data may be an unnecessary signal, and the integrated track has an actual true trajectory unless continuity is ensured. Since there is still a possibility of detachment, the selected high-precision sensor track is selected.
On the other hand, when the track quality value TQ is equal to or greater than the predetermined threshold Th _TQ (step ST31), the identity between the integrated track calculated by the tracking processing unit 13 and the selected high-precision sensor track is determined.

航跡選択出力部１５は、検定値ε４を算出すると、下記の式（１７）に示すように、その検定値ε４を所定の閾値ε_thと比較し、その検定値ε４が閾値ε_th以上であれば、統合航跡と選抜高精度センサ航跡の同一性を認定しない。
一方、その検定値ε４が閾値ε_th未満であれば、統合航跡と選抜高精度センサ航跡の同一性を認定する。

That is, the wake selection output unit 15 includes the integrated wake estimated value x _{t, FUSE} hat, the integrated wake estimated error covariance matrix P _{t, FUSE} , the selected high accuracy sensor wake estimated value x _{t, SNSOUT} hat, the selected high accuracy. The test value ε4 is calculated by substituting the estimated error covariance matrix P _{t, SNSOUT} of the sensor track into the following equation (16) by chi-square test.

When calculating the test value ε4, the wake selection output unit 15 compares the test value ε4 with a predetermined threshold value ε _th as shown in the following equation (17), and if the test value ε4 is equal to or greater than the threshold value ε _th. For example, the identity of the integrated track and the selected high-precision sensor track is not recognized.
On the other hand, the test value ε4 is less than the threshold epsilon _th, certifies the identity of the selected high-precision sensor track and integration track.

If the test value ε4 is less than the threshold value ε _th (step ST33), the wake selection output unit 15 selects the selected high accuracy sensor track because the identity of the integrated track and the selected high accuracy sensor track is recognized (step ST33). ST32).
On the other hand, if the test value ε4 is equal to or greater than the threshold value ε _th (step ST33), since the identity of the integrated track and the selected high-precision sensor track is not recognized, the integrated track calculated by the tracking processing unit 13 is selected ( Step ST34).
When the identity of the integrated track and the selected high-precision sensor track is not recognized, as shown in FIG. 16, it is considered that the integrated track in which continuity is ensured is correct.

  As apparent from the above, according to the second embodiment, the track selection output unit 15 is integrated by the tracking processing unit 13 based on the correlation determination results of the tracking processing units 13, 14a, 14b, and 14c. Since one track is selected from the high accuracy sensor tracks calculated by the track and tracking processing units 14a, 14b, and 14c, the low accuracy sensor 11 and the high accuracy sensors 12a, 12b, 12c is used to produce an effect that the target position can be estimated with high accuracy.

  In the second embodiment, an example of application to an air traffic control support system has been shown. However, any device that estimates a target position using a plurality of sensors having different observation accuracy may be used. Needless to say, it may be applied to other systems.

Embodiment 3 FIG.
FIG. 17 is a block diagram showing a target tracking device according to Embodiment 3 of the present invention. In the figure, the same reference numerals as those in FIG.
The observation accuracy calculation unit 21 is composed of, for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like, and determines the positional relationship of the high-precision sensors 12a, 12b, 12c with respect to the target, the reliability of sensor data, and the like. Considering this, a process for calculating the observation accuracy of the high-precision sensors 12a, 12b, and 12c, which changes every moment, is performed. The observation accuracy calculation unit 21 constitutes observation accuracy calculation means.

The wake selection output unit 22 is composed of, for example, a semiconductor integrated circuit mounted with a CPU or a one-chip microcomputer, and selects a target wake by the same method as the wake selection output unit 15 of FIG. Before selecting a target track, a process for recognizing the level of observation accuracy of the high-precision sensors 12a, 12b, and 12c is performed with reference to the calculation result of the observation accuracy calculation unit 21.
That is, in the wake selection output unit 15 of FIG. 10, as shown in FIG. 12, the high accuracy sensor 12a always has the highest observation accuracy, the high accuracy sensor 12b has the highest observation accuracy, and then the high accuracy sensor 12c. However, the observation accuracy of the high-precision sensors 12a, 12b, and 12c is not determined only by the accuracy of the sensor itself, but if the positional relationship between the target and the sensor changes, the observation accuracy Therefore, the difference is that the level of observation accuracy of the high-precision sensors 12a, 12b, and 12c is recognized with reference to the calculation result of the observation accuracy calculation unit 21.
Therefore, depending on conditions, it is conceivable that the observation accuracy of the high accuracy sensor 12c is the highest and the observation accuracy of the high accuracy sensor 12a is the lowest.
The wake selection output unit 22 constitutes wake selection means.

In the example of FIG. 17, the low-precision sensor 11, high-precision sensors 12a, 12b, and 12c, the tracking processing unit 13, the tracking processing units 14a, 14b, and 14c, the observation accuracy calculating unit 21, and the wake selection that are components of the target tracking device. Although each of the output units 22 is assumed to be configured with dedicated hardware, all or part of the target tracking device may be configured with a computer.
For example, when a part of the target tracking device (for example, a part excluding the low accuracy sensor 11 and the high accuracy sensors 12a, 12b, and 12c) is configured by a computer, the tracking processing units 13, 14a, 14b, and 14c, the observation accuracy What is necessary is just to store the program describing the processing content of the calculation part 21 and the track selection output part 22 in the memory of a computer, and to run the program stored in the memory of the said computer.

Next, the operation will be described.
Compared to the second embodiment, since the observation accuracy calculation unit 21 is implemented and the track selection output unit 22 is provided instead of the track selection output unit 15, the observation accuracy is calculated here. Only the processing contents of the calculation unit 21 and the wake selection output unit 22 will be described.

ただし、Ｒ_TSはレーダと目標間の距離、σ_AZはレーダの方位角観測精度である。
例えば、目標に対する視線方向の観測精度と、視線方向に対して直交する方向の観測精度を比較して、最悪値を高精度センサ１２ａ，１２ｂ，１２ｃの観測精度σ_RADARとすればよい。 The observation accuracy calculation unit 21 calculates the observation accuracy of the high accuracy sensors 12a, 12b, and 12c that change from moment to moment in consideration of the positional relationship of the high accuracy sensors 12a, 12b, and 12c with respect to the target and the reliability of the sensor data. To do.
For example, when the high-precision sensors 12a, 12b, and 12c are radars, and the high-precision sensor data that is the observation values of the high-precision sensors 12a, 12b, and 12c is SSR data, the observation accuracy in the line-of-sight direction with respect to the target is the distance of the radar. This is the observation accuracy, and the observation accuracy in the direction orthogonal to the line-of-sight direction can be calculated as in the following equation (18).

However, R _TS is the distance between the radar and the target, sigma _AZ is the azimuthal angle observation accuracy of the radar.
For example, the observation accuracy in the gaze direction with respect to the target may be compared with the observation accuracy in the direction orthogonal to the gaze direction, and the worst value may be set as the observation accuracy σ _RADAR of the high-precision sensors 12a, 12b, and 12c.

For example, when the high-precision sensor data that is the observation values of the high-precision sensors 12a, 12b, and 12c is WAM, the accuracy reduction rate DOP (Dilution of Precision) calculated in advance according to the positional relationship between the target and each receiving station. ) And the distance difference observation accuracy σ _dR , the observation accuracy σ _WAM of the high-precision sensors 12a, 12b, and 12c can be calculated as in the following equation (19).

For example, when the high-precision sensor data that is the observation values of the high-precision sensors 12a, 12b, and 12c is ADS-B data, the observation accuracy of the high-precision sensors 12a, 12b, and 12c is calculated using the data performance parameter. be able to.
As the performance parameter, there are NACp (Navigation Accuracy Category for Position), NIC (Navigation Integrity Category), and the like. For example, the range of twice the error standard deviation σ _ADS-B (2σ) is known from NACp.
Note that, if the sensor data does not satisfy the criteria, the sensor data may not be used or the priority of the sensor data may be lowered using a NIC that represents data reliability.

The wake selection output unit 22 selects a target track in the same manner as the wake selection output unit 15 in FIG. 10, but refers to the calculation result of the observation accuracy calculation unit 21 before selecting the target track. The high and low accuracy of the high accuracy sensors 12a, 12b and 12c is recognized.
That is, as shown in FIG. 12, the observation accuracy of the high-precision sensors 12a, 12b, and 12c is high-precision sensor 12a> high-precision sensor 12b> high-precision sensor 12c. The observation accuracy of the high-precision sensors 12a, 12b, 12c changes from moment to moment according to the positional relationship between the 12b, 12c, etc. Therefore, referring to the calculation result of the observation accuracy calculation unit 21, the high-precision sensors 12a, 12b, Recognize the level of observation accuracy of 12c.

For example, if the calculation result of the observation accuracy calculation unit 21 indicates that the high accuracy sensor 12b> the high accuracy sensor 12c> the high accuracy sensor 12a, the high accuracy sensor track calculated by the tracking processing unit 14a is “ "3 high-precision sensor track", the high-precision sensor track calculated by the tracking processing unit 14b is "first high-precision sensor track", and the high-precision sensor track calculated by the tracking processing unit 14c is "second high-precision sensor track" To handle as.
For example, if the calculation result of the observation accuracy calculation unit 21 indicates that the high accuracy sensor 12c> the high accuracy sensor 12a> the high accuracy sensor 12b, the high accuracy sensor track calculated by the tracking processing unit 14a is used. “Second high-precision sensor track”, the high-precision sensor track calculated by the tracking processing unit 14 b is “third high-precision sensor track”, and the high-precision sensor track calculated by the tracking processing unit 14 c is “first high-precision sensor track”. Handle as “wake”.

  When the wake selection output unit 22 recognizes the level of observation accuracy of the high accuracy sensors 12a, 12b, and 12c and determines the handling of the high accuracy sensor wake calculated by the tracking processing units 14a, 14b, and 14c, the wake track will be described below. The wake selection process is the same as that of the wake selection output unit 15 in FIG.

  As apparent from the above, according to the third embodiment, the high-precision sensor 12a that changes every moment in consideration of the positional relationship of the high-precision sensors 12a, 12b, and 12c with respect to the target, the reliability of the sensor data, and the like. , 12b, 12c is provided, and the wake selection output unit 22 refers to the calculation result of the observation accuracy calculation unit 21 to determine the observation accuracy of the high accuracy sensors 12a, 12b, 12c. Since it is configured to recognize the height, even if the observation accuracy of the high-precision sensors 12a, 12b, 12c changes from moment to moment according to the positional relationship of the high-precision sensors 12a, 12b, 12c with respect to the target, etc. There is an effect that the track of the target can be selected.

  In the present invention, within the scope of the invention, any combination of the embodiments, or any modification of any component in each embodiment, or omission of any component in each embodiment is possible. .

  DESCRIPTION OF SYMBOLS 1 Low precision sensor (1st sensor), 2 High precision sensor (2nd sensor), 3 Tracking process part (1st wake calculation means) 4 Tracking process part (2nd wake calculation means), 5 wakes Selection output unit, 6 correlation determination result input unit, 7 continuity determination unit (continuity determination unit), 8 identity determination unit (identity determination unit), 9 track selection unit (track selection unit), 11 low accuracy sensor ( First sensor), 12a, 12b, 12c High-precision sensor (second sensor), 13 Tracking processing unit (first wake calculation means), 14a, 14b, 14c Tracking processing unit (second wake calculation means) , 15 Track selection output unit (track selection unit), 21 Observation accuracy calculation unit (observation accuracy calculation unit), 22 Track selection output unit (track selection unit).

Claims (13)

A first sensor for observing a target position;
A second sensor for observing the target position with higher observation accuracy than the first sensor;
Using the observation result of the first sensor and the observation result of the second sensor, the track of the target is calculated, and the track of the target calculated before the current calculation and the observation of the first sensor are calculated. First wake calculation means for determining a correlation with the result;
A second track for calculating the track of the target using the observation result of the second sensor and determining a correlation between the track of the target calculated before the current calculation and the observation result of the second sensor. Wake calculation means,
Continuity determination means for determining continuity of the target track calculated by the first track calculation means from the determination result of the first track calculation means;
Identity determining means for determining the identity of the target track calculated by the first track calculating means and the target track calculated by the second track calculating means;
Based on the determination result of the second wake calculation means, the determination result of the continuity determination means, and the determination result of the identity determination means, the target wake calculated by the first wake calculation means, or the above A target tracking device comprising: a track selecting unit that selects a track of the target calculated by the second track calculating unit.   The wake selecting means selects the target wake calculated by the second wake calculating means if the determination result of the second wake calculating means indicates that they are correlated. The target tracking device described.   When the wake selection means indicates that the determination result of the second wake calculation means is not correlated, if the determination result of the continuity determination means indicates that there is no continuity of the wake, The target tracking device according to claim 2, wherein the target track calculated by the track calculation means is selected.   If the determination result of the continuity determination means indicates that there is continuity of the wake, if the determination result of the identity determination means indicates that there is identity, the wake selection means calculates the second wake If the target track calculated by the means is selected, and the determination result of the identity determination unit indicates that there is no identity, the target track calculated by the first track calculation unit is selected. The target tracking device according to claim 3, wherein   The identity determination means performs a chi-square test to determine the identity of the target track calculated by the first track calculation means and the target track calculated by the second track calculation means. The target tracking device according to any one of claims 1 to 4, wherein the target tracking device is characterized.   The first and second wake calculation means use the sensor observation results, and when the time when the target wake is calculated exceeds a certain time from the sensor observation time, the sensor observation time 6. The track of the target is calculated using a prediction error covariance matrix included in the predicted value of the target after a predetermined time from 6. Target tracking device. A first sensor for observing a target position;
A plurality of second sensors having observation accuracy higher than that of the first sensor and observing the target position with different observation accuracy;
The wake of the target is calculated using the observation results of the first sensor and the observation results of the plurality of second sensors, and the target wake calculated before the current calculation and the first sensor are calculated. First wake calculation means for determining a correlation with the observation results of
Using the observation result of the second sensor, a track of the target is calculated, and a plurality of determinations are made to determine a correlation between the target track calculated before the current calculation and the observation result of the second sensor. A second wake calculation means;
Based on the determination result of the first wake calculation means and the determination result of the plurality of second wake calculation means, the target wake calculated by the first wake calculation means and the plurality of second wake calculation A track tracking means comprising track selection means for selecting one track from the track of the target calculated by the means.   The wake selection means uses the observation result of the second sensor having the highest observation accuracy among the plurality of second wake calculation means, and the determination result of the second wake calculation means that calculates the target wake. 8. The target tracking device according to claim 7, wherein the target track calculated by the second track calculation means is selected if the two indicate that they are correlated.   The wake selection means is calculated by the second wake calculation means when there is a second wake calculation means indicating that the determination result is not correlated among the plurality of second wake calculation means. Among the second wake calculation means indicating that the target wake and the determination result are correlated, the second wake calculation means using the observation result of the second sensor with the highest observation accuracy. The target track and the determination result calculated by the second track calculation means indicating that the determination result is not correlated based on the determination result based on the determination result. Among the second wake calculation means indicating that the two are correlated, the target wake calculated by the second wake calculation means using the observation result of the second sensor having the highest observation accuracy is used. A contract characterized by selecting one wake from among Target tracking device of claim 8, wherein.   The wake selection means, when the determination results of all the second wake calculation means indicate that there is no correlation, the target wake calculated when the target wake is calculated by the second wake calculation means. The sum of the diagonal terms of the estimation error covariance matrix included in the estimated value is calculated, and the second wake calculation means having the smallest sum of the diagonal terms among the plurality of second wake calculation means. And the identity of the track of the target calculated by the second track calculation means and the track of the target calculated by the first track calculation means is determined. 10. The target track calculated by the second track calculation unit is selected, and if the identity is not recognized, the target track calculated by the first track calculation unit is selected. Target tracking device. In consideration of the positional relationship of each second sensor with respect to the target, an observation accuracy calculation means for calculating the observation accuracy of each second sensor is provided,
11. The wake selection unit recognizes the level of observation accuracy between the plurality of second sensors with reference to the calculation result of the observation accuracy calculation unit. Item 1. The target tracking device according to item 1. The first wake calculation means uses the observation result of the first sensor that observes the target position, and the observation result of the second sensor that observes the target position with higher observation accuracy than the first sensor. A first wake calculation processing step for calculating a wake of the target and determining a correlation between the wake of the target calculated before the current calculation and the observation result of the first sensor;
The second wake calculation means calculates the target wake using the observation result of the second sensor, and calculates the target wake calculated before the current calculation and the observation result of the second sensor. A second wake calculation processing step for determining a correlation with
A continuity determining means for determining the continuity of the target track calculated in the first track calculation processing step from the determination result in the first track calculation processing step;
An identity determination means for determining the identity between the target track calculated in the first track calculation processing step and the target track calculated in the second track calculation processing step; ,
The wake selection unit is configured to perform the first wake calculation process based on the determination result in the second wake calculation process step, the determination result in the continuity determination process step, and the determination result in the identity determination process step. A target tracking method comprising: a target track calculated in a step or a track selection processing step for selecting a target track calculated in the second track calculation processing step. The first wake calculation means has an observation result of the first sensor observing the target position and an observation accuracy higher than the observation accuracy of the first sensor, and the target position with different observation accuracy. The target track is calculated using the observation results of the plurality of second sensors observing the image, and the correlation between the target track calculated before the current calculation and the observation result of the first sensor is calculated. A first wake calculation processing step for determining
A plurality of second wake calculation means calculate the target wake using the observation result of the second sensor, and calculate the target wake calculated before the current calculation and the second sensor. A second wake calculation processing step for determining a correlation with the observation result;
The wake selection means, based on the determination result in the first wake calculation processing step and the determination result in the second wake calculation processing step, the target wake calculated in the first wake calculation processing step and A target tracking method comprising: a track selection processing step of selecting one track from the track of the target calculated in the second track calculation processing step.

Images