JP2000221265A - Target correlation integrating device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve correlation integration performance and at the same time reduce the processing load of a device by providing a correlation gate threshold controller or the like for calculating the threshold of a correlation gate. SOLUTION: A target observation device position element-inputting device 3 inputs the position relationship between first and second target observation devices 1 and 2, and target observation information input equipment 4 calculates the smoothing value or the like of target position and speed based on target observation information being obtained successively from each of the target observation devices 1 and 2. A target observation device noise level controller 15 calculates a threshold for detecting a target from the noise level of the target observation devices 1 and 2. A correlation gate threshold controller 16 calculates a correlation gate threshold for determining the spread of a correlation gate according to the noise level. Correlation gate volume calculation equipment 17 calculates the volume of the correlation gate from a prediction error covariance, an observation error covariance, and a correlation gate threshold. New target observation frequency calculation equipment 8 calculates a new target observation frequency.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving object such as an airplane, a target observing device for detecting and analyzing the target by radio waves, infrared rays, and the like. The present invention relates to a target correlation integrating device for estimating motion parameters (position, speed, etc.) of the target.

[0002]

2. Description of the Related Art A target correlation integrating device is a device in which a target is observed by a plurality of target observation devices, and whether or not target observation information observed by each target observation device is the same target is determined. This is a device that integrates observation information and estimates the target movement parameters (position, speed, etc.). FIG. 7 shows a situation in which target correlation integration is performed. A plurality of targets are observed by the first target observation device and the second target observation device, and the target observation information is input to the target correlation integration device. Is shown.

FIG. 14 shows an example of a configuration diagram of a conventional target correlation integrating device, and FIG. 15 shows a processing procedure of the conventional target correlation integrating device. 1 is a first target observation device that outputs the position and the like of a target and an unnecessary signal as target observation information, 2 is a second target observation device arranged at a position different from that of the first target observation device 1, and 3 is A target observation device position specification input device 4 for inputting a positional relationship between the first target observation device 1 and the second target observation device 2, and target observation information from the first target observation device 1 and a second target Observation device 2
For converting the target observation information from the target observation information into a common coordinate system for performing correlation integration based on the position data of each target observation device 2
The target observation information input devices 5 corresponding to 0, 21 and 22 determine whether or not the target observation information from each of the target observation devices 1 and 2 is to be correlated with the already managed target already managed.
The motion specification correlator 6 corresponding to the processing 24 for extracting the target information to be correlated has a “managed target”, an “unnecessary signal”, and a “new target” for the target observation information extracted as the correlation target. A hypothesis generator corresponding to a process 25 for adding three discriminant elements and generating a hypothesis, and a hypothesis generator 7 corresponding to a process 26 for inputting a hypothesis reliability calculation parameter used for calculating the reliability of the hypothesis. The calculation parameter input unit 8 is a hypothesis reliability calculator corresponding to a process 27 for calculating the reliability of the hypothesis based on the hypothesis reliability calculation parameter, the target observation information, and the information of the managed target, and 9 is the hypothesis reliability calculator. The hypothesis reduction parameter input unit 10 corresponding to the processing 28 for inputting a hypothesis reduction parameter that gives a condition for deleting a low-reliability one, sets the accumulated hypothesis with the hypothesis reduction parameter. Hypothesis reducer corresponding to the processing 29 for deleting unreliable hypothesis according the reduction conditions, 1
Reference numeral 1 denotes a target information integration determiner corresponding to a process 30 for determining whether or not to integrate target observation information associated with a managed target based on a reduced hypothesis. Reference numeral 12 denotes a managed target determined to be integrated. And a target motion specification smoother 13 corresponding to a process 31 for calculating a smoothed value of a target position / velocity and a smoothed error covariance matrix of the target observation information based on the Kalman filter theory, and 13 is output from the target observation devices 1 and 2. This is a target motion specification predictor corresponding to the processing 23 for calculating the predicted value of the target position / velocity of the managed target at the latest time of the target observation information and the prediction error covariance matrix.

When the hypothesis reliability calculator 8 calculates the reliability of the hypothesis that the input target observation information is an unnecessary signal and the reliability of the hypothesis that the input target observation information is a new target, the target within the correlation gate is calculated. An unnecessary signal occurrence frequency for estimating a ratio of what target is an unnecessary signal and what target is a new target among the number of observation information is used. The unnecessary signal occurrence frequency and the new target occurrence frequency are input to the hypothesis reliability calculation parameter input device 7 as fixed values.

The hypothesis reducer 10 is for suppressing an increase in the number of hypotheses generated each time target observation information is input in time series from each target observation device. The mainstream method is to delete a hypothesis with low reliability or leave only a high hypothesis using a hypothesis reliability. As an example, some reduction methods are explained. Set the total number of hypotheses allowed in all hypotheses and set a hypothesis number limit method that leaves only hypotheses with high hypothesis reliability within that range, or set a threshold value for hypothesis reliability and compare with that threshold to trust Set the hypothesis reliability limit method that deletes only the hypotheses that have a high degree of confidence and remove the other hypotheses, and set the cumulative threshold of the hypothesis reliability, and set the total reliability of the hypotheses in the order of high or low reliability. There is a hypothesis reliability cumulative limit method in which a hypothesis exceeding the threshold value is deleted or left by comparing with the calculated cumulative threshold value.

FIGS. 11A and 11B are diagrams showing a process of generating a correlation gate generated in the process 24 and an example of determining whether or not a target exists. First, the position where the managed target exists is predicted from the past position and speed of the target, and furthermore, the error of the target position at the current time is calculated in consideration of the error of the past target position / speed and the increase of the error due to the transition of time. Estimate as spatial range. This is the prediction error covariance shown in (a) in the figure, which is the predicted existence range of the managed target. On the contrary,
Since the target information observed by the target observation device also has an observation error, the existence range is also estimated in the observation target. This is the observation error covariance in (b) in the figure, which is the predicted existence range of the observation target. As described above, the range in which the managed target and the observation target exist at the same time is the sum of the range in which the managed target exists and the range in which the observation target exists, and this is the correlation gate in FIG.

FIG. 15 shows a processing procedure of a conventional target correlation integrating apparatus. In a conventional correlation integrating method using target observation information, target information of the first observation from any target observation apparatus is processed in processing 20. In step 21, based on the observation position of the observation target, the initial value of the smoothed value indicating the target position / velocity and the initial value of the smoothed error covariance matrix indicating the spatial error evaluation amount of the target position / velocity are calculated based on the Kalman filter theory. Then, in process 22, motion parameters such as a target observation position from the target observation device and observation parameters such as observation accuracy are input as target observation information, and in process 23 the latest managed target at the latest time is already managed. A predicted error covariance matrix indicating the predicted value of the target position / velocity and the error evaluation amount of the target position / velocity is calculated, and target observation information is generated in process 24 from observation parameters such as the prediction error covariance and observation accuracy. Target scope It is determined whether or not the target observation information is present in the correlation gate indicating the target observation information, and the target observation information to be correlated with the managed target is extracted by the determination. , An unnecessary signal, or a new target hypothesis is generated, and set values such as a new target observation frequency and an unnecessary signal observation frequency used for calculating the reliability of the hypothesis generated in the process 25 in the process 26 are read, In a process 27, the reliability of the hypothesis generated in the process 25 is calculated based on the set value read in the process 26, the target observation information, and the information on the managed target. In a process 28, information on the hypothesis reduction method is read. In step 29, the hypothesis is reduced based on the hypothesis reduction information read in step 28 and the hypothesis reliability calculated in step 27. Based on the hypothesis reduced in step 29 in step 30, the managed target and 1 In step 31, the target observation information associated with the target observation information is determined. In step 31, the smoothed value indicating the target position / speed of the managed target and the target observation information associated in step 30 and the smoothed error indicating the error evaluation amount of the target position / speed are shared. The variance matrix is calculated based on the Kalman filter theory, and this series of flows is repeated until the correlation integration is completed in the process 32.

[0008]

In a target correlation integration in a system in which a plurality of target observation devices are arranged at a wide range of different positions, when each target observation device observes a dense target performing a formation flight, the target The number of targets to be observed may differ for each target observation device depending on the resolution performance of the observation device, the observation position, and the formation of the target. In the conventional target correlation integration device, a hypothesis generator 6 generates a hypothesis for target observation information input from each target observation device, and the hypothesis reliability calculator 8 inputs the hypothesis to a hypothesis reliability calculation parameter input device 7. Since the hypothesis reliability was calculated using the new target observation frequency and the unnecessary signal observation frequency that are fixed values, there is a possibility that the inaccurate hypothesis reliability may be calculated for the input of the above-mentioned different number of observation targets. Higher, hypothesis reducer 1
In order to reduce the hypothesis based on the hypothesis reliability at 0, the hypothesis reduction performance that affects the target correlation integration performance is reduced, and as a result, the target correlation integration performance is reduced. Furthermore, in each target observation device, generally, clutter and the like are suppressed by signal processing in order to minimize generation of unnecessary signals in observation of the target.However, a target flying in a low altitude where noise level fluctuation is large, For targets with small effective reflection areas of radio waves, such as stealth aircraft,
If the noise level rises equivalently, the threshold for target detection will increase, and the target cannot be detected, or the target will not approach unless the target is close to the target observation device and the reception level is higher than the noise level. Detection was difficult. As a result, the input of the target observation information to the target correlation integration device becomes unstable, which is a factor of degrading the target correlation integration performance. Also, as the noise level increases, a large number of unnecessary signals are mixed in the correlation gate generated by the motion specification correlator 5, and the number of hypotheses generated by the hypothesis generator 6 is increased. This is also a factor that increases the processing load of the device. FIG.
Fig. 7 shows the relationship between the false alarm probability and the detection probability when the noise level increases. As shown in the figure, if the threshold is determined so that the false alarm probability becomes constant in accordance with the increase in the noise level, the detection probability decreases, and it becomes difficult to detect the target.

SUMMARY OF THE INVENTION The present invention has been made to solve such a problem, and is based on a hypothesis generated for target observation information of different observation target numbers input from target observation devices arranged at a plurality of different positions. And calculate the hypothesis reliability corresponding to changes in the situation, monitor the noise level of the target observation device based on the target observation information input from the target observation device and the correlation integration result, and set the threshold for target detection. The present invention relates to a target correlation integration device capable of controlling the spread of correlation gates for controlling and further reducing the processing load of the device in accordance with the noise level after threshold control.

[0010]

According to a first aspect of the present invention, there is provided a target correlation integrating apparatus comprising: a correlation gate volume calculator for calculating a volume of a correlation gate based on correlation gate specifications from a motion specification correlator; A new target observation frequency calculator for calculating a new target observation frequency from the managed target number, the number of observation targets in the correlation gate, and the correlation gate volume before one sampling from the source correlator; and the new target observation frequency and hypothesis reliability. The hypothesis reliability calculator that calculates the reliability of the hypothesis from the unnecessary signal observation frequency from the calculation parameters and the target observation information, and manages and displays the target information from the target observation information input device and the target motion specification smoother,
A target information management / display that can check the smoothed value and the noise level of the target observation device, a target observation device noise level controller that controls the threshold for target detection according to the noise level of the target observation device, and a noise level Correlation gate threshold controller that controls the spread of correlation gate according to the target observation device with different resolution performance and new target observation to calculate accurate hypothesis reliability corresponding to change in target formation shape The frequency is used to determine the number of target observation information per unit volume from the number of target observation information present in the correlation gate of the target observation information observed by the target observation device, and target observation per unit volume between multiple target observation devices Variations in the difference in the number of information are obtained by unifying based on the least squares method, and changes in the observation state of the target observation device in an unnecessary signal environment are monitored. By monitoring the noise level of the target and automatically controlling the threshold for target detection, the signal from the target is detected together with unnecessary signals such as clutter, and the correlation gate is controlled according to the noise level after the threshold control. By controlling the threshold value for determining the spread, the number of unnecessary signals generated in the correlation gate is reduced when the noise level increases, thereby suppressing the increase in the number of hypotheses generated and reducing the processing load on the apparatus. .

According to a second aspect of the present invention, there is provided a target correlation integrating apparatus comprising: a correlation gate volume calculator for calculating a volume of a correlation gate based on correlation gate specifications from a motion specification correlator; An unnecessary signal observation frequency calculator that calculates an unnecessary signal observation frequency from the managed target number before sampling, the observation target number in the correlation gate, and the correlation gate volume, and an unnecessary signal observation frequency and a hypothesis reliability calculation parameter input unit. Hypothesis reliability calculator that calculates hypothesis reliability from new target observation frequency and target observation information, manages and displays target information from target observation information input unit and target motion specification smoother, and calculates smoothed values and target observations Target information management that can check the noise level of the equipment
Equipped with an indicator and a target observation device noise level controller that controls the threshold for target detection according to the noise level of the target observation device, and responds to changes in the observation state of the target observation device in an unnecessary signal environment. Unnecessary signal observation frequency for calculating accurate hypothesis reliability is calculated from the number of target observation information present in the correlation gate among the target observation information observed by the target observation device, and the number of target observation information per unit volume is calculated. Further, the variation in the number of target observation information per unit volume for each sampling of one target observation device is obtained by unifying based on the least squares method, and the noise caused by the change in the observation state of the target observation device under an unnecessary signal environment is obtained. By monitoring the level and automatically controlling the threshold for target detection, By detecting the signal and controlling the threshold value that determines the spread of the correlation gate according to the noise level after threshold control, the number of unnecessary signals in the correlation gate when the noise level increases is reduced. Thus, the number of generated hypotheses is suppressed from increasing, and the processing load on the device is reduced.

A target correlation integrating device according to a third aspect of the present invention is a correlation gate volume calculator for calculating the volume of the correlation gate based on the correlation gate data from the motion data correlator, and the one from the motion data correlator. A new target observation frequency calculator for calculating a new target observation frequency from the number of managed targets before sampling, the number of observation targets in the correlation gate, and the volume of the correlation gate, and the number of managed targets one sample before from the motion specification correlator An unnecessary signal observation frequency calculator that calculates an unnecessary signal observation frequency from the number of observation targets in the correlation gate and the correlation gate volume; and calculates the reliability of a hypothesis from the target observation information based on the new target observation frequency and the unnecessary signal observation frequency. Manages target information from the hypothesis reliability calculator, target observation information input device, and target motion specification smoother.
Equipped with a target information management / display that can display and check the smoothed value and the noise level of the target observation device, and a target observation device noise level controller that controls the threshold for target detection according to the noise level of the target observation device Changes in target observation device and target formation shape with different resolution performance,
The new target observation frequency and unnecessary signal observation frequency for calculating accurate hypothesis reliability corresponding to the change of the observation information of the target observation device in the unnecessary signal environment are correlated with the target observation information observed by the target observation device. Based on the least squares method, the difference in the difference in the number of target observation information per unit volume from the number of target observation information existing in the In addition to monitoring the noise level for changes in the observation information of the target observation device and automatically controlling the threshold for target detection, it detects signals from the target together with unnecessary signals such as clutter. Further, by controlling the threshold value for determining the spread of the correlation gate according to the noise level after the threshold control, Zureberu will reduce unnecessary signal number in the correlation gate in the case of increasing, the increase in the number of hypotheses to be generated is suppressed to reduce the processing load of the apparatus.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIG. 7 is a diagram showing a premise of the first embodiment of the present invention. In FIG. 7, reference numeral 1 denotes a first target observation device, 2 denotes a second target observation device, 39 denotes an observation target, and 40 denotes a target correlation. An integrated device,
This shows a situation where two target observation devices observe three targets.

FIG. 8 shows a situation in which the two target observation devices observe targets in time series in the situation shown in FIG. 7 and output the results to the target correlation integration device.

FIG. 9 is a diagram showing a situation where each target observation device actually observes a target in the situations shown in FIGS. 7 and 8. In the figure, it is assumed that the first target observation device can observe two targets. It is also assumed that the second target observation device has observed three targets. When the number of targets already managed is 2, one of the three targets observed by the second target observation device is hypothesized as a new target. In order to determine the reliability of the hypothesis, the frequency of observation of a new target in space is required. The observation frequency is defined as the target number of observations per unit volume. Therefore, the volume of the correlation gate indicating the expected existence range of the target used in determining the correlation is calculated, and the new target observation frequency is calculated from the number of targets assumed to be new targets in the correlation gate. In addition, when the number of targets observed between two target observation devices fluctuates in time series, the estimation of the new target number is
First, the target number existing in the space is estimated by the least squares method, and the target number is obtained from the difference between the target number already managed and the target number existing in the estimated space. As described above, the new target observation frequency can be calculated flexibly even when there is a difference in the number of observation targets among a plurality of target observation devices or when the number of observation targets varies in a time series.

FIG. 4 is a diagram showing a processing procedure according to the first embodiment of the present invention. In FIG. 4, reference numeral 22 denotes input of target observation information from a target observation device, reference numeral 23 denotes a prediction value and prediction error covariance matrix calculation, 24 is a correlation determination between the observation target and the managed target, 25 is a hypothesis generation of the target observation information, 26 is a hypothesis reliability calculation parameter read, 27 is a hypothesis reliability calculation of the target observation information, 28 is a hypothesis reduction parameter read, 29 is Hypothesis reduction, 30: integrated judgment of observation target and managed target, 3
1 is a smoothed value and smoothed error covariance matrix calculation, 32 is correlation integration end determination, 33 is target information management / display, 34 is a noise level calculation of the target observation device, 35 is a correlation gate threshold value calculation, 36 is a correlation gate Volume calculation 37 is a new target observation frequency calculation. In process 20, target observation information of the new target obtained from the first target observation device at the first sampling and 2
At the sampling, the target observation information of the new target obtained from the second target observation device is input, and the target position / velocity is obtained based on the target observation information (the situation shown in FIG. 8) obtained in order from each target observation device in process 21. The initial value of the smoothed error covariance indicating the spatial error evaluation amount of the target position / velocity and the smoothed value of the target position / velocity is calculated. , And processing 23
Calculates the predicted value of the target position / velocity at the latest time of the managed target calculated at the second sampling and the prediction error covariance indicating the error evaluation amount of the target position / speed. The correlation between the target and the target is calculated using the prediction error covariance centered on the predicted position of the target and the observation error covariance that indicates the error of the target observation information and the correlation gate threshold that determines the spread of the correlation gate. It is determined whether or not the correlation gate exists in the correlation gate indicating the predicted existence range.
In step 6, the volume of the correlation gate is calculated from the prediction error covariance, the observation error covariance, and the correlation gate threshold. In step 37, a new value indicating the frequency at which the new target is included in a plurality of target observation information existing in the correlation gate. The target observation frequency is calculated using the target observation information number in the correlation gate and the correlation gate volume, and the processing 25
A hypothesis is generated as to whether the target observation information in the correlation gate is a managed target, a new target, or an unnecessary signal, and is a hypothesis reliability calculation parameter for calculating the reliability of the hypothesis generated in the process 26. The signal observation frequency is read, the hypothesis reliability is calculated from the target observation information and the content of the hypothesis, the new target observation frequency, and the unnecessary signal observation frequency in a process 27, and a hypothesis reduction parameter for reducing a plurality of hypotheses generated in the process 28. Is read, the hypothesis is reduced using the hypothesis contents, the hypothesis reliability, and the hypothesis reduction parameter in a process 29, and the target observation information associated with the managed target in a one-to-one correspondence with the hypothesis content after the reduction in a process 30 is integrated. A determination is made, and a smoothed value and a smoothed error covariance of the position / velocity of the management target at the current time are calculated based on the Kalman filter theory in a process 31.
Manages the smoothed value and smoothed error covariance of the position / velocity of the management target calculated in step 3 and the target observation information input from the target observation device, and displays the smoothed value of the position of the management target and the observation position of the target observation information. Then, a threshold for target detection is calculated from the noise level of the target observation device in a process 34 and sent to the target observation device, and a correlation gate threshold for determining the spread of the correlation gate in accordance with the noise level in a process 35. This sequence is repeated until the value is calculated and the correlation integration is completed in process 32.

FIG. 1 is a block diagram showing a first embodiment of the present invention. In FIG. 1, reference numeral 1 denotes a first target observation device for outputting the position and the like of a target and an unnecessary signal as target observation information; A second target observation device arranged at a position different from that of the first target observation device, and a target observation device position specification input device for inputting a positional relationship between the first target observation device and the second target observation device. And 4, target observation information of the new target obtained from the first target observation device at the first sampling and 2
At the sampling, the target observation information of the new target obtained from the second target observation device is input, and the smoothed value of the target position / velocity is obtained based on the target observation information (the situation shown in FIG. 8) sequentially obtained from each target observation device. And calculating the initial value of the smoothed error covariance and inputting the target observation information of the managed target again from the first target observation device at the third sampling.
A target observation information input unit corresponding to 2 and a correlation between the target observation information and the managed target, and an observation error covariance and correlation indicating an error between the prediction error covariance and the target observation information centering on the predicted position of the managed target. A motion specification correlator corresponding to the processing 24 for determining whether or not the target prediction existence range is generated in the correlation gate indicating the target prediction existence range generated by using the correlation gate threshold value for determining the spread of the gate; A hypothesis generator corresponding to a process 25 for generating a hypothesis whether the inner target observation information is a managed target, a new target, or an unnecessary signal; and 7 is a hypothesis reliability calculation parameter for calculating the reliability of the generated hypothesis. The hypothesis reliability calculation parameter input device 8 corresponding to the process 26 for reading the unnecessary signal observation frequency is a process 2 for calculating the hypothesis reliability from the target observation information, the content of the hypothesis, the new target observation frequency, and the unnecessary signal observation frequency. , 9 is a hypothesis reduction parameter input unit corresponding to a process 28 for reading a hypothesis reduction parameter for reducing a plurality of generated hypotheses, and 10 is a hypothesis content, a hypothesis reliability, and a hypothesis reduction parameter. Is a hypothesis reducer corresponding to a process 29 for reducing a hypothesis by using a target, and 11 is a target corresponding to a process 30 for performing an integrated determination of target observation information that is associated one-to-one with a managed target according to the content of the reduced hypothesis. The information integration determiner 12 is a target motion specification smoother corresponding to the processing 31 for calculating the smoothed value and the smoothed error covariance of the position / velocity of the management target at the current time based on the Kalman filter theory, and 13 is calculated at the second sampling. It corresponds to the process 23 of calculating the predicted value of the target position / velocity at the current time of the already-managed target and the prediction error covariance obtained by estimating the error of the predicted value. The target motion specification predictor 14 manages the calculated smoothed value and smoothed error covariance of the position and velocity of the management target and the target observation information input from the target observation device, and further calculates the smoothed value and the position of the management target. A target information management / display device 15 corresponding to a process 33 for displaying the observation position of the target observation information, and 15 corresponds to a process 34 for calculating a threshold for target detection from the noise level of the target observation device and sending the threshold to the target observation device. A target observation apparatus noise level controller, 16 is a correlation gate threshold controller corresponding to a process 35 for calculating a correlation gate threshold for determining the spread of the correlation gate according to the noise level, 17 is a prediction error The correlation gate volume calculator 18 corresponding to the processing 36 for calculating the volume of the correlation gate from the covariance and the observation error covariance and the threshold value of the correlation gate. New target observation frequency calculation corresponding to the process 37 of calculating a new target observation frequency indicating the frequency at which a new target is included in a plurality of target observation information existing in the target using the number of target observation information in the correlation gate and the correlation gate volume It is a vessel.

Here, the calculation of the threshold for detecting the target with respect to the noise level of the 15 target observation devices will be described. The noise level of the target observation device is determined by the false alarm probability and the detection probability. FIG. 12 shows the relationship between the false alarm probability and the detection probability when the noise level increases. The relationship between the noise level and the threshold is represented by equation (1). The noise level is calculated from a signal received by the target observation device. The coefficient K1 in the equation (1) is determined by the false alarm probability.

[0019]

(Equation 1)

The calculation of the correlation gate threshold value for the noise level of the 16 target observation devices will be described.
The correlation gate threshold is a random variable that determines the spread of the correlation gate. The spread of the correlation gate is represented by the sum of the prediction error covariance centered on the predicted position of the managed target and the observation error covariance indicating the error of the target observation information. It is defined by the chi-square distribution. Therefore, by determining the probability of the target existing in the correlation gate, the correlation gate threshold, which is a random variable of the chi-square distribution, can be determined. FIG. 13 shows the relationship between the change in the threshold value of the correlation gate, the target existence probability in the correlation gate, and the spread of the correlation gate. The target existence probability in the correlation gate is associated with the increase / decrease of the noise level, and the equation is such that the correlation gate is reduced when the noise level increases. Equation (2) shows the relationship between the target existence probability in the correlation gate and the noise level. The noise level is calculated from a signal received by the target observation device. The coefficient K2 in equation (2) is set as an arbitrary parameter.

[0021]

(Equation 2)

The calculation of the seventeen correlation gate volumes will now be described. The correlation gate volume is calculated from the correlation gate spread and the correlation gate threshold. The relationship between the correlation gate volume, the correlation gate threshold value, and the spread of the correlation gate is expressed by Expression (3).

[0023]

(Equation 3)

The calculation of the eighteen new target observation frequencies will be described. The basic principle of calculating the new target observation frequency is that when multiple target observation devices separated from each other observe the same target,
Since a difference occurs in the number of target observation information due to the target formation shape and the positional relationship and resolution of the target observation device, the difference number is used as a new target number. However, since the observation frequency of the new target is required as a parameter for calculating the hypothesis reliability, the spatial existence frequency of the new target is used as the new target observation frequency using the volume of the correlation gate where the target observation information exists. It has established. When the target observation information is alternately input to the target correlation integration device from the two target observation devices as shown in FIG. 8, the calculation formula of the new target observation frequency is expressed by Expressions (4) and (5). Is done. The principle of the new target observation frequency calculation is that the number of observation targets alternately input from two target observation devices is estimated by the least squares method, and the new target number is estimated from the difference from the managed target number. The number of new targets per unit volume (new target observation frequency) is calculated from the volume of existing correlation gates. Further, the spread of the correlation gate changes according to the correlation gate threshold value calculated by the 16 correlation gate threshold value controllers, so that the target existence probability in the correlation gate also changes. In the state where the target existence probability in the correlation gate is decreasing, the number of observation targets in the correlation gate decreases stochastically from the true value. to correct.

[0025]

(Equation 4)

Embodiment 2 FIG. FIG. 7 is a diagram showing a premise of the second embodiment of the present invention. In the figure, reference numeral 1 denotes a first target observation device, 2 denotes a second target observation device, 39 denotes an observation target, and 40 denotes a target correlation. It is an integrated device, and represents a situation where two target observation devices observe three targets.

FIG. 8 shows a situation in which the two target observation devices observe targets in time series in the situation shown in FIG. 7 and output the results to the target correlation integration device.

FIG. 10 is a diagram showing a situation where each target observation device actually observes a target in the situations shown in FIGS. 7 and 8. In the figure, it is assumed that the first target observation device is observing the target and signals other than the target. Also,
It is assumed that the second target observation device can observe three targets.
When the number of targets already managed is 3, three or more targets among the targets observed by the first target observation device are hypothesized as unnecessary signals. What is important here is that the emphasis is placed on the target observation number of the first target observation device out of the two target observation devices. That is, when the number of observation targets fluctuates in one target observation device, it is estimated that unnecessary signals are mixed with signals from the targets. Also,
In order to determine the reliability of a hypothesis, the frequency of observation of unnecessary signals in space is required. The observation frequency is defined as the target number of observations per unit volume. Therefore, the volume of the correlation gate indicating the expected existence range of the target used in determining the correlation is calculated, and the unnecessary signal observation frequency is calculated from the target number assumed to be an unnecessary signal in the correlation gate. Also, 2
When the number of targets observed between two target observation devices fluctuates in time series, the number of unnecessary signals is estimated by estimating the number of targets existing in space by the least squares method for each target observation device, and is already managed. It is determined from the difference between the target number that exists and the target number that exists in the estimated space. As described above, even when there is a difference in the number of unnecessary signal observations among a plurality of target observation devices, or when the number of unnecessary signals fluctuates in a time series, it is possible to flexibly calculate an unnecessary signal observation frequency.

FIG. 5 is a diagram showing a processing procedure according to the second embodiment of the present invention. In FIG. 5, reference numeral 22 denotes a target observation information input from the target observation device, 23 denotes a prediction value and prediction error covariance matrix calculation, 24 is a correlation determination between the observation target and the managed target, 25 is a hypothesis generation of the target observation information, 26 is a hypothesis reliability calculation parameter read, 27 is a hypothesis reliability calculation of the target observation information, 28 is a hypothesis reduction parameter read, 29 is Hypothesis reduction, 30: integrated judgment of observation target and managed target, 3
1 is a smoothed value and smoothed error covariance matrix calculation, 32 is correlation integration end determination, 33 is target information management / display, 34 is a noise level calculation of the target observation device, 35 is a correlation gate threshold value calculation, 36 is a correlation gate A volume calculation 38 is an unnecessary signal observation frequency calculation. In the processing 20, target observation information of a new target obtained from the first target observation apparatus at the first sampling and 2
At the sampling, the target observation information of the new target obtained from the second target observation device is input, and the target position / velocity is obtained based on the target observation information (the situation shown in FIG. 8) obtained in order from each target observation device in process 21. The initial value of the smoothed error covariance indicating the spatial error evaluation amount of the target position / velocity and the smoothed value of the target position / velocity is calculated. , And processing 23
Calculates the predicted value of the target position / velocity at the latest time of the managed target calculated at the second sampling and the prediction error covariance indicating the error evaluation amount of the target position / speed. The correlation between the target and the target is calculated using the prediction error covariance centered on the predicted position of the target and the observation error covariance that indicates the error of the target observation information and the correlation gate threshold that determines the spread of the correlation gate. It is determined whether or not the correlation gate exists in the correlation gate indicating the predicted existence range.
In step 6, the correlation gate volume is calculated from the prediction error covariance, the observation error covariance, and the correlation gate threshold, and in step 38, the unnecessary signal indicating the frequency at which the unnecessary signal is included in a plurality of target observation information existing in the correlation gate is unnecessary. The signal observation frequency is calculated using the number of target observation information in the correlation gate and the correlation gate volume, and processing 25 is performed.
To generate a hypothesis whether the target observation information in the correlation gate is a managed target, a new target, or an unnecessary signal, and a new hypothesis reliability calculation parameter for calculating the reliability of the hypothesis generated in the process 26. The target observation frequency is read, the hypothesis reliability is calculated from the target observation information, the hypothesis content, the new target observation frequency, and the unnecessary signal observation frequency in processing 27, and a hypothesis reduction parameter for reducing the plurality of hypotheses generated in processing 28. Is read, the hypothesis is reduced using the hypothesis content, the hypothesis reliability, and the hypothesis reduction parameter in a process 29, and the target observation information associated with the managed target one-to-one according to the hypothesis content after the reduction in a process 30 is integrated. A determination is made, and a smoothed value and a smoothed error covariance of the position / velocity of the management target at the current time are calculated based on the Kalman filter theory in a process 31.
Manages the smoothed value and smoothed error covariance of the position and velocity of the management target calculated in step 2 and the target observation information input from the target observation device, and displays the smoothed value of the position of the management target and the observation position of the target observation information. Then, a threshold for target detection is calculated from the noise level of the target observation device in a process 34 and sent to the target observation device, and a correlation gate threshold for determining the spread of the correlation gate in accordance with the noise level in a process 35. This sequence is repeated until the value is calculated and the correlation integration is completed in process 32.

FIG. 2 is a block diagram showing a second embodiment of the present invention. In FIG. 2, reference numeral 1 denotes a first target observation device which outputs the position and the like of a target and an unnecessary signal as target observation information; A second target observation device arranged at a position different from that of the first target observation device, and a target observation device position specification input device for inputting a positional relationship between the first target observation device and the second target observation device. And 4, target observation information of the new target obtained from the first target observation device at the first sampling and 2
At the sampling, the target observation information of the new target obtained from the second target observation device is input, and the smoothed value of the target position / velocity is obtained based on the target observation information (the situation shown in FIG. 8) sequentially obtained from each target observation device. And calculating the initial value of the smoothed error covariance and inputting the target observation information of the managed target again from the first target observation device at the third sampling.
A target observation information input unit corresponding to 2 and a correlation between the target observation information and the managed target, and an observation error covariance and correlation indicating an error between the prediction error covariance and the target observation information centering on the predicted position of the managed target. A motion specification correlator corresponding to the processing 24 for determining whether or not the target prediction existence range is generated in the correlation gate indicating the target prediction existence range generated by using the correlation gate threshold value for determining the spread of the gate; A hypothesis generator corresponding to a process 25 for generating a hypothesis whether the inner target observation information is a managed target, a new target, or an unnecessary signal; and 7 is a hypothesis reliability calculation parameter for calculating the reliability of the generated hypothesis. The hypothesis reliability calculation parameter input unit 8 corresponding to the process 26 for reading the new target observation frequency is a process 2 for calculating the hypothesis reliability from the target observation information, the hypothesis contents, the new target observation frequency, and the unnecessary signal observation frequency. , 9 is a hypothesis reduction parameter input unit corresponding to a process 28 for reading a hypothesis reduction parameter for reducing a plurality of generated hypotheses, and 10 is a hypothesis content, a hypothesis reliability, and a hypothesis reduction parameter. Is a hypothesis reducer corresponding to a process 29 for reducing a hypothesis by using a target, and 11 is a target corresponding to a process 30 for performing an integrated determination of target observation information that is associated one-to-one with a managed target according to the content of the reduced hypothesis. The information integration determiner 12 is a target motion specification smoother corresponding to the processing 31 for calculating the smoothed value and the smoothed error covariance of the position / velocity of the management target at the current time based on the Kalman filter theory, and 13 is calculated at the second sampling. It corresponds to the process 23 of calculating the predicted value of the target position / velocity at the current time of the already-managed target and the prediction error covariance obtained by estimating the error of the predicted value. The target motion specification predictor 14 manages the calculated smoothed value and smoothed error covariance of the position and velocity of the management target and the target observation information input from the target observation device, and further calculates the smoothed value and the position of the management target. A target information management / display device 15 corresponding to a process 33 for displaying the observation position of the target observation information, and 15 corresponds to a process 34 for calculating a threshold for target detection from the noise level of the target observation device and sending the threshold to the target observation device. A target observation apparatus noise level controller, 16 is a correlation gate threshold controller corresponding to a process 35 for calculating a correlation gate threshold for determining the spread of the correlation gate according to the noise level, 17 is a prediction error A correlation gate volume calculator corresponding to a process 36 for calculating the volume of the correlation gate from the covariance and the observation error covariance and the threshold value of the correlation gate. Unnecessary signal observation frequency calculation corresponding to the process 38 of calculating the unnecessary signal observation frequency indicating the frequency at which the unnecessary signal is included in a plurality of target observation information existing in the target using the number of target observation information in the correlation gate and the correlation gate volume It is a vessel.

Here, the calculation of the threshold for detecting the target with respect to the noise level of the 15 target observation devices will be described. The noise level of the target observation device is determined by the false alarm probability and the detection probability. FIG. 12 shows the relationship between the false alarm probability and the detection probability when the noise level increases. The relationship between the noise level and the threshold is represented by the above equation (1).

Next, the calculation of the correlation gate threshold value for the noise level of the 16 target observation devices will be described.
The correlation gate threshold is a random variable that determines the spread of the correlation gate. The spread of the correlation gate is represented by the sum of the prediction error covariance centered on the predicted position of the managed target and the observation error covariance indicating the error of the target observation information. It is defined by the chi-square distribution. Therefore, by determining the probability of the target existing in the correlation gate, the correlation gate threshold, which is a random variable of the chi-square distribution, can be determined. FIG. 13 shows the relationship between the change in the threshold value of the correlation gate, the target existence probability in the correlation gate, and the spread of the correlation gate. The target existence probability in the correlation gate is associated with the increase / decrease of the noise level, and the equation is such that the correlation gate is reduced when the noise level increases. The relationship between the target existence probability in the correlation gate and the noise level is expressed by the above equation (2).

The calculation of the seventeen correlation gate volumes will now be described. The correlation gate volume is calculated from the correlation gate spread and the correlation gate threshold. The relationship between the correlation gate volume, the threshold value of the correlation gate, and the spread of the correlation gate is expressed by the above equation (3).

The calculation of the unnecessary signal observation frequency of 19 will be described. The fundamental principle of the unnecessary signal observation frequency calculation is that if the target observation device continuously observes the same target, the number of target observation information for each observation will differ due to changes in the signal processing performance of the target observation device and the observation environment. The difference in the number of information is used as the number of unnecessary signals. However, since the frequency of observation of unnecessary signals is required as a parameter for calculating the hypothesis reliability, the spatial presence frequency of unnecessary signals is determined as the unnecessary signal observation frequency using the volume of the correlation gate where the target observation information exists. It has established. When the target observation information is alternately input to the target correlation integration device from the two target observation devices as shown in FIG. 8, the calculation formula of the unnecessary signal observation frequency is expressed by Expressions (6) and (7). Is done. The principle of the unnecessary signal observation frequency calculation is to estimate the number of observation targets input in time series from one target observation device by the least square method, and further estimate the number of unnecessary signals from the difference from the already managed target number. The number of unnecessary signals per unit volume (frequency of unnecessary signal observation) is calculated based on the volume of the correlation gate in which is present. Also, 1
Since the spread of the correlation gate changes according to the correlation gate threshold value calculated by the correlation gate threshold value controller of No. 6,
The target existence probability in the correlation gate also changes. In the state where the target existence probability in the correlation gate is decreasing, the number of observation targets in the correlation gate decreases stochastically from the true value. to correct.

[0035]

(Equation 5)

Embodiment 3 FIG. 7 is a diagram showing a premise of Embodiment 3 of the present invention.
Is the first target observation device, 2 is the second target observation device, 3
Reference numeral 9 denotes an observation target, and reference numeral 40 denotes a target correlation integration device, which represents a situation where two target observation devices observe three targets.

FIG. 8 shows a situation in which the two target observation devices observe targets in time series in the situation shown in FIG. 7 and output the results to the target correlation integration device.

FIG. 9 is a diagram showing a situation where each target observation device actually observes a target in the situations shown in FIGS. 7 and 8. In the figure, it is assumed that the first target observation device can observe two targets. It is also assumed that the second target observation device has observed three targets. When the number of targets already managed is 2, one of the three targets observed by the second target observation device is hypothesized as a new target. In order to determine the reliability of the hypothesis, the frequency of observation of a new target in space is required. The observation frequency is defined as the target number of observations per unit volume. Therefore, the volume of the correlation gate indicating the expected existence range of the target used in determining the correlation is calculated, and the new target observation frequency is calculated from the number of targets assumed to be new targets in the correlation gate. In addition, when the number of targets observed between two target observation devices fluctuates in time series, the estimation of the new target number is
First, the target number existing in the space is estimated by the least squares method, and the target number is obtained from the difference between the target number already managed and the target number existing in the estimated space.

FIG. 10 is a diagram showing a situation where each target observation device actually observes a target in the situations shown in FIGS. 7 and 8. In the figure, it is assumed that the first target observation device is observing the target and signals other than the target. Also,
It is assumed that the second target observation device can observe three targets.
When the number of targets already managed is 3, three or more targets among the targets observed by the first target observation device are hypothesized as unnecessary signals. What is important here is that the emphasis is placed on the target observation number of the first target observation device out of the two target observation devices. That is, when the number of observation targets fluctuates in one target observation device, it is estimated that unnecessary signals are mixed with signals from the targets. Also,
In order to determine the reliability of a hypothesis, the frequency of observation of unnecessary signals in space is required. The observation frequency is defined as the target number of observations per unit volume. Therefore, the volume of the correlation gate indicating the expected existence range of the target used in determining the correlation is calculated, and the unnecessary signal observation frequency is calculated from the target number assumed to be an unnecessary signal in the correlation gate. Also, 2
When the number of targets observed between two target observation devices fluctuates in time series, the number of unnecessary signals is estimated by estimating the number of targets existing in space by the least squares method for each target observation device, and is already managed. It is determined from the difference between the target number existing and the target number existing in the estimated space. From the above, when there is a difference in the number of observation targets among a plurality of target observation devices, when the number of observation targets fluctuates in time series, when there is a difference in the number of observations of unnecessary signals among a plurality of target observation devices, Even when the number of unnecessary signals fluctuates in a time series, the new target observation frequency and the unnecessary signal observation frequency can be calculated flexibly.

FIG. 6 is a diagram showing a processing procedure according to the third embodiment of the present invention. In FIG. 6, reference numeral 22 denotes input of target observation information from a target observation device, reference numeral 23 denotes a prediction value and prediction error covariance matrix calculation, 24 is a correlation judgment between the observation target and the already-managed target, 25 is a hypothesis generation of the target observation information, 27 is a hypothesis reliability calculation of the target observation information, 28 is a hypothesis reduction parameter read, 29 is a hypothesis reduction, and 30 is an hypothesis reduction and 30 is the hypothesis reduction. The integrated judgment of the management target is performed.
2 is a correlation integration end determination, 33 is target information management / display, 3
4 is a noise level calculation of the target observation device, 35 is a correlation gate threshold value calculation, 36 is a correlation gate volume calculation, 37 is a new target observation frequency calculation, and 38 is an unnecessary signal observation frequency calculation. The target observation information of the new target obtained from the first target observation device and the target observation information of the new target obtained from the second target observation device at the second sampling are input to Based on the obtained target observation information (the situation shown in FIG. 8), the initial value of the smoothed value of the target position / velocity and the smoothed error covariance indicating the spatial error evaluation amount of the target position / velocity are calculated.
The target observation information of the managed target is input again from the first target observation device at the third sampling at the third sampling, and the predicted value of the target position / velocity and the target at the latest time of the managed target calculated at the second sampling at the processing 23 are input. The prediction error covariance indicating the position / velocity error evaluation amount is calculated, and the correlation between the target observation information and the managed target is calculated in process 24. The error between the prediction error covariance and the target observation information is centered on the predicted position of the managed target. It is determined whether or not the observation error covariance is present in the correlation gate indicating the target prediction existence range generated using the correlation gate threshold value for determining the spread of the correlation gate and the correlation error indicating the spread of the correlation gate. Calculate the volume of the correlation gate from the variance and the observation error covariance and the correlation gate threshold, and in a process 37, a new target observation indicating the frequency at which the new target is included in a plurality of target observation information present in the correlation gate The degree is calculated using the number of target observation information in the correlation gate and the correlation gate volume, and the unnecessary signal observation frequency indicating the frequency at which unnecessary signals are included in a plurality of target observation information existing in the correlation gate in processing 38 is calculated in the correlation gate. It is calculated using the number of target observation information and the correlation gate volume, and in step 25, whether the target observation information in the correlation gate is a managed target or a new target,
A hypothesis of an unnecessary signal is generated, a hypothesis reliability is calculated from a target observation information, a hypothesis content, a new target observation frequency and an unnecessary signal observation frequency in a process 27, and a plurality of hypotheses generated in the process 28 are reduced. Read hypothesis reduction parameter and process 2
In step 9, the hypothesis is reduced using the hypothesis content, the hypothesis reliability, and the hypothesis reduction parameter. In step 30, the integrated determination of the target observation information that is associated with the managed target in a one-to-one manner is performed according to the reduced hypothesis content. In a process 31, a smoothed value and a smoothed error of the position / speed of the management target at the current time are calculated based on the Kalman filter theory, and a smoothed value and the smoothed error of the position / speed of the management target calculated in the process 33 are calculated. The target observation information input from the observation device is managed, the smoothed value of the position of the management target and the observation position of the target observation information are displayed, and a threshold for target detection is calculated from the noise level of the target observation device in process. At step 35, and calculates a correlation gate threshold value for determining the spread of the correlation gate according to the noise level at step 35. Until the completion correlation integration 2 repeats this series of flows.

FIG. 3 is a block diagram showing a third embodiment of the present invention. In FIG. 3, reference numeral 1 denotes a first target observation device for outputting the position and the like of a target and an unnecessary signal as target observation information; A second target observation device arranged at a position different from that of the first target observation device, and a target observation device position specification input device for inputting a positional relationship between the first target observation device and the second target observation device. And 4, target observation information of the new target obtained from the first target observation device at the first sampling and 2
At the sampling, the target observation information of the new target obtained from the second target observation device is input, and the smoothed value of the target position / velocity is obtained based on the target observation information (the situation shown in FIG. 8) sequentially obtained from each target observation device. And calculating the initial value of the smoothed error covariance and inputting the target observation information of the managed target again from the first target observation device at the third sampling.
A target observation information input unit corresponding to 2 and a correlation between the target observation information and the managed target, and an observation error covariance and correlation indicating an error between the prediction error covariance and the target observation information centering on the predicted position of the managed target. A motion specification correlator corresponding to the processing 24 for determining whether or not the target prediction existence range is generated in the correlation gate indicating the target prediction existence range generated by using the correlation gate threshold value for determining the spread of the gate; A hypothesis generator 25 corresponding to a process 25 for generating a hypothesis of whether the inner target observation information is a managed target, a new target, or an unnecessary signal. The hypothesis generator 8 uses the target observation information, the hypothesis content, the new target observation frequency, and the unnecessary signal observation frequency. Hypothesis reliability calculator corresponding to processing 27 for calculating hypothesis reliability, 9
Is a hypothesis reduction parameter input unit corresponding to a process 28 for reading hypothesis reduction parameters for reducing a plurality of generated hypotheses, and 10 is a process 29 for reducing a hypothesis using the hypothesis contents, the hypothesis reliability and the hypothesis reduction parameter. , A target information integration determiner 11 corresponding to a process 30 for performing integrated determination of target observation information that is associated one-to-one with managed targets in accordance with the hypothesis content after reduction, and 12 is a target information integration determiner at the current time. A target motion specification smoother corresponding to a process 31 for calculating a smoothed value and a smoothed error covariance of the position and speed of the management target based on the Kalman filter theory, and 13 is a target position at the current time of the managed target calculated at the second sampling. A target motion specification predictor 14 corresponding to a process 23 for calculating a predicted error of the speed and a predicted error covariance estimating an error of the predicted value; Processing 33 for managing the smoothed value and smoothed error covariance of the position / velocity of the management target and the target observation information input from the target observation device, and displaying the smoothed value of the position of the management target and the observation position of the target observation information 33 15 is a target observation device noise level controller corresponding to a process 34 for calculating a threshold for target detection from the noise level of the target observation device and sending the threshold to the target observation device, and 16 The correlation gate threshold controller 17 corresponding to the processing 35 for calculating the correlation gate threshold for determining the spread of the correlation gate according to the noise level, the prediction error covariance and the observation error covariance and the correlation gate 17 The correlation gate volume calculator 18 corresponding to the processing 36 for calculating the volume of the correlation gate from the threshold value includes a plurality of target observation information existing in the correlation gate. , A new target observation frequency calculator corresponding to the process 37 of calculating the new target observation frequency indicating the frequency at which the new target is included using the number of target observation information in the correlation gate and the correlation gate volume, and 19 is present in the correlation gate. This is an unnecessary signal observation frequency calculator corresponding to the processing 38 of calculating the unnecessary signal observation frequency indicating the frequency at which unnecessary signals are included in a plurality of target observation information using the number of target observation information in the correlation gate and the correlation gate volume.

Here, the calculation of the threshold for detecting the target with respect to the noise level of the 15 target observation devices will be described. The noise level of the target observation device is determined by the false alarm probability and the detection probability. FIG. 12 shows the relationship between the false alarm probability and the detection probability when the noise level increases. The relationship between the noise level and the threshold is represented by the above equation (1).

The calculation of the correlation gate threshold value for the noise level of the 16 target observation devices will be described.
The correlation gate threshold is a random variable that determines the spread of the correlation gate. The spread of the correlation gate is represented by the sum of the prediction error covariance centered on the predicted position of the managed target and the observation error covariance indicating the error of the target observation information. It is defined by the chi-square distribution. Therefore, by determining the probability of the target existing in the correlation gate, the correlation gate threshold, which is a random variable of the chi-square distribution, can be determined. FIG. 13 shows the relationship between the change in the threshold value of the correlation gate, the target existence probability in the correlation gate, and the spread of the correlation gate. The target existence probability in the correlation gate is associated with the increase / decrease of the noise level, and the equation is such that the correlation gate is reduced when the noise level increases. The relationship between the target existence probability in the correlation gate and the noise level is expressed by the above equation (2).

The calculation of the seventeen correlation gate volumes will now be described. The correlation gate volume is calculated from the correlation gate spread and the correlation gate threshold. The relationship between the correlation gate volume, the threshold value of the correlation gate, and the spread of the correlation gate is expressed by the above equation (3).

The calculation of 18 new target observation frequencies and 19 unnecessary signal observation frequencies will be described. The basic principle of calculating the new target observation frequency is that when multiple target observation devices separated from each other observe the same target, the number of target observation information will differ due to the target formation shape and the positional relationship and resolution of the target observation device, The difference number is set as a new target number. However, since the observation frequency of the new target is required as a parameter for calculating the hypothesis reliability, the spatial existence frequency of the new target is used as the new target observation frequency using the volume of the correlation gate where the target observation information exists. It has established. When the target observation information is alternately input to the target correlation integration device from the two target observation devices as shown in FIG. 8, the calculation formula of the new target observation frequency is expressed by Expressions (8) and (9). Is done. The principle of the new target observation frequency calculation is that the number of observation targets alternately input from two target observation devices is estimated by the least squares method, and the new target number is estimated from the difference from the managed target number. The number of new targets per unit volume (new target observation frequency) is calculated from the volume of existing correlation gates. The basic principle of the unnecessary signal observation frequency calculation is that if the target observation device continuously observes the same target, the number of target observation information differs for each observation due to changes in the signal processing performance of the target observation device and the observation environment. The difference in the number of information is used as the number of unnecessary signals. However, since the frequency of observation of unnecessary signals is required as a parameter for calculating the hypothesis reliability, the spatial presence frequency of unnecessary signals is determined as the unnecessary signal observation frequency using the volume of the correlation gate where the target observation information exists. It has established. As shown in FIG. 8, when the target observation information is alternately input from the two target observation devices to the target correlation integration device, the calculation formula of the unnecessary signal observation frequency is represented by Expressions (10) and (11). expressed. The principle of the unnecessary signal observation frequency calculation is to estimate the number of observation targets input in time series from one target observation device by the least square method, and further estimate the number of unnecessary signals from the difference from the already managed target number. The number of unnecessary signals per unit volume (frequency of unnecessary signal observation) is calculated based on the volume of the correlation gate in which is present. Also,
Since the spread of the correlation gate changes according to the correlation gate threshold value calculated by the 16 correlation gate threshold value controllers, the target existence probability in the correlation gate also changes. In the state where the target existence probability in the correlation gate is decreasing, the number of observation targets in the correlation gate decreases stochastically from the true value.
The number of observation targets is corrected with the target existence probability in the correlation gate determined from the threshold value.

[0046]

(Equation 6)

[0047]

According to the first aspect, a new target observation frequency calculator is provided in the conventional target correlation integration device,
Since the frequency of new target observations at the time of new target occurrence can be accurately grasped, highly reliable hypothesis reliability can be obtained.As a result of accurate hypothesis reduction by the hypothesis reducer using the hypothesis reliability, correlation integration performance is improved. Can be improved. further,
By providing a target information management / display unit and a target observation device noise level controller, it is possible to appropriately and automatically control the noise level. Targets that fly in low altitudes where noise level fluctuations are large and difficult to detect, and stealth The target detection performance is improved for targets with a small effective reflection area of radio waves such as aircraft, and the input of target observation information to the target correlation integration device is stabilized. As a result, correlation integration performance can be improved. . In addition, by providing a correlation gate threshold controller, it is possible to reduce the number of unnecessary signals in the correlation gate when the noise level increases, and as a result, the processing of the target correlation integration device accompanying the decrease in the number of generation hypotheses. The load can be reduced.

According to the second aspect of the present invention, by providing an unnecessary signal observation frequency calculator in the conventional target correlation integration device, the frequency of unnecessary signal observation when an unnecessary signal is generated can be accurately grasped. The degree of correlation can be obtained, and as a result of accurate hypothesis reduction by the hypothesis reducer using the hypothesis reliability, correlation integration performance can be improved. further,
By providing a target information management / display unit and a target observation device noise level controller, it is possible to appropriately and automatically control the noise level. Targets that fly in low altitudes where noise level fluctuations are large and difficult to detect, and stealth The target detection performance is improved for targets with a small effective reflection area of radio waves such as aircraft, and the input of target observation information to the target correlation integration device is stabilized. As a result, correlation integration performance can be improved. . In addition, by providing a correlation gate threshold controller, it is possible to reduce the number of unnecessary signals in the correlation gate when the noise level increases, and as a result, the processing of the target correlation integration device accompanying the decrease in the number of generation hypotheses. The load can be reduced.

According to the third aspect of the present invention, a new target observation frequency calculator and an unnecessary signal observation frequency calculator are provided in the conventional target correlation integration device, so that the new target observation frequency when a new target is generated and the unnecessary signal , It is possible to obtain a highly reliable hypothesis
As a result of accurate hypothesis reduction by the hypothesis reducer using the hypothesis reliability, correlation integration performance can be improved. Furthermore, by providing a target information management / display device and a target observation device noise level controller, it is possible to appropriately and automatically control the noise level. The target detection performance is improved for targets with small effective reflection areas of radio waves, such as stealth machines, and the input of target observation information to the target correlation integration device is stabilized. As a result, the correlation integration performance is improved. Can be. In addition, by providing a correlation gate threshold controller, it is possible to reduce the number of unnecessary signals in the correlation gate when the noise level increases, and as a result, the processing of the target correlation integration device accompanying the decrease in the number of generation hypotheses. The load can be reduced.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing Embodiment 1 of a target correlation integration device according to the present invention.

FIG. 2 is a configuration diagram showing a second embodiment of the target correlation integration apparatus according to the present invention.

FIG. 3 is a configuration diagram showing a third embodiment of the target correlation integrating apparatus according to the present invention;

FIG. 4 is a diagram showing a processing procedure according to the first embodiment of the target correlation integrating apparatus according to the present invention;

FIG. 5 is a diagram showing a processing procedure of a target correlation integrating apparatus according to a second embodiment of the present invention;

FIG. 6 is a diagram showing a processing procedure of a third embodiment of the target correlation integrating apparatus according to the present invention.

FIG. 7 is a diagram showing a situation on which the present invention is based.

FIG. 8 is a diagram showing an input order of target observation information from two target observation devices to a target correlation integration device.

FIG. 9 is a diagram illustrating the order in which target observation information is input to the target correlation integration device when different target numbers are observed by two target observation devices.

FIG. 10 is a diagram showing the order in which target observation information in which different target numbers are observed by two target observation devices and in which unnecessary signals are mixed is input to the target correlation integration device.

FIG. 11 is a diagram showing a correlation gate generation process and an example of inside / outside correlation gate determination.

FIG. 12 is a diagram showing a relationship between a false alarm probability and a detection probability associated with an increase in the noise level of the target observation device.

FIG. 13 is a diagram illustrating a relationship between a target existence probability in a correlation gate and a spread of the correlation gate according to a change in the threshold value of the correlation gate.

FIG. 14 is a configuration diagram showing an embodiment of a conventional target correlation integration device.

FIG. 15 is a diagram showing a processing procedure of an embodiment of a conventional target correlation integration device.

[Explanation of symbols]

1 first target observation device, 2 second target observation device, 3
Target observation device position specification input device, 4 target observation information input device, 5 motion specification correlator, 6 hypothesis generator, 7 hypothesis reliability calculation parameter input device, 8 hypothesis reliability calculator,
9 hypothesis reduction parameter input device, 10 hypothesis reduction device, 1
1 target information integration determiner, 12 target motion specification smoother,
13 target motion specification predictor, 14 target information management / display, 15 target observation device noise level controller, 16 correlation gate threshold controller, 17 correlation gate volume calculator, 1
8 New target observation frequency calculator, 19 Unwanted signal observation frequency calculator, 20 Target observation information input for first observation, 21 Initial setting of smoothed value and smooth error covariance matrix, 22 Target observation information input from target observation device, 23 Predicted value and prediction error covariance matrix calculation, 24 Correlation judgment of observed target and managed target, 25 Generation of hypothesis of target observation information, 26 Reading of hypothesis reliability calculation parameter, 27 Calculation of hypothesis reliability of target observation information, 28 Hypothesis reduction Read parameters, 29
Hypothesis reduction, integrated judgment of 30 observation targets and managed targets,
31 Smooth value and smooth error covariance matrix calculation, 32 Correlation integration termination judgment, 33 Target information management / display, 34 Noise level calculation of target observation device, 35 Correlation gate threshold value calculation, 36 Correlation gate volume calculation, 37 New target Observation frequency calculation, 38 Unwanted signal observation frequency calculation, 39 observation targets, 40 target correlation integration device.

Claims (3)

[Claims] 1. A method for outputting target observation information such as a position from a target and an unnecessary signal other than the target and observation status such as an observation time, and changing a threshold of target detection with respect to a noise level of the unnecessary signal. Target observation equipment,
A target observation device position specification input device for inputting a positional relationship between the plurality of target observation devices, and target information such as observation position data from the target observation device for each target observation device of the target observation device position specification input device. Target observation information input device that converts the coordinate system of the target to the coordinate system of the correlation integration process, and the target motion information that calculates the target prediction information such as the position and velocity at the current observation time from the position and velocity at the observation time of the managed target. Managed target prediction information and target observation using a managed gate target prediction information from the target motion specification predictor and a correlation gate generated based on target observation information from the target observation information input device. A motion specification correlator for determining correlation of information; and a plurality of observation targets including unnecessary signals determined to be correlated with the managed target by the motion specification correlator. And the hypothesis generator to make a target of the hypothesis,
A hypothesis reliability calculation parameter input device for inputting setting conditions such as unnecessary signal observation frequency, a correlation gate volume calculator for calculating a volume of a correlation gate based on correlation gate specifications from the motion specification correlator, New target observation frequency based on the number of managed targets from the motion specification correlator, the number of target observation information in the correlation gate for each target observation information observed by the target observation device, and the correlation gate volume from the correlation gate volume calculator Is calculated by the hypothesis generator based on the unnecessary signal observation frequency and the like from the hypothesis reliability calculation parameter input device and the new target observation frequency from the new target observation frequency calculator. A hypothesis reliability calculator for calculating the reliability of each hypothesis, a hypothesis reduction parameter input device for inputting a hypothesis reduction condition, a hypothesis reduction condition from the hypothesis reduction parameter input device, and A hypothesis reducer that reduces the hypotheses generated by the hypothesis generator based on the hypothesis reliability from the hypothesis reliability calculator; and a managed target based on the hypothesis reliability and the content of the hypothesis reduced by the hypothesis reducer. And a target information integration determiner for determining the combination of the target and the observation target, and a target motion for calculating a smoothing gain from the managed target prediction information and the target observation information and calculating a smoothed value of the managed target prediction information and the target observation information. A specification information smoother, a target information management / display for managing and displaying a smoothed value calculated by the target movement specification smoother and target information input from the target observation device, and the target information management / display A target observation device noise level controller for automatically detecting a noise level of the target observation device and calculating a target detection threshold, and a noise level detected by the target observation device noise level controller. Target correlation integration apparatus comprising the correlation gate threshold controller for calculating a threshold noise level than the correlation gates. 2. A method for outputting target observation information such as a position from a target and an unnecessary signal other than the target and observation status such as an observation time, and changing a threshold of target detection with respect to a noise level of the unnecessary signal. Target observation equipment,
A target observation device position specification input device for inputting a positional relationship between the plurality of target observation devices, and target information such as observation position data from the target observation device for each target observation device of the target observation device position specification input device. Target observation information input device that converts the coordinate system of the target to the coordinate system of the correlation integration process, and the target motion information that calculates the target prediction information such as the position and velocity at the current observation time from the position and velocity at the observation time of the managed target. Managed target prediction information and target observation using a managed gate target prediction information from the target motion specification predictor and a correlation gate generated based on target observation information from the target observation information input device. A motion specification correlator for determining correlation of information; and a plurality of observation targets including unnecessary signals determined to be correlated with the managed target by the motion specification correlator. And the hypothesis generator to make a target of the hypothesis,
A hypothesis reliability calculation parameter input device for inputting setting conditions such as a new target observation frequency, a correlation gate volume calculator for calculating a correlation gate volume based on correlation gate specifications from the motion specification correlator, Unwanted signal observation frequency based on the number of managed targets from the motion specification correlator, the number of target observation information in the correlation gate for each target observation information observed by the target observation device, and the correlation gate volume from the correlation gate volume calculator Unnecessary signal observation frequency calculator, the new target observation frequency from the hypothesis reliability calculation parameter input device and the unnecessary signal observation frequency from the unnecessary signal observation frequency calculator are generated by the hypothesis generator. A hypothesis reliability calculator for calculating the reliability of each hypothesis, a hypothesis reduction parameter input device for inputting a hypothesis reduction condition, a hypothesis reduction condition from the hypothesis reduction parameter input device, and A hypothesis reducer that reduces the hypotheses generated by the hypothesis generator based on the hypothesis reliability from the hypothesis reliability calculator; and a managed target based on the hypothesis reliability and the content of the hypothesis reduced by the hypothesis reducer. And a target information integration determiner for determining the combination of the target and the observation target, and a target motion for calculating a smoothing gain from the managed target prediction information and the target observation information and calculating a smoothed value of the managed target prediction information and the target observation information. A specification information smoother, a target information management / display for managing and displaying a smoothed value calculated by the target movement specification smoother and target information input from the target observation device, and the target information management / display A target observation device noise level controller for automatically detecting a noise level of the target observation device and calculating a target detection threshold, and a noise level detected by the target observation device noise level controller. Target correlation integration apparatus comprising the correlation gate threshold controller for calculating a threshold noise level than the correlation gates. 3. A method for outputting target observation information such as a position from a target and an unnecessary signal other than the target and an observation state such as an observation time, and changing a threshold of target detection with respect to a noise level of the unnecessary signal. Target observation equipment,
A target observation device position specification input device for inputting a positional relationship between the plurality of target observation devices, and target information such as observation position data from the target observation device for each target observation device of the target observation device position specification input device. Target observation information input device that converts the coordinate system of the target to the coordinate system of the correlation integration process, and the target motion information that calculates the target prediction information such as the position and velocity at the current observation time from the position and velocity at the observation time of the managed target. Managed target prediction information and target observation using a managed gate target prediction information from the target motion specification predictor and a correlation gate generated based on target observation information from the target observation information input device. A motion specification correlator for determining correlation of information; and a plurality of observation targets including unnecessary signals determined to be correlated with the managed target by the motion specification correlator. And the hypothesis generator to make a target of the hypothesis,
A correlation gate volume calculator that calculates the volume of the correlation gate based on the correlation gate data from the motion data correlator; and the managed target number from the motion data correlator and the target observation device. A new target observation frequency calculator for calculating a new target observation frequency from the number of target observation information in the correlation gate for each target observation information and the correlation gate volume from the correlation gate volume calculator; Unwanted signal observation frequency calculation for calculating an unnecessary signal observation frequency from the number of target observation information in the correlation gate for each target observation information observed by the target observation device and the correlation gate volume from the correlation gate volume calculator And the reliability of each hypothesis generated by the hypothesis generator based on the new target observation frequency from the new target observation frequency calculator and the unnecessary signal observation frequency from the unnecessary signal observation frequency calculator. A hypothesis reliability calculator, a hypothesis reduction parameter input device for inputting a hypothesis reduction condition, and a hypothesis reduction condition from the hypothesis reduction parameter input device and a hypothesis reliability from the hypothesis reliability calculator. A hypothesis reducer for reducing a hypothesis generated by a hypothesis generator, and a target information integration determiner for determining a combination of a managed target and an observation target based on the content of the hypothesis reduced by the hypothesis reducer and the hypothesis reliability A target motion specification smoother that calculates a smoothed gain from the managed target prediction information and the target observation information to calculate a smoothed value of the managed target prediction information and the target observation information,
A target information management / display for managing and displaying the smoothed value calculated by the target motion specification smoother and the target information input from the target observation device, and a target information management / display from the target information management / display. A target observation device noise level controller that automatically detects a noise level and calculates a threshold for target detection; and a correlation gate that calculates a threshold value of the correlation gate from the noise level detected by the target observation device noise level controller. A target correlation integration device comprising a threshold controller.