JPWO2019167268A1 - Target monitoring device and target monitoring system - Google Patents

Abstract

The target monitoring device estimates the state of the target based on a plurality of reception signals obtained by a plurality of reception antennas that respectively receive radio signals that have arrived from the target via a plurality of communication paths. The target monitoring device detects a movement of a target by monitoring a arrival time difference or arrival frequency difference included in the plurality of measurement values by a measurement unit (61) that sequentially calculates a plurality of measurement values based on the plurality of received signals. A state detection unit (62A) for performing the positioning calculation and a calculation unit (63A, 64A, 65) for performing positioning calculation and tracking calculation. When it is detected that there is no movement of the target, the calculation units (63A, 64A, 65) execute positioning calculation to calculate a state estimation value indicating the estimated position of the target, and move from the stationary state of the target to the moving state. When a change to is detected, a tracking calculation is executed to successively calculate a state estimation value indicating both the estimated position and the moving state of the target.

Description

  The present invention relates to a technique for estimating a state such as a position of a radio wave transmission source based on a radio wave received from a radio wave transmission source via a plurality of satellites.

  There is known a technique for estimating the position of a radio wave transmission source using a plurality of satellites. For example, Non-Patent Document 1 below describes a positioning method based on a time-of-arrival (Time-Differences-Of-Arrival, TDOAs) between radio waves received from a radio wave source via a plurality of satellites, and a method of receiving the received radio waves. And a positioning method based on an arrival frequency difference (Frequency-Differences-Of-Arrivals, FDOAs) and a time-of-arrival difference (TDOAs) indicating the Doppler frequency difference between the transmitted radio waves.

  In a target monitoring system that employs such a positioning method, a radio wave transmitted from a target as a radio wave transmission source passes through transponders (Transponders) mounted on a plurality of satellites, respectively, and then transmits to a plurality of ground stations. Received by the receiving antenna. The positioning calculator measures the arrival time difference (TDOAs) between the received radio waves obtained by the plurality of receiving antennas, or measures the arrival frequency difference (FDOAs) and the arrival time difference (TDOAs) between the received radio waves. Then, based on the measured arrival time difference, or the measured arrival frequency difference and arrival time difference, the positioning calculator calculates the position of the radio wave source under the assumption that the radio wave source exists on the earth's surface. Can be estimated. Since the orbital positions of the plurality of satellites are different from each other, a difference in propagation time occurs between the received radio waves. This makes it possible to measure the time difference of arrival (TDOAs) between the received radio waves. Further, even if the plurality of satellites are geostationary satellites, since the satellites slightly move with respect to the ground surface, a difference in Doppler frequency occurs between the received radio waves. As a result, it is possible to measure the arrival frequency difference (FDOAs) between the received radio waves.

K. C. Ho and Y. T. Chan, "Geolocation of a known altitude object from TDOA and FDOA measurements," IEEE Transactions on Aerospace and Electronic Systems, Vol. 33, Issue 3, pp. 770-783, 1997.

  When monitoring the state of a movable radio wave transmission source (for example, a ship), in the conventional target monitoring system, the target radio wave transmission source is in a stationary state (in the case of a ship, the ship is anchored). The monitoring may be started from the state (state). At this time, when the target is in a stationary state, a positioning operation for estimating the instantaneous position of the target is executed, and after the target starts moving, a tracking operation for tracking the target is executed. When tracking a moving target, it is necessary to estimate not only the position of the moving target that changes every moment, but also the moving speed at each time. When the state of the target changes from the stationary state to the moving state, it is necessary to switch the positioning calculation to the tracking calculation. However, in the conventional target monitoring system, it is not possible to accurately determine the timing at which the positioning calculation should be switched to the tracking calculation in accordance with the change, and the tracking of the moving target may fail.

  In view of the above, it is an object of the present invention to accurately determine a timing at which the positioning calculation should be switched to the tracking calculation in accordance with a change in the state of a target which is a radio wave source, and to achieve a high tracking accuracy with respect to the target. The object is to provide a monitoring device and a target monitoring system.

  A target monitoring device according to one embodiment of the present invention provides a target monitoring device based on a plurality of reception signals obtained by a plurality of reception antennas each receiving a radio signal arriving from a target that is a radio wave transmission source via three or more communication paths. A target monitoring apparatus for estimating a state of a target, wherein a plurality of measurement values indicating an arrival time difference between the reception signals or an arrival time difference and an arrival frequency difference between the reception signals are sequentially calculated based on the plurality of reception signals. A measuring unit that monitors the time variation of at least one of the plurality of measured values to detect the presence or absence of the movement of the target, and the target detecting unit detects no movement of the target. When detecting that the plurality of measured values, the computing unit performs a positioning operation using the plurality of measurement values to calculate a first state estimated value indicating the estimated position of the target, the computing unit, Condition When a state change of the target from the stationary state to the moving state is detected by the detection unit, a tracking operation is performed to calculate a second state estimated value indicating both the estimated position and the moving state of the target. It is characterized by.

  ADVANTAGE OF THE INVENTION According to this invention, the timing which should switch a positioning calculation to a tracking calculation according to a state change from a stationary state of a target to a moving state can be determined with high accuracy. Therefore, high tracking accuracy for the target can be ensured.

1 is a diagram schematically showing a configuration of a target monitoring system according to a first embodiment of the present invention. FIG. 3 is a block diagram illustrating a configuration example of a receiver according to the first embodiment. FIG. 3 is a block diagram illustrating a schematic configuration of a monitoring processing unit according to the first embodiment. FIG. 3 is a block diagram schematically illustrating a configuration example of a state detection unit according to the first embodiment. FIG. 5A and FIG. 5B are graphs illustrating an example of the time variation of the smoothed value. It is a flowchart which shows an example of the procedure of the iterative estimation process by a Kalman filter roughly. FIG. 3 is a block diagram schematically illustrating a configuration of a calculation unit including a positioning calculation unit, a tracking calculation unit, and a data transfer unit according to the first embodiment. FIG. 3 is a block diagram illustrating an example of a hardware configuration that implements a function of a monitoring processing unit. FIG. 13 is a block diagram schematically showing a configuration of a calculation unit including a positioning calculation unit, a tracking calculation unit, and a data transfer unit according to a second embodiment of the present invention. FIG. 13 is a block diagram illustrating a schematic configuration of a monitoring processing unit according to a third embodiment of the present invention. FIG. 14 is a block diagram schematically showing a configuration of a calculation unit including a positioning calculation unit, a tracking calculation unit, and a data transfer unit according to Embodiment 3. FIG. 14 is a block diagram schematically showing a configuration of a calculation unit including a positioning calculation unit, a tracking calculation unit, and a data transfer unit according to a fourth embodiment of the present invention. FIG. 17 is a block diagram illustrating a schematic configuration of a monitoring processing unit according to a fifth embodiment of the present invention. 26 is a block diagram illustrating a schematic configuration of a state detection unit according to Embodiment 5. FIG. FIG. 21 is a block diagram illustrating a schematic configuration of a monitoring processing unit according to a sixth embodiment of the present invention. FIG. 21 is a block diagram illustrating a schematic configuration of a state detection unit according to a sixth embodiment.

  Hereinafter, various embodiments according to the present invention will be described in detail with reference to the drawings. Note that components denoted by the same reference numerals throughout the drawings have the same configuration and the same function.

Embodiment 1 FIG.
FIG. 1 is a diagram schematically showing a configuration of a target monitoring system 1 according to a first embodiment of the present invention. As shown in FIG. 1, the target monitoring system 1 receives radio signals arriving via a plurality of communication paths from a target Tgt, which is a radio wave transmission source in a monitoring area SA, and outputs received signals S1, S2, and S3. And a target monitoring device 2 that estimates the state of the target Tgt based on the received signals S1, S2, and S3. In the present embodiment, three artificial satellites St1, St2, St3 (hereinafter simply referred to as “satellite St1, St2, St3”) are respectively provided on three communication paths between the receiving antennas Rx1, Rx2, Rx3 and the target Tgt. .) Exists. Transponders (repeaters) mounted on the satellites St1, St2, St3 receive the radio signals transmitted from the target Tgt, amplify the received radio signals to generate downlink signals, and generate these downlink signals. To the receiving antennas Rx1, Rx2, Rx3 of the target monitoring system 1 respectively. The satellites St1, St2, and St2 are assumed to be geostationary satellites put into geostationary orbit. The target monitoring system according to the present embodiment is not limited to the three satellites St1, St2, and St3, and can receive downlink signals from two or N satellites (N is an integer of 4 or more). 1 may be changed.

  The target Tgt is a movable radio wave transmission source (a moving object such as a ship, an aircraft, or a vehicle), and has a function of transmitting a radio signal in a specific frequency band to a satellite. However, a radio signal transmitted from the target Tgt may interfere with a transmission radio wave of a terrestrial wireless communication network or a satellite communication network to cause a communication failure. In such a case, there is a need to specify the position of a target Tgt for transmitting a radio signal serving as an interference wave. There is also a need to specify the position of the target Tgt based on a search and rescue beacon signal transmitted from the target Tgt. The target monitoring device 2 of the present embodiment can detect whether or not the target Tgt is moving, that is, whether the target Tgt is in a stationary state or a moving state, based on the received signals S1 to S3. When the target Tgt is in the stationary state, the target monitoring device 2 can sequentially calculate a state estimation value indicating the estimated position (stationary position) of the target Tgt. When the target Tgt is in the moving state, the target monitoring device 2 2 can sequentially calculate an estimated position (tracking position) and a state estimated value indicating a moving state (for example, speed) of the target Tgt.

  As shown in FIG. 1, the target monitoring device 2 converts a reception signal S1, S2, S3 in a high frequency band into digital reception signals D1, D2, D3, and a digital reception signal D1, D2, D3. The monitoring processing unit 11A sequentially calculates a state estimated value indicating the state of the target Tgt based on the calculated state estimated value, and the display unit 12 displays image information based on the calculated state estimated value.

  FIG. 2 is a block diagram illustrating a configuration example of the receiver 10 according to the first embodiment. The receiver 10 shown in FIG. 2 includes a local oscillator 20 that outputs a local oscillation signal for frequency conversion, and band-pass filters 31, 41, and 51 that perform filter processing on received signals S1, S2, and S3 in a high-frequency band, respectively. Down converters 32, 42, and 52 for performing frequency conversion on high-frequency outputs F1, F2, and F3 of band-pass filters 31, 41, and 51 using local oscillation signals, respectively, and A / D converters (ADCs) 33 and 43; 53. The down converters 32, 42, and 52 are frequency converters that mix the high-frequency outputs F1, F2, and F3 with local oscillation signals and generate analog signals C1, C2, and C3 in a lower frequency band. The ADCs 33, 43, and 53 convert the analog signals C1, C2, and C3 into digital reception signals D1, D2, and D3, respectively, and output the digital reception signals D1, D2, and D3 to the monitoring processing unit 11A. Each of digital reception signals D1, D2, and D3 is a complex signal including an amplitude component and a phase component.

FIG. 3 is a block diagram illustrating a schematic configuration of the monitoring processing unit 11A according to the first embodiment. The monitoring processing unit 11A shown in FIG. 3 includes a measurement unit 61 that sequentially calculates the measured values of the arrival time difference and the arrival frequency difference of the radio wave based on the digital reception signals D1, D2, and D3, and a monitoring unit based on one of these measurement values. A state detection unit 62A that detects the presence or absence of movement of the target Tgt, and a state that indicates the estimated position of the target Tgt by executing positioning calculation using the measured value when it is detected that there is no movement of the target Tgt. A positioning calculation unit 63A that sequentially calculates an estimated value, and when it is detected that there is a movement of the target Tgt, a tracking calculation using the measured value is executed to estimate an estimated position and a moving state (for example, estimated speed) of the target Tgt. And a data transfer for transferring the calculation result data from one of the positioning calculation unit 63A and the tracking calculation unit 64A to the other. 65, a data storage unit 67 for storing data (satellite orbit information OD, monitoring area information AD, and target swing information SD) necessary for processing in the state detection unit 62A, and an external server (not shown). A data acquisition unit 68 for acquiring data. Here, k is an integer that indicates the time t _k.

  Further, the monitoring processing unit 11A includes an output control unit 66 that generates image information DD based on the state estimation value calculated by the positioning calculation unit 63A and the tracking calculation unit 64A. The output control unit 66 outputs the image information DD to the display 12 such as a liquid crystal display or an organic EL display. As the image information DD, for example, alphanumeric information indicating the coordinate value of the estimated position of the target Tgt, map information visually indicating the estimated position, alphanumeric information indicating the value of the estimated speed of the target Tgt, and the estimation thereof There is a graph that visually represents the transition of speed. The user can grasp the state of the target Tgt by looking at the image information DD displayed on the display 12.

Based on the digital reception signals D1, D2, and D3, the measurement unit 61 calculates the arrival time difference TDOA ₁₂ (k) between the reception signals S1 and S2, the arrival time difference TDOA ₂₃ (k) between the reception signals S2 and S3, and the reception signal S3. The arrival time difference TDOA ₃₁ (k) between S1, the arrival frequency difference FDOA ₁₂ (k) between the reception signals S1 and S2, the arrival frequency difference FDOA ₂₃ (k) between the reception signals S2 and S3, and the reception signals S3 and S1 It has a function of sequentially calculating the measured value of the incoming frequency difference FDOA ₃₁ (k). As a specific method of calculating the measurement value, a known method may be employed, and is not particularly limited. For example, a two-dimensional cross-correlation between the digital reception signals Di and Dj in the time direction and the frequency direction is calculated, a peak appearing in the calculated cross-correlation is detected, and an arrival time difference TDOA _ij (k) corresponding to the peak is calculated. And a calculation method of obtaining a set of arrival frequencies FDOA _ij (k). Here, i and j are integers in the range of 1 to 3 (i ≠ j).

The measurement unit 61 also measures the arrival time difference TDOA ₁₂ (k), TDOA ₂₃ (k), TDOA ₃₁ (k) and the arrival frequency difference FDOA ₁₂ (k), FDOA ₂₃ (k), and FDOA ₃₁ (k). Is supplied to the state detection unit 62A, the positioning calculation unit 63A, and the tracking calculation unit 64A.

  The state detection unit 62A monitors at least one of the measurement values sequentially supplied from the measurement unit 61, and constantly monitors the time variation of the measurement value of the monitoring target to detect whether or not the target Tgt has moved. The detection signal ES indicating the detection result is supplied to the positioning calculation unit 63A, the tracking calculation unit 64A, and the data transfer unit 65.

FIG. 4 is a block diagram schematically illustrating a configuration example of the state detection unit 62A according to the first embodiment. First, the configuration and operation of the state detection unit 62A when the arrival time difference TDOA _ij (k) is selected as a monitoring target will be described.

The state detection unit 62A illustrated in FIG. 4 sets a set of a smoothed value calculation unit 81 that performs a filtering process on the time-series data of the measurement values to be monitored to calculate a smoothed value <TDOA> _k , and thresholds TH1 and TH2. The threshold value setting unit 82A and the smoothed value <TDOA> _k are compared with a numerical range Δ (TH1, TH2) having the lower limit of the threshold value TH1 and the upper limit of the threshold value TH2 to determine whether the state of the target Tgt is the stationary state or the moving state. And a state determination unit 83 that determines which one of them is used. Here, k is an integer that indicates the time t _k. The state determination unit 83 provides the determination output unit 89 with a determination result ES1 indicating whether the state of the target Tgt is a stationary state or a moving state. For example, the N arrival time differences TDOA _ij (k−N + 1), TDOA _ij (k−N + 2),..., TDOA _ij (k−1), and TDOA _ij (k) are subjected to filtering (for example, averaging). by the applied, it is possible to calculate the smoothed value _{<TDOA> k} at time _{t k.} The value of the arrival time difference TDOA _ij (k) has a variation due to a measurement error. By using the smoothed value <TDOA> _k , it is possible to reduce the influence of such variations.

State determining unit 83 constantly monitors the time variation of smoothed value _{<TDOA> k,} when the smoothed value _{<TDOA> k} falls continuously for a period of time T1, the numerical range delta (TH1, TH2) in, It is determined that the target Tgt is in a stationary state. FIG. 5A is a graph illustrating an example of a temporal change of the smoothed value <TDOA> _k . As shown in the graph of FIG. 5A, (when t _{<t m)} when the target Tgt is stationary, smoothed value _{<TDOA> k} varies to fit within the numerical range between thresholds TH1, TH2 .

When the smoothed value <TDOA> _k changes from within the numerical range Δ (TH1, TH2) to outside the numerical range Δ (TH1, TH2), the state determination unit 83 determines that the state change of the target Tgt from the stationary state to the moving state has occurred. It is determined that it has occurred. In the example of FIG. 5A, smoothed value _{<TDOA> k,} since fluctuates to exceed the threshold value TH2 of the upper at time _{t m,} the state determination section 83, at time _{t m} and the state change of the target Tgt occurs Can be determined.

Regarding the threshold values TH1 and TH2, the width of the numerical range Δ (TH1, TH2) is set as small as possible so that a change in the smoothed value <TDOA> _k due to a change in the state of the target Tgt from the stationary state to the moving state can be detected quickly. On the other hand, it is desirable to increase the width of the numerical range Δ (TH1, TH2) so as not to detect a change in the smoothed value <TDOA> _k due to a factor other than the state change. Factors other than the state change mainly include movements of the satellites Sti and Stj. Therefore, it is possible to set appropriate thresholds TH1 and TH2 in consideration of the fluctuation of the smoothed value <TDOA> _k due to the change in the state of the target Tgt and the fluctuation of the smoothed value <TDOA> _k due to the movement of the satellites Sti and Stj. desirable.

  The threshold setting unit 82A sets a set of thresholds TH1 and TH2 shown in the following state determination formulas (1) and (2) based on the satellite orbit information OD and the monitoring area information AD supplied from the data storage unit 67. be able to.

ここで、ＴＤＯＡ_ｍｉｎは、ｉ番目の衛星Ｓｔｉおよびｊ番目の衛星Ｓｔｊ（ｉ≠ｊ）の動きを考慮して設定された最小値であり、ＴＤＯＡ_ｍａｘは、衛星Ｓｔｉ，Ｓｔｊの動きを考慮して設定された最大値である。また、σ_ＴＤＯＡは、到来時間差の計測誤差の標準偏差である。３σ_ＴＤＯＡは、到来時間差の計測誤差を考慮して標準偏差の３倍相当のマージンを与えるものである。

Here, TDOA _min is a minimum value set in consideration of the movement of the i-th satellite Sti and the j-th satellite Stj (i ≠ j), and TDOA _max is considered in consideration of the movement of the satellites Sti and Stj. This is the maximum value set. Σ _TDOA is the standard deviation of the measurement error of the arrival time difference. The 3σ _TDOA gives a margin equivalent to three times the standard deviation in consideration of the measurement error of the arrival time difference.

When detecting that one of the state determination formulas (1) and (2) is satisfied, the state determination unit 83 determines that the state change of the target Tgt from the stationary state to the moving state has occurred. it can. The minimum value TDOA _min and the maximum value TDOA _max can be calculated using the satellite orbit information OD and the monitoring area information AD. The monitoring area information AD is coordinate information that defines the monitoring area SA shown in FIG. The satellite orbit information OD may be any orbit element information necessary for calculating the orbit of the satellites Sti and Stj, or an orbit calculation value calculated by an orbit calculation algorithm using the orbit element information.

For example, the threshold setting unit 82A determines that a radio wave transmission source exists at a plurality of points Ψ ₁ , Ψ ₂ ,..., _{Ｌ L} (L is a positive integer) within the monitoring area SA defined in advance by the monitoring area information AD. under the assumption, the satellite Sti, point [psi ₁ by using the latest trajectory calculation value of Stj, Ψ _2, ..., Ψ respectively _L corresponding TDOA _{TDOA (Ψ 1, k),} TDOA k (Ψ 2, k _{), ..., TDOA (Ψ L} , k) to predict the (k is an integer that indicates the time _{t k).} Threshold setting unit 82A is able to determine the minimum and maximum values of the predicted arrival time difference _{TDOA (Ψ 1, k) ~TDOA} (Ψ L, k) as the minimum value _{TDOA min} and a maximum value _{TDOA max} respectively it can. Since the orbits of the satellites Sti and Stj have a 24-hour period, the threshold setting unit 82A calculates the minimum value TDOA _min (k) and the maximum value TDOA _max (k) for 24 hours in advance using the latest orbit calculation value. That is enough. Also, the threshold setting unit 82A selects an arrival time difference close to the smoothed value <TDOA> _k from the predicted arrival time differences TDOA (Ψ ₁ , k) to TDOA (Ψ _L , k), and selects the selected arrival time difference. The minimum value TDOA _min and the maximum value TDOA _max may be determined based on the arrival time difference.

Next, when the state of the target Tgt changes from the moving state to the stationary state, the fluctuation range of the smoothed value <TDOA> _k falls within a certain range. In this case, when the state determination unit 83 detects that the smoothed value <TDOA> _k is continuously within the numerical range Δ (TH1, TH2) during the certain period T1, the state determination unit 83 stops moving from the moving state of the target Tgt. It is determined that a state change to a state has occurred. The time length of the fixed period T1 is a parameter setting value. FIG. 5B is a graph illustrating another example of the temporal variation of the smoothed value <TDOA> _k . As shown in the graph of FIG. 5B, the variation width of the smoothed value _{<TDOA> k} will have after the time _{t d} decreased, during time _{t d} ~ time _{t s,} smoothed value _{<TDOA> k} are numerical range delta (TH1 , TH2). In this case, the state determination unit 83 may determine that the state change from the moving state of the target Tgt the stationary state occurs at time t _s.

  More specifically, when the state determination unit 83 continuously detects that both of the following state determination formulas (3) and (4) are satisfied during the fixed period T1, the state determination unit 83 determines the moving state of the target Tgt. It can be determined that the state change from the state to the stationary state has occurred.

Next, the configuration and operation of the state detection unit 62A when the arrival frequency difference FDOA _ij (k) is selected as a monitoring target will be described with reference to FIG.

The state detection unit 62A shown in FIG. 4 performs a filtering process on the time-series data of the measurement values to be monitored to calculate a smoothed value <FDOA> _k , and a set of threshold values TH3 and TH4. The threshold value setting unit 86A to be set and the smoothed value <FDOA> _k are compared with a numerical range Δ (TH3, TH4) _where the threshold value TH3 is a lower limit and the threshold value TH4 is an upper limit, and the state of the target Tgt is a stationary state or a moving state. And a state determination unit 87 that determines which of Here, k is an integer that indicates the time t _k. For example, a filter processing (for example, averaging) is performed on the measured values of N arrival frequency differences FDOA _ij (k−N + 1), FDOA _ij (k−N + 2),..., FDOA _ij (k−1), and FDOA _ij (k). ) by the applied, it is possible to calculate the smoothed value _{<FDOA> k} at time _{t k.} The value of the arrival frequency difference FDOA _ij (k) has a variation due to a measurement error. By using the smoothed value <FDOA> _k , it is possible to reduce the influence of such variations.

State determination unit 87 constantly monitors the time variation of smoothed value _{<FDOA> k,} while smoothed value _{<FDOA> k} is constant period T2, the numerical range Δ when fit to continue (TH3, TH4) in, The target Tgt can be determined to be in a stationary state. Further, when the smoothed value <FDOA> _k changes from within the numerical range Δ (TH3, TH4) to outside the numerical range Δ (TH3, TH4), the state determination unit 87 sets the state of the target Tgt from the stationary state to the moving state. It can be determined that a change has occurred.

Regarding the threshold values TH3 and TH4, the width of the numerical range Δ (TH3, TH4) is set as small as possible so that the fluctuation of the smoothed value <FDOA> _k due to the change in the state of the target Tgt from the stationary state to the moving state can be detected quickly. On the other hand, it is desirable to increase the width of the numerical range Δ (TH3, TH4) so as not to detect the fluctuation of the smoothed value <FDOA> _k due to factors other than the state change. Factors other than the state change mainly include the movements of the satellites Sti and Stj and the Doppler effect due to fluctuations such as a change in the attitude of the target Tgt. For example, when the target Tgt is a ship, the ship may be shaken (for example, pitching and rolling) due to the influence of a wave or wind, and the Doppler effect resulting therefrom may occur. Therefore, a variation of smoothed values _{<FDOA> k} by the state change of the target Tgt, satellite Sti, the variation of the smoothed value _{<FDOA> k} due to the motion of Stj, the variation of the smoothed value _{<FDOA> k} according to upset the target Tgt In consideration of the above, it is desirable to set appropriate thresholds TH3 and TH4.

  The threshold value setting unit 86A determines the threshold values TH3 and TH3 shown in the following state determination formulas (5) and (6) based on the satellite orbit information OD, the monitoring area information AD and the target swing information SD supplied from the data storage unit 67. A set of TH4 can be set.

ここで、ＦＤＯＡ_ｍｉｎは、ｉ番目の衛星Ｓｔｉおよびｊ番目の衛星Ｓｔｊ（ｉ≠ｊ）の動きを考慮して設定された最小値であり、ＦＤＯＡ_ｍａｘは、衛星Ｓｔｉ，Ｓｔｊの動きを考慮して設定された最大値である。また、σ_ＦＤＯＡは、到来周波数差の計測誤差の標準偏差である。３σ_ＦＤＯＡは、到来周波数差の計測誤差を考慮して標準偏差の３倍相当のマージンを与えるものである。さらに、σ_{ｍｏｔｉｏｎ}は、目標揺動情報ＳＤに基づき、目標Ｔｇｔの動揺によるドップラー効果を考慮して設定された値であり、当該ドップラー効果に起因する到来周波数差の揺らぎを示す値である。たとえば、目標Ｔｇｔが船舶の場合、海洋上の多数の地点における波に関するデータが提供されている。データ取得部６８は、外部サーバ（図示せず）から、このような波に関するデータを目標揺動情報ＳＤとして取得することができるので、閾値設定部８６Ａは、目標揺動情報ＳＤに基づき、任意の地点における船舶の動揺の見積もりを実行して到来周波数差の揺らぎを示す値σ_{ｍｏｔｉｏｎ}を算出することができる。

Here, FDOA _min is a minimum value set in consideration of the movement of the i-th satellite Sti and the j-th satellite Stj (i ≠ j), and FDOA _max is set in consideration of the movement of the satellites Sti and Stj. This is the maximum value set. Further, σ _FDOA is a standard deviation of a measurement error of an arrival frequency difference. The 3σ _FDOA gives a margin equivalent to three times the standard deviation in consideration of the measurement error of the arrival frequency difference. Further, σ _motion is a value set in consideration of the Doppler effect due to the fluctuation of the target Tgt based on the target fluctuation information SD, and is a value indicating fluctuation of an incoming frequency difference caused by the Doppler effect. For example, if the target Tgt is a ship, data is provided on waves at a number of points on the ocean. The data acquisition unit 68 can acquire data on such a wave as the target swing information SD from an external server (not shown). The estimation of the motion of the ship at the point is performed, and the value σ _motion indicating the fluctuation of the arrival frequency difference can be calculated.

When detecting that one of the state determination formulas (5) and (6) is satisfied, the threshold value setting unit 86A may determine that the state change of the target Tgt from the stationary state to the moving state has occurred. it can. Similarly to the case where the minimum value TDOA _min and the maximum value TDOA _max are calculated, the minimum value FDOA _min and the maximum value FDOA _max can be calculated using the satellite orbit information OD and the monitoring area information AD. Further, the fluctuation σ _motion can be calculated based on the target fluctuation information SD including the dispersion value of the fluctuation speed of the target Tgt under the assumption that the position of the target Tgt and the positions of the satellites Sti and Stj are invariable. is there. Given the variance of the fluctuation speed of the target Tgt, the margin 3σ _motion can be calculated.

On the other hand, when the state change of the target Tgt from the moving state to the stationary state occurs, the fluctuation range of the smoothed value <FDOA> _k falls within a certain range. In this case, when the state determination unit 87 detects that the smoothed value <FDOA> _k is continuously within the numerical range Δ (TH3, TH4) during the certain period T2, the state determination unit 87 stops moving from the moving state of the target Tgt. It is determined that a state change to a state has occurred. The time length of the fixed period T2 is a parameter setting value. More specifically, when the state determination unit 87 continuously detects that both of the following state determination formulas (7) and (8) are satisfied during the fixed period T2, the state determination unit 87 moves the target Tgt. It can be determined that the state change from the state to the stationary state has occurred.

  As described above, the state determination unit 83 determines whether the target Tgt is moving, that is, whether the target Tgt is in the stationary state or the moving state, and outputs the determination result ES1 to the determination output unit 89. Give to. The determination output unit 89 supplies a detection signal ES indicating the determination result ES1 to the positioning calculation unit 63A, the tracking calculation unit 64A, and the data transfer unit 65, as shown in FIG.

  Next, before describing the positioning calculation unit 63A, the tracking calculation unit 64A, and the tracking calculation unit 64A in detail, the Kalman filter and the nonlinear least square method used in the first embodiment and the second to fourth embodiments described below will be described. I do. Kalman filter is a kind of filtering theory.

The purpose of the Kalman filter, the observed value series z ₁ that mixing of the noise was at time t _k, ..., when z _k-1, z k is obtained, a stationary target Tgt by using the corresponding observation value series and a state space model It is to estimate physical quantities indicating the state and the moving state. The state space model includes a motion model (state equation) of the target Tgt and an observation model (observation equation). The motion model is given by the following equation (9).

ここで、ｘ_ｋは、時刻ｔ_ｋにおけるｎ行１列の状態ベクトルであり、観測不能な真の状態を表している。ｎは２以上の整数である。ｗ_ｋは、時刻ｔ_ｋにおけるｎ行１列の雑音ベクトルであり、Ｆ_ｋは、時刻ｔ_ｋにおけるｎ行ｎ列の状態遷移行列である。後述するように、状態ベクトルｘ_ｋの成分は、測位演算に適用される場合と追尾演算で適用される場合とで異なる。雑音ベクトルｗ_ｋの平均は零ベクトルとし、雑音ベクトルｗ_ｋの共分散行列はＱ_ｋで表現されるものとする。

Here, x _k is the state vector of n rows and one column at a time t _k, represent unobservable true state. n is an integer of 2 or more. w _k is the noise vector n -by-1 at time _{t k, F k} is the state transition matrix of n rows and n columns at time _{t k.} As described later, the components of the state vector xk are different between a case _where the components are applied to the positioning calculation and a case _where the components are applied to the tracking calculation. The mean of the noise vector w _k is a zero vector, and the covariance matrix of the noise vector w _k is represented by Q _k .

  On the other hand, the observation model is given by the following equation (10).

ここで、ｚ_ｋは、時刻ｔ_ｋにおけるｍ行１列の観測ベクトルであり、ｖ_ｋは、時刻ｔ_ｋにおけるｍ行１列の雑音ベクトルであり、ｍは正整数である。雑音ベクトルｖ_ｋの平均は零ベクトルとし、雑音ベクトルｖ_ｋの共分散行列はＲ_ｋで表現されるものとする。また、ｈ［ｘ_ｋ］は、状態ベクトルｘ_ｋに関する観測関数であり、ｍ行１列のベクトルである。ｍ＝１のとき、ｚ_ｋ，ｖ_ｋ，ｈ［ｘ_ｋ］は、スカラー量となる。

Here, _{z k} is the observation vector of m rows and one column at a time _{t k, v k} is the noise vector of m rows and one column at a time _{t k,} m is a positive integer. The average of the noise vector v _k is a zero vector, and the covariance matrix of the noise vector v _k is represented by R _k . H [x _k ] is an observation function relating to the state vector x _k and is a vector of m rows and 1 column. When m = 1, z _k , v _k , h [x _k ] are scalar quantities.

The estimation process by the Kalman filter (iterative estimation process) is realized by repeatedly executing a prediction process, a smoothing process (update process), and an outlier determination process. Hereinafter, x (k | p) denote the vector of n rows and one column consisting of state estimation value at time t _k at the time of time t _p.

  In the prediction process, a prior state estimation vector x (k | k-1) and a prior error covariance matrix P (k | k-1) are calculated according to the following equations (11) and (12).

ここで、Ｆ_ｋ ^Ｔは、状態遷移行列Ｆ_ｋに対する転置行列である。

Here, F _k ^T is a transposed matrix for the state transition matrix F _k .

  In the smoothing processing (update processing), the state estimation vector x (k | k) and the posterior error covariance matrix P (k | k) are calculated according to the following equations (13) and (14).

ここで、ｅ_ｋは、観測値と状態推定値との間の残差からなる観測残差ベクトルである。また、Ｓ_ｋは観測残差共分散行列、Ｋ_ｋはカルマンゲイン、Ｈ_ｋは観測行列である。観測残差ベクトルｅ_ｋ、観測残差共分散行列Ｓ_ｋおよびカルマンゲインＫ_ｋは、それぞれ、次式（１５），（１６），（１７）で与えられる。

Here, _ek is an observation residual vector composed of a residual between the observation value and the state estimation value. Further, S _k is an observation residual covariance matrix, K _k is a Kalman gain, and H _k is an observation matrix. The observed residual vector e _k , the observed residual covariance matrix S _k and the Kalman gain K _k are given by the following equations (15), (16) and (17), respectively.

The observation matrix H _k is defined by a Jacobian matrix expressed by the following equation (18).

Further, after the prediction processing, an outlier determination processing for determining whether or not the observed value is an outlier value is executed. In the outlier determination processing, first, a weighted residual sum of squares D _k using the observed residual covariance matrix S _k defined by the following equation (20) is calculated based on the following equation (19).

Equation (19), by using the observation residual vector _{e k} represented by the above formula (15) can be expressed as the following equation (21).

  Thereafter, it is determined whether the following determination formula (22) is satisfied.

ここで、γは、カイ二乗分布に従うゲートサイズパラメータである。

Here, γ is a gate size parameter according to a chi-square distribution.

  When the determination formula (22) holds, it is determined that the observed value is not an outlier, and a smoothing process (update process) is performed. On the other hand, when the determination formula (22) does not hold, the observation value is determined to be an outlier, and thus the smoothing process is not performed. In this case, the state estimation value x (k | k) and the posterior error covariance matrix P (k | k) are calculated according to the following equation (23) instead of the above equations (13) and (14).

FIG. 6 is a flowchart schematically illustrating an example of a procedure of an iterative estimation process using a Kalman filter. First, in step ST11, initialization is performed. That is, state estimation vector x (0 | 0) and posterior error covariance matrix P (0 | 0) at time t ₀ (k = 0) are provided.

Next, the process waits until a set of observation values at time t ₁ (k = 1), that is, an observation vector z _k is obtained (NO in step ST12). When observation vector _{z k} is obtained (YES in step ST12), the time _t k (k = 1) the prediction process is performed for (step ST13). As a result, a prior state estimation vector x (k | k-1) and a prior error covariance matrix P (k | k-1) are obtained. Subsequently, the outlier determination processing is executed (step ST14). As a result, when the observed value in the observation vector z _k is determined not to be outliers (NO in step ST15), the smoothing process is executed (step ST20). Thus, the time _t k (k = 1) in the state estimate x (k | k) and post-error covariance matrix P (k | k) is calculated.

After the smoothing process (step ST20), it is determined whether a convergence condition is satisfied (step ST21). Convergence condition, each time t _k at the state estimate x is a condition that | | (k k) is satisfied when it is estimated that sufficiently converged (k k) and post the error covariance matrix P. If the convergence condition is not satisfied (NO in step ST21), the prior state estimation vector x (k | k-1) and the prior error covariance matrix P (k | k-1) are used as the state estimation values calculated in step ST20. x (k | k) and the posterior error covariance matrix P (k | k) are respectively substituted (step ST22). Next, the smoothing process (step ST20) is performed again. For time _{t k,} (YES in step ST21) when the final convergence condition is satisfied, the next step S24 is executed. By repeatedly executing the smoothing process (the cyclic process) until the convergence condition is satisfied, a highly accurate state estimation value x (k | k) and a posterior error covariance matrix P (k | k) are obtained. be able to.

Convergence condition in step ST21, for example, if the smoothing process is repeatedly executed the number of times set in advance for each time t _k, or, for each time t _k, the last j-th to the calculated state estimate x It can be satisfied when the norm between _j (k | k) and the state estimation value _xj-1 (k | k) calculated at the _j- 1th time is equal to or smaller than the threshold (j Is the number of repetitions of the smoothing process in the cyclic process). Alternatively, the norm between the latest posterior error covariance matrix P _j (k | k) calculated at the j-th time and the posterior error covariance matrix P _j-1 (k | k) calculated at the j−1-th time Is smaller than or equal to the threshold and the norm between the state estimation values x _j (k | k) and x _j−1 (k | k) is smaller than or equal to the threshold, the convergence condition of step ST21 may be satisfied.

  On the other hand, when it is determined that the observed value is an outlier (YES in step ST15), the smoothing process in step ST20 is not performed, and the state estimation value x (k | k) and the state estimated value x (k | k) are calculated according to the above equation (23). A prior error covariance matrix P (k | k) is calculated (step ST16).

After step ST16 or step ST21, the procedure is performed iteratively for when it is determined that the predetermined repetition condition is satisfied (YES in step ST24), the next time _t k (k = 2) (After steps ST25 and ST12). Finally, when it is determined that the repetition condition is not satisfied (NO in step ST24), the repetition estimation process ends.

  Next, the nonlinear least squares method will be described.

  First, it is assumed that the observation model (observation equation) of the following equation (24) is established.

ここで、ｙ_ｋは、時刻ｔ_ｋにおけるｐ行１列の観測ベクトルであり、η_ｋは、時刻ｔ_ｋにおけるｐ行１列の雑音ベクトルであり、ｐは２以上の整数である。雑音ベクトルη_ｋの平均は零ベクトルとし、雑音ベクトルη_ｋの共分散行列（観測雑音共分散行列）はΩ_ｋで表現されるものとする。また、φ_ｋ［χ］は、状態ベクトルχに関する観測関数であり、ｐ行１列のベクトルである。

Here, y _k is the observation vector of p rows and one column at a time t _k, the eta _k is a noise vector of p rows and one column at a time t _k, p is an integer of 2 or more. The mean of the noise vector η _k is a zero vector, and the covariance matrix (observation noise covariance matrix) of the noise vector η _k is represented by Ω _k . Φ _k [χ] is an observation function related to the state vector χ, and is a vector of p rows and 1 column.

  For K measurement results (K is an integer of 2 or more), an evaluation function J (χ) for evaluating the magnitude of the observation residual is defined by the following equation (25).

The evaluation function J (χ) represents the sum of the squares of the distance between the observation vector y _k and the vector φ _k [χ] in a noise environment (referred to as “Mahalanobis distance”). .

The purpose of the nonlinear least squares method is to find an approximate solution χ _{min of} χ that minimizes the evaluation function J (χ) as shown in the following equation (26).

It is assumed that the observation function φ _k [χ] is linear and can be expressed using the matrix Γ _k as shown in the following equation (27).

At this time, the approximate solution _{ｍｉｎ min} by the nonlinear least squares method is given by the following equation (28).

On the other hand, when the observation function φ _k [χ] is nonlinear, an approximate solution _{ｍｉｎ min} can be obtained by a recursive least-squares (RLS) algorithm. For example, it is assumed that the observation function φ _k [χ] can be approximately expressed using a Jacobian matrix (function matrix) represented by the following equation (29).

  In the RLS algorithm, for example, the following convergence equations (30), (31), and (32) can be used (m is the number of repetitions in recursive processing).

The procedure of the recursive processing based on the RLS algorithm is as follows. First, initialization is performed. That is, an initial value χ ₀ (m = 0) of a solution candidate is set. Next, for the number of repetitions m = 1, 2,..., Calculations based on the above equations (30), (31), (32) are repeatedly executed. At this time, the recursive processing ends normally when the repetition end condition represented by the following equation (33) is satisfied.

ここで、εは、微少な正の実数であり、Ｍは、繰り返し回数の上限値である。最終的に、反復終了条件を満たす解χ_ｍが、近似解χ_ｍｉｎとして与えられる。

Here, ε is a small positive real number, and M is the upper limit of the number of repetitions. Finally, a solution を 満 た す_m satisfying the iteration termination condition is given as an approximate solution 近似_min .

  Next, the configuration and operation of the positioning calculation unit 63A will be described.

The positioning calculation unit 63A selects a set of two or more measurement values from the measurement values of the arrival time differences TDOA ₁₂ (k), TDOA ₂₃ (k), and TDOA ₃₁ (k) supplied from the measurement unit 61, and It has a function of sequentially calculating a state estimation vector x _po (k) indicating an estimated position of the target Tgt by executing a positioning operation using the selected set of measurement values as an observation vector. These measurement values are values obtained in synchronization with each other or almost in synchronization with each other. The positioning calculation unit 63A can select a set of measurement values of the arrival time differences TDOA ₁₂ (k) and TDOA ₂₃ (k) obtained in synchronization with each other, and execute a positioning calculation using this set as an observation vector. .

Alternatively, the positioning operation unit 63A, measurement unit 61 TDOA supplied from _{TDOA 12 (k), TDOA 23} (k), TDOA 31 (k) and arriving frequency difference _{FDOA 12 (k), FDOA 23} (k), At least one measurement value of the arrival time difference and at least one arrival frequency difference is selected from the measurement values of the FDOA ₃₁ (k), and a positioning operation is performed using the selected set of measurement values as an observation vector. Thus, a state estimation vector x _po (k) indicating the estimated position of the target Tgt is sequentially calculated. These measurement values are values obtained in synchronization with each other or almost in synchronization with each other. For example, the positioning calculation unit 63A can select a set of measurement values of the arrival time difference TDOA ₁₂ (k) and the arrival frequency difference FDOA ₁₂ (k), and execute positioning calculation using this set as an observation vector.

  FIG. 7 is a block diagram schematically illustrating a configuration of a calculation unit including a positioning calculation unit 63A, a tracking calculation unit 64A, and a data transfer unit 65 according to the first embodiment.

As shown in FIG. 7, the positioning calculation unit 63A has a single positioning unit 91A and an iterative estimation unit 92A. The single positioning unit 91A calculates a positioning vector POS _k indicating the position of the target Tgt by executing a positioning operation by the non-linear least squares method using the selected set of measurement values as an observation vector, and calculates the positioning vector POS _k Is output to the iterative estimation unit 92A. The iterative estimating unit 92A calculates a state estimation vector x _po (k) indicating the estimated position of the target Tgt by executing an iterative estimation process using a Kalman filter using the positioning vector POS _k .

More specifically, the single positioning unit 91A can be above the recursive least square method in accordance with (RLS) algorithm, it executes the recursive process using a set of the selected measured value as an observation vector y _k. For _example, TDOA 12 _(k), if the set of measurements of _TDOA 23 (k) is selected as the observation vector _{y k,} the observation vector _{y k} is given by the following equation (34).

At this time, the observation function φ _k [χ _{= p tgt]} is at time _{t k,} is given by the following equation (35).

ここで、ｐ_ｔｇｔは、地球の重心を原点とする、目標Ｔｇｔの位置座標を示す位置ベクトルであり、ｐ_１（ｋ）は、時刻ｔ_ｋでの衛星Ｓｔ１の位置座標を示す位置ベクトルであり、ｐ_２（ｋ）は、時刻ｔ_ｋでの衛星Ｓｔ２の位置座標を示す位置ベクトルである。

_{Here, p tgt} is the origin of the center of gravity of the earth, a position vector indicating the position coordinates of the target _{Tgt, p} 1 (k) is located at a position vector indicating the position coordinates of the satellites St1 at time _{t k} , _p 2 (k) is a position vector indicating the position coordinates of the satellite St2 at time _{t k.}

The observation noise covariance matrix Ω _k is given by the following equation (36).

The single positioning unit 91A executes the recursive process by the above-described RLS algorithm on the measurement results of a plurality of times, and sets the approximate solution _{ｍｉｎ min} that minimizes the evaluation function J (χ = _ptgt ) as the positioning vector POS _k. Can be calculated. The single positioning unit 91A supplies the positioning vector POS _k to the iterative estimation unit 92A.

ここで、Φ_ｋ［χ］は、上式（２９）で示されるヤコビアン行列である。An error covariance matrix 表 す representing a theoretical positioning error in the single positioning operation is given by the following equation (36A).

Here, Φ _k [χ] is a Jacobian matrix expressed by the above equation (29).

Next, the iterative estimating unit 92A executes an iterative estimating process by a Kalman filter using the state space model represented by the above equations (9) and (10). In the motion model (state equation) applied to the iterative estimation unit 92A, the state vector x _k , the state transition matrix F _k, and the covariance matrix Q _k are, for example, expressed by the following equations (37), (38), and (39). Given.

ここで、ＬＮ_ｋ，ＬＴ_ｋは、目標Ｔｇｔの真の位置を示す経度および緯度である。目標Ｔｇｔは静止状態にあるので、状態遷移行列Ｆ_ｋは単位行列である。また、目標Ｔｇｔの運動に不確かさがなく、雑音ベクトルｗ_ｋは零ベクトルとなる。よって、共分散行列Ｑ_ｋは零行列である。

Here, LN _k and LT _k are longitude and latitude indicating the true position of the target Tgt. Since the target Tgt is at rest, a state transition matrix F _k is the identity matrix. In addition, there is no uncertainty in the movement of the target Tgt, noise vector w _k is the zero vector. Therefore, the covariance matrix Q _k is a zero matrix.

In the observation model (observation equation) applied to the iterative estimation unit 92A, the observation vector z _k , the observation matrix H _k, and the covariance matrix R _k are, for example, the following equations (40), (41), (42) Given by

ここで、式（４０）におけるＯＬＮ_ｋ，ＯＬＴ_ｋは、目標Ｔｇｔの観測位置を示す経度および緯度であり、式（４２）におけるＤＯＰは、精度劣化度（Ｄｉｌｕｔｉｏｎ ｏｆ Ｐｒｅｃｉｓｉｏｎ）と呼ばれ、本実施の形態では、単測位演算における理論的な測位誤差式によって与えられる。具体的には、ＤＯＰは、上式（３６Ａ）の誤差共分散行列Λに設定されればよい。

Here, OLN _k and OLT _k in the equation (40) are the longitude and latitude indicating the observation position of the target Tgt, and the DOP in the equation (42) is called a degree of precision degradation (Dilution of Precision). Is given by a theoretical positioning error equation in a single positioning operation. Specifically, the DOP may be set to the error covariance matrix の of the above equation (36A).

Iterative estimation unit 92A executes the iterative estimation process by the Kalman filter using the state space model (Fig. 6), the time t _k for each of the state estimate vector x (k | k) and post-error covariance matrix P ( k | k) can be calculated. The iterative estimation unit 92A outputs the state estimation vector x (k | k) as the state estimation vector x _po (k) indicating the estimated position of the target Tgt. As described above, the single positioning unit 91A can calculate the positioning vector POS _k with high accuracy by executing the positioning operation by the non-linear least square method using the measurement results (cumulative measurement results) a plurality of times. The estimating unit 92A can generate a highly accurate positioning calculation result based on the highly accurate positioning vector POS _k .

  Next, the configuration and operation of the tracking calculation unit 64A will be described.

The tracking calculation unit 64A selects a set of two or more measurement values from the measurement values of the arrival time differences TDOA ₁₂ (k), TDOA ₂₃ (k), and TDOA ₃₁ (k) supplied from the measurement unit 61, and A function of sequentially calculating a state estimation vector x _trk (k) indicating an estimated position and a moving state (for example, estimated speed) of the target Tgt by executing a tracking operation using the selected set of measured values as an observation vector. . Alternatively, the tracking calculation unit 64A may calculate the arrival time difference TDOA ₁₂ (k), TDOA ₂₃ (k), TDOA ₃₁ (k) and the arrival frequency difference FDOA ₁₂ (k), FDOA ₂₃ (k) supplied from the measurement unit 61. At least one measurement value of the arrival time difference and at least one arrival frequency difference is selected from the measurement values of the FDOA ₃₁ (k), and a tracking operation using the selected set of measurement values as an observation vector is performed. Accordingly, a function of sequentially calculating a state estimation vector x _trk (k) indicating an estimated position and a moving state of the target Tgt is provided.

More specifically, the tracking calculation section 64A has a single positioning section 93A and an iterative estimation section 94A as shown in FIG. The configuration and function of single positioning section 93A are the same as those of single positioning section 91A. The iterative estimating unit 94A executes, as a tracking operation, an iterative estimating process using a Kalman filter using the positioning vector POS _k generated by the single positioning unit 93A, thereby obtaining a state estimation vector x _trk indicating the estimated position and moving state of the target Tgt. (K) can be calculated.

The iterative estimating unit 94A executes an iterative estimating process by a Kalman filter using the state space model represented by the above equations (9) and (10) as a tracking operation. In the motion model (state equation) applied to the iterative estimation unit 94A, the state vector x _k , the state transition matrix F _k and the covariance matrix Q _k are, for example, expressed by the following equations (43), (44), and (45). Given.

ここで、ＬＮ_ｋ，ＬＴ_ｋは、目標Ｔｇｔの真の位置を示す経度および緯度、ＶＬＮ_ｋは経度方向の速度、ＬＴ_ｋは緯度方向の速度であり、Ｔ＝ｔ_ｋ−ｔ_ｋ−１である。また、Ｉ_２は２行２列の単位行列であり、ｑはパワースペクトラム密度と呼ばれ、運動モデルの不確かさを表す設定パラメータ値である。目標Ｔｇｔは移動状態にあるので、状態遷移行列Ｆ_ｋは、等速直進運動を想定して設定されている。

_Here, LN k, LT _k is the longitude and latitude indicating the true position of the target Tgt, VLN _k longitude direction of the velocity, LT _k is the speed of _latitudinal, at _T = t _{k -t k-1} is there. Also, I ₂ is the identity matrix of two rows and two columns, q is called a power spectrum density, a setting parameter value representing the uncertainty of motion model. Since the target Tgt is in moving state, the state transition matrix F _k is set assuming constant velocity rectilinear motion.

In the observation model (observation equation) applied to the iterative estimation unit 94A, the observation vector z _k , the observation matrix H _k, and the covariance matrix R _k are, for example, the following equations (46), (47), (48) Given by

ここで、ＯＬＮ_ｋ，ＯＬＴ_ｋは、目標Ｔｇｔの観測位置を示す経度および緯度であり、式（４８）におけるＤＯＰは、精度劣化度（Ｄｉｌｕｔｉｏｎ ｏｆ Ｐｒｅｃｉｓｉｏｎ）と呼ばれ、本実施の形態では、単測位演算における理論的な測位誤差式によって与えられる。具体的には、ＤＯＰは、上式（３６Ａ）の誤差共分散行列Λに設定されればよい。

Here, OLN _k and OLT _k are the longitude and latitude indicating the observation position of the target Tgt, and the DOP in equation (48) is called the degree of precision degradation (Dilution of Precision). It is given by a theoretical positioning error formula in the positioning calculation. Specifically, the DOP may be set to the error covariance matrix の of the above equation (36A).

Iterative estimation unit 94A executes the iterative estimation process by the Kalman filter using the state space model (Fig. 6), the time t _k for each of the state estimate vector x (k | k) and post-error covariance matrix P ( k | k) can be calculated. The iterative estimation unit 94A outputs the state estimation vector x (k | k) as a state estimation vector x _trk (k) indicating the estimated position and the moving state of the target Tgt.

  Next, operations of the positioning calculation unit 63A, the tracking calculation unit 64A, and the data transfer unit 65 when the state detection unit 62A detects a change in the state of the target Tgt will be described.

When the state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimation unit 92A of the positioning calculation unit 63A determines the state estimation vector x _po (k) that is the latest positioning calculation result in accordance with the detection signal ES. ) (= X (k | k)) and the posterior error covariance matrix P _po (k) (= P (k | k)) are supplied to the tracking calculation unit 64A via the data transfer unit 65.

At this time, the iterative estimation unit 94A of the tracking calculation unit 64A determines the state estimation vector x (k) based on the state estimation vector x _po (k) and the posterior error covariance matrix P _po (k) supplied from the positioning calculation unit 63A. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). Then, the iterative estimating unit 94A executes an iterative estimating process (FIG. 6) using a Kalman filter using the initial values x (0 | 0), P (0 | 0) and the positioning vector POS _k , thereby obtaining the time t _k for each of the state estimation vector _{x trk} the (k) can be calculated.

  In this case, the initial values x (0 | 0) and P (0 | 0) used in the iterative estimating unit 94A can be set, for example, as shown in the following equations (49) and (50).

ここで、０_２×１は２行１列の零行列、０_２×２は２行２列の零行列である。式（４９）の状態推定ベクトルｘ（０｜０）は、目標Ｔｇｔの初期速度が零となるように設定されている。また、式（５０）の初期値Ｐ（０｜０）は、推定速度成分に誤差はないとの仮定に基づいて設定されている。

Here, 0 _{2 × 1} is a 2-row, 1-column zero matrix, and 0 _{2 × 2} is a 2-row, 2-column zero matrix. The state estimation vector x (0 | 0) in Expression (49) is set such that the initial speed of the target Tgt becomes zero. Further, the initial value P (0 | 0) in Expression (50) is set based on the assumption that there is no error in the estimated speed component.

On the other hand, when a state change of the target Tgt from the moving state to the stationary state is detected, the iterative estimating unit 94A of the tracking calculating unit 64A determines the state estimation vector x _trk which is the latest tracking calculation result according to the detection signal ES. (K) (= x (k | k)) and the posterior error covariance matrix P _trk (k) (= P (k | k)) are supplied to the positioning calculation unit 63A via the data transfer unit 65.

At this time, the iterative estimation unit 92A in the positioning calculation unit 63A determines the state estimation vector x (based on the state estimation vector x _trk (k) and the posterior error covariance matrix P _trk (k) supplied from the tracking calculation unit 64A. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). Then, the iterative estimating unit 92A executes an iterative estimating process (FIG. 6) using a Kalman filter using the initial values x (0 | 0), P (0 | 0) and the positioning vector POS _k , thereby obtaining the time t _k for each of the state estimation vector _x po a (k) can be calculated.

  In this case, the initial values x (0 | 0) and P (0 | 0) used in the iterative estimation unit 92A can be set, for example, as shown in the following equations (51) and (52).

ここで、［ｘ_ｔｒｋ（ｋ）］_１，１は、状態推定ベクトルｘ_ｔｒｋ（ｋ）の１行１列目成分（１番目成分）、［ｘ_ｔｒｋ（ｋ）］_２，１は、状態推定ベクトルｘ_ｔｒｋ（ｋ）の２行１列目成分（２番目成分）である。また、［Ｐ_ｔｒｋ（ｋ）］_１，１は、事後誤差共分散行列Ｐ_ｔｒｋ（ｋ）の１行１列目成分、［Ｐ_ｔｒｋ（ｋ）］_２，１は、事後誤差共分散行列Ｐ_ｔｒｋ（ｋ）の２行１列目成分、［Ｐ_ｔｒｋ（ｋ）］_１，２は、事後誤差共分散行列Ｐ_ｔｒｋ（ｋ）の１行２列目成分、［Ｐ_ｔｒｋ（ｋ）］_２，２は、事後誤差共分散行列Ｐ_ｔｒｋ（ｋ）の２行２列目成分である。

Here, [x _trk (k)] _1,1 is the first row, first column component (first component) of the state estimation vector x _trk (k), and [x _trk (k)] _2,1 is the state estimation vector x _trk (k). This is the second row, first column component (second component) of the vector x _trk (k). [P _trk (k)] _1,1 is the first row and first column component of the posterior error covariance matrix P _trk (k), and [P _trk (k)] _2,1 is the posterior error covariance matrix P the second row and first column component of _{trk (k), [P trk} (k)] 1,2 is first row second column component of posterior error covariance matrix _{P trk (k), [P} trk (k)] 2 _{, 2} are the second row and second column components of the posterior error covariance matrix P _trk (k).

As described above, when the state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimating unit 94A of the tracking calculating unit 64A sets the latest positioning calculation result x _po (k), P _po ( The tracking calculation using the measured value can be performed using k) as an initial value. Conversely, when a state change from the moving state of the target Tgt to the stationary state is detected, the iterative estimating unit 92A of the positioning calculation unit 63A calculates the latest tracking calculation results x _trk (k) and P _trk (k). It is possible to execute a positioning calculation using the measurement value as an initial value.

  The hardware configuration of the above-described monitoring processing unit 11A is, for example, a single or multiple semiconductor integrated circuit such as a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array). It should just be realized. Alternatively, the hardware configuration of the monitoring processing unit 11A includes a single unit including an arithmetic device such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit) that executes a program code of software or firmware read from the nonvolatile memory. Alternatively, it may be realized by a plurality of processors. The hardware configuration of the monitoring processing unit 11A may be realized by one or more processors including a combination of a semiconductor integrated circuit such as a DSP and an arithmetic device such as a CPU.

  FIG. 8 is a block diagram illustrating a schematic configuration of a signal processing device 70 that is a hardware configuration example that implements the function of the monitoring processing unit 11A. The signal processing device 70 illustrated in FIG. 8 includes a processor 71, a memory 72, an input interface unit 73, an output interface unit 74, a communication circuit 75, and a signal path 76. The signal path 76 is a bus for mutually connecting the processor 71, the memory 72, the input interface unit 73, the output interface unit 74, and the communication circuit 75. The input interface unit 73 has a function of transferring digital reception signals D1, D2, and D3 input from the outside to the processor 71 via the signal path 76. The processor 71 can calculate a state estimation value indicating the stationary state and the moving state of the target Tgt by executing the positioning operation and the tracking operation based on the transferred digital reception signals D1, D2, and D3. Further, the processor 71 can generate image information DD based on the calculated state estimation value and output the image information DD to the outside via the signal path 75 and the output interface unit 74.

  Here, the communication circuit 75 is a circuit that communicates with an external server (not shown) and receives data necessary for processing in the state detection unit 62A. The memory 72 is a data storage area used when the processor 71 executes digital signal processing. When the processor 71 incorporates an arithmetic device such as a CPU, the memory 72 has a data storage area for storing program codes of software or firmware executed by the processor 71. As the memory 72, for example, a semiconductor memory such as a ROM (Read Only Memory) and an SDRAM (Synchronous Dynamic Random Access Memory) can be used.

  As described above, in the target monitoring device 2 according to the first embodiment, the state detection unit 62A monitors the time variation of at least one of the plurality of measurement values to determine whether the target Tgt has moved. Since the detection signal ES indicating the detection result is supplied to the positioning calculation unit 63A, the tracking calculation unit 64A, and the data transfer unit 65, the positioning calculation unit 63A and the tracking calculation unit 64A should switch the positioning calculation to the tracking calculation. The timing or the timing at which the tracking calculation should be switched to the positioning calculation can be determined with high accuracy. Therefore, the monitoring processing unit 11A can execute highly accurate calculation according to a state change of the target Tgt from the stationary state to the moving state or a state change of the target Tgt from the moving state to the stationary state.

When a state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimation unit 94A of the tracking calculation unit 64A sets the latest positioning calculation result x _po (k) and P _po (k) to the initial values. And a tracking calculation using a set of measurement values can be performed. Therefore, even if the positioning calculation is switched to the tracking calculation, high tracking accuracy with respect to the target Tgt can be ensured.

Embodiment 2 FIG.
Next, a second embodiment according to the present invention will be described. FIG. 9 is a block diagram schematically showing a configuration of a calculation unit including a positioning calculation unit 63B, a tracking calculation unit 64B, and a data transfer unit 65 according to Embodiment 2 of the present invention. The configuration of the target monitoring system and the target monitoring device (including the monitoring processing unit) of the present embodiment is different from the positioning calculation unit 63A and the tracking calculation unit 64A of FIG. 3 in that the positioning calculation unit 63B and the tracking calculation unit of FIG. The configuration is the same as that of the target monitoring system 1 and the target monitoring device 2 of the first embodiment, except that the configuration includes a unit 64B.

The positioning calculation section 63B of the present embodiment has an iterative estimation section 92B as shown in FIG. The iterative estimating unit 92B selects a set of two or more measured values from the measured values of the arrival time differences TDOA ₁₂ (k), TDOA ₂₃ (k), and TDOA ₃₁ (k) supplied from the measuring unit 61, and It has a function of sequentially calculating a state estimation vector x _po (k) indicating an estimated position of the target Tgt by executing iterative estimation processing by a Kalman filter using the selected set of measurement values as an observation vector as a positioning operation. These measurement values are values obtained in synchronization with each other or almost in synchronization with each other. For example, the iterative estimating unit 92B selects measurement values of the arrival time differences TDOA ₁₂ (k) and TDOA ₂₃ (k) obtained in synchronization with each other, and executes a positioning operation using a set of these measurement values as an observation vector. be able to.

Alternatively, the iterative estimating unit 92 </ b> B receives the arrival time difference TDOA ₁₂ (k), TDOA ₂₃ (k), TDOA ₃₁ (k) and the arrival frequency difference FDOA ₁₂ (k), FDOA ₂₃ (k) supplied from the measurement unit 61. Iterative estimation processing by a Kalman filter using at least one arrival time difference and at least one arrival frequency difference measurement value among the measurement values of the FDOA ₃₁ (k) and using the selected set of measurement values as an observation vector Is executed as a positioning calculation, thereby sequentially calculating a state estimation vector x _po (k) indicating the estimated position of the target Tgt. These measurement values are values obtained in synchronization with each other or almost in synchronization with each other. For example, iterative estimation section 92B selects measurement values of arrival time difference TDOA ₁₂ (k) and arrival frequency difference FDOA ₁₂ (k) obtained in synchronization with each other, and performs a positioning operation using a set of these measurement values as an observation vector. Can be performed.

More specifically, the iterative estimating unit 92B of the positioning operation unit 63B can execute the iterative estimation process by the Kalman filter as the positioning operation using the state space model represented by the above equations (9) and (10). In the motion model (state equation) applied to the iterative estimation unit 92B, the state vector x _k , the state transition matrix F _k, and the covariance matrix Q _k are, for example, expressed by the following equations (53), (54), and (55). Given.

ここで、ＬＮ_ｋ，ＬＴ_ｋは、目標Ｔｇｔの真の位置を示す経度および緯度である。目標Ｔｇｔは静止状態にあるので、状態遷移行列Ｆ_ｋは単位行列である。また、目標Ｔｇｔの運動に不確かさがなく、雑音ベクトルｗ_ｋは零ベクトルとなる。よって、共分散行列Ｑ_ｋは零行列である。

Here, LN _k and LT _k are longitude and latitude indicating the true position of the target Tgt. Since the target Tgt is at rest, a state transition matrix F _k is the identity matrix. In addition, there is no uncertainty in the movement of the target Tgt, noise vector w _k is the zero vector. Therefore, the covariance matrix Q _k is a zero matrix.

In the observation model (observation equation) applied to the iterative estimation unit 92B, the observation vector z _k , the observation function h [x _k ], and the covariance matrix R _k are, for example, the following equations (56), (57), (58).

ここで、観測関数ｈ［ｘ_ｋ］は、上式（１８）に従い、ヤコビアン行列Ｈ_ｋに変換される。

Here, the observation function h [x _k ] is converted into a Jacobian matrix H _k according to the above equation (18).

Iterative estimation unit 92B executes the iterative estimation process (Figure 6) by the Kalman filter using the state space model, the time t _k for each of the state estimate vector x (k | k) and post-error covariance matrix P ( k | k) can be calculated. The iterative estimation unit 92B outputs the state estimation vector x (k | k) as the state estimation vector x _po (k) indicating the estimated position of the target Tgt.

On the other hand, the tracking calculation unit 64B has an iterative estimation unit 94B as shown in FIG. The iterative estimation unit 94B selects a set of two or more measurement values from the measurement values of the arrival time differences TDOA ₁₂ (k), TDOA ₂₃ (k), and TDOA ₃₁ (k) supplied from the measurement unit 61, and A function of sequentially calculating a state estimation vector x _trk (k) indicating an estimated position and a moving state (for example, estimated speed) of the target Tgt by executing a tracking operation using the selected set of measured values as an observation vector. . The selected measurement values are values obtained in synchronization with each other or almost in synchronization with each other. Alternatively, the tracking calculation unit 64B may calculate the arrival time differences TDOA ₁₂ (k), TDOA ₂₃ (k), TDOA ₃₁ (k) and the arrival frequency differences FDOA ₁₂ (k), FDOA ₂₃ (k) supplied from the measurement unit 61. At least one measurement value of the arrival time difference and at least one arrival frequency difference is selected from the measurement values of the FDOA ₃₁ (k), and a tracking operation using the selected set of measurement values as an observation vector is performed. Accordingly, a state estimation vector x _trk (k) indicating the estimated position and the moving state of the target Tgt is sequentially calculated. The selected measurement values are values obtained in synchronization with each other or almost in synchronization with each other.

More specifically, the iterative estimating unit 94B executes the iterative estimating process by the Kalman filter using the state space model represented by the above equations (9) and (10) as a tracking operation. In the motion model (state equation) applied to the iterative estimation unit 94B, the state vector x _k , the state transition matrix F _k, and the covariance matrix Q _k are _represented by, for example, the following equations (59), (60), and (61). Given.

ここで、ＬＮ_ｋ，ＬＴ_ｋは、目標Ｔｇｔの真の位置を示す経度および緯度、ＶＬＮ_ｋは経度方向の速度、ＬＴ_ｋは緯度方向の速度であり、Ｔ＝ｔ_ｋ−ｔ_ｋ−１である。また、Ｉ_２は２行２列の単位行列であり、ｑはパワースペクトラム密度と呼ばれ、運動モデルの不確かさを表す設定パラメータ値である。

_Here, LN k, LT _k is the longitude and latitude indicating the true position of the target Tgt, VLN _k longitude direction of the velocity, LT _k is the speed of _latitudinal, at _T = t _{k -t k-1} is there. Also, I ₂ is the identity matrix of two rows and two columns, q is called a power spectrum density, a setting parameter value representing the uncertainty of motion model.

In the observation model (observation equation) applied to the iterative estimation unit 92B, the observation vector z _k , the observation function h [x _k ], and the covariance matrix R _k are, for example, the following equations (62), (63), (64).

ここで、観測関数ｈ［ｘ_ｋ］は、上式（１８）に従い、ヤコビアン行列Ｈ_ｋに変換される。

Here, the observation function h [x _k ] is converted into a Jacobian matrix H _k according to the above equation (18).

Iterative estimation unit 94B executes the iterative estimation process by the Kalman filter using the state space model (Fig. 6), the time t _k for each of the state estimate vector x (k | k) and post-error covariance matrix P ( k | k) can be calculated. The iterative estimating unit 94B outputs the state estimation vector x (k | k) as a state estimation vector x _trk (k) indicating the estimated position and the moving state of the target Tgt.

  Next, operations of the positioning calculation unit 63B, the tracking calculation unit 64B, and the data transfer unit 65 when a change in the state of the target Tgt is detected will be described.

When the state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimation unit 92B of the positioning calculation unit 63B determines the state estimation vector x _po (k) that is the latest positioning calculation result according to the detection signal ES. ) (= X (k | k)) and the posterior error covariance matrix P _po (k) (= P (k | k)) are supplied to the tracking operation unit 64B via the data transfer unit 65.

At this time, the iterative estimation unit 94B of the tracking calculation unit 64B determines the state estimation vector x (k) based on the state estimation vector x _po (k) and the posterior error covariance matrix P _po (k) supplied from the positioning calculation unit 63B. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). The iterative estimation unit 94B, they initial value x (0 | 0), P | by executing (0 0) and the observation vector and the iterative estimation process by the Kalman filter using (6), each time _{t k} The state estimation vector x _trk (k) can be calculated. In this case, the initial values x (0 | 0) and P (0 | 0) used in the iterative estimating unit 92B can be set as shown in the above equations (49) and (50), for example.

On the other hand, when the state change from the moving state of the target Tgt to the stationary state is detected, the iterative estimating unit 94B of the tracking calculating unit 64B determines the state estimation vector x _trk which is the latest tracking calculation result according to the detection signal ES. (K) (= x (k | k)) and the posterior error covariance matrix P _trk (k) (= P (k | k)) are supplied to the positioning calculation unit 63B via the data transfer unit 65.

At this time, the iterative estimation unit 92B of the positioning calculation unit 63B determines the state estimation vector x (k) based on the state estimation vector x _trk (k) and the posterior error covariance matrix P _trk (k) supplied from the tracking calculation unit 64B. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). The iterative estimation unit 92B, they initial value x (0 | 0), P | by executing (0 0) and the observation vector and the iterative estimation process by the Kalman filter using (6), each time _{t k} Can be calculated as the state estimation vector x _po (k). In this case, the initial values x (0 | 0) and P (0 | 0) used in the iterative estimation unit 92B can be set, for example, as shown in the following equations (51) and (52).

As described above, the target monitoring device according to the second embodiment includes the state detection unit 62A similarly to the target monitoring device 2 according to the first embodiment, so that the positioning calculation unit 63B and the tracking calculation unit 64B perform the positioning calculation. The timing to switch to the tracking calculation or the timing to switch from the tracking calculation to the positioning calculation can be determined with high accuracy. Therefore, the monitoring processing unit according to the second embodiment can execute highly accurate calculation according to a state change of the target Tgt from the stationary state to the moving state or a state change of the target Tgt from the moving state to the stationary state. it can. When a state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimating unit 94B of the tracking calculating unit 64B sets the latest positioning calculation results x _po (k) and P _po (k) to the initial values. And a tracking operation using a set of measured values can be performed. Therefore, even if the positioning calculation is switched to the tracking calculation, high tracking accuracy with respect to the target Tgt can be ensured.

Embodiment 3 FIG.
Next, a third embodiment according to the present invention will be described. In the first and second embodiments, a plurality of measurement values to be used as components of the observation vector for each time _{t k} (e.g., the measurement value of TDOA _{TDOA 12 (k), TDOA 23} (k)) are synchronized with each other Or observations obtained almost synchronously with each other. On the other hand, in the present embodiment, the plurality of measurement values are observation values obtained asynchronously with each other, in other words, observation values that cannot be obtained simultaneously at the same time.

For example, when there are three receiving antennas (receiving stations), the measured values of the arrival time differences TDOA ₁₂ (k) and TDOA ₂₃ (k) can be obtained simultaneously in a synchronized manner. On the other hand, when only two reception antennas (reception stations) can be used, the antenna pointing is used to obtain a set of the measurement value between the satellites St1 and St2 and the measurement value between the satellites St2 and St3. You need to switch directions. However, if a time lag occurs in the switching, the two types of measurement values may be obtained asynchronously with each other. Further, even if the measured values of the arrival time differences TDOA ₁₂ (k) and TDOA ₂₃ (k) are obtained at the same time, there is a possibility that one of these measured values is missing for some reason such as when it is determined to be an outlier. is there. The monitoring processing unit according to the present embodiment can execute the positioning calculation and the tracking calculation using a plurality of measurement values obtained asynchronously.

FIG. 10 is a block diagram illustrating a schematic configuration of the monitoring processing unit 11C according to the third embodiment of the present invention. The configuration of the target monitoring system according to the present embodiment is the same as the configuration of the target monitoring system 1 according to the first embodiment except that a monitoring processing unit 11C shown in FIG. 10 is used instead of the monitoring processing unit 11A shown in FIG. It is. The monitoring processing unit 11C includes a measuring unit 61C for sequentially calculating the measurement values of the radio wave arrival time difference TDOA _ij (k) and the arrival frequency difference FDOA _ij (k) based on the digital reception signals D1, D2, and D3. The arrival time difference TDOA _ij (k) is one of the arrival time differences TDOA ₁₂ (k) and TDOA ₂₃ (k), and the arrival frequency difference FDOA _ij (k) is the arrival frequency difference FDOA ₁₂ (k) or FDOA _23. (K). For example, the measurement value is obtained in the arrival time difference at time _{t k1 TDOA} 12 _(k1), the time _{t k2} when the measurement value of TDOA _TDOA 23 (k2) to (≠ _{t k1)} is obtained, TDOA _TDOA 12 (k1 ), TDOA ₂₃ (k2) are observed values obtained asynchronously with each other.

Further, the monitoring processing unit 11C constantly monitors the time variation of any one of the measured values of the arrival time difference TDOA _ij (k) and the arrival frequency difference FDOA _ij (k) to determine whether or not the target Tgt has moved. The state detection unit 62A that detects the position of the target Tgt, and the position calculation that sequentially calculates the state estimated value indicating the estimated position of the target Tgt by executing the position calculation using the measured value when the movement of the target Tgt is detected. When it is detected that there is a movement of the target Tgt, the unit 63C performs a tracking operation using the measured value to obtain a state estimated value indicating both the estimated position and the moving state (for example, estimated speed) of the target Tgt. A tracking operation unit 64C for calculating, a data transfer unit 65 for transferring the data of the operation result from one of the positioning operation unit 63C and the tracking operation unit 64C to the other, and a process performed by the state detection unit 62A. And a data storage section 67 for storing data necessary for

FIG. 11 is a block diagram schematically illustrating a configuration of a calculation unit including a positioning calculation unit 63C, a tracking calculation unit 64C, and a data transfer unit 65 according to the third embodiment. The positioning calculation section 63C of the present embodiment has an iterative estimation section 92C. The iterative estimating unit 92C uses a Kalman filter using a set of measured values (k1 ≠ k2) of the arrival time differences TDOA ₁₂ (k1) and TDOA ₂₃ (k2) obtained asynchronously by the measuring unit 61 as a positioning operation. When executed, it has a function of calculating a state estimation vector x _po (k) indicating the estimated position of the target Tgt.

Alternatively, iterative estimation section 92C performs iterative estimation by a Kalman filter using a set of measurement values (k1 ≠ k2) of arrival time difference TDOA ₁₂ (k1) and arrival frequency difference FDOA ₂₃ (k2) obtained asynchronously by measurement section 61. By executing the processing as a positioning operation, the state estimation vector x _po (k) indicating the estimated position of the target Tgt may be calculated.

More specifically, the iterative estimating unit 92C of the positioning operation unit 63C can execute the iterative estimation process using the Kalman filter as the positioning operation using the state space model represented by the above equations (9) and (10). In the motion model (state equation) applied to the iterative estimation unit 92C, the state vector x _k , the state transition matrix F _k, and the covariance matrix Q _k are _represented by, for example, the following equations (65), (66), and (67). Given.

ここで、ＬＮ_ｋ，ＬＴ_ｋは、目標Ｔｇｔの真の位置を示す経度および緯度である。目標Ｔｇｔは静止状態にあるので、状態遷移行列Ｆ_ｋは単位行列である。また、目標Ｔｇｔの運動に不確かさがなく、雑音ベクトルｗ_ｋは零ベクトルとなる。よって、共分散行列Ｑ_ｋは零行列である。

Here, LN _k and LT _k are longitude and latitude indicating the true position of the target Tgt. Since the target Tgt is at rest, a state transition matrix F _k is the identity matrix. In addition, there is no uncertainty in the movement of the target Tgt, noise vector w _k is the zero vector. Therefore, the covariance matrix Q _k is a zero matrix.

Further, in the observation model (observation equation) applied to the iterative estimation unit 92C, the observation value z _k , the observation function h [x _k ], and the covariance R _k are, for example, the following expressions (68), (69), (69) 70).

ここで、式（６８），（６９）における（ｉ，ｊ）は、各時刻ｔ_ｋで（１，２）または（２，３）のいずれかである。たとえば、時刻ｔ_ｋ１では、到来時間差ＴＤＯＡ_１２（ｋ１）の観測値（計測値）が得られ（ｉ＝１；ｊ＝２）、時刻ｔ_ｋ２ではＴＤＯＡ_２３（ｋ２）の観測値（計測値）が得られる（ｉ＝２；ｊ＝３）。

Here, equation (68), (i, j) in (69) is either at each time _{t k} (1, 2) or (2,3). For example, at time t _k1 , an observation value (measurement value) of arrival time difference TDOA ₁₂ (k1) is obtained (i = 1; j = 2), and at time t _k2 , an observation value (measurement value) of TDOA ₂₃ (k2). Is obtained (i = 2; j = 3).

Iterative estimation unit 92C executes the iterative estimation process (Figure 6) by the Kalman filter using the state space model, the time t _k for each of the state estimate vector x (k | k) and post-error covariance matrix P ( k | k) can be calculated. The iterative estimation unit 92C outputs the state estimation vector x (k | k) as the state estimation vector x _po (k) indicating the estimated position of the target Tgt.

On the other hand, the tracking calculation unit 64C has an iterative estimation unit 94C as shown in FIG. The iterative estimating unit 94C uses the Kalman filter using a set of measured values (k1 ≠ k2) of the arrival time differences TDOA ₁₂ (k1) and TDOA ₂₃ (k2) obtained asynchronously by the measuring unit 61 as a tracking operation. By executing, a state estimation vector x _trk (k) indicating an estimated position and a moving state (for example, estimated speed) of the target Tgt is sequentially calculated.

Alternatively, the tracking calculation unit 64C performs iterative estimation by a Kalman filter using a set of measurement values (k1 ≠ k2) of the arrival time difference TDOA ₁₂ (k1) and the arrival frequency difference FDOA ₂₃ (k2) obtained asynchronously by the measurement unit 61. By executing the processing as a tracking operation, a state estimation vector x _trk (k) indicating the estimated position and the moving state (for example, the estimated speed) of the target Tgt may be calculated.

More specifically, the iterative estimating unit 94C executes the iterative estimating process by the Kalman filter using the state space model represented by the above equations (9) and (10) as a tracking operation. In the motion model (state equation) applied to the iterative estimation unit 94C, the state vector x _k , the state transition matrix F _k, and the covariance matrix Q _k are _represented by, for example, the following equations (71), (72), and (73). Given.

ここで、ＬＮ_ｋ，ＬＴ_ｋは、目標Ｔｇｔの真の位置を示す経度および緯度、ＶＬＮ_ｋは経度方向の速度、ＬＴ_ｋは緯度方向の速度であり、Ｔ＝ｔ_ｋ−ｔ_ｋ−１である。また、Ｉ_２は２行２列の単位行列であり、ｑはパワースペクトラム密度と呼ばれ、運動モデルの不確かさを表す設定パラメータ値である。

_Here, LN k, LT _k is the longitude and latitude indicating the true position of the target Tgt, VLN _k longitude direction of the velocity, LT _k is the speed of _latitudinal, at _T = t _{k -t k-1} is there. Also, I ₂ is the identity matrix of two rows and two columns, q is called a power spectrum density, a setting parameter value representing the uncertainty of motion model.

In the observation model (observation equation) applied to the iterative estimation unit 94C, the observation value z _k , the observation function h [x _k ], and the covariance R _k are, for example, the following equations (74), (75), (75) 76).

ここで、式（７４），（７５）における（ｉ，ｊ）は、各時刻ｔ_ｋで（１，２）または（２，３）のいずれかである。たとえば、時刻ｔ_ｋ１では、到来時間差ＴＤＯＡ_１２（ｋ１）の計測値が得られ（ｉ＝１；ｊ＝２）、時刻ｔ_ｋ２ではＴＤＯＡ_２３（ｋ２）の計測値が得られる（ｉ＝２；ｊ＝３）。

Here, equation (74), (i, j) in (75) is either at each time _{t k} (1, 2) or (2,3). For example, at time t _k1 , a measured value of arrival time difference TDOA ₁₂ (k1) is obtained (i = 1; j = 2), and at time t _k2 , a measured value of TDOA ₂₃ (k2) is obtained (i = 2; j = 3).

Iterative estimation unit 94C executes the iterative estimation process by the Kalman filter using the state space model (Fig. 6), the time t _k for each of the state estimate vector x (k | k) and post-error covariance matrix P ( k | k) can be calculated. The iterative estimation unit 94C outputs the state estimation vector x (k | k) as a state estimation vector x _trk (k) indicating the estimated position and the moving state of the target Tgt.

  Next, operations of the positioning calculation unit 63C, the tracking calculation unit 64C, and the data transfer unit 65 when a change in the state of the target Tgt is detected will be described.

When the state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimation unit 92C of the positioning calculation unit 63C determines the state estimation vector x _po (k) that is the latest positioning calculation result according to the detection signal ES. ) (= X (k | k)) and the posterior error covariance matrix P _po (k) (= P (k | k)) are supplied to the tracking calculation unit 64C via the data transfer unit 65.

At this time, the iterative estimation unit 94C of the tracking calculation unit 64C determines the state estimation vector x (k) based on the state estimation vector x _po (k) and the posterior error covariance matrix P _po (k) supplied from the positioning calculation unit 63C. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). Then, the iterative estimating unit 94C executes an iterative estimating process (FIG. 6) using a Kalman filter using the initial values x (0 | 0), P (0 | 0) and the observed value (measured value). time _{t k} for each of the state estimation vector _{x trk} the (k) can be calculated. In this case, the initial values x (0 | 0) and P (0 | 0) used in the iterative estimating unit 92B can be set as shown in the above equations (49) and (50), for example.

On the other hand, when a state change of the target Tgt from the moving state to the stationary state is detected, the iterative estimation unit 94C of the tracking calculation unit 64C determines the state estimation vector x _trk which is the latest tracking calculation result according to the detection signal ES. (K) (= x (k | k)) and the posterior error covariance matrix P _trk (k) (= P (k | k)) are supplied to the positioning calculation unit 63C via the data transfer unit 65.

At this time, the iterative estimation unit 92C of the positioning calculation unit 63C determines the state estimation vector x (k) based on the state estimation vector x _trk (k) and the posterior error covariance matrix P _trk (k) supplied from the tracking calculation unit 64C. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). Then, the iterative estimation unit 92C executes an iterative estimation process (FIG. 6) using a Kalman filter using the initial values x (0 | 0), P (0 | 0) and the observed value (measured value). time _{t k} for each of the state estimation vector _x po a (k) can be calculated. In this case, the initial values x (0 | 0) and P (0 | 0) used in the iterative estimation unit 92B can be set, for example, as shown in the following equations (51) and (52).

As described above, the target monitoring device according to the third embodiment includes the state detection unit 62A similarly to the target monitoring device 2 according to the first embodiment, so that the positioning calculation unit 63C and the tracking calculation unit 64C perform the positioning calculation. The timing to switch to the tracking calculation or the timing to switch from the tracking calculation to the positioning calculation can be determined with high accuracy. Therefore, the monitoring processing unit 11C according to the third embodiment executes highly accurate calculation according to a state change of the target Tgt from the stationary state to the moving state or a state change of the target Tgt from the moving state to the stationary state. Can be. When a state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimating unit 94C of the tracking calculating unit 64C sets the latest positioning calculation result x _po (k) and P _po (k) to the initial values. And a tracking operation using the measured value can be performed. Therefore, even if the positioning calculation is switched to the tracking calculation, high tracking accuracy with respect to the target Tgt can be ensured.

Embodiment 4 FIG.
Next, a fourth embodiment according to the present invention will be described. In the third embodiment, the positioning operation section 63C performs sequentially positioning calculation whenever the measured values in the respective times t _k are input. On the other hand, the positioning calculation according to the fourth embodiment is a batch-type positioning calculation that collectively processes a plurality of measured values.

  FIG. 12 is a block diagram schematically showing a configuration of a calculation unit including positioning calculation unit 63D, tracking calculation unit 64C, and data transfer unit 65 according to Embodiment 4 of the present invention. The configuration of the target monitoring system and the target monitoring device according to the present embodiment is the same as that of the target monitoring system and the target monitoring system according to the third embodiment except that a positioning operation unit 63D in FIG. 10 is provided instead of the positioning operation unit 63C in FIG. The configuration is the same as that of the target monitoring device.

Measurement values of the arrival time difference TDOA _ij (k) and the arrival frequency difference FDOA _ij (k) are input to the positioning calculation unit 63D of the present embodiment from the measurement unit 61C (FIG. 10). The arrival time difference TDOA _ij (k) is one of the arrival time differences TDOA ₁₂ (k) and TDOA ₂₃ (k), and the arrival frequency difference FDOA _ij (k) is the arrival frequency difference FDOA ₁₂ (k) or FDOA _23. (K). For example, the measurement value is obtained in the arrival time difference at time _{t k1 TDOA} 12 _(k1), the time _{t k2} when the measurement value of TDOA _TDOA 23 (k2) to (≠ _{t k1)} is obtained, TDOA _TDOA 12 (k1 ), TDOA ₂₃ (k2) are observed values obtained asynchronously with each other.

The positioning calculation unit 63D of the present embodiment has a buffer memory 100 and a batch-type repetition estimation unit 92D as shown in FIG. The buffer memory 100 has a capacity to store at least K-1 past measured values from time tk _{-K + 1} to time tk _-1 . Batch iterative estimation unit 92D, each time the current measured value at time t _k can be obtained, read out past measurement value of the K-1 times from the buffer memory 100. Then, the batch-type iterative estimating unit 92D executes the batch processing by the non-linear least square method using the past measurement value and the current measurement value collectively as the positioning operation, thereby calculating the estimated position of the target Tgt. The calculated state estimation vector x _po (k) can be calculated.

More specifically, the batch-type iterative estimating unit 92D can execute a recursive process using the observed value y _k shown in the following equation (77) according to the above-described recursive least squares (RLS) algorithm.

At this time, the observation function phi _k at time _{t k [χ = p tgt]} is given by the following equation (78).

ここで、ｐ_ｔｇｔは、地球の重心を原点とする、目標Ｔｇｔの位置座標を示す位置ベクトルであり、ｐ_ｉ（ｋ）は、時刻ｔ_ｋでの衛星Ｓｔｉの位置座標を示す位置ベクトルであり、ｐ_ｊ（ｋ）は、時刻ｔ_ｋでの衛星Ｓｔｊの位置座標を示す位置ベクトルである。

_{Here, p tgt} is the origin of the center of gravity of the earth, a position vector indicating the position coordinates of the target _{Tgt, p} i (k) is located at a position vector indicating the position coordinates of the satellite Sti at time _{t k} , _p j (k) is a position vector indicating the position coordinates of the satellite Stj at time _{t k.}

The observation noise covariance Ω _k is given by the following equation (79).

Batch iterative estimation unit 92D executes the recursive process by RLS algorithm described above for K times the measurement results, an approximate solution chi _min evaluating function J a (χ _{= p tgt)} to minimize target Tgt Can be calculated as a state estimation vector x _po (k) indicating the estimated position of.

  Next, operations of the positioning calculation unit 63D, the tracking calculation unit 64C, and the data transfer unit 65 when a change in the state of the target Tgt is detected will be described.

When the state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimating unit 92D of the positioning calculation unit 63D determines the state estimation vector x _po (k) that is the latest positioning calculation result in accordance with the detection signal ES. ) (= X (k | k)) and the posterior error covariance matrix P _po (k) (= P (k | k)) are supplied to the tracking calculation unit 64C via the data transfer unit 65.

At this time, the iterative estimation unit 94C of the tracking calculation unit 64C calculates the state estimation vector x (k) based on the state estimation vector x _po (k) and the posterior error covariance matrix P _po (k) supplied from the positioning calculation unit 63D. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). Then, the iterative estimating unit 94C executes an iterative estimating process (FIG. 6) using a Kalman filter using the initial values x (0 | 0), P (0 | 0) and the observed value (measured value). time _{t k} for each of the state estimation vector _{x trk} the (k) can be calculated.

On the other hand, when a state change of the target Tgt from the moving state to the stationary state is detected, the iterative estimation unit 94C of the tracking calculation unit 64C determines the state estimation vector x _trk which is the latest tracking calculation result according to the detection signal ES. (K) (= x (k | k)) and the posterior error covariance matrix P _trk (k) (= P (k | k)) are supplied to the positioning calculation unit 63D via the data transfer unit 65.

At this time, the iterative estimation unit 92D of the positioning calculation unit 63D determines the state estimation vector x (k) based on the state estimation vector x _trk (k) and the posterior error covariance matrix P _trk (k) supplied from the tracking calculation unit 64C. k | k) and initial values x (0 | 0) and P (0 | 0) of the posterior error covariance matrix P (k | k). Then, the iterative estimation unit 92D executes an iterative estimation process (FIG. 6) using a Kalman filter using the initial values x (0 | 0), P (0 | 0) and the observed value (measured value). time _{t k} for each of the state estimation vector _x po a (k) can be calculated. In this case, the initial values x (0 | 0) and P (0 | 0) used in the iterative estimation unit 92D can be set, for example, as shown in the following equations (51) and (52).

As described above, the target monitoring device according to the fourth embodiment includes the state detection unit 62A similarly to the target monitoring device 2 according to the first embodiment, so that the positioning calculation unit 63D and the tracking calculation unit 64C perform the positioning calculation. The timing to switch to the tracking calculation or the timing to switch from the tracking calculation to the positioning calculation can be determined with high accuracy. Therefore, the monitoring processing unit according to the fourth embodiment can execute highly accurate calculation according to a state change of the target Tgt from the stationary state to the moving state or a state change of the target Tgt from the moving state to the stationary state. it can. When a state change of the target Tgt from the stationary state to the moving state is detected, the iterative estimating unit 94C of the tracking calculating unit 64C sets the latest positioning calculation result x _po (k) and P _po (k) to the initial values. And a tracking operation using the measured value can be performed. Therefore, even if the positioning calculation is switched to the tracking calculation, high tracking accuracy with respect to the target Tgt can be ensured.

Embodiment 5 FIG.
Next, a fifth embodiment according to the present invention will be described. In the state detection unit 62A (FIG. 4) of the first embodiment, the threshold setting unit 82A sets the thresholds TH1 and TH2 using the monitoring area information AD that defines the monitoring area SA. The threshold values TH3 and TH4 are set using the information AD. In the fifth and sixth embodiments, threshold values TH1, TH2, TH3, and TH4 are set using the latest positioning calculation result or the latest tracking calculation result instead of monitoring area information AD.

FIG. 13 is a block diagram illustrating a schematic configuration of the monitoring processing unit 11E according to the fifth embodiment. The configuration of the monitoring processing unit 11E is the same as that of the monitoring processing unit 11A according to the first embodiment except that the monitoring processing unit 11E has a state detection unit 62E and delay elements 78 and 79 in FIG. 13 instead of the state detection unit 62A in FIG. Is the same. As shown in FIG. 13, one delay element 78, the state estimate vector x _{po (k-1)} a condition that is positioning calculation result at the current time t _k of one time before than time t _k-1 The delay element 79 supplies the state estimation vector x _trk (k-1), which is the tracking operation result at time tk _-1 , to the state detection unit 62E.

FIG. 14 is a block diagram illustrating a schematic configuration of a state detection unit 62E according to the fifth embodiment. The configuration of the state detection unit 62E is the same as the configuration of the above-described state detection unit 62A, except that threshold value setting units 82A and 86A in FIG. 4 are replaced with threshold value setting units 82E and 86E in FIG. Based on one of the positioning calculation result x _po (k-1) or the tracking calculation result x _trk (k-1) and the satellite orbit information OD, the threshold value setting unit 82E calculates the above equation (1) or (2). A set of indicated threshold values TH1 and TH2 can be set. The threshold setting unit 86E calculates the above equation based on one of the positioning calculation result x _po (k-1) or the tracking calculation result x _trk (k-1), the satellite orbit information OD, and the target swing information SD. A set of thresholds TH3 and TH4 shown in (5) and (6) can be set.

As described above, in the fifth embodiment, the threshold values TH1, TH2, TH3, and TH4 are set using the state estimation vectors x _po (k-1) and x _trk (k-1) for the target Tgt. , It is possible to determine a highly reliable numerical range Δ (TH1, TH2), Δ (TH3, TH4) for detecting a change in the state of the target Tgt.

Embodiment 6 FIG.
Next, a sixth embodiment according to the present invention will be described. FIG. 15 is a block diagram illustrating a schematic configuration of the monitoring processing unit 11F according to the sixth embodiment. The configuration of the monitoring processing unit 11F is the same as that of the monitoring processing unit of the first embodiment except that the monitoring processing unit 11F includes a state detection unit 62F and delay elements 78, 79, 91, and 92 in FIG. 15 instead of the state detection unit 62A in FIG. The configuration is the same as 11A. As shown in FIG. 15, the delay elements 78,91 are positioning calculation result at the current than the time _{t k} of one time before the time _{t k-1} state estimate vector _x po (k-1) and post The error covariance matrix P _po (k-1) is supplied to the state detection unit 62E, and the delay elements 79 and 92 calculate the state estimation vector x _trk (k-1) and the tracking operation result at the time tk _-1. The posterior error covariance matrix P _trk (k−1) is supplied to the state detection unit 62E.

  FIG. 16 is a block diagram illustrating a schematic configuration of a state detection unit 62F according to the sixth embodiment. The configuration of the state detection unit 62F is the same as that of the state detection unit 62A except that threshold value setting units 82E and 86E and error ellipse calculation units 84 and 88 in FIG. 16 are provided instead of the threshold value setting units 82A and 86A in FIG. The configuration is the same.

The error ellipse calculation unit 84 converts the region data determined by the error ellipse represented by the following equation (80) into the threshold setting units 82F and 86F based on the positioning operation results x _po (k-1) and P _po (k-1). To supply.

ここで、ζは、設定パラメータである。誤差楕円とは、測位結果の誤差範囲を示す確率的な範囲のことであり、測位位置を中心とする楕円状の領域内に目標Ｔｇｔの真の位置があることを意味する。

Here, ζ is a setting parameter. The error ellipse is a stochastic range indicating an error range of the positioning result, and means that the true position of the target Tgt is located within an elliptical area centered on the positioning position.

On the other hand, based on the tracking operation results x _trk (k−1) and P _trk (k−1), the error ellipse calculation unit 88 _{converts the} region data determined by the error ellipse represented by the following equation (81) into the threshold setting unit 82F , 86F.

ここで、ζ’は、設定パラメータである。

Here, ζ ′ is a setting parameter.

  The threshold value setting unit 82F is a set of threshold values TH1 and TH2 shown in the above equations (1) and (2) based on the area data determined by the error ellipse shown in the equation (80) or (81) and the satellite orbit information OD. Can be set. Further, the threshold value setting unit 86F calculates the above equations (5) and (6) based on the area data determined by the error ellipse shown by the equation (80) or (81), the satellite orbit information OD, and the target swing information SD. A set of indicated threshold values TH3 and TH4 can be set.

  As described above, in the sixth embodiment, since the thresholds TH1, TH2, TH3, and TH4 are set using the area data determined by the error ellipse, the reliability for detecting the state change of the target Tgt is high. It is possible to determine the numerical ranges Δ (TH1, TH2), Δ (TH3, TH4). Also, there is an advantage that it is not necessary to set the area data artificially.

Modifications of the first to sixth embodiments.
As described above, various embodiments according to the present invention have been described with reference to the drawings. However, the above embodiments are merely examples of the present invention, and various embodiments other than these embodiments can be adopted.

  For example, a monitor having a combination of an arbitrary positioning operation unit selected from the positioning operation units 63A to 63D, an arbitrary tracking operation unit selected from the tracking operation units 64A to 63C, and a data transfer unit 65 It is possible to configure a processing unit.

  The hardware configuration of each monitoring processing unit in the second to fifth embodiments is, for example, one or more processors having a semiconductor integrated circuit such as a DSP, an ASIC, or an FPGA, as in the first embodiment. It should just be realized. Alternatively, the hardware configuration of the monitoring processing unit 11A may be realized by a single or a plurality of processors including an arithmetic device such as a CPU or a GPU that executes a program code of software or firmware read from a nonvolatile memory. . The hardware configuration of each monitoring processing unit may be realized by one or more processors including a combination of a semiconductor integrated circuit such as a DSP and an arithmetic device such as a CPU. Furthermore, the hardware configuration of each monitoring processing unit according to the second to fifth embodiments may be realized by the signal processing device 70 illustrated in FIG.

  Within the scope of the present invention, any combination of the above-described first to sixth embodiments, modification of any component of each embodiment, or omission of any component of each embodiment is possible.

  A target monitoring device and a target monitoring system according to the present invention use a plurality of satellites to estimate, with high accuracy, a position and a moving speed of a target which is a movable radio wave source such as a ship, an aircraft or a vehicle. Suitable for use in, for example, a target tracking system and a search and rescue system.

  St1 to St3 satellites, Rx1 to Rx2 receiving antenna, SA monitoring area, Tgt target, 1 target monitoring system, 2 target monitoring devices, 10 receivers, 11A, 11C, 11E, 11F monitoring processing unit, 12 display, 20 local Oscillator, 31, 41, 51 band-pass filter, 32, 42, 52 down converter, 33, 43, 53 A / D converter (ADC), 61, 61C measuring unit, 62A, 62E state detecting unit, 63A, 63B, 63C, 63D positioning operation unit, 64A, 64B, 64C tracking operation unit, 65 data transfer unit, 66 output control unit, 67 data storage unit, 70 signal processing unit, 71 processor, 72 memory, 73 input interface unit, 74 output interface Part, 75 communication circuit, 76 signal path, 78, 79 delay element, 8 Smoothed value calculation unit, 82A, 82B, 82H, 82E, 82F threshold value setting unit, 83 state determination unit, 84 error ellipse calculation unit, 85 smoothed value calculation unit, 86A, 86B, 86H, 86E, 86F threshold value setting unit, 87 state Judgment unit, 88 error ellipse calculation unit, 89 judgment output unit, 91A single positioning unit, 92A repetition estimation unit, 92B repetition estimation unit, 92C repetition estimation unit, 92D batch-type repetition estimation unit, 93A single positioning unit, 94A repetition estimation unit , 94B iterative estimator, 94C iterative estimator, 100 buffer memory, 101 buffer memory.

A target monitoring device according to one embodiment of the present invention includes a plurality of reception antennas which receive radio signals arriving from a target serving as a radio wave transmission source via three or more communication paths including a plurality of satellites , respectively. A target monitoring apparatus for estimating a state of the target based on a received signal, wherein a plurality of arrival time differences between the received signals or a plurality of arrival time differences and arrival frequency differences between the received signals are indicated based on the plurality of received signals. A measurement unit that sequentially calculates the measurement values of the plurality of measurement values, a state detection unit that monitors the time variation of at least one of the plurality of measurement values to detect the presence or absence of the movement of the target, and the state detection unit. When it is detected that there is no movement of the target, a position calculation using the plurality of measurement values is executed to calculate a first state estimation value indicating the estimated position of the target, and the state detection unit Yo When the state changes to the moving state from the stationary state of the target is detected, and a calculation unit for calculating a second state estimate running tracking calculation indicates both the estimated position and movement status of the target The state detection unit includes: one of coordinate information that defines a monitoring area, an estimated position of the target calculated by the positioning calculation or the tracking calculation, or an error ellipse, and an orbit calculation value of each of the plurality of satellites. And the state detection unit performs a filtering process on the at least one measured value to calculate a smoothed value, and the smoothed value changes from the numerical value range to outside the numerical value range. Then, it is determined that a state change of the target from a stationary state to a moving state has occurred .

Claims (19)

A target monitoring device for estimating the state of a target based on a plurality of reception signals obtained by a plurality of reception antennas respectively receiving radio signals arriving via three or more communication paths from a target which is a radio wave transmission source. hand,
A measurement unit that sequentially calculates a plurality of measurement values indicating the arrival time difference between the reception signals, or the arrival time difference and the arrival frequency difference between the reception signals, based on the plurality of reception signals,
A state detection unit that monitors a time variation of at least one measurement value of the plurality of measurement values and detects whether or not the target moves;
When the state detection unit detects that the target does not move, an operation of executing a positioning operation using the plurality of measurement values to calculate a first state estimated value indicating the estimated position of the target Department and
When the state detection unit detects a change in the state of the target from a stationary state to a moving state, the calculation unit performs a tracking calculation to indicate both the estimated position and the moving state of the target. Calculate state estimates,
A target monitoring device, characterized in that:   2. The target monitoring device according to claim 1, wherein the state detection unit performs a filtering process on the at least one measured value to calculate a smoothed value, and calculates the smoothed value from the set numerical value range. 3. A target monitoring apparatus characterized in that it is determined that a state change of the target from a stationary state to a moving state has occurred when the target has moved out of the range.   3. The target monitoring device according to claim 2, wherein the state detection unit is configured such that, after the smoothed value changes from outside the numerical range to within the numerical range, the smoothed value remains within the numerical range for a certain period of time. A target monitoring device characterized in that it is determined that a state change from a moving state of the target to a stationary state has occurred when the target is continuously stopped. The target monitoring device according to claim 2 or 3, wherein
The plurality of communication paths include a plurality of satellites,
The state detection unit is configured to calculate one of the coordinate information defining the monitoring area, the estimated position of the target calculated by the positioning calculation or the tracking calculation, or the error ellipse, and the orbit calculation value of each of the plurality of satellites. A target monitoring device, wherein the numerical value range is set based on the target value.   3. The target monitoring device according to claim 1, wherein the calculation unit is configured to obtain the target by the positioning calculation when the state detection unit detects a state change of the target from a stationary state to a moving state. 4. A target monitoring apparatus, wherein the tracking operation is executed using a latest calculation result obtained as an initial value and using a plurality of measurement values calculated by the measurement unit.   4. The target monitoring device according to claim 3, wherein when the state detection unit detects a state change from a moving state of the target to a stationary state, the calculation unit calculates the latest obtained by the tracking calculation. 5. A target monitoring device, wherein the positioning calculation is performed using a calculation result as an initial value and using a plurality of measurement values calculated by the measurement unit. The target monitoring device according to claim 1 or 2, wherein:
The arithmetic unit includes:
A single positioning unit that calculates a positioning vector by performing an operation according to a non-linear least squares method using the set of the plurality of measurement values as an observation vector,
A target monitoring apparatus comprising: an iterative estimating unit configured to execute an iterative estimation process using a Kalman filter using the positioning vector to calculate the first state estimation value. The target monitoring device according to claim 1 or 2, wherein:
The plurality of measurement values are observation values obtained in synchronization with each other at each time or substantially in synchronization with each other,
The target monitoring device, wherein the calculation unit executes, as the positioning calculation, an iterative estimation process using a Kalman filter using the set of the plurality of measurement values as an observation vector. The target monitoring device according to claim 1 or 2, wherein:
The plurality of measurement values are observation values obtained asynchronously with each other,
The target monitoring device according to claim 1, wherein the calculation unit performs an iterative estimation process using a Kalman filter using the plurality of measurement values as the positioning calculation. The target monitoring device according to claim 1 or 2, wherein:
The plurality of measurement values are observation values obtained asynchronously with each other,
The target monitoring device, wherein the calculation unit executes a batch process by a non-linear least squares method using a plurality of times of the plurality of measurement values collectively as the positioning calculation.   8. The target monitoring device according to claim 7, wherein the iterative estimation process using the Kalman filter includes an outlier determination process for determining whether the positioning vector is an outlier.   9. The target monitoring apparatus according to claim 8, wherein the iterative estimation processing by the Kalman filter includes an outlier determination processing for determining whether the observation vector is an outlier. The target monitoring device according to claim 1 or 2, wherein:
The arithmetic unit includes:
A single positioning unit that calculates a positioning vector by performing an operation according to a non-linear least squares method using the set of the plurality of measurement values as an observation vector,
A target monitoring apparatus comprising: an iterative estimating unit that performs iterative estimation processing by a Kalman filter using the positioning vector to calculate the second state estimation value. The target monitoring device according to claim 1 or 2, wherein:
The plurality of measurement values are observation values obtained in synchronization with each other at each time or substantially in synchronization with each other,
The target monitoring device, wherein the calculation unit performs an iterative estimation process using a Kalman filter using the plurality of measurement values as the tracking calculation. The target monitoring device according to claim 1 or 2, wherein:
The plurality of measurement values are observation values obtained asynchronously with each other,
The target monitoring device, wherein the calculation unit performs an iterative estimation process using a Kalman filter using the plurality of measurement values as the tracking calculation.   14. The target monitoring device according to claim 13, wherein the iterative estimation process using the Kalman filter includes an outlier determination process for determining whether the positioning vector is an outlier.   15. The target monitoring device according to claim 14, wherein the iterative estimation process using the Kalman filter includes an outlier determination process for determining whether the plurality of measured values are outliers.   16. The target monitoring device according to claim 15, wherein the iterative estimation process using the Kalman filter includes an outlier determination process for determining whether the plurality of measured values are outliers. A target monitoring device according to any one of claims 1 to 18,
A target monitoring system, comprising: a plurality of receiving antennas that output the plurality of reception signals.

Images