JP2006003256A - Target tracking device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a target tracking device which attains high estimation precision in such a state that a plurality of targets separate from already tracked targets. <P>SOLUTION: The target tracking device is equipped with: an observation information extracting section which includes a range information calculating section, a singular/plural number judging section, a singular number angle information calculating section, a plural number angle information calculating section, and an observation result outputting section; and a tracking processing section which includes a data memory section, a track estimating section and a track evaluating section. The tracking processing section further includes a hypothesis qualifying section which only registers a track hypothesis obtained by updating an existing main track from track hypotheses generated in the track evaluating section, in the case where a judgement result obtained by the singular/plural number judging section represents a singular number, and which registers both the track hypothesis obtained by updating the existing main track and a track hypothesis corresponding to an appearing new target separating from the existing main track from track hypotheses generated in the track evaluating section in the case where the judgement result represents a plural number. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to target tracking for estimating a target trajectory based on an observation value obtained from a radar. In particular, in a situation where unnecessary signals are simultaneously observed, the target tracking is separated from a target state on the way and a new one is obtained. The present invention relates to a target tracking device that can accurately and efficiently process tracking of a flying object in which a target is generated.

  In target tracking in which a track that is a target trajectory is estimated using an observation value obtained from a sensor, it is necessary to determine whether the observation value is from a target or an unnecessary signal. If it is determined that the observed value is from a target, then whether it is from a target that has already been tracked (if there are multiple existing targets, which is one of them) It is determined whether or not the target has occurred. The position at the corresponding time is predicted from the speed of the existing target, a spatial region (gate) centered on the predicted value is set, and it is determined whether or not there is a target in the gate.

  The size of the gate is set so that the probability that the observed value from the tracking target can be observed in the gate is high, for example, 0.9 or more. Observations that have entered this gate are likely to be from existing targets, and observations that have not entered any gate are likely to be unused signals or new targets.

  For observed values determined to have been obtained from existing wakes, the true value of the target at the time the observed values were obtained according to the Kalman filter calculation using the position information of the observed values and the predicted values of the existing wakes. Calculate position and velocity estimates. However, if there are multiple observations in a gate, there is a risk that if one of the targets to be tracked is determined, the target tracking will fail if it is wrong. Becomes higher.

  Therefore, considering multiple allocation methods (hypotheses), determining the correlation between the observed value and the target for each hypothesis, and finally leaving the best hypothesis, more accurate target tracking is possible. There is a target tracking technique based on this (see, for example, Patent Document 1). In this method, the processing value is improved by processing the observation value, the wake that is a combination of the observation value, and the hypothesis that is the combination of the wake for each cluster that does not share the observation value.

  On the other hand, in the observation by the radar, an observation information extraction process for calculating the target position and velocity information by processing the received signal before the observation value is obtained is necessary. At this stage, as a method for calculating angle information from the received signal, there is a monopulse angle measurement method.

  Furthermore, the angle calculation process is performed while calculating whether the signal source is a single target or multiple targets by calculating the real part of the ratio of the sum beam and difference beam of the signal for each antenna element, which is calculated in this angle measurement method. There is a sensor signal processing system to perform (for example, see Non-Patent Document 1). By using this single determination result, it is possible to determine whether to make a single observation value or a plurality of observation values to be input for tracking.

Japanese Patent No. 3145893 (page 14, FIG. 1) Document "Two-step angle measurement method for in-vehicle radar" (The IEICE Transactions Vol. J86-B No. 8)

  However, the prior art has the following problems. When observing a plurality of targets approaching the sensor from a distance, there may occur a case where what is observed as one target in the distance due to the resolution of the sensor is observed as a plurality of targets from a certain point in time. When such a target is to be tracked and maintained, one or more other targets are separated from the target to be maintained and must be tracked together with the original tracked target. However, in such a case, there is a problem that only one of the separated targets can be tracked only by the tracking maintenance function in the conventional target tracking technique.

  In the tracking method having both the tracking maintenance function and the initialization function, any one of the separated targets is treated as an existing tracking target, and the other is treated as a new target. The wake that started as a new wake grows from the temporary wake and is promoted to the main wake. This method makes it possible to track everything after separation, but the tracking initialization method does not assume separation from the existing main wake, and therefore promotes an incorrect wake to the main wake. The possibility increases.

  On the other hand, in the conventional sensor signal processing system, the estimation result regarding the target number calculated at the time of monopulse angle measurement cannot be reflected in the hypothesis configuration of the tracking process for tracking the separation target. Therefore, it is easy to cause a tracking error and a possibility that a false track is started. Furthermore, in the conventional sensor signal processing system, the hypothetical configuration of the tracking process for tracking the separation target cannot be reflected in the target number estimation threshold in monopulse angle measurement. Therefore, an error in target number estimation is likely to occur.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to obtain a target tracking device that can achieve high estimation accuracy in a situation where a plurality of targets are separated from a tracked target.

  A target tracking device according to the present invention includes an observation information extraction unit that generates observation value information from received signals for tracking targets from a plurality of antenna elements, and a tracking processing unit that updates a target track based on the observation value information. The observation information extraction unit includes a distance information calculation unit that extracts target distance information based on reception signals for the tracking targets from a plurality of antenna elements, and each antenna element. A singular / multiple determination unit for comparing whether the ratio of the difference signal to the sum signal of the received signals is equal to or greater than a predetermined threshold and determining whether the target is singular or plural in the adjacent observation region; A single angle information calculation unit that calculates target angle information for a target that is determined to be singular by the singular / multiple determination unit, and a target that is determined to be plural by the singular / multiple determination unit. Multiple angle information calculation unit for calculating degree information, distance information calculated by distance information calculation unit, single determination result determined by single multiple determination unit, single angle information calculation unit and multiple angle information calculation unit An observation result output unit that generates and outputs observation value information including the measured angle information, and the tracking processing unit stores data relating to the target track updated every predetermined sampling period Based on the observation value information from the observation information extraction unit and the data on the target track updated at the previous sampling time stored in the data storage unit, the data on the target track at the current sampling time is generated. The trajectory estimator that updates data related to the target track in the data storage unit and the data related to the target track are combined to generate a wake hypothesis, and the generated temporary A tracking evaluation unit that calculates the reliability of the single target when the single determination result is determined to be singular based on the single determination result determined by the single determination unit And leave only the wake hypothesis by updating the existing main wake from the wake hypothesis generated by the trajectory evaluation unit, and if it is determined that there are multiple single determination results, determine multiple targets, A hypothesis limiting unit that leaves a wake hypothesis by updating the existing main wake from the wake hypotheses generated by the trajectory evaluation unit, and a wake hypothesis that corresponds to the new target separated from the existing main wake. It is provided.

  According to the present invention, it is possible to limit a hypothesis based on a single determination result of a target number determined by an observation information extraction unit, and a target that can achieve improvement in tracking accuracy and reduction in calculation load A tracking device can be obtained.

Embodiment 1 FIG.
DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of a target tracking device according to the invention will be described with reference to the drawings. FIG. 1 is a configuration diagram of a target tracking device according to Embodiment 1 of the present invention.

  The target tracking device according to the first embodiment includes an observation information extraction unit 100 and a tracking processing unit 200. Here, the observation information extraction unit 100 includes a distance information calculation unit 101, a single plural determination unit 102, a single angle information calculation unit 103, a multiple angle information calculation unit 104, and an observation result output unit 105.

  The distance information calculation unit 101 in the observation information extraction unit 100 is a calculation unit that calculates target distance information based on reception signals from a plurality of antenna elements installed in a target observation apparatus (not shown). The singular / multiple determination unit 102 is a determination unit that determines whether the observation region has a single target or a plurality of targets based on the sum signal and the difference signal of the received signals for each antenna element.

  The singular angle information calculation unit 103 is a calculation unit that calculates target angle information by the monopulse method when the single / multiple determination unit 102 determines that the target is singular. The multiple angle information calculation unit 104 is a calculation unit that calculates angle information of multiple targets when the single multiple determination unit 102 determines that there are multiple targets. As this calculation method, for example, the maximum likelihood estimation method described in the document “Two-step angle measurement method for in-vehicle radar” (Journal of the Institute of Electronics, Information and Communication Engineers, Research Bulletin Vol. J86-B No. 8) is used. Can be mentioned.

  The observation result output unit 105 includes, for each predetermined sampling period, the distance between the observation values calculated by the distance information calculation unit 101, the single determination result determined by the single plural determination unit 102, the single angle information calculation unit 103, and The output processing unit generates observation value information including the angle information of the observation values calculated by the multiple angle information calculation unit 104 and outputs the observation value information to the tracking processing unit 200.

  On the other hand, the tracking processing unit 200 includes a trajectory estimation unit 201, a trajectory evaluation unit 202, a hypothesis limiting unit 203, and a data storage unit 204. The trajectory estimation unit 201 in the tracking processing unit 200 uses the information on the target track generated based on the past observation values and the information on the current observation values to update the target track and generate a new track. It is the estimation means which performs. At that time, the motion parameters of the latest wake are calculated using the Kalman filter.

  The trajectory evaluation unit 202 is an evaluation unit that generates a hypothesis that is a combination of the hypotheses based on the wake generated by the trajectory estimation unit 201 and calculates the reliability. The hypothesis limiting unit 203 is a limiting unit that selects hypotheses generated by the trajectory evaluation unit 202 and narrows down hypotheses. Further, the data storage unit 204 is a storage unit that stores data necessary for the tracking process in advance and accumulates data related to the target track updated every time the measurement value is sampled.

  Next, the flow of observation value extraction processing by the observation information extraction unit 100 and target tracking processing by the tracking processing unit 200 will be described with reference to flowcharts. First, observation value extraction processing by the observation information extraction unit 100 will be described. FIG. 2 is a flowchart of the observation value extraction process in the first embodiment of the present invention. It is characterized in that the single / multiple determination result of the single / multiple determination unit 102 is output to the tracking processing unit 200 as observation value information.

  The distance information calculation unit 101 in the observation information extraction unit 100 reads received signals from a plurality of antenna elements installed in a target observation apparatus (not shown) (step S201), and based on the received signals, the target distance Information is calculated (step S202). Next, the singular / multiple determination unit 102 determines whether there is a single target or multiple targets in the observation region based on the received signals received by the multiple antenna elements (step S203). The singular / multiple determination unit 102 applies the following expression as this determination. Here, Re means a real part.

In the above equation, Σ _m and Δ _m are a sum beam and a difference beam of signals for each antenna element. Nth is a threshold value for determination. The singular / plurality determination unit 102 determines that there are a plurality of targets when the above inequality is satisfied, and determines that the target is singular when the inequality is not satisfied.

  When it is determined that the target is singular, the singular angle information calculation unit 103 calculates target angle information by the monopulse angle measurement method (step S204). On the other hand, when it is determined that there are a plurality of targets, the multiple angle information calculation unit 104 calculates angle information of the multiple targets using a maximum likelihood estimation method or the like (step S205).

  Then, the observation result output unit 105, for each predetermined sampling period, the distance of the observation value calculated by the distance information calculation unit 101 in step S201, and the single determination result determined by the single plural determination unit 102 in step S202. The observation value information including the angle information of the observation value calculated by the single angle information calculation unit 103 in step S203 or the angle information of the observation value calculated by the multiple angle information calculation unit 104 in step S204 is generated, and tracking processing is performed. The data is output to the unit 200 (step S206).

  As described above, the observation information extraction unit 100 generates the observation value information based on the received signal from the target observation device at every predetermined sampling period by the series of processes of steps S201 to S206, and the tracking processing unit 200. Can be output.

  Next, target tracking processing in the tracking processing unit 200 will be described. FIG. 3 is a flowchart of the target tracking process in the first embodiment of the present invention. The flowchart of FIG. 3 shows a series of processes for reading observation value information output from the observation information extraction unit 100 in a certain sampling period and generating / updating the latest track data. It is characterized in that hypothesis limitation is performed using the single / multiple determination result received as part of the observation value information.

  The trajectory estimation unit 201 in the tracking processing unit 200 reads the observation value information from the observation information extraction unit 100, and obtains the single / double determination result and the data of the distance and angle of the observation value (step S301). Next, the trajectory estimation unit 201 performs in-gate determination on the observed value, and determines which cluster the observed value belongs to (step S302).

  In this in-gate determination, the trajectory estimation unit 201 calculates a gate based on data relating to the existing track stored in the data storage unit 204, and performs clustering by examining the correlation with the newly read observation value. . This process will be described with reference to FIGS. FIG. 4 is a diagram showing a state of an existing wake in the first embodiment of the present invention. In a certain cluster, five observed values 00, 10, 11, 20, and 21 are read by three samplings from time t = 0 to t = 2.

From the five observed value states shown in FIG. 4, the following six types of wakes T1 to T6 can be expressed as a set of observed values.
T1: 10-20
T2: 10-21
T3: 11-20
T4: 11-21
T5: 20
T6: 21

Further, the hypothesis for adopting the wake can be expressed as, for example, the following six types of hypotheses H1 to H6 as “ID: {wake set} reliability” by the wake T1 to T6 pair and the reliability thereof. Here, rel represents the reliability of each hypothesis.
H1: {T1, T4} rel1
H2: {T1, T6} rel2
H3: {T2, T5} rel3
H4: {T2} rel4
H5: {T5} rel5
H6: {T2, T3} rel6

  Next, FIG. 5 is a diagram showing the relationship between the existing track and the latest observed value in the first embodiment of the present invention. The observation values 30, 31, 32, and 33 are newly read as the observation values at the sampling time t = 3 in the previous step S301. The trajectory estimation unit 201 calculates a gate for the observation value 20 at the sampling time t = 2 and a gate for the observation value 21 at the sampling time t = 2 based on the tracks T1 to T6.

  The trajectory estimation unit 201 performs clustering of the new observed values 30 to 33 by examining the correlation between the calculated gate and each position of the new observed value. In FIG. 5, it is assumed that the following clustering has been performed. It is assumed that the new observation values 30 and 31 enter the gate of the observation value 20, that is, the gates of the wakes T1, T3, and T5. Further, it is assumed that the new observed value 32 enters the gate of the observed value 21, that is, the gates of the wakes T2, T4, and T6. Further, it is assumed that the new observation value 33 has not entered any wake gate.

Next, the trajectory estimation unit 201 creates a wake at the sampling time t = 3 based on the correlation between the gate and the new observation value as shown in FIG. 5 (step S303). The following three types of wakes are created.
(1) Updated track: A track created by adding the observation value entered in the gate to the existing track (2) New track: A track starting from the new observation value at the sampling time (3) Memory track track : Tracks that do not have correlated observations at the corresponding time for existing tracks

In the example of FIG. 5, the trajectory estimation unit 201 first creates the following nine types of wakes as the updated wakes based on the existing wakes T1 to T6 and the result of clustering the new observation values 30 to 33. .
T11: 10-20-30
T12: 10-20-31
T21: 10-21-32
T31: 11-20-30
T32: 11-20-31
T41: 11-21-32
T51: 20-30
T52: 20-31
T61: 21-32

Further, the trajectory estimation unit 201 creates the following three types of wakes as new wakes.
T71: 30
T72: 31
T73: 32

Furthermore, the trajectory estimation unit 201 creates the following six types of wakes as memory track wakes.
T81: 10-20
T82: 10-21
T83: 11-20
T84: 11-21
T85: 20
T86: 21

  Here, the wake can be divided into a temporary wake and a main wake. The provisional wake is a wake at a stage where it is determined whether or not the target is being tracked, and the wake is a wake that is determined to be tracking the target. It is a temporary wake at the stage where it has just been generated as a new wake, and then continues to be updated until it is recognized as the main wake.

  In step S303, for the observed value that has no correlation with any cluster, a new cluster is generated starting from the observed value. In the example of FIG. 5, since the observed value 33 has no correlation with any wake, a new cluster starting from this observed value 33 is generated.

  Further, in step S303, when an equivalence relationship described later occurs between a plurality of clusters due to the observed value at this time, processing for integrating the clusters is performed. Through a series of processes as shown in steps S301 to S303, the trajectory estimation unit 201 updates the existing track, generates a new track, and creates a new cluster based on the observation values read at each sampling time. Can do. That is, the update track, the new track, and the memory track track are stored as update data in the data storage unit 24 at every predetermined sampling period.

  Next, processing in the trajectory evaluation unit 202 will be described. First, the trajectory evaluation unit 202 generates an in-gate determination matrix based on various data created by the trajectory estimation unit 201 (step S304). This in-gate determination matrix shows a list of interpretation possibilities for the observation values correlated with each cluster.

  FIG. 6 is a diagram showing an in-gate determination matrix according to Embodiment 1 of the present invention. Each row of the in-gate determination matrix indicates an observation value correlated with the cluster at the corresponding time. In FIG. 6, new observed values 30, 31, and 32 at the sampling time t = 3 correspond to the first to third rows, respectively. Regarding the columns, the first column indicates an unnecessary signal, the second column to the seventh column indicate existing tracks T1 to T6, and the last three columns indicate a new track. ing.

  Since all observation values have the possibility of being unnecessary signals, the elements in the first column are all 1. The second to seventh columns correspond to the updated tracks generated from the existing tracks T1 to T6 and the new observed values 30 to 32. There are nine portions where 1 is set in the second column to the seventh column, and these correspond to each of the nine types of update tracks T11 to T61 described above.

  For example, for the second column, the observed values correlated with the existing track T1 are 30, 31, so the first and second rows are 1 and the third row is 0. That is, 1 in the second column and first row corresponds to the update track T11, and 1 in the second column and second row corresponds to the update track T12. Further, the last three columns correspond to the previously described new tracks T71 to T73, and these three new tracks can only correspond to their respective starting points, and thus form a unit matrix.

Next, the trajectory evaluation unit 202 generates a wake correlation matrix based on the in-gate determination matrix (step S305). Here, the wake correlation matrix is all matrices that satisfy the following three conditions, and is a combination of interpretations of observation values that are consistent with each other.
Condition 1) The element of the in-gate determination matrix that corresponds to the same position as the element that becomes 1 in the wake correlation matrix is always 1.
Condition 2) The number of elements that are 1 in each row of the wake correlation matrix is only one.
Condition 3) At most one element is 1 in each column of the wake correlation matrix. However, the first column may have any number of elements.

  FIG. 7 is a diagram showing a wake correlation matrix in the first embodiment of the present invention. As an example, five types of wake correlation matrices that satisfy the above three conditions are shown.

  Next, the trajectory evaluation unit 202 updates the existing hypothesis with the wake correlation matrix (step S306). In updating each hypothesis, all the wake correlation matrices except for the wake correlation matrix assuming the existence of wakes other than the wake included in the hypothesis are used. For example, the tracks H1 and T4 are referred to in the hypothesis H1 described above. In the wake correlation matrices (1) to (5) in FIG. 7, since (2) and (3) assume the existence of the wake T2, they are not used for updating the hypothesis H1.

When H1 is updated by the wake correlation matrix (1) in FIG. 7, this indicates that all the observation values at the sampling time t = 3 are unnecessary signals. It becomes.
H11: {T81, T84}

Further, when H1 is updated by the wake correlation matrix (4) in FIG. 7, the following H12 is obtained using the updated wake.
H12: {T11, T41}
Furthermore, when H1 is updated by the wake correlation matrix (5) in FIG. 7, the following H13 is obtained using the updated wake.
H13: {T81, T72, T41}

  In this way, the trajectory evaluation unit 202 updates an existing hypothesis with a combination of a possible hypothesis and a wake correlation matrix, and generates a new hypothesis. These hypotheses generated every sampling period are also accumulated in the data storage unit 24.

  Next, the trajectory evaluation unit 202 executes a semi-optimization process for reducing hypothesis groups by deleting hypotheses with low reliability and integrating similar hypotheses, and further executes a cluster separation process associated therewith (step S307). ).

Here, there are various methods for the semi-optimization processing, and the following methods are known.
Method 1) A threshold is set for the reliability, and all hypotheses having a reliability less than that are deleted.
Method 2) An upper limit is set for the number of hypotheses, only the set number of hypotheses are left in order of higher reliability, and others are deleted.
Method 3) The hypotheses having the same correlation contents of observation values for the past N times are integrated.

When executing the cluster separation process, the trajectory evaluation unit 202 checks the sharing of observation values between wakes as follows. In a cluster, as a general rule, all wakes that share some of the observation values that make up a wake must belong to the same cluster. For example,
Ta: 11-20-30
Tb: 21-32
Tc: 20-32
When three wakes exist, Ta and Tc sharing the observation value 20 and Tb and Tc sharing the observation value 32 constitute the same cluster.

  In this situation, considering the case where the wake Tc is deleted by the quasi-optimization, the remaining two wakes Ta and Tb do not share the observation value, and thus can form a cluster independently.

Next, the trajectory evaluation unit 202 performs a main wake determination process that promotes a tentative wake that is very likely to be tracking a target from the tentative wake group to the main wake. (Step S308). Examples of the condition for determining that “the possibility of tracking the target is extremely high” include the following conditions.
Condition) The wake is included in all remaining hypotheses.
Through a series of processes as shown in steps S304 to S308, the trajectory evaluation unit 202 can generate a hypothesis and determine the main wake based on the wake generated by the trajectory estimation unit 201.

  Next, processing in the hypothesis limiting unit 203 will be described. The hypothesis limiting unit 203 selects the hypothesis group generated by the trajectory evaluation unit 202 (step S309). If it is determined that the target is singular based on the single determination result acquired in step S301, a hypothesis consisting only of the track by updating the existing main track is left, and the rest is deleted. Further, when it is determined that there are a plurality of targets based on the single determination result acquired in step S301, the hypothesis limiting unit 203 leaves a track by updating the existing main track and corresponds to a separation target. Update hypotheses consisting of temporary tracks or hypotheses consisting of new targets. Further, one of the obtained observation values is set as an unnecessary signal, and the hypothesis that the target number is regarded as singular is deleted.

  In the above-described hypothesis limiting process in step S309, the case where the single determination result at the current sampling time is used has been described. However, the single determination result at the past sampling time is further used, and several samples are continuously obtained. The hypothesis limiting process can also be performed by determining that the target is a single target or a plurality of targets. As a result, the reliability of the determination processing as a plurality of targets can be improved.

  FIG. 8 is a flowchart of the overall processing of the target tracking device according to the first embodiment of the present invention, and shows the overall processing that connects the flowcharts of FIGS. 2 and 3. Steps S801 to S806 in FIG. 8 correspond to steps S201 to S206 in FIG. 2 illustrating the processing of the observation information extraction unit 100.

  Step S807 in FIG. 8 collectively shows steps S301 to S303 in FIG. 3 showing the processing of the trajectory estimation unit 201. Step S808 in FIG. 8 collectively shows steps S304 to S308 in FIG. 3 showing the processing of the trajectory evaluation unit 202. Further, step S809 in FIG. 8 corresponds to step S309 in FIG. 3 showing the processing of the hypothesis limiting unit 203.

  The processing content of FIG. 8 is the same as that of FIG. 2 and FIG. By repeating the series of operations in steps S801 to S809 in FIG. 8 at predetermined sampling times, target tracking processing based on observation value information that changes from time to time is performed.

  According to the first embodiment, based on the target number single determination result determined by the observation information extraction unit, the tracking processing unit performs a hypothesis limiting process to make a reliable hypothesis without creating an incorrect hypothesis. Therefore, it is possible to obtain a target tracking device that can be limited to only a highly-priced track and can achieve improvement in tracking accuracy and reduction in calculation load.

Embodiment 2. FIG.
FIG. 9 is a configuration diagram of the target tracking device according to Embodiment 2 of the present invention. Compared with the configuration of FIG. 1 in the first embodiment, the configuration of FIG. 9 in the second embodiment uses a hypothesis reliability correction unit 205 instead of the hypothesis limitation unit 203. The details of the processing will be described focusing on this difference.

  The hypothesis reliability correction unit 205 in the tracking processing unit 200 has a function of correcting the hypothesis reliability and limiting the hypothesis according to the noise amplitude distribution. This specific process will be described with reference to the flowchart of FIG. FIG. 10 is a flowchart of processing of the entire target tracking device according to Embodiment 2 of the present invention. A description will be given mainly of processing different from steps S801 to S809 in FIG.

  Steps S1001 to S1006 corresponding to the processing executed by the observation information extraction unit 100 are the same as steps S801 to 806 of the first embodiment, and the observation information extraction unit 100 is based on the observation signal read for each sampling period. Observation value information is generated and output to the tracking processing unit 200. However, the observation result output unit 105 generates data including the threshold used by the single / multiple determination unit 102 in the single / multiple determination process as the observation value information, and outputs the data to the tracking processing unit 200.

  Next, as in steps S807 and 808 of the first embodiment, the trajectory estimation unit 201 in the tracking processing unit 200 performs trajectory estimation (step S1007), and the trajectory evaluation unit 202 further performs trajectory evaluation (step S1008). )I do. Next, the hypothesis reliability correction unit 205 selects a hypothesis group generated by the trajectory evaluation unit 202 based on a single determination result and a threshold value included in the observation information received from the observation information extraction unit 100. (Step S1009).

  As in the first embodiment, the single / multiple determination unit 102 performs single / multiple determination using the following determination formula.

On the other hand, the data storage unit 204 uses the probability P _false that the determination formula exceeds the threshold value Nth even if the target is singular (that is, the target value is singular using the threshold value Nth). Regardless of the probability of misjudging as a plurality, regardless of the threshold), the data is associated in advance with the threshold value. The probability P _false associated with this threshold value can be uniquely defined in advance from the distribution of the noise components of the observation signal determined in conjunction with the threshold value.

Further, the data storage unit 204 has in advance the target detection probability P _TARGET at the threshold value Nth as data associated with the threshold value. The probability P _TARGET associated with this threshold value can also be uniquely defined in advance from the distribution of the noise components of the observation signal determined in conjunction with the threshold value.

Then, the hypothesis reliability correction unit 205 extracts the probability P _false and the probability P _TARGET corresponding to the threshold received as part of the observation information from the data storage unit 204, and the reliability of the hypothesis generated by the trajectory evaluation unit 202 _Is weighted with a probability P _false and a probability P _TARGET .

The hypothesis reliability correction unit 205 performs the following process for each hypothesis when it is determined that there are a plurality of targets as a single determination result. The hypothesis that one of a plurality of adjacent observed values is assumed to be the already tracked target and the other unnecessary signal is regarded as an unnecessary signal is updated by multiplying its reliability by P _false . In addition, a hypothesis that one of a plurality of adjacent observed values is assumed to be a tracked target and the other is regarded as a target signal generated separately is updated by multiplying its reliability by P _TARGET .

  According to the second embodiment, in the hypothesis generation, even when there is a possibility that the target separation is erroneously determined due to the high noise level, the reliability of the hypothesis is weighted with the probability according to the level of the noise amplitude. Therefore, it becomes possible to limit the hypotheses. As a result, an accurate target separation determination can be performed, and a target tracking device that can achieve improvement in tracking accuracy and reduction in computation load can be obtained.

Embodiment 3 FIG.
FIG. 11 is a configuration diagram of the target tracking device according to Embodiment 3 of the present invention. Compared with the configuration of FIG. 9 in the second embodiment, the configuration of FIG. 11 in the third embodiment further includes a hypothesis evaluation unit 300. Further, the singular / multiple determination unit 102 is characterized by performing single / multiple determination in the next sampling period using a threshold value calculated at a predetermined sampling time by the hypothesis evaluation unit 300. The details of the processing will be described focusing on these differences.

  The hypothesis evaluation unit 300 includes a target number estimation unit 301 and a threshold value determination unit 302. The target number estimating unit 301 calculates an expected value of the target number based on the hypothesis calculated by the hypothesis limiting unit 203. Further, the threshold value determination unit 302 adjusts the determination threshold value used for single / multiple determination based on the expected value of the target number estimated by the target number estimation unit 301, and calculates a new threshold value.

  On the other hand, the single / multiple determination unit 102 in the observation information extraction unit 100 adopts the calculated new threshold value and performs single / multiple determination processing in the next sampling period. This specific process will be described with reference to the flowchart of FIG. FIG. 12 is a flowchart of processing of the entire target tracking device according to Embodiment 3 of the present invention. A series of processing from step S1201 to step S1209 is basically the same as steps S801 to S809 of the first embodiment, and steps S1210 and S1211 are added as new processing. This different process will be mainly described.

By a series of processing from step S1201 to step S1209, selection of hypothesis groups at the current sampling time is performed. Next, the target number estimation unit 301 in the hypothesis evaluation unit 300 calculates an expected value of the target number according to the following procedure based on the hypothesis calculated by the hypothesis limiting unit 203. Target-count-estimation unit 301, a reliability rel _i for each hypothesis, the track number _{n i} of each hypothesis, calculates the expected value _{N TARGET} target speed according to the following equation (step S1210).

Next, the threshold determination unit 302 adjusts the determination threshold used by the single / multiple determination unit 102 for the single determination process based on the calculated target number expected value N _TARGET (step S1211). Specifically, the threshold value determination unit 302 reduces the threshold value by a predetermined amount when the target number expectation value N _TARGET is smaller than the target number expectation value calculated at the previous sampling time. The threshold value is adjusted. In addition, the threshold value determination unit 302 performs adjustment to return to the original threshold value when the decrease in the target number expected value N _TARGET is suppressed.

  The single / multiple determination unit 102 in the observation information extraction unit 100 employs the threshold newly calculated by the threshold determination unit 302 in the single determination process at the next sampling time. Thus, by adjusting the threshold value for each sampling period, it is possible to easily detect a plurality of targets.

  According to the third embodiment, the function of the hypothesis evaluation unit can change the single determination threshold for the target number of the observation information extraction unit according to the target number estimated by the tracking processing unit, and changes for each sampling period. Based on the detection result, the single / multiple determination can be performed more accurately.

  In addition, although the threshold value determination part demonstrated the case where a threshold value was changed based on comparison of a target number, it is also possible to change a threshold value based on the distance between targets. For example, when a plurality of targets having a short distance between targets cannot be detected, the plurality of targets can be easily detected by lowering the threshold value.

Embodiment 4 FIG.
FIG. 13 is a configuration diagram of the target tracking device according to Embodiment 4 of the present invention. Compared with the configuration of FIG. 11 in the third embodiment, the configuration of FIG. 13 in the fourth embodiment has a configuration in which the hypothesis limiting unit 203 is eliminated. Even in such a configuration, it is possible to detect the target while changing the single determination threshold for each sampling period.

  The hypothesis evaluation unit 300 determines a new threshold based on the hypothesis evaluated by the trajectory evaluation unit 202 without performing the hypothesis limitation process by the hypothesis limitation unit 203. FIG. 14 is a flowchart of processing of the entire target tracking device according to Embodiment 4 of the present invention. The series of processes in steps S1401 to S1410 in FIG. 14 is obtained by removing the hypothesis limiting process in step S1209 from the series of processes in steps S1201 to S1211 in FIG.

  Further, the threshold value determination unit 302 can perform the following new threshold value determination process in the threshold value determination process step in step S1211 of FIG. 12 or step S1410 of FIG.

Specifically, the threshold value determination unit 302 reduces the threshold value by a predetermined amount when the target number expectation value N _TARGET is smaller than the target number expectation value calculated at the previous sampling time. The threshold value is adjusted. This facilitates detection of multiple targets. Further, when there is a track of the separation target in all hypotheses maintained by the tracking processing unit 200, the determination threshold used by the singular / multiple determination unit 102 is set to 0 so that it can be determined as a multiple target unconditionally. To do.

  According to the fourth embodiment, even in the configuration without the hypothesis limiting unit, the single determination threshold of the target number of the observation information extraction unit is changed according to the target number estimated by the tracking processing unit by the function of the hypothesis evaluation unit. Thus, the single / multiple determination can be performed more accurately based on the detection result that changes at each sampling period. Furthermore, when the existence of the separation target is recognized in the hypothesis maintained by the tracking processing unit, accurate observation value information can be obtained by determining as a plurality of targets unconditionally.

  Note that the process of making the threshold variable using the hypothesis evaluation unit can be similarly applied to the configuration of the second embodiment.

Embodiment 5. FIG.
FIG. 15 is a configuration diagram of the target tracking device according to the fifth embodiment of the present invention. Compared with the configuration of FIG. 13 in the fourth embodiment, the configuration of FIG. 15 in the fifth embodiment eliminates the target number estimation unit 301 and newly provides a signal intensity storage unit 106 in the observation information extraction unit 100. It has been. Even in such a configuration, it is possible to detect the target while changing the single determination threshold for each sampling period.

  In the tracking processing unit 200, the threshold determination unit 302 according to the fifth embodiment adjusts the determination threshold based on the observation value information in which the main track is correlated in the past. The signal intensity calculation itself is performed based on the received signal simultaneously with the single / multiple determination unit 102 by the single / multiple determination unit 102 in the observation information extraction unit 100 and stored in the signal strength storage unit 106. The threshold value determination unit 302 receives data related to observation values with which the main track has been correlated in the past from the tracking processing unit 200, and extracts signal strengths corresponding to these observation values from the signal strength storage unit 106. Further, signal intensities that do not correspond to these observed values are extracted from the signal intensity storage unit 106 as noise components. Based on these signal intensity data, the threshold value determination unit 302 determines the threshold value used by the single / multiple determination unit 102 using the following equation.

Here, n _Δ bar, which is a numerator on the right side, is an average value of noise, and Σ _m bar, which is a numerator on the right side, is an average value of signal strengths of observation values whose wakes are correlated in the past. This equation is based on the fact that the following holds approximately when considering noise in the discriminant of the target single / multiple determination.

Here, n _Σ and n _Δ are sum beam and difference beam errors caused by noise.

  FIG. 16 is a flowchart of processing of the entire target tracking device according to the fifth embodiment of the present invention. The processing from step S1601 to S1608 in FIG. 16 is basically the same as the processing from step S1401 to S1408 in FIG. However, the single / multiple determination unit 102 is different from the single / multiple determination unit 102 in step S1603 in that the signal strength is further calculated and stored in the signal strength storage unit 106. The threshold value determination unit 302 calculates a new threshold value for each sampling period according to the above-described equation using the average value of noise and the average value of the signal intensity of the observation values with which the track is correlated in the past (step S1609).

  On the other hand, in the singular / multiple determination process of step S1603, the singular / multiple determination unit 102 executes threshold processing at the next sampling time using the new threshold calculated by the threshold determination unit 302 in step 1609.

  According to the fifth embodiment, the threshold value for target single / multiple determination is calculated from the signal intensity of the correlated observation value by tracking, so that single / multiple determination can be performed more accurately.

It is a block diagram of the target tracking apparatus in Embodiment 1 of this invention. It is a flowchart of the observation value extraction process in Embodiment 1 of this invention. It is a flowchart of the target tracking process in Embodiment 1 of this invention. It is a figure which shows the state of the existing wake in Embodiment 1 of this invention. It is a figure which shows the relationship between the existing track and the latest observed value in Embodiment 1 of this invention. It is the figure which showed the determination matrix in a gate in Embodiment 1 of this invention. It is the figure which showed the wake correlation matrix in Embodiment 1 of this invention. It is a flowchart of the process of the whole target tracking apparatus in Embodiment 1 of this invention. It is a block diagram of the target tracking apparatus in Embodiment 2 of this invention. It is a flowchart of the process of the whole target tracking apparatus in Embodiment 2 of this invention. It is a block diagram of the target tracking apparatus in Embodiment 3 of this invention. It is a flowchart of the process of the whole target tracking apparatus in Embodiment 3 of this invention. It is a block diagram of the target tracking apparatus in Embodiment 4 of this invention. It is a flowchart of the process of the whole target tracking apparatus in Embodiment 4 of this invention. It is a block diagram of the target tracking apparatus in Embodiment 5 of this invention. It is a flowchart of the process of the whole target tracking apparatus in Embodiment 5 of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 Observation information extraction part, 101 Distance information calculation part, 102 Single number determination part, 103 Single angle information calculation part, 104 Multiple angle information calculation part, 105 Observation result output part, 106 Signal strength memory | storage part, 200 Tracking process part, 201 Trajectory estimation unit, 202 Trajectory evaluation unit, 203 Hypothesis limitation unit, 204 Data storage unit, 205 Hypothesis reliability correction unit, 300 Hypothesis evaluation unit, 301 Target number estimation unit, 302 Threshold determination unit

Claims (6)

An observation information extraction unit that generates observation value information from received signals for tracking targets from a plurality of antenna elements;
A target tracking device for tracking a specific target, comprising: a tracking processing unit that updates a target track based on the observation value information;
The observation information extraction unit
A distance information calculation unit for extracting target distance information based on received signals for tracking targets from a plurality of antenna elements;
A singular for comparing whether the ratio of the difference signal to the sum signal of the received signals for each antenna element is equal to or greater than a predetermined threshold and determining whether the target is singular or plural in the adjacent observation region A multiple determination unit;
A single angle information calculation unit for calculating angle information of the target with respect to the target determined to be single by the single plural determination unit;
A plurality of angle information calculating unit for calculating angle information of a plurality of adjacent targets with respect to the target determined to be plural by the single plural determining unit;
The distance information calculated by the distance information calculation unit, the single / multiple determination result determined by the single / multiple determination unit, and the angle information calculated by the single angle information calculation unit and the multiple angle information calculation unit. An observation result output unit that generates and outputs observation value information, and
The tracking processing unit
A data storage unit for accumulating data relating to the target track updated every predetermined sampling period;
Based on the observation value information from the observation information extraction unit and data on the target track updated at the previous sampling time stored in the data storage unit, data on the target track at the current sampling time is generated A trajectory estimation unit that updates data related to the target track in the data storage unit;
A trajectory evaluation unit for generating a wake hypothesis by combining data on the target wake, and calculating the reliability of the generated hypothesis,
Have
The tracking processing unit determines a singular target when the single determination result is determined to be singular based on the single determination result determined by the single / multiple determination unit, and is generated by the trajectory evaluation unit. Only the wake hypothesis by updating the existing main wake is left out of the wake hypothesis, and when it is determined that the single determination result is plural, it is determined as a plural target, and the trajectory evaluation unit generates the It further comprises a hypothesis limiting unit that leaves a wake hypothesis by updating an existing main wake from among the wake hypotheses and a wake hypothesis corresponding to a new target separated from the existing main wake. Target tracking device. The target tracking device according to claim 1,
The hypothesis limiting unit determines that there are a plurality of targets when the state in which it is determined that there are a plurality of single determination results continuously occurs at a plurality of sampling periods, and the wake generated by the trajectory evaluation unit A target tracking device characterized by leaving a wake hypothesis by updating an existing main wake from among the hypotheses and a wake hypothesis corresponding to a new target separated from the existing main wake. An observation information extraction unit that generates observation value information from received signals for tracking targets from a plurality of antenna elements;
A target tracking device for tracking a specific target, comprising: a tracking processing unit that updates a target track based on the observation value information;
The observation information extraction unit
A distance information calculation unit for extracting target distance information based on received signals for tracking targets from a plurality of antenna elements;
A singular for comparing whether the ratio of the difference signal to the sum signal of the received signals for each antenna element is equal to or greater than a predetermined threshold and determining whether the target is singular or plural in the adjacent observation region A multiple determination unit;
A single angle information calculation unit for calculating angle information of the target with respect to the target determined to be single by the single plural determination unit;
A plurality of angle information calculating unit for calculating angle information of a plurality of adjacent targets with respect to the target determined to be plural by the single plural determining unit;
The distance information calculated by the distance information calculation unit, the single / multiple determination result determined by the single / multiple determination unit, and the angle information calculated by the single angle information calculation unit and the multiple angle information calculation unit. An observation result output unit that generates and outputs observation value information, and
The tracking processing unit
A data storage unit for accumulating data relating to the target track updated every predetermined sampling period;
Based on the observation value information from the observation information extraction unit and data on the target track updated at the previous sampling time stored in the data storage unit, data on the target track at the current sampling time is generated A trajectory estimation unit that updates data related to the target track in the data storage unit;
A trajectory evaluation unit for generating a wake hypothesis by combining data on the target wake, and calculating the reliability of the generated hypothesis,
Have
The observation result output unit generates and outputs observation value information further including a threshold value used for single determination by the single plural determination unit,
The data storage unit further includes data associating a probability that a target is erroneously determined as a plurality in a single determination process and a threshold, and data associating a detection probability of the target with a threshold,
The tracking processing unit determines the reliability of the hypothesis generated by the trajectory evaluation unit based on the single determination result determined by the single / multiple determination unit and the threshold value used for single / multiple determination by the single / multiple determination unit. On the other hand, the target tracking device further comprising a hypothesis reliability correction unit that corrects the reliability by multiplying the probability of erroneous determination or the detection probability. In the target tracking device according to any one of claims 1 to 3,
A hypothesis evaluation unit that calculates a new threshold value for each predetermined sampling period;
The hypothesis evaluation unit
A target number estimation unit that calculates an expected value of the target number based on the number of hypotheses generated by the tracking processing unit and the respective reliability levels;
Threshold determination for calculating a new threshold value that is changed by a predetermined amount with respect to the threshold value used in the singular / multiple determination unit based on a change in the expected value of the target number calculated for each predetermined sampling period And
The single / multiple determination unit performs single / multiple determination processing using the new threshold value calculated by the threshold value determination unit. The target tracking device according to claim 4,
The threshold determination unit sets a new threshold as 0 when there is a separation target track in all hypotheses maintained by the tracking processing unit,
The singular / multiple determination unit unconditionally performs a plurality of determination processes when the new threshold is set to 0 by the threshold determination unit. In the target tracking device according to any one of claims 1 to 3,
A hypothesis evaluation unit that calculates a new threshold value for each predetermined sampling period;
The observation information extraction unit further includes a signal strength storage unit that stores the signal strength of the received signal,
The hypothesis evaluation unit receives data related to the observed value with which the main track has been correlated in the past from the tracking processing unit, extracts the signal intensity corresponding to the observed value from the signal intensity storage unit, and based on the extracted signal intensity A threshold value determination unit for calculating a new threshold value,
The single / multiple determination unit calculates a signal strength of a received signal for each predetermined sampling period and stores the signal strength in the signal strength storage unit, and uses the new threshold value calculated by the threshold value determination unit. The target tracking device characterized by performing.

Images