JP4970077B2 - Motion specification estimation device - Google Patents

Description

  The present invention uses a plurality of sensors such as radar sensors to input observation information such as a distance difference or a Doppler frequency difference between a movable body such as an aircraft, a ship, or a vehicle and any two of the plurality of sensors. The present invention relates to a motion parameter estimation device that estimates motion parameters such as the true position and velocity of a moving object based on these observation information.

  In conventional motion specification estimation devices that estimate target motion specifications based on observations of distance differences from targets and Doppler frequency differences from multiple sensors, including radar sensors, in general, multiple sensors synchronize with each other. The observation information obtained was used. Here, the synchronization means that the target is observed at the same time among a plurality of sensors. Asynchronous means that targets are observed at different times among a plurality of sensors.

ここで、ｚ_i,j（ｔ）は時刻ｔにおけるセンサｉ及びセンサｊ間の目標との距離差（又は時刻ｔにおけるセンサｉ及びセンサｊ間の目標とのドップラ周波数差）の観測値である。ｈ_i,j（ｘ（ｔ））は時刻ｔにおけるセンサｉ及びセンサｊ間の目標との距離差の真値を表す観測式であり、ｌ_i,j（ｘ（ｔ））は時刻ｔにおけるセンサｉ及びセンサｊ間の目標とのドップラ周波数差の真値を表す観測式である。ｖ_i,j（ｔ）は、時刻ｔにおけるセンサｉ及びセンサｊ間の目標との距離差の観測誤差（又は時刻ｔにおけるセンサｉ及びセンサｊ間の目標とのドップラ周波数差の観測誤差）である。 A conventional motion specification estimation device disclosed in Non-Patent Document 1 will be described.
Here, for the sake of simplicity of explanation, an observation model of an observation apparatus composed of a plurality of sensors is assumed as in the following equation (1).

Here, z _{i, j} (t) is an observation value of the distance difference from the target between the sensor i and the sensor j at the time t (or the Doppler frequency difference between the sensor i and the sensor j at the time t). . h _{i, j} (x (t)) is an observation formula representing the true value of the distance difference between the sensor i and sensor j at time t, and l _{i, j} (x (t)) is at time t. It is an observation formula showing the true value of the Doppler frequency difference with the target between sensor i and sensor j. v _{i, j} (t) is an observation error of a distance difference between the sensor i and the sensor j at the time t (or an observation error of a Doppler frequency difference between the sensor i and the sensor j at the time t). is there.

For example, it is assumed that an observation result of a distance difference from the target position between the sensor i and the sensor j in the xy two-dimensional orthogonal coordinate system is obtained by the observation device. At this time, the true value h _{i, j} (x (t)) of the observed value is defined as the following formula (2) and the following formula (4). When the observation result of the Doppler frequency difference between the sensor i and the sensor j and the target position is obtained by the observation device, the true value l _{i, j} (x (t)) of the observation value is expressed by the following equation (3 ), The following formula (2) and the following formula (5).

Incidentally, s _i (t), s _j (t), the sensor i at time t, respectively, represent the position vector of the j, s _i (t) dot and s _j (t) dot (hereinafter, the process of the application documents (Represented in terms of reading) indicates the speeds of the sensors i and j at time t. Further, x (t) is a target state vector represented by the above equation (6), and represents the motion specification obtained by the motion specification estimation device.

Under the assumption of the observation model as described above, the conventional motion specification estimation apparatus estimates the state vector by the following algorithm.
When there are three or more sensors, two or more equations of the above equation (2) (or the above equation (3)) can be obtained, and therefore the state vector can be obtained by solving these simultaneous equations. However, due to the influence of observation errors and the like, the above equation (1) is actually obtained, so a solution cannot be obtained uniquely. Therefore, the following method is adopted.

  First, an appropriate initial value is determined for the target state vector of the above formula (2) (or the above formula (3)), and the initial value of the left side of the above formula (2) (or the above formula (3)) is given. . Next, when an observed value of an actual distance difference (or Doppler frequency difference) is obtained by a plurality of sensors, an error signal from the initial value is obtained.

Here, the case where the observed value of the actual distance difference is obtained by a plurality of sensors will be described. From the above equation (2), the relationships of the following equations (7) to (10) can be derived. From these, since the relationship between the true value of the distance difference and its error signal and the correction amount of the target state vector is obtained, the above equation (2) is approximated by the initial value of the target state vector (the following equation (7) )), And the correction amount δx underbar to the true value (hereinafter referred to as “reading” due to the processing of the application document) is determined.

When the above equation (7) is solved using the least square method, the correction amount δx underbar is given by the following equation (11). This solves said Formula (10).

  When the above processing is repeated until the correction amount δx underbar converges, the state vector is estimated. That is, when it is determined that the convergence is determined, a target state vector is obtained by adding a correction amount to the current target state vector. On the other hand, when it is not regarded as convergence determination, a correction amount is added to the state vector, and the process according to the above-described algorithm is repeated. Note that the same algorithm as described above can also be applied to the case where the observed value of the Doppler frequency difference is obtained.

K. C. HO: "An Accurate Algebraic Solution for Moving Source Location Using TDOA and FDOA Measurements" IEEE Trans. On Signal Processing, Vol.52, No.9, Sep. 2004

  In a conventional motion specification estimation device, a target motion specification is estimated on the assumption that a plurality of sensors observe at the same time. However, in reality, it is assumed that the observation times are often different among a plurality of sensors due to a difference in observation timing, a transmission delay, or the like. Therefore, in the conventional motion specification estimation device, it is difficult to estimate the target motion specification, and there is a problem that it is not possible to shift to processing such as target tracking, which is subsequent processing. For this reason, it has been desired that the target motion specifications can be estimated even when a distance difference or a Doppler frequency difference is obtained at different times.

  The present invention has been made to solve the above-described problems, and provides a motion specification estimation device capable of estimating a target motion specification even when a plurality of sensors observe asynchronously. With the goal.

  The motion specification estimating apparatus according to the present invention includes a receiving unit having a plurality of sensors for observing a target motion, and the sensor pair based on observation results at the same time by an arbitrary sensor pair among the plurality of sensors of the receiving unit. The detection means for detecting the distance difference between the sensors and the target and the observation value of the speed difference, and the distance difference between the sensors of each sensor pair detected from the observation results at different times by a plurality of sensor pairs In addition, each sensor pair at each time calculated by reflecting the time difference between the observed value of the speed difference and the reference value for the initial value of the target motion specification at the reference time according to the target equation of motion assumed in advance. By calculating the correction amount of the target motion specifications from the difference between the distance difference between the sensors and the target and the rough estimated value of the speed difference, and correcting the motion specifications until the difference converges by the convergence calculation, the reference time Goal luck in Those comprising an asynchronous positioning means for estimating the specifications.

  According to the present invention, according to the observed value of the distance difference and the speed difference between the sensors of each sensor pair detected from the observation results at different times by the plurality of sensor pairs, and the target equation of motion assumed in advance. Approximate estimated values of the distance difference and speed difference between the sensor of each sensor pair at each time calculated by reflecting the time difference from the reference time with respect to the initial estimated value of the target motion specifications at the reference time The amount of correction of the target motion specifications is calculated from the difference between and the target motion specifications at the reference time are estimated by correcting the motion specifications until the difference converges by the convergence calculation. Asynchronous, i.e., it is possible to estimate the motion parameters of the target even from the results observed at different times, and it is possible to track the target at an early stage using the motion parameters.

Embodiment 1 FIG.
FIG. 1 is a diagram showing the configuration of a motion specification estimation apparatus according to Embodiment 1 of the present invention. The motion specification estimation apparatus according to the first embodiment includes a receiving unit 1, a distance difference / speed difference observation value detecting unit (detecting unit) 2, and an asynchronous positioning unit 4. The receiving means 1 includes sensors 11 to 1M (M is an integer greater than 1), and each sensor 11 to 1M is between a target and any two sensors among the sensors 11 to 1M (in the example of FIG. 11 and the sensors 12,..., The distance difference between the sensor 1 (M-1) and the sensor 1M) and the Doppler frequency difference are observed.

  The distance difference / speed difference observation value detection means 2 includes distance difference / speed difference observation value detection units 21 to 2N (N is an integer of M−1). The distance difference / velocity difference observation value detection units 21 to 2N detect the observation results of the distance difference between the sensors and the Doppler frequency difference for each of the two arbitrary sensors from the sensors 11 to 1M constituting the reception unit 1. The speed difference obtained from the distance difference between the two sensors and the Doppler frequency difference and the observation time are output.

  Asynchronous positioning means 4 inputs a speed difference obtained from a distance difference with a target between any two sensors in receiving means 1 and a Doppler frequency difference outputted from distance difference / speed difference observation value detecting means 2, Estimate the target state vector taking into account asynchrony. As shown in FIG. 1, the asynchronous positioning means 4 includes an initial state amount setting unit 41, time difference calculation units 421 to 42N, distance difference / speed difference approximate value calculation units 431 to 43N, and correction amount calculation units 441 to 44N. A modified state vector calculation unit 45 and a convergence determination unit 46.

  The initial state quantity setting unit 41 uses an appropriate initial value of the target state vector and a target motion specification as an initial value used when the correction state vector calculation unit 45 calculates the target state vector correction amount by the least square method. Set the reference time to estimate. The time difference calculation units 421 to 42N calculate the time difference between the reference time set by the initial state quantity setting unit 41 and the observation time.

  The distance difference / speed difference approximate value calculation units 431 to 43N are the time difference calculated by the time difference calculation units 421 to 42N and the target initial value (or convergence determination unit) at the reference time set by the initial state quantity setting unit 41. The approximate value along the equation of motion assumed in advance is calculated using the correction amount from 46 as an initial value.

  The correction amount calculation units 441 to 44N calculate the difference between the approximate value calculated by the distance difference / speed difference approximate value calculation units 431 to 43N and the observation value obtained from the distance difference / speed difference observation value detection means 2. The correction amount of the target state vector is output using this difference. The correction state vector calculation unit 45 inputs the correction amount from the correction amount calculation units 441 to 44N, and corrects the target state vector using the least square method until the difference converges.

  The convergence determination unit 46 feeds back the correction state vector to the distance difference / speed difference approximate value calculation units 431 to 43N according to the comparison result between the correction amount by the correction state vector calculation unit 45 and a predetermined threshold value, or performs convergence calculation. And the corrected state vector is output as a positioning result.

  In the motion specification estimation apparatus according to the present invention, the observation information of the distance difference and the speed difference obtained asynchronously is applied to a previously assumed motion equation (for example, constant velocity linear motion), and a reference time (target motion specification) is obtained. The target state vector (position, velocity, etc.) is estimated as the target motion specifications at the specified time at which it is desired to estimate. For this estimation, for example, a least square method is used.

In the following description, the target motion model is assumed as shown in the following equation (12).
However, t is a reference time, and x (t) is a state vector representing a true value of a target motion specification at the reference time t. ΔT is a time difference from the reference time t.

For example, assuming that the basic motion of the target is constant velocity linear motion and estimating the target position and velocity in the xy two-dimensional orthogonal coordinate system, the target state vector x (t) and the motion model are as follows: It is defined as equation (13) and equation (14) below. However, T indicates transposition of vectors and matrices, and dots indicate time differentiation. I _{n × n} is an _{n × n} unit matrix, and On _{× n} is an n _{× n} zero matrix.

Next, from the above formula (1) and the above formula (12), the following formula (15) is defined for the distance difference between the two sensors and the target.

For example, if the target is an aircraft as shown in FIG. 2 and the target is an observation value obtained by a total of four observations including the reference time t at different times t _{1 to} t ₃ , the above formula ( The following formula (16) is derived from 15). The time t _k (k = 1 to 3) is a time Δt _k before the reference time t from the following equation (17).

  The estimated value of the state vector at the reference time t (represented by a thick square and line at time t in FIG. 2) is calculated from the above equation (16) using convergence calculation by the least square method. In addition, the convergence calculation by the least square method performs the same process as the conventional motion specification estimation apparatus disclosed in Non-Patent Document 1, for example.

Next, the operation will be described.
Observed values of the distance difference and Doppler frequency difference between any two sensors among the sensors 11 to 1M are output from the receiving unit 1 to the distance difference / velocity difference observed value detecting unit 2. In the distance difference / speed difference observation value detection means 2, the distance difference / speed difference observation value detection units 21 to 2N input the observation values obtained by any two sensors in the reception means 1. In the example of FIG. 1, among the sensors 11 to 1M of the receiving unit 1, the sensor outputs of the sensors 11 and 12 are input to the distance difference / speed difference observation value detection unit 21, and the sensor outputs of the sensors 12 and 13 are the distances. The difference / speed difference observation value detection unit 22 is input, and similarly, the outputs of the two sensors are sequentially input to the distance difference / speed difference observation value detection unit.

  The distance difference / velocity difference observation value detection units 21 to 2N detect the distance difference between the two sensors and the target based on the detection results respectively input from the two sensors 11 to 1M. The speed difference is obtained from the Doppler frequency difference from the target. Thereafter, the distance difference / speed difference observation value detection units 21 to 2N output the distance difference and the speed difference between the two sensors to the correction amount calculation units 441 to 44N in the asynchronous positioning means 4, respectively. The sensor observation times are also output to the time difference calculation units 421 to 42N in the asynchronous positioning means 4, respectively.

The initial state quantity setting unit 41 in the asynchronous positioning means 4 outputs the reference time t to the time difference calculation units 421 to 42N, respectively, and calculates the initial value x (t) of the target state vector as a distance difference / speed difference approximate value. It outputs to each of the parts 431-43N. Time difference calculating section 421~42N uses the a reference time t and observing the time, the equation (17) by calculating the time difference Delta] t _k between the reference time t measurement time t _k, the distance difference / velocity difference schematic calculating It outputs to each of the parts 431-43N.

The distance difference / speed difference approximate value calculation units 431 to 43N use the initial value of the target state vector x (t) input from the initial state quantity setting unit 41 and the time difference input from the time difference calculation units 421 to 42N. According to the above equation (12), the approximate value of the distance difference or speed difference between the target state vector x (t) and the target between the sensors i, j at different times (corresponding to g _{i, j} in the above equation (16)). Are output to the correction amount calculation units 441 to 44N, respectively.

  As described above, the distance difference / speed difference approximate value calculation units 431 to 43N are the initial values of the target motion specifications at the reference time t according to the above equations (12) to (17) that are target motion equations assumed in advance. Reflecting the time difference ΔT with respect to the reference time t with respect to x (t), approximate values (schematic estimated values) of the distance difference and the speed difference between the sensors of each sensor pair at each time are calculated.

In the correction amount calculation units 441 to 44N, the approximate value g _{i, j} from the distance difference / speed difference approximate value calculation units 431 to 43N and the observation value z _{i, j} from the distance difference / speed difference observation value detection units 21 to 2N _j (observed value of distance difference and speed difference) and an error between the approximate value g _{i, j} and the observed value z _{i, j} (distance difference error Δh _{i, j} , speed difference error Δl _{i, j} ) And outputs the correction amount of the target state vector calculated using this error. The correction state vector calculation unit 45 inputs correction amounts at different times from the correction amount calculation units 441 to 44N, performs convergence calculation by the least square method, and calculates the correction amount of the target state vector and the correction amount to the state vector. The modified state vector reflected in is output.

  In the convergence determination unit 46, the correction state vector and the correction amount are input. If the correction amount is equal to or greater than the threshold value, the correction state vector is converted into the distance difference / speed difference outline instead of the initial value x (t) of the target state vector. It feeds back to the input of the value calculation parts 431-43N, and repeats the said process. If the correction amount is less than the threshold value, the correction state vector is output as a positioning result.

  As described above, according to the first embodiment, the receiving unit 1 having the plurality of sensors 11 to 1M for observing the target motion and the arbitrary sensor pair among the plurality of sensors 11 to 1M of the receiving unit 1 are used. Distance difference / velocity difference observation value detecting means 2 for detecting an observation value of a distance difference and a speed difference between targets of the sensor pair from the observation results at the same time, and a plurality of sensor pairs at different times The reference time relative to the initial value of the target motion specification at the reference time according to the observed distance difference and speed difference between the sensors of each sensor pair detected from the observation results and the target motion equation assumed in advance. The amount of correction of the target motion parameters is calculated based on the difference between the distance difference from the target of each sensor pair and the approximate value of the speed difference at each time calculated to reflect the time difference between Until the difference converges by calculation Asynchronous positioning means 4 for estimating the target motion specification at the reference time by correcting the motion specification is provided, so that the observation time differs among a plurality of sensors due to a difference in observation timing, transmission delay, etc. Even if it becomes asynchronous, it is possible to estimate the motion parameters of the target with high reliability.

  In the first embodiment, the case of estimating the target position and speed in the xy two-dimensional orthogonal coordinate system has been shown. However, the target position and speed in three dimensions are estimated by increasing the unknowns. You can also. In this case, the target state vector can be estimated by increasing the number of observation values.

Embodiment 2. FIG.
FIG. 3 is a diagram showing the configuration of a motion specification estimation apparatus according to Embodiment 2 of the present invention. In FIG. 3, the basic configuration of the motion specification estimation apparatus according to the second embodiment is the same as that shown in FIG. 1 of the first embodiment, but the asynchronous positioning means 5 performs distance difference / velocity difference observation. The difference is that the observation accuracy is input from the value detection units 21 to 2N.

  The correction state vector calculation unit 55 of the asynchronous positioning means 5 considers the observation accuracy input from the distance difference / velocity difference observation value detection units 21 to 2N, and performs the minimum two corrections on the correction amounts from the correction amount calculation units 541 to 54N. Estimate the target motion parameters using multiplication. That is, when calculating the correction amount of the state vector, the corrected state vector calculation unit 55 adds the observation accuracy of the distance difference and the Doppler frequency difference as a weighting term in the above equation (11).

  The initial state quantity setting unit 51 uses an appropriate initial value of the target state vector and a target motion specification as an initial value used when the correction state vector calculation unit 55 calculates the target state vector correction amount by the least square method. Set the reference time to estimate. The time difference calculation units 521 to 52N calculate the time difference between the reference time set by the initial state quantity setting unit 51 and the observation time.

  The distance difference / speed difference approximate value calculation units 531 to 53N use the time difference calculated by the time difference calculation units 521 to 52N and the initial value of the target at the reference time set by the initial state quantity setting unit 51 (or , Using the correction amount from the convergence determination unit 56 as an initial value), an approximate value is calculated according to a previously assumed equation of motion.

The convergence determination unit 56 feeds back the correction state vector to the distance difference / speed difference approximate value calculation units 531 to 53N according to the comparison result between the correction amount by the correction state vector calculation unit 55 and a predetermined threshold value, or performs convergence calculation. And the corrected state vector is output as a positioning result.
Since other configurations are the same as or correspond to those in FIG. 1, the same reference numerals are given and description thereof is omitted.

Next, the operation will be described.
Operations different from those of the first embodiment will be mainly described.
The correction state vector calculation unit 55 calculates the observation amount accuracy of the state vector input from the correction amount calculation units 541 to 54N and the distance difference and Doppler frequency difference input from the distance difference / speed difference observation value detection units 21 to 2N. The correction amount of the target state vector and the correction state vector reflecting this in the state vector are output by the least square method using the observation accuracy. For example, the target state vector correction amount and the corrected state vector are calculated according to an equation weighted by the observation accuracy S as in the following equation (18).

  The convergence determination unit 56 inputs the correction state vector and the correction amount from the correction state vector calculation unit 55. If the correction amount is equal to or greater than the threshold, the convergence state vector is replaced with the initial value x (t) of the target state vector. It feeds back to the input of the distance difference / speed difference approximate value calculation units 531 to 53N and repeats the above processing. If the correction amount is less than the threshold value, the correction state vector is output as a positioning result.

  As described above, according to the second embodiment, the asynchronous positioning means 5 weights the observation accuracy of the sensors 11 to 1M to the error (difference) between the observation value and the approximate value (approximate estimated value). Since the correction amount of the target motion specification is calculated, the target state vector can be estimated even if the observed values are asynchronous. In addition, more reliable estimation can be performed by taking observation accuracy into consideration.

Embodiment 3 FIG.
FIG. 4 is a diagram showing a configuration of a motion specification estimation apparatus according to Embodiment 3 of the present invention. As shown in FIG. 4, the motion specification estimation apparatus according to Embodiment 3 includes a receiving unit 1, a distance difference / speed difference observation value detection unit 2, a temporary positioning / initial value calculation unit 6, and an asynchronous positioning unit 7. In FIG. 4, the same reference numerals are given to the same or corresponding components as in FIG. 1, and description thereof will be omitted.

  The provisional positioning / initial value calculation means 6 includes an initial state quantity setting unit 61, distance difference / speed difference approximate value calculation units 621 and 622, correction amount calculation units 631 and 632, a correction state vector calculation unit 64, and a convergence determination unit 65. Is provided. The initial state quantity setting unit 61 uses an appropriate initial value of the target state vector and a target motion specification as an initial value used when the correction state vector calculation unit 64 calculates the correction amount of the target state vector by the least square method. Set the reference time to estimate.

  The distance difference / speed difference approximate value calculation units 621 and 622 use the initial value of the target state vector at the reference time set by the initial state amount setting unit 61 (or the initial correction amount from the convergence determination unit 65). Using the value as a value), an approximate value is calculated according to a previously assumed equation of motion. The correction amount calculation units 631 and 632 are the approximate values calculated by the distance difference / speed difference approximate value calculation units 621 and 622 and the observation values (sensors 11, 22) obtained from the distance difference / speed difference observation value detection units 21 and 22. 12 and an observation value obtained by the sensors 12 and 13), and the correction amount of the target state vector calculated using the difference is output.

  The correction state vector calculation unit 64 receives the correction amount from the correction amount calculation units 631 and 632 and corrects the target state vector using the least squares method until the difference converges. The convergence determination unit 65 feeds back the correction state vector to the distance difference / speed difference approximate value calculation units 621 and 622 according to the comparison result between the correction amount input from the correction state vector calculation unit 64 and a predetermined threshold value. A corrected state vector is output as an initial value of the asynchronous positioning means 7.

  The asynchronous positioning means 7 includes time difference calculation units 711 to 71N, distance difference / speed difference approximate value calculation units 721 to 72N, correction amount calculation units 731 to 73N, a correction state vector calculation unit 74, and a convergence determination unit 75. The above-described embodiment except that the positioning result by the positioning / initial value calculating means 6 is used as an initial value used when the correction amount of the target state vector is calculated by the least square method, and the initial state amount setting unit is omitted. It operates in the same way as 1.

5, the sensor 11 and the sensor 12 at time t _3, the distance difference and the Doppler frequency difference by the sensor 12 and the sensor 13 at time t ₄ is a diagram showing a case where obtained. In the case of estimating the target position and velocity in the xy two-dimensional orthogonal coordinate system, in the third embodiment, the intersection of isochronous lines (hyperbola) of detection results by different sensor sets as shown in FIG. 5 is measured. The target position and speed at the intersection are set as initial values of the target state vector in the asynchronous positioning means 7.

For example, the hyperbola indicated by the broken line in FIG. 5 indicates the time difference at which radar radio waves (or sound waves) radiated from the sensor 11 and sensor 12 to the target at time t ₃ are reflected by the target and reach the sensor 11 and sensor 12. A hyperbola (a distance difference between the target and the sensor 11 and the sensor 12 is a constant hyperbola with a constant distance difference between the target and the sensor 11 and the sensor 12 or a Doppler between the target and the sensor 11 and the sensor 12. An equal Doppler frequency difference hyperbola with a constant frequency difference).

In addition, the hyperbola indicated by the solid line in FIG. 5 is a hyperbola based on the distance difference between the two sensors detected using the time difference at which the reflected wave reaches the sensor 12 and the sensor 13 at time t ₄ and the Doppler frequency difference. (Equal distance difference hyperbola with a constant distance difference between the sensor 12 and the sensor 13 or a constant Doppler frequency difference hyperbola with a constant Doppler frequency difference between the sensor 12 and the sensor 13).

In the third embodiment, the distance difference and the Doppler frequency difference are observed by the sensor 11 and the sensor 12 at the time t ₃ , the distance difference and the Doppler frequency difference are observed by the sensor 12 and the sensor 13 at the time t ₄ , and the time is set as the reference time. By setting t ₄ and calculating the target state vector from the observed values of each sensor set, the target state vector at the intersection of these hyperbolic curves is calculated, and the state vector is calculated in the amount of correction by the asynchronous positioning means 7. Use as initial value.

Next, the operation will be described.
In the temporary positioning / initial value calculating means 6, the initial state quantity setting unit 61 sets the reference time and the initial value of the target state vector at the reference time in the distance difference / speed difference approximate value calculating units 621 and 622. The distance difference / speed difference approximate value calculation units 621 and 622 calculate the approximate value of the distance difference and speed difference between the target and each sensor set from the initial value of the target state vector input from the initial state quantity setting unit 61. It outputs to the correction amount calculation parts 631 and 632, respectively.

In the correction amount calculation units 631 and 632, the approximate values from the distance difference / speed difference approximate value calculation units 621 and 622 and the observed values from the distance difference / speed difference observation value detection units 21 and 22 (the target between the two sensors). The velocity difference obtained from the distance difference and the Doppler frequency difference) is input, and the error (distance difference error Δh _{i, j} , speed difference error Δl _{i, j} ) between this approximate value and the observed value is calculated. The correction amount of the target state vector calculated using is output. The correction state vector calculation unit 64 receives the correction amounts from the correction amount calculation units 631 and 632, and performs convergence calculation by the least square method to reflect the correction amount of the target state vector and the correction amount in the state vector. Output the modified state vector.

  The convergence determination unit 65 receives the corrected state vector and the correction amount from the corrected state vector calculation unit 64. If the correction amount is equal to or greater than the threshold value, the convergence state vector is converted into a distance difference / Feedback is made to the inputs of the speed difference approximate value calculation units 621 and 622, and the above process is repeated.

If the correction amount is less than the threshold, the convergence determination unit 65 inputs a target state vector (target position, speed, etc.) reflecting the correction amount from the correction state vector calculation unit 64, and the latest observation time (FIG. 5 outputs the target positioning result at the time t ₄ ) to the asynchronous positioning means 7. Specifically, the time t ₄ is output as the reference time to the time difference calculation units 711 to 71N of the asynchronous positioning means 7, and the target correction state vector is set as the target state vector correction amount in the asynchronous positioning means 7 using the least square method. Is output to the distance difference / velocity difference approximate value calculation units 721 to 72N as initial values at the time of calculation.

  When the reference time and the initial value of the target state vector are set from the temporary positioning / initial value calculating means 6, the asynchronous positioning means 7 estimates the target positioning result by the same processing as in the first embodiment.

  As described above, according to the third embodiment, the positioning result of the target calculated based on the observation values observed by the different sensor sets at two different times by the temporary positioning / initial value calculating means 6 is obtained by the asynchronous positioning. Since it is used as the initial value of the modified state vector calculation process by means 7, the observed values at different times can be regarded as if they were observed at the same time, and even if the observed values are asynchronous, the reliable target motion Specifications can be estimated.

Embodiment 4 FIG.
FIG. 6 is a diagram showing a configuration of a motion specification estimation apparatus according to Embodiment 4 of the present invention. As shown in FIG. 6, the motion specification estimation apparatus according to the fourth embodiment includes a receiving unit 1, a distance difference / speed difference observation value detection unit 2, a temporary positioning / initial value calculation unit 8, and an asynchronous positioning unit 7. In FIG. 6, the same or corresponding components as those in FIGS. 1 and 4 are denoted by the same reference numerals and description thereof is omitted.

  The temporary positioning / initial value calculation means 8 includes an initial state quantity setting unit 81, tracking filter units 821, 822, distance difference / speed difference approximate value calculation units 831, 832, correction amount calculation units 841, 842, and a correction state vector calculation unit. 85 and a convergence determination unit 86. The initial state quantity setting unit 81 uses an appropriate initial value of the target state vector and various target motion parameters as initial values used when the correction state vector calculation unit 85 calculates the correction amount of the target state vector by the least square method. Set the reference time to estimate the origin.

  The tracking filter units 821 and 822 are velocity differences obtained from observation values (distance difference and Doppler frequency difference) obtained by a combination of sensors in the receiving means 1 (in the example of FIG. 6, sensors 11 and 12 and sensors 12 and 13). ) And the speed difference obtained from the distance difference between the sensors in the sensor set and the Doppler frequency difference at a predetermined time is estimated. Distance difference / speed difference approximate value calculation units 831 and 832 use the target initial value at the reference time set by initial state quantity setting unit 81 (or use the correction amount from convergence determination unit 86 as the initial value). And a rough value of the target motion specification according to the motion equation assumed in advance is calculated.

  The correction amount calculation unit 841 calculates the approximate value of the target motion calculated by the distance difference / speed difference approximate value calculation unit 831 and the observed value obtained by the tracking filter unit 821 (the target between the sensors 11 and 12). A difference with a distance difference and an estimated value of a speed difference determined by a Doppler frequency difference is calculated, and a correction amount of the target state vector calculated using the difference is output. The correction amount calculation unit 842 calculates the approximate value of the target motion calculated by the distance difference / speed difference approximate value calculation unit 832 and the observed value obtained by the tracking filter unit 822 (the target between the sensors 12 and 13). A difference with a distance difference and an estimated value of a speed difference determined by a Doppler frequency difference is calculated, and a correction amount of the target state vector calculated using the difference is output.

  The correction state vector calculation unit 85 corrects the target state vector using the least square method until the correction amounts from the correction amount calculation units 841 and 842 converge. The convergence determination unit 86 feeds back the correction state vector to the distance difference / speed difference approximate value calculation units 831 and 832 according to the comparison result between the correction amount by the correction state vector calculation unit 85 and a predetermined threshold value, or performs convergence calculation. And the correction state vector is output as an initial value for correction amount calculation by the asynchronous positioning means 7.

FIG. 7 shows a case where the distance difference and the Doppler frequency difference are observed by the sensors 11 and 12 at times t ₁ and t ₂ , and the distance difference and the Doppler frequency difference are observed by the sensors 12 and 13 at time t ₃ . FIG. In the case of estimating the position and speed of a target in an xy two-dimensional orthogonal coordinate system, in the fourth embodiment, the target is tracked with any one of the sensor combinations (the speed based on the distance difference from the target and the Doppler frequency difference). The time is synchronized by estimating the difference) and performing prediction processing for the time when the observed value is obtained by the other sensor combination.

For example, in FIG. 7, when the information regarding the distance difference and the speed difference is obtained by the sensor 12 and the sensor 13 at the time t ₃ , the information regarding the distance difference and the speed difference obtained by the combination of the sensor 11 and the sensor 12 is The prediction process in ₃ is performed. As a result, as shown in FIG. 7, the intersection of the hyperbola of the observation values by the sensors 11 and 12 and the hyperbola of the observation values by the sensors 12 and 13 at time t ₃ is determined. In the fourth embodiment, the target state vector at this intersection is used as an initial value for correction amount calculation in the asynchronous positioning means 7.

Next, the operation will be described.
In the temporary positioning / initial value calculation means 8, the initial state quantity setting unit 81 sets a reference time in the tracking filter units 821 and 822, and the initial value of the target state vector at the reference time is calculated as a distance difference / speed difference approximate value calculation unit. 831 and 832.

The tracking filter unit 821 obtains an observed value (a distance difference between a target between two sensors and a speed difference due to a Doppler frequency difference) obtained from the detection result of the sensors 11 and 12 from the distance difference / speed difference observation value detection unit 21 and its The observation time is input, the time difference between the reference time set by the initial state quantity setting unit 81 and the observation time is calculated, and the time difference and the observation value input from the distance difference / speed difference observation value detection unit 21 are calculated. Is used to track the target at a certain time (time t ₁ , t ₂ in the example of FIG. 7) (estimated distance difference and speed difference between the sensors 11 and 12 with the target).

In the tracking filter unit 822, the observation value obtained from the detection results of the sensors 12 and 13 and its observation time (time t _{3 in} the example of FIG. 7) are input from the distance difference / velocity difference observation value detection unit 22 to initialize the initial value. The time difference between the reference time set in the state quantity setting unit 81 and the observation time is calculated, and the target at time t ₃ is calculated using the time difference and the observation value input from the distance difference / speed difference observation value detection unit 22. Tracking processing (estimating distance difference and speed difference from the target between the two sensors).

As described above, when the distance difference from the target and the speed difference due to the Doppler frequency difference are obtained by the sensors 12 and 13 at time t ₃ , the tracking filter unit 821 determines the distance difference from the target by the sensors 11 and 12 and the Doppler. The prediction process at the time t ₃ is executed for the speed difference due to the frequency difference. Thereby, the intersection point of the isochronous difference line of the observed values by the sensors 11 and 12 indicated by the one-dot broken line in FIG. 7 and the isochronous difference line of the observed values by the sensors 12 and 13 indicated by the solid line in FIG. In other words, the distance difference and the Doppler frequency difference by sensors 12 and 13 is obtained, to obtain a prediction value of the observation by the sensor 11 and 12 at time t _3, the distance difference and the Doppler from the target at the same time (time t ₃₎ The positioning process is performed assuming that the speed difference obtained from the frequency difference is obtained. However, the combination of sensors to be predicted is one or more.

  The distance difference / speed difference approximate value calculation units 831 and 832 use the initial value of the target state vector input from the initial state amount setting unit 81 to calculate the distance difference and speed difference between the two sensors of each sensor set. The approximate value is calculated and output to the correction amount calculation units 841 and 842, respectively.

In the correction amount calculation units 841 and 842, the approximate values from the distance difference / speed difference approximate value calculation units 831 and 832 and the observation values input from the tracking filter units 821 and 822 (the distance difference from the target and the Doppler frequency difference are obtained. (Estimated value of speed difference) is input, the error between the approximate value and the observed value (distance difference error Δh _{i, j} , speed difference error Δl _{i, j} ) is calculated, and the target state calculated using this error Output vector correction amount. The correction state vector calculation unit 85 inputs the correction amounts from the correction amount calculation units 841 and 842, and performs convergence calculation by the least square method to reflect the target state vector correction amount and the correction amount in the state vector. Output the modified state vector.

  Convergence determining unit 86 inputs the corrected state vector and its correction amount from corrected state vector calculating unit 85, and if the correction amount is equal to or greater than the threshold, the corrected state vector is replaced with the distance difference / The speed difference approximate value calculation units 831 and 832 are fed back and the above process is repeated.

If the correction amount is less than the threshold, the convergence determination unit 86 inputs a target state vector (target position, speed, etc.) reflecting the correction amount from the correction state vector calculation unit 85, and the time shown in FIG. The target positioning result at t ₃ is output to the asynchronous positioning means 7. Specifically, the time t ₃ is output as a reference time to the time difference calculation units 711 to 71N of the asynchronous positioning means 7, and the target correction state vector is set as the target state vector correction amount in the asynchronous positioning means 7 using the least square method. Is output to the distance difference / velocity difference approximate value calculation units 721 to 72N as initial values at the time of calculation.

  When the reference time and the initial value of the target state vector are set from the temporary positioning / initial value calculating unit 8, the asynchronous positioning unit 7 estimates the target positioning result by the same processing as in the first embodiment.

  As described above, according to the fourth embodiment, the provisional positioning / initial value calculation means 8 tracks the speed difference obtained from the distance difference from the target and the Doppler frequency difference obtained from the detection results of the sensors 11 and 12. If a speed difference obtained from the distance difference from the target and the Doppler frequency difference is obtained from the detection results of different sensor sets (sensors 12 and 13) at different times, the sensor set (sensor 12, 13), the target position and velocity are calculated and set as the initial value for calculating the target state vector by the least squares method. Therefore, observed values at different times are observed at the same time. It is possible to estimate the motion parameters of the target with high reliability even if the observed values are asynchronous.

Embodiment 5 FIG.
FIG. 8 is a diagram showing a configuration of a motion specification estimation apparatus according to Embodiment 5 of the present invention. As shown in FIG. 8, the motion specification estimation apparatus according to the fifth embodiment includes a reception unit 1, a distance difference / speed difference observation value detection unit 2, and an asynchronous positioning unit 9. In FIG. 8, the same or corresponding components as those in FIG.

  The asynchronous positioning means 9 includes an initial state quantity setting unit 91, tracking filter units 921 to 92N, distance difference / speed difference approximate value calculation units 931 to 93N, correction amount calculation units 941 to 94N, a correction state vector calculation unit 95, and a convergence. A determination unit 96 is provided.

  The initial state quantity setting unit 91 uses an appropriate initial value of the target state vector and a target motion specification as an initial value used when the correction state vector calculation unit 95 calculates the target state vector correction amount by the least square method. Set the reference time to estimate. The tracking filter units 921 to 92N input the distance difference / velocity difference observation value detection units 21 to 2N distance difference and Doppler frequency difference based on the detection result obtained by the combination of the sensors in the receiving means 1, and perform the tracking process. To do.

  The distance difference / speed difference approximate value calculation units 931 to 93N use the target initial value at the reference time set by the initial state amount setting unit 91 (or use the correction amount from the convergence determination unit 96 as the initial value). And an approximate value in accordance with a previously assumed equation of motion is calculated. The correction amount calculation units 941 to 94N use the difference between the approximate value calculated by the distance difference / velocity difference approximate value calculation units 931 to 93N and the observed value obtained from the tracking filter units 921 to 92N, as a target state vector. The amount of correction is calculated.

  The correction state vector calculation unit 95 inputs the correction amount from the correction amount calculation units 941 to 94N, and corrects the target state vector using the least square method until the difference converges. The convergence determination unit 96 feeds back the corrected state vector to the distance difference / velocity difference approximate value calculating units 931 to 93N according to the comparison result between the correction amount by the corrected state vector calculating unit 95 and a predetermined threshold, or performs convergence calculation. And the corrected state vector is output as a positioning result.

Next, the operation will be described.
The initial state quantity setting unit 91 sets the reference time in the tracking filter units 921 to 92N, and sets the initial value of the target state vector at the reference time to the distance difference / speed difference approximate value calculation units 931 to 93N.

  The tracking filter units 921 to 92N are observation values (speed difference due to distance difference from target and Doppler frequency difference) obtained from the detection results of the sensor sets from the distance difference / speed difference observation value detection units 21 to 2N and their observation times. Are input, the time difference between the reference time set by the initial state quantity setting unit 91 and the observation time is calculated, and the time difference and the observation value input from the distance difference / speed difference observation value detection unit are used. The target is then tracked (estimated distance difference and speed difference between the sensors of each sensor set).

  Among the tracking filter units 921 to 92N, a certain tracking filter unit is in an initial state when an observation value is obtained from a distance difference / velocity difference observation value detection unit with a different sensor set by another tracking filter unit at a certain time. The time difference between the reference time set in the quantity setting unit 91 and the observation time is calculated, and using this time difference and the observation value input from the distance difference / velocity difference observation value detection unit, the difference from the target at the time is calculated. The distance difference and the speed difference are predicted (the distance difference from the target and the speed difference are estimated).

  For example, when the tracking filter unit 922-92N obtains a distance difference and speed difference from the target by the detection results of the sensors 12, 13-1 (N-1), 1N at a certain time, the tracking filter unit 921 By obtaining the predicted value, the positioning process is performed assuming that the speed difference obtained by the distance difference from the target and the Doppler frequency difference is obtained at the same time. However, the combination of sensors to be predicted is one or more.

  The distance difference / speed difference approximate value calculation units 931 to 93N calculate the approximate values of the distance difference and the speed difference between the target and each sensor set using the initial value of the target state vector input from the initial state quantity setting unit 91. And output to the correction amount calculation units 941 to 94N, respectively.

In the correction amount calculation units 941 to 94N, the approximate values from the distance difference / speed difference approximate value calculation units 931 to 93N and the observation values input from the tracking filter units 921 to 92N (estimated values of distance difference and speed difference from the target) To calculate the error between the approximate value and the observed value (distance difference error Δh _{i, j} , speed difference error Δl _{i, j} ), and output the correction amount of the target state vector calculated using this error To do. In the correction state vector calculation unit 95, the correction amount from the correction amount calculation units 941 to 94N is input, and the convergence calculation by the least square method is performed to reflect the correction amount of the target state vector and the correction amount in the state vector. Output the modified state vector.

  Convergence determining unit 96 inputs the corrected state vector and the correction amount from corrected state vector calculating unit 95, and if the correction amount is equal to or greater than the threshold, the corrected state vector is replaced with the distance difference / Feedback is made to the input of the approximate speed difference calculation units 931 to 93N, and the above process is repeated. When the correction amount is less than the threshold value, the convergence determination unit 96 inputs a target state vector (target position, speed, etc.) reflecting the correction amount from the correction state vector calculation unit 95, and as a target positioning result. Output.

  As described above, according to the fifth embodiment, the asynchronous positioning means 9 tracks the speed difference obtained from the distance difference from the target and the Doppler frequency difference obtained from the detection result of the sensor set, and performs another time. In this case, when the distance difference and speed difference from the target are obtained from the detection results of different sensor sets, the movement specifications of the target are estimated after performing time alignment with respect to the sensor set based on the tracking result. The observed values at the time can be regarded as if they were observed at the same time, and even if the observed values are asynchronous, it is possible to estimate the motion parameters of the target with high reliability.

Embodiment 6 FIG.
FIG. 9 is a diagram showing the configuration of a motion specification estimation apparatus according to Embodiment 6 of the present invention. As shown in FIG. 9, the motion specification estimation apparatus according to Embodiment 6 includes a receiving unit 1, a distance difference / velocity difference observation value detecting unit 2, and an asynchronous positioning unit 10. In FIG. 9, components that are the same as or equivalent to those in FIG.

  Asynchronous positioning means 10 includes initial state quantity setting unit 101, time difference calculation units 1021 to 102N, distance difference / speed difference approximate value calculation units 1031 to 103N, correction amount calculation units 1041 to 104N, correction state vector calculation unit 105, convergence A determination unit 106 and a convergence calculation count unit 107 are provided. Configurations other than the convergence determination unit 106 and the convergence calculation counting unit 107 are the initial state quantity setting unit 41, time difference calculation units 421 to 42N, distance difference / speed difference approximate value calculation unit shown in FIG. 1 of the first embodiment. 431 to 43N, the correction amount calculation units 441 to 44N, and the correction state vector calculation unit 45 operate in the same manner.

  The convergence determination unit 106 calculates the distance difference between the correction state vectors according to the comparison result between the correction amount by the correction state vector calculation unit 105 and a predetermined threshold until the convergence calculation counting unit 107 counts the calculation number limit value. / Feedback to the approximate speed difference calculation units 1031 to 103N, or the convergence calculation is terminated and the corrected state vector is output as the positioning result.

  The convergence calculation counting unit 107 counts the number of convergence calculations until the correction state vector is fed back to the distance difference / speed difference approximate value calculation units 1031 to 103N by the convergence determination unit 106 until a predetermined calculation number limit value is reached. . Here, the calculation number limit value is a convergence that gives an optimal calculation load and estimation accuracy determined according to the calculation performance of the computer that realizes the motion specification estimation device according to Embodiment 6 and the sampling rate of the detection data, etc. The number of operations.

Next, the operation will be described.
Hereinafter, operations unique to the motion specification estimation apparatus according to Embodiment 6 will be mainly described.
First, when the convergence determination unit 106 performs the convergence calculation, the convergence calculation counting unit 107 initializes the count value to 0 and sets the calculation count limit value as the count limit value. As in the first embodiment, when the correction amount is input from the correction state vector calculation unit 105, the convergence determination unit 106 compares the correction amount with a predetermined threshold value. At this time, if the correction amount is larger than the threshold and does not converge, the convergence calculation counting unit 107 increases the count value by one.

  If the correction amount does not converge even when the count value reaches the calculation limit value, the convergence calculation counting unit 107 transmits a calculation end signal to the convergence determining unit 106. Thereby, the convergence determination part 106 outputs the correction state vector (correction state vector which added correction amount) reflecting the correction amount at the time of the completion of calculation as a target positioning result. When the correction amount converges before the count value reaches the calculation limit value, the correction state vector to which the converged correction amount is added is output as a target positioning result, and the convergence calculation counting unit 107 stops counting. To do.

  As described above, according to the sixth embodiment, when the convergence determination unit 106 reaches the predetermined number of calculation limit value by the convergence calculation counting unit 107, the convergence calculation is terminated and the motion at that time is reached. Since the specification estimation value is output, the number of times of the convergence calculation by the corrected state vector calculation unit 105 is limited, so that the execution exceeding the performance of the motion specification estimation device according to the sixth embodiment is suppressed, and Since the calculation is completed at a predetermined time, the time for convergence calculation can be shortened.

  In the sixth embodiment, the case where the convergence calculation counting unit 107 is applied to the configuration according to the first embodiment has been described. However, the same may be applied to the configurations of the second to fifth embodiments. Can be obtained.

Embodiment 7 FIG.
FIG. 10 is a diagram showing the configuration of a motion specification estimation apparatus according to Embodiment 7 of the present invention. As shown in FIG. 10, the motion specification estimation apparatus according to the seventh embodiment includes a receiving unit 1, a distance difference observation value detecting unit 2a, and an asynchronous positioning unit 12a. In FIG. 10, the same reference numerals are given to the same or corresponding components as those in FIG.

  The distance difference observation value detection means 2a includes distance difference observation value detection units 111 to 11N (N is an integer of M−1). The distance difference observation value detection units 111 to 11N detect the distance difference between the two sensors from the sensors 11 to 1M constituting the receiving unit 1 for each of the two sensors, and the distance difference between the two sensors and the observation thereof. Output the time.

  As the sensors 11 to 1M according to the seventh embodiment, for example, a radar sensor that calculates a distance difference from a target from a time difference when a reflected radio wave of a radar radio wave radiated to the target arrives can be considered. In addition, when calculating the Doppler frequency difference as shown in the above formula (3), the initial value of the target position is required, but the target position is the distance from the target as shown in the above formula (2). An observation of the difference is required. Therefore, in the seventh embodiment, the calculation process is simplified by estimating the target state vector using only the distance difference between the target between any two sensors without calculating the Doppler frequency difference. .

  The asynchronous positioning means 12a receives the distance difference between any two sensors in the receiving means 1 output from the distance difference observation value detecting means 2a, and estimates the target state vector in consideration of asynchrony. As shown in FIG. 10, the asynchronous positioning means 12a includes an initial state quantity setting unit 121, time difference calculation units 1221 to 122N, distance difference approximate value calculation units 1231 to 123N, correction amount calculation units 1241 to 124N, a correction state vector. A calculation unit 125 and a convergence determination unit 126 are provided.

  The initial state quantity setting unit 121 uses an appropriate initial value of the target state vector and target motion specifications as initial values used when the correction state vector calculation unit 125 calculates the target state vector correction amount by the least square method. Set the reference time to estimate. The time difference calculation units 1221 to 122N calculate the time difference between the reference time set by the initial state quantity setting unit 121 and the observation time.

  The distance difference approximate value calculation units 1231 to 123N use the time difference calculated by the time difference calculation units 1221 to 122N and the target initial value at the reference time set by the initial state quantity setting unit 121 (or convergence determination). Using the correction amount from the unit 126 as an initial value), an approximate value is calculated according to a previously assumed equation of motion.

  The correction amount calculation units 1241 to 124N calculate the difference between the approximate value calculated by the distance difference approximate value calculation units 1231 to 123N and the observation value obtained from the distance difference observation value detection means 2a, and use this difference. The calculated correction amount is output. The correction state vector calculation unit 125 receives the correction amount from the correction amount calculation units 1241 to 124N, and corrects the target state vector using the least squares method until the difference converges.

  The convergence determination unit 126 feeds back the correction state vector to the distance difference approximate value calculation units 1231 to 123N according to the comparison result between the correction amount by the correction state vector calculation unit 125 and a predetermined threshold value, or ends the convergence calculation. The corrected state vector is output as a positioning result.

Next, the operation will be described.
The observed value of the distance difference from the target between any two sensors among the sensors 11 to 1M is output from the receiving means 1 to the distance difference observed value detecting means 2a. In the distance difference observation value detection means 2a, the distance difference observation value detection units 111 to 11N input observation values obtained by arbitrary two sensors in the reception means 1. In the example of FIG. 10, among the sensors 11 to 1M of the receiving unit 1, the sensor outputs of the sensors 11 and 12 are input to the distance difference observation value detection unit 111, and the sensor outputs of the sensors 12 and 13 are the distance difference observation values. Similarly, the outputs of the two sensors are sequentially input to the distance difference observation value detection unit.

  The distance difference observation value detection units 111 to 11N detect the distance difference between the two sensors and the target based on the detection results respectively input from the two sensors 11 to 1M. Thereafter, the distance difference observation value detection units 111 to 11N output the distance difference between the two sensors and the target to the correction amount calculation units 1231 to 123N in the asynchronous positioning means 12a, respectively, and the observation times of the sensors are also asynchronous. It outputs to the time difference calculation parts 1221 to 122N in the positioning means 12a.

The initial state quantity setting unit 121 in the asynchronous positioning means 12a outputs the reference time t to the time difference calculation units 1221 to 122N, and the initial value x (t) of the target state vector is the distance difference approximate value calculation unit 1231 to 121. To 123N. Time difference calculating section 1221~122N uses the reference time t and the observation time, the above equation (17) calculates the time difference Delta] t _k between the reference time t and the observation time t _k, distance difference approximate value calculating unit 1231～ To 123N.

In the distance difference approximate value calculation units 1231 to 123N, the initial value of the target state vector x (t) input from the initial state quantity setting unit 121 and the time difference input from the time difference calculation units 1221 to 122N are used. 12) to calculate the approximate value g _{i, j} of the distance difference between the target state vector x (t) (target position, velocity, etc.) and the sensor i, j at different times according to 12). (Corresponding to g _{i, j} in the above equation (16)).

The correction amount calculation units 1241 to 124N include the approximate value g _{i, j} from the distance difference approximate value calculation units 1231 to 123N and the observation value z _{i, j} from the distance difference observation value detection units 111 to 11N (observation of the distance difference). Value) is calculated _, an error (distance difference error Δh _{i, j} ) between the approximate value g _{i, j} and the observed value z _{i, j} is calculated, and a correction amount calculated using this error is output. The correction state vector calculation unit 125 inputs the correction amount from the correction amount calculation units 1241 to 124N, performs a convergence calculation by the least square method, and corrects the target state vector correction amount and the correction amount reflected in the state vector. A state vector (a state vector to which the correction amount is added) is output.

  In the convergence determination unit 126, the correction state vector and the correction amount are input. If the correction amount is equal to or greater than the threshold, the correction state vector is converted into the distance difference approximate value calculation unit instead of the initial value x (t) of the target state vector. The above processing is repeated by feeding back to the inputs of 1231-123N. If the correction amount is less than the threshold value, the correction state vector is output as a positioning result.

  As described above, according to the seventh embodiment, only the observation value of the distance difference from the target between any two sensors is input from the sensors 11 to 1M, and the target state vector is estimated based on the distance difference. Therefore, even if the Doppler frequency difference cannot be obtained, the target motion specification can be estimated. Thereby, the calculation process for obtaining the Doppler frequency difference can be omitted, and the same effect as in the first embodiment can be obtained with a simple configuration.

Embodiment 8 FIG.
FIG. 11 is a diagram showing the configuration of a motion specification estimation apparatus according to Embodiment 8 of the present invention. As shown in FIG. 11, the basic configuration of the motion specification estimation apparatus according to the eighth embodiment is the same as that shown in FIG. 10 of the seventh embodiment, but the asynchronous positioning means 13a performs distance difference observation. The difference is that the observation accuracy is input from the value detectors 111 to 11N.

  The correction state vector calculation unit 135 in the asynchronous positioning means 13a uses the least square method for the correction amounts from the correction amount calculation units 1341 to 134N in consideration of the observation accuracy from the distance difference observation value detection units 111 to 11N. To estimate the target motion specifications. In other words, the correction state vector calculation unit 135 adds the observation accuracy of the distance difference as a weighting term in the above equation (11) when calculating the correction amount of the state vector.

  The initial state quantity setting unit 131 uses an appropriate initial value of the target state vector and the target motion specifications as initial values used when the correction state vector calculation unit 135 calculates the target state vector correction amount by the least square method. Set the reference time to estimate. The time difference calculation units 1321 to 132N calculate the time difference between the reference time set by the initial state quantity setting unit 131 and the observation time.

  The distance difference approximate value calculation units 1331 to 133N use the time difference calculated by the time difference calculation units 1321 to 132N and the target initial value at the reference time set by the initial state quantity setting unit 131 (or convergence determination). Using the correction amount from the unit 136 as an initial value), an approximate value is calculated according to the motion equation assumed in advance.

  The convergence determination unit 136 outputs the correction state vector to the distance difference approximate value calculation units 1331 to 133N when the correction amount by the correction state vector calculation unit 135 is larger than the threshold value, and ends the convergence calculation when the correction amount is smaller than the threshold value. Then, the corrected state vector is output as a positioning result. The other configurations are the same as or correspond to those in FIG.

Next, the operation will be described.
Operations different from those of the seventh embodiment are mainly described.
The correction state vector calculation unit 135 uses the correction amount of the state vector input from the correction amount calculation units 1341 to 134N and the observation accuracy of the distance difference input from the distance difference observation value detection units 1331 to 133N to observe the distance difference. A correction amount of the target state vector and a correction state vector reflecting this in the state vector are output by the least square method weighted with accuracy. For example, the target state vector correction amount and the corrected state vector are calculated according to an equation weighted by the observation accuracy S as in the above equation (18).

  Convergence determining unit 136 inputs the corrected state vector and the correction amount from corrected state vector calculating unit 135, and if the correction amount is equal to or greater than the threshold, the corrected state vector is replaced with the initial value x (t) of the target state vector. The distance difference approximate value calculation units 1331 to 133N are fed back to the input and the above process is repeated. If the correction amount is less than the threshold value, the correction state vector is output as a positioning result.

  As described above, according to the eighth embodiment, the target state vector can be estimated even if the observation values are asynchronous. In addition, since only the distance difference from the target between any two sensors may be used as observation information, even if the sensor does not provide a Doppler frequency difference as in the seventh embodiment, the target motion specifications are estimated. be able to. Thereby, the calculation process for obtaining the Doppler frequency difference can be omitted, and the same effect as in the first embodiment can be obtained with a simple configuration. In addition, more reliable estimation can be performed by considering the observation accuracy of the distance difference.

Embodiment 9 FIG.
FIG. 12 is a diagram showing the configuration of a motion specification estimation apparatus according to Embodiment 9 of the present invention. As shown in FIG. 12, the motion specification estimating apparatus according to the ninth embodiment includes a receiving unit 1, a distance difference observation value detecting unit 2a, a temporary positioning / initial value calculating unit 14a, and an asynchronous positioning unit 15a. In FIG. 12, the same reference numerals are given to the same or corresponding components as those in FIG.

  The temporary positioning / initial value calculating means 14a includes an initial state quantity setting unit 141, a distance difference approximate value calculating units 1421, 1422, correction amount calculating units 1431 and 1432, a corrected state vector calculating unit 144, a convergence determining unit 145, a memory 146, and An initial value calculation unit 147 is provided. The initial state quantity setting unit 141 includes an appropriate initial value of the target state vector and target motion specifications as initial values used when the correction state vector calculation unit 144 calculates the correction amount of the target state vector by the least square method. Set the reference time to estimate.

  The distance difference approximate value calculation units 1421 and 1422 use the target initial value at the reference time set by the initial state amount setting unit 141 (or use the correction amount from the convergence determination unit 145 as the initial value). Approximate values are calculated according to the motion equation assumed in advance. The correction amount calculation units 1431 and 1432 are the approximate values calculated by the distance difference approximate value calculation units 1421 and 1422 and the observation values obtained from the distance difference observation value detection units 111 and 112 (sensors 11 and 12 and sensors 12 and 13, respectively). And the amount of correction of the target state vector calculated using this difference is output.

  The correction state vector calculation unit 144 receives the correction amounts from the correction amount calculation units 1431 and 1432 and corrects the target state vector using the least squares method until the difference converges. The convergence determination unit 145 feeds back the corrected state vector to the distance difference approximate value calculating units 1421 and 1422 according to the comparison result between the correction amount by the corrected state vector calculating unit 144 and a predetermined threshold value, or ends the convergence calculation. The corrected state vector is output to the memory 146.

The memory 146 stores a target positioning result (target state vector) determined to have been converged by the convergence determination unit 145. The initial value calculation unit 147 is a target position x (t ₁ ) hat, x (t ₂ ) that is a positioning result at two different times t ₁ and t ₂ represented by the following formula (19) and the following formula (20). ) An initial speed x dot (expressed as a reading due to the processing of the application document) is obtained from the hat (represented as a reading due to the processing of the application document) according to the following equation (21). The target positioning result at the latest time is output as an initial value to the asynchronous positioning means 15a.

  Asynchronous positioning means 15a includes time difference calculation units 1511 to 151N, distance difference approximate value calculation units 1521 to 152N, correction amount calculation units 1531 to 153N, correction state vector calculation unit 154, and convergence determination unit 155. The positioning result by the initial value calculating means 14a is used as an initial value used when calculating the target state vector correction amount by the least square method, and the same as in the eighth embodiment except that the initial state amount setting unit is omitted. To work.

13, sensor 11 and sensor 12 at time t _1, the sensor 12 and the sensor 13 at time t _2, the sensor 11 and the sensor 12 at time t _3, the distance difference is obtained by the sensor 12 and the sensor 13 at time t ₄ It is a figure which shows a case. In the case of estimating the target position and velocity in the xy two-dimensional orthogonal coordinate system, in the ninth embodiment, the intersection of isochronous lines (hyperbola) of detection results by different sensor sets as shown in FIG. 13 is measured. The target position and speed at the intersection are set as the initial value of the target state vector in the asynchronous positioning means 15a.

For example, a hyperbola indicated by a broken line in FIG. 13 indicates two sensors detected using a time difference at which the radar radio waves radiated to the target at the times t ₁ and t ₃ are reflected by the target and reach the sensors 11 and 12. It is a hyperbola (equal distance difference hyperbola with a constant distance difference between the sensor 11 and the sensor 12 with respect to the target) due to the distance difference between the target and the target. Further, the hyperbola indicated by the solid line in FIG. 13 is a hyperbola (due to a difference in distance from the target between the two sensors detected using the time difference at which the reflected radio waves reach the sensors 12 and 13 at times t ₂ and t ₄ ( The equidistant difference hyperbola with a constant distance difference between the sensor 12 and the sensor 13 with respect to the target.

In the ninth embodiment, the distance difference is observed by the sensor 11 and the sensor 12 at the times t ₁ and t ₃ , the distance difference is observed by the sensor 12 and the sensor 13 at the times t ₂ and t ₄ , and the time t _{2 is used} as the reference time. , T ₄ and calculating the target state vector from the observation values of each sensor set, the target state vector at the intersection of these hyperbolic curves is calculated, and the correction amount is calculated by the asynchronous positioning means 15a. Used as the initial value in.

Next, the operation will be described.
In the temporary positioning / initial value calculating means 14a, the initial state quantity setting unit 141 sets the reference time and the initial value of the target state vector at the reference time in the distance difference approximate value calculating units 1421 and 1422. The distance difference approximate value calculation units 1421 and 1422 calculate an approximate value of the distance difference between the two sensors of each sensor set from the initial value of the target state vector input from the initial state amount setting unit 141 and the correction amount. It outputs to the calculation parts 1431 and 1432, respectively.

In the correction amount calculation units 1431 and 1432, the approximate values from the distance difference approximate value calculation units 1421 and 1422 and the observation values from the distance difference observation value detection units 111 and 112 (distance difference between the targets between the two sensors) are input. Then, an error (distance difference error Δh _{i, j} ) between the approximate value and the observed value is calculated, and the correction amount of the target state vector calculated using this error is output. The correction state vector calculation unit 144 receives the correction amounts from the correction amount calculation units 1431 and 1432, performs a convergence calculation by the least square method, and corrects the target state vector correction amount and the correction amount in the state vector. Output a state vector.

  The convergence determination unit 145 receives the corrected state vector and the correction amount from the corrected state vector calculation unit 144. If the correction amount is equal to or greater than the threshold value, the convergence determination unit 145 uses the corrected state vector instead of the initial value of the target state vector. It feeds back to the input of the value calculation parts 1421, 1422 and repeats the above process.

If the correction amount is less than the threshold value, the convergence determination unit 145 inputs a target state vector (target position, speed, etc.) reflecting the correction amount from the correction state vector calculation unit 144, and the latest observation time (FIG. 13, the time t ₁ and t ₂ that are close in time are the time t ₂ , and the times t ₃ and t ₄ are sequentially output to the memory 146 as target positioning results at the time t ₄ ). Hold.

The initial value calculation unit 147 defines target position x (t ₂ ) hat and x (t ₄ ) hat as positioning results at _two times t ₂ and t ₄ by the above formulas (19) and (20). By substituting these into equation (21) above, the target initial velocity x dot is obtained. The target positioning result at the latest time is output as an initial value to the asynchronous positioning means 15a. Specifically, the time t ₄ as a reference time and outputs the time difference calculating portion 1511~151N asynchronous positioning means 15a, the correction state vector of the target, the method of least squares correction amount of the target state vector in an asynchronous positioning means 15a Is output to the distance difference approximate value calculation units 1521 to 152N as initial values for the calculation.

  Asynchronous positioning means 15a estimates the target positioning result by the same processing as in the seventh embodiment when the reference time and the initial value of the target state vector are set from temporary positioning / initial value calculating means 14a.

  As described above, according to the ninth embodiment, the temporary positioning / initial value calculating means 14a is calculated based on the distance difference between the two sensors and the target observed by different sensor sets at two different times. Since the target positioning result is used as the initial value of the corrected state vector calculation process by the asynchronous positioning means 15a, the observed values at different times can be regarded as if they were observed at the same time, and the observed values are asynchronous. However, it is possible to estimate the motion specifications of the target with high reliability. In addition, since only the distance difference from the target between any two sensors may be used as observation information, even if the sensor does not provide a Doppler frequency difference as in the seventh embodiment, the target motion specifications are estimated. be able to.

Embodiment 10 FIG.
FIG. 14 is a diagram showing a configuration of a motion specification estimation apparatus according to Embodiment 10 of the present invention. As shown in FIG. 14, the motion specification estimation apparatus according to the tenth embodiment includes a receiving means 1, a distance difference observation value detecting means 2a, a temporary positioning / initial value calculating means 16a, and an asynchronous positioning means 15a. In FIG. 14, the same or corresponding components as those in FIGS. 1, 6, and 10 are denoted by the same reference numerals, and description thereof is omitted.

  The temporary positioning / initial value calculation means 16a includes an initial state quantity setting unit 161, tracking filter units 1621 and 1622, distance difference approximate value calculation units 1631 and 1632, correction amount calculation units 1641 and 1642, a correction state vector calculation unit 165, and a convergence. A determination unit 166, a memory 167, and an initial value calculation unit 168 are provided. The initial state quantity setting unit 161 uses an appropriate initial value of the target state vector and target motion parameters as initial values used when the correction state vector calculation unit 165 calculates the correction amount of the target state vector by the least square method. Set the reference time to estimate the origin.

  The tracking filter units 1621 and 1622 input observation values (distance difference) obtained by a combination of sensors in the receiving means 1 (in the example of FIG. 14, the sensors 11 and 12 and the sensors 12 and 13), and from this, a predetermined time is obtained. The distance difference between the two sensors in the sensor set and the target is estimated. The distance difference approximate value calculation units 1631 and 1632 use the target initial value at the reference time set by the initial state amount setting unit 161 (or use the correction amount from the convergence determination unit 166 as the initial value). Approximate values of target motion specifications along a pre-assumed equation of motion are calculated.

  The correction amount calculation unit 1641 calculates the approximate value of the target motion calculated by the distance difference approximate value calculation unit 1631 and the estimated value obtained by the tracking filter unit 1621 (distance difference between the target between the sensors 11 and 12). And the amount of correction of the target state vector calculated using this difference is output. The correction amount calculation unit 1642 is configured to calculate the approximate value of the target motion calculated by the distance difference approximate value calculation unit 1632 and the estimated value obtained by the tracking filter unit 1622 (distance difference between the target between the sensors 12 and 13). And the amount of correction of the target state vector calculated using this difference is output.

  The correction state vector calculation unit 165 receives the correction amounts from the correction amount calculation units 1641 and 1642, and corrects the target state vector using the least square method until the difference converges. The convergence determination unit 166 feeds back the correction state vector to the distance difference approximate value calculation units 1631 and 1632 or ends the convergence calculation according to the comparison result between the correction amount by the correction state vector calculation unit 165 and a predetermined threshold. The corrected state vector is output to the memory 167.

  The memory 167 stores a target positioning result (target state vector) determined to have been converged by the convergence determination unit 166. The initial value calculation unit 168 obtains the target initial speed according to the above equation (21) from the target position, which is the positioning result at two different times represented by the above equation (19) and the above equation (20). The target positioning result at the latest time is output as an initial value to the asynchronous positioning means 15a.

FIG. 15 is a diagram showing a case where a distance difference is observed between the sensor 11 and the sensor 12 at times t ₁ and t ₂ , and a distance difference is observed between the sensor 12 and the sensor 13 at times t ₃ and t ₄ . In the case of estimating a target position and velocity in an xy two-dimensional orthogonal coordinate system, in Embodiment 10, a target obtained by any one of the sensor combinations is tracked (estimated distance difference from the target), and Time synchronization is achieved by performing a prediction process for a time at which a distance difference is obtained by the other sensor combination.

For example, in Figure 15, when the distance difference is obtained by the sensor 12 and the sensor 13 at time t _3, carrying out the prediction processing the distance difference due to the combination of the sensor 11 and the sensor 12 at time t _3, t _4. As a result, as shown in FIG. 15, the intersection of the hyperbola (solid line) of the observation values by the sensors 11 and 12 and the hyperbola (dotted line) of the observation values by the sensors 12 and 13 at time t ₃ is determined. In addition, the intersection of the hyperbola (solid line) of the observation values obtained by the sensors 11 and 12 and the hyperbola (one-dot broken line) of the observation values obtained by the sensors 12 and 13 at time t ₄ is determined. In the tenth embodiment, the target state vector at these intersections is used as the initial value for calculating the correction amount in the asynchronous positioning means 15a.

Next, the operation will be described.
In the temporary positioning / initial value calculation means 16a, the initial state quantity setting unit 161 sets the reference time in the tracking filter units 1621 and 1622, and the initial value of the target state vector at the reference time is calculated as the distance difference approximate value calculation units 1631 and 1632. Set to.

The tracking filter unit 1621 receives the observation value (distance difference from the target) obtained from the detection results of the sensors 11 and 12 and the observation time from the distance difference observation value detection unit 111, and the initial state quantity setting unit 161 The time difference between the set reference time and the observation time is calculated, and the observation value input from the time difference and the distance difference observation value detection unit 111 is used, and at a certain time (times t ₁ and t ₂ in FIG. 15). The target is tracked (a distance difference between the sensor 11 and the target is estimated).

  The tracking filter unit 1622 inputs the observation value and the observation time obtained from the detection results of the sensors 12 and 13 from the distance difference observation value detection unit 112, and the reference time and observation set in the initial state quantity setting unit 161. The time difference from the time is calculated, and the target tracking process (estimating the distance difference between the two sensors) is performed using the time difference and the observation value input from the distance difference observation value detection unit 112.

In this way, when the distance difference from the target is obtained by the sensors 12 and 13 at time t ₃ , the tracking filter unit 1621 executes the prediction process at time t ₃ for the distance difference from the target by the sensors 11 and 12. To do. Thereby, the intersection point of the isochronous difference line of the observed values by the sensors 11 and 12 indicated by the one-dot broken line in FIG. 15 and the isochronous difference line of the observed values by the sensors 12 and 13 indicated by the solid line in FIG. That is, by obtaining the predicted value of the observation by the sensors 11 and 12 at the times t ₃ and t _{4 at which the} distance difference is obtained by the sensors 12 and 13, the distance difference from the target can be obtained at the same time (time t ₃ and t ₄ ). The positioning process is performed assuming that it is obtained. However, the combination of sensors to be predicted is one or more.

  The distance difference approximate value calculation units 1631 and 1632 calculate the approximate value of the distance difference between the two sensors of each sensor set with the target using the initial value of the target state vector input from the initial state quantity setting unit 161. The correction values are output to the correction amount calculation units 1641 and 1642, respectively.

In the correction amount calculation units 1641 and 1642, the approximate values from the distance difference approximate value calculation units 1631 and 1632 and the estimated values input from the tracking filter units 1621 and 1622 (distance difference between the targets between the two sensors) are input. An error between the value and the observed value (distance difference error Δh _{i, j} ) is calculated, and a correction amount of the target state vector calculated using this error is output. In the correction state vector calculation unit 165, the correction amount is input from the correction amount calculation units 1641 and 1642, the convergence calculation by the least square method is performed, and the correction amount of the target state vector and the correction amount reflected in the state vector Output a state vector.

  The convergence determination unit 166 receives the correction state vector and the correction amount from the correction state vector calculation unit 165. If the correction amount is equal to or greater than the threshold value, the convergence determination unit 166 uses the correction state vector instead of the initial value of the target state vector as an approximate distance difference. The above processing is repeated by feeding back to the inputs of the value calculation units 1631 and 1632.

If the correction amount is less than the threshold, the convergence determination unit 166 inputs a target state vector (target position, speed, etc.) reflecting the correction amount from the correction state vector calculation unit 165, and the time shown in FIG. The values are sequentially output to the memory 167 as target positioning results at t ₃ and t ₄ and held.

The initial value calculation unit 168 defines target position x (t ₃ ) hat and x (t ₄ ) hat, which are the positioning results at the two times t ₃ and t ₄ , according to the above equations (19) and (20). By substituting these into equation (21) above, the target initial velocity x dot is obtained. The target positioning result at the latest time is output as an initial value to the asynchronous positioning means 15a. Specifically, the time t ₄ as a reference time and outputs the time difference calculating portion 1511~151N asynchronous positioning means 15a, the correction state vector of the target, the method of least squares correction amount of the target state vector in an asynchronous positioning means 15a Is output to the distance difference approximate value calculation units 1521 to 152N as initial values for the calculation.

  Asynchronous positioning means 15a estimates the target positioning result by the same process as in the seventh embodiment when the reference time and the initial value of the target state vector are set from temporary positioning / initial value calculating means 16a.

  As described above, according to the tenth embodiment, the provisional positioning / initial value calculation means 16a tracks the distance difference from the target obtained from the detection results of the sensors 11 and 12, and at another time, When the distance difference from the target is obtained from the detection results of different sensor sets (sensors 12 and 13), the time alignment for the sensor set (sensors 12 and 13) is performed based on the tracking result, and then the target position, Since the velocity is calculated and set as the initial value of the target state vector calculation by the least squares method, observations at different times can be regarded as if they were observed at the same time, even if the observations are asynchronous It is possible to estimate the motion specifications of the target with high reliability. In addition, since only the distance difference from the target between any two sensors may be used as observation information, even if the sensor does not provide a Doppler frequency difference as in the seventh embodiment, the target motion specifications are estimated. be able to.

Embodiment 11 FIG.
FIG. 16 is a diagram showing a configuration of a motion specification estimation apparatus according to Embodiment 11 of the present invention. As shown in FIG. 16, the motion specification estimating apparatus according to the eleventh embodiment includes a receiving means 1, a distance difference observation value detecting means 2a, and an asynchronous positioning means 17a. In FIG. 16, the same or corresponding components as those in FIGS. 1, 8, and 10 are denoted by the same reference numerals, and description thereof is omitted.

  The asynchronous positioning means 17a includes an initial state quantity setting unit 171, tracking filter units 1721 to 172N, distance difference approximate value calculation units 1731 to 173N, correction amount calculation units 1741 to 174N, a correction state vector calculation unit 175, and a convergence determination unit 176. Is provided.

  The initial state quantity setting unit 171 uses an appropriate initial value of the target state vector and a target motion specification as an initial value used when the correction state vector calculation unit 175 calculates the target state vector correction amount by the least square method. Set the reference time to estimate. The tracking filter units 1721 to 172N input a distance difference obtained by a combination of sensors in the receiving unit 1 and perform a tracking process.

  The distance difference approximate value calculation units 1731 to 173N use the initial value of the target at the reference time set by the initial state quantity setting unit 171 (or use the correction amount from the convergence determination unit 176 as the initial value). Approximate values are calculated according to the motion equation assumed in advance.

  The correction amount calculation unit 1741 calculates a difference between the approximate value calculated by the distance difference approximate value calculation unit 1731 and the observed value obtained from the tracking filter unit 172 (observed value by the sensors 11 and 12). The correction amount of the target state vector calculated by using this is output. The correction amount calculation units 1742 to 174N calculate the difference between the approximate value calculated by the distance difference approximate value calculation units 1732 to 173N and the observation value obtained from the distance difference observation value detection units 112 to 11N, and calculates the difference. The correction amount of the target state vector calculated by using this is output.

  The correction state vector calculation unit 175 receives the correction amount from the correction amount calculation units 1741 to 174N, and corrects the target state vector using the least squares method until the difference converges. The convergence determination unit 176 outputs the corrected state vector to the distance difference approximate value calculation units 1731 to 173N when the correction amount by the correction state vector calculation unit 175 is larger than the threshold value, and ends the convergence calculation when the correction amount is smaller than the threshold value. Then, the corrected state vector is output as a positioning result.

Next, the operation will be described.
The initial state quantity setting unit 171 sets the reference time in the tracking filter units 1721 to 172N, and sets the initial value of the target state vector at the reference time in the distance difference approximate value calculation units 1731 to 173N.

  The tracking filter units 1721 to 172N respectively input observation values (distance difference with the target between the two sensors) obtained from the detection results of the sensor sets and their observation times from the distance difference observation value detection units 111 to 11N. A time difference between the reference time set by the initial state quantity setting unit 171 and the observation time is calculated, and the target is tracked using the time difference and the observation value input from the distance difference observation value detection unit (each Estimate the distance difference between the two sensors of the sensor set and the target).

  Among the tracking filter units 1721 to 172N, a certain tracking filter unit is an initial state quantity setting unit when an observation value is obtained from a distance difference observation value detection unit with a different sensor set by another tracking filter unit at a certain time. The time difference between the reference time set to 171 and the observation time is calculated, and the distance difference from the target at the time is predicted using the time difference and the observation value input from the distance difference observation value detection unit ( Estimate the distance difference between the two sensors and the target.

  For example, when the tracking filter units 1722 to 172N obtain a distance difference from the target in the detection results of the sensors 12, 13-1 (N-1), 1N at a certain time, the tracking filter unit 1721 predicts at that time. By obtaining the value, it is assumed that a distance difference from the target has been obtained at the same time, and the positioning process is performed. However, the combination of sensors to be predicted is one or more.

  The distance difference approximate value calculation units 1731 to 173N calculate the approximate value of the distance difference between the target and each sensor set by using the initial value of the target state vector input from the initial state amount setting unit 171, and the correction amount calculation unit Output to 1741 to 174N, respectively.

In the correction amount calculation units 1741 to 174N, the approximate value from the distance difference approximate value calculation units 1731 to 173N and the estimated value (distance difference from the target) input from the tracking filter units 1721 to 172N are input, and the approximate value and the observed value are input. Error (distance difference error Δh _{i, j} ) is calculated, and the target state vector correction amount calculated using this error is output. The correction state vector calculation unit 175 receives the correction amount from the correction amount calculation units 1741 to 174N, performs a convergence calculation by the least square method, and corrects the target state vector correction amount and the correction amount in the state vector. Output a state vector.

  The convergence determination unit 176 receives the corrected state vector and the correction amount from the corrected state vector calculation unit 175. If the correction amount is equal to or greater than the threshold value, the convergence determination unit 176 uses the corrected state vector instead of the initial value of the target state vector. Feedback is made to the inputs of the value calculation units 1731 to 173N, and the above processing is repeated. If the correction amount is less than the threshold value, the convergence determination unit 176 inputs a target state vector (target position, speed, etc.) reflecting the correction amount from the correction state vector calculation unit 175 and outputs it as a target positioning result. Output.

  As described above, according to the eleventh embodiment, the asynchronous positioning means 17a performs the tracking process on the distance difference from the target obtained from the detection result of the sensor set, and the detection result of the different sensor set at another time. If the distance difference from the target is obtained, the time adjustment for the sensor set is performed based on the tracking result, and then the motion specifications of the target are estimated. Therefore, the observed values at different times are observed at the same time. As a result, even if the observed values are asynchronous, it is possible to estimate the motion parameters of the target with high reliability. In addition, since only the distance difference from the target between any two sensors may be used as observation information, even if the sensor does not provide a Doppler frequency difference as in the seventh embodiment, the target motion specifications are estimated. be able to.

Embodiment 12 FIG.
FIG. 17 is a diagram showing the configuration of a motion specification estimation apparatus according to Embodiment 12 of the present invention. As shown in FIG. 17, the motion specification estimation apparatus according to the twelfth embodiment includes a receiving means 1, a distance difference observation value detecting means 2a, and an asynchronous positioning means 18a. In FIG. 17, the same or corresponding components as those in FIGS. 1, 9, and 10 are denoted by the same reference numerals and description thereof is omitted.

  Asynchronous positioning means 18a includes initial state quantity setting unit 181, time difference calculation units 1821 to 182N, distance difference approximate value calculation units 1831 to 183N, correction amount calculation units 1841 to 184N, correction state vector calculation unit 185, and convergence determination unit 186. And a convergence calculation counting unit 187. Configurations other than the convergence determination unit 186 and the convergence calculation count unit 187 are the initial state quantity setting unit 121, the time difference calculation units 1221 to 122N, and the distance difference approximate value calculation units 1231 to 123N shown in FIG. 10 of the seventh embodiment. The correction amount calculation units 1241 to 124N and the correction state vector calculation unit 125 operate in the same manner.

  Until the convergence calculation counting unit 187 counts the calculation number limit value, the convergence determination unit 186 determines the distance difference between the correction state vectors according to the comparison result between the correction amount by the correction state vector calculation unit 185 and a predetermined threshold value. The approximate value calculation units 1831 to 183N are fed back or the convergence calculation is terminated and the corrected state vector is output as the positioning result.

  The convergence calculation counting unit 187 counts the number of convergence calculations until the convergence determination unit 186 feeds back the corrected state vector to the distance difference approximate value calculation units 1831 to 183N until a predetermined calculation frequency limit value is reached. Here, the calculation number limit value is a convergence that gives an optimal calculation load and estimation accuracy determined according to the calculation performance of the computer realizing the motion specification estimation device according to the twelfth embodiment and the sampling rate of detected data and the like. The number of operations.

Next, the operation will be described.
Hereinafter, operations unique to the motion specification estimation apparatus according to the twelfth embodiment will be mainly described.
First, when the convergence determination unit 186 performs the convergence calculation, the convergence calculation counting unit 187 initializes the count value to 0, and sets the calculation frequency limit value as the limit value of the count. As in the first embodiment, when the correction amount is input from the correction state vector calculation unit 185, the convergence determination unit 186 compares the correction amount with a predetermined threshold value. At this time, when the correction amount is larger than the threshold and does not converge, the convergence calculation counting unit 187 increments the count value by one.

  When the correction amount does not converge even when the count value reaches the calculation frequency limit value, the convergence calculation counting unit 187 transmits a calculation end signal to the convergence determining unit 186. Thereby, the convergence determination unit 186 outputs a correction state vector (correction state vector with the correction amount added) reflecting the correction amount at the end of the calculation as a target positioning result. If the correction amount converges before the count value reaches the calculation count limit value, the correction state vector to which the converged correction amount is added is output as a target positioning result, and the convergence calculation counting unit 187 stops counting. To do.

  As described above, according to the twelfth embodiment, the convergence determination unit 186 counts the number of times the convergence calculation is performed by the convergence calculation counting unit 187, and the correction amount does not converge even when the calculation number limit value is reached in advance. In this case, since the correction state vector reflecting the correction amount at that time is used as the target positioning result, the number of convergence calculations by the correction state vector calculation unit 185 is limited. The execution of the calculation exceeding the performance of the estimation device is suppressed, and the calculation is completed in a predetermined time, so that the time for the convergence calculation can be shortened. In addition, since only the distance difference from the target between any two sensors may be used as the observation information, even if the sensor cannot obtain the Doppler frequency difference as in the seventh embodiment, the target motion specifications can be estimated. .

  In the twelfth embodiment, the case where the convergence calculation counting unit 187 is applied to the configuration according to the seventh embodiment has been described. However, the same may be applied to the configurations of the eighth to eleventh embodiments. Can be obtained.

It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 1 of this invention. It is a diagram showing a case where four observations before the reference time T～t ₃ for the target such as an aircraft is obtained. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 2 of this invention. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 3 of this invention. Sensors 11 and 12 at time t _3, the distance difference and the Doppler frequency difference by sensors 12 and 13 at time t ₄ is a diagram showing a case where obtained. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 4 of this invention. Time t _1, t ₂ to sensors 11 and 12, the distance difference and the Doppler frequency difference by sensors 12 and 13 at time t ₃ is a diagram showing a case where obtained. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 5 of this invention. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 6 of this invention. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 7 of this invention. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 8 of this invention. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 9 of this invention. Sensors 11 and 12 at time t _1, sensors 12 and 13 to the time t _2, the sensors 11 and 12 at time t _3, a diagram illustrating a case where the distance difference is obtained by the sensors 12 and 13 at time t _4. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 10 of this invention. Time t _1, t ₂ to the sensor 11 and 12 is a diagram showing a case where the distance difference is obtained by the sensors 12 and 13 at time t _3, t _4. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 11 of this invention. It is a figure which shows the structure of the motion specification estimation apparatus by Embodiment 12 of this invention.

Explanation of symbols

  1 receiving means, 2 distance difference / speed difference observed value detecting means, 2a distance difference observed value detecting means (detecting means), 4, 5, 7, 9, 10, 12a, 13a, 15a, 17a, 18a asynchronous positioning means, 8, 14a, 16a Temporary positioning / initial value calculation means, 11-1M sensor, 21-2N distance difference / velocity difference observation value detection unit, 41, 61, 81, 91, 141, 161, 181 initial state quantity setting unit, 421 to 42N, 521 to 52N, 711 to 71N, 1021 to 102N, 1221 to 122N, 1321 to 132N, 1511 to 151N, 1821 to 182N, time difference calculation unit, 431 to 43N, 531 to 53N, 721 to 72N, 831, 832,931-93N, 1031-103N, 1231-123N Distance difference / speed difference approximate value calculation unit, 1331-133N, 142 , 1422, 1521 to 152N, 1631, 1632, 1731 to 173N, 1831 to 183N Distance difference approximate value calculation unit, 441 to 44N, 541 to 54N, 731 to 73N, 841, 842, 941 to 94N, 1041 to 104N, 1241 -124N, 1341-134N, 1431, 1432, 1531-153N, 1641, 1642, 1741-174N, 1841-184N Correction amount calculation unit, 45, 55, 64, 74, 85, 95, 106, 125, 135, 144 , 154, 165, 175, 185 Correction state vector calculation unit, 46, 56, 65, 75, 86, 96, 126, 136, 145, 155, 166, 176, 186 Convergence determination unit, 821, 822, 921-92N , 1621, 1622, 1721-172N Filter unit, 107,187 convergence calculation counting unit, 111 to 11N distance difference observations detector, 146,167 memory, 147,168,171 initial value calculation section.

Claims (15)

Receiving means having a plurality of sensors for observing the movement of the target;
Detecting means for detecting an observation value of a distance difference and a speed difference from a target between the sensors of the sensor pair from an observation result at the same time by an arbitrary sensor pair among the plurality of sensors of the receiving means;
The target motion at the reference time according to the observed distance difference and speed difference between the targets of each sensor pair detected from the observation results at different times by a plurality of sensor pairs, and the target equation of motion assumed in advance. The target is determined based on the difference between the distance difference between the sensor and the target between the sensors at each time and the rough estimated value of the speed difference calculated by reflecting the time difference from the reference time with respect to the initial value of the specifications. Motion parameter estimation comprising an asynchronous positioning means that calculates a motion parameter correction amount and corrects the motion parameter until the difference is converged by a convergence operation, and estimates the target motion parameter at the reference time apparatus.   Asynchronous positioning means uses the least squares method to calculate the amount of correction of the target motion parameters based on the difference between the observed value and the rough estimate, and then corrects the target motion parameters until the difference converges by the convergence calculation. The motion specification estimation apparatus according to claim 1, wherein the motion specification of a target at a time is estimated.   The asynchronous positioning means calculates the correction amount of the target motion specification by weighting the observation accuracy of the sensor with respect to the difference between the observed value and the approximate estimated value. Motion specification estimation device. While tracking the distance difference and speed difference between the targets of the first sensor pair detected from the observation result of the first sensor pair among the plurality of sensors, the tracking sensor is different from the first sensor pair. When a distance difference and a speed difference between the second sensor pair and a target between the sensors of the second sensor pair are detected from the observation result of the second sensor pair at the same time, the second sensor pair observes the tracking process result. A tracking filter unit for predicting a distance difference and a speed difference between a target between the sensors of the first sensor pair at a time,
The asynchronous positioning means performs a convergence calculation using the distance difference and the speed difference between the target of the first and second sensor pairs simulated in the observation at the same time by the prediction processing of the tracking filter unit as observation values. 2. The motion specification estimation apparatus according to claim 1, wherein the motion specification of the target at the reference time is estimated.   Positioning is performed assuming that the observed values of the distance difference and the speed difference between the targets of each sensor pair obtained from the observation results at different times of the plurality of sensor pairs are observed at the same time, The provisional positioning / initial value calculating means for setting the value of the target motion specification at the time as the initial value of the target motion specification at the reference time used in the convergence calculation of the asynchronous positioning means is provided. Item motion estimation apparatus according to Item 1. While tracking the distance difference and speed difference between the targets of the first sensor pair detected from the observation result of the first sensor pair among the plurality of sensors, the tracking sensor is different from the first sensor pair. When a distance difference and a speed difference between the second sensor pair and a target between the sensors of the second sensor pair are detected from the observation result of the second sensor pair at the same time, the second sensor pair observes the tracking process result. A tracking filter unit for predicting a distance difference and a speed difference between a target between the sensors of the first sensor pair at a time,
The temporary positioning / initial value calculation means is an observation value of a distance difference and a speed difference from a target between the sensors in the first and second sensor pairs simulated by observation of the same time by the prediction processing of the tracking filter unit. 6. The convergence calculation is performed as follows, and the target motion specification at the estimated reference time is set as an initial value of the target motion specification at the reference time used in the convergence calculation of the asynchronous positioning means. Motion specification estimation device. Receiving means having a plurality of sensors for observing the movement of the target;
Detecting means for detecting an observation value of a distance difference between a target between the sensors of the sensor pair from an observation result at the same time by an arbitrary sensor pair among the plurality of sensors of the receiving means;
The observed value of the distance difference between the targets of each sensor pair detected from the observation results at different times by a plurality of sensor pairs and the target motion parameters at the reference time according to the target motion equation assumed in advance. Based on the difference between the initial value and the rough estimated value of the distance difference between the target of each sensor pair at each time calculated by reflecting the time difference from the reference time, A motion specification estimation apparatus comprising asynchronous positioning means for calculating a correction amount and correcting a motion specification until a difference is converged by a convergence calculation to estimate a target motion specification at the reference time. Asynchronous positioning means uses the least squares method to calculate the amount of correction of the target motion parameters based on the difference between the observed value and the rough estimate, and then corrects the target motion parameters until the difference converges by the convergence calculation. The motion specification estimation apparatus according to claim 7, wherein the motion specification of a target at a time is estimated. The asynchronous positioning means calculates the correction amount of the target motion specification by weighting the observation accuracy of the sensor with respect to the difference between the observed value and the approximate estimated value. Motion specification estimation device. A tracking process is performed for a difference in distance between the sensor of the first sensor pair detected from the observation result of the first sensor pair among the plurality of sensors, and a second different from the first sensor pair. When a distance difference between the sensor of the second sensor pair and the target between the sensors is detected from the observation result at the same time by the sensor pair, the first at the observation time by the second sensor pair is obtained from the tracking processing result. A tracking filter unit for predicting a distance difference between the target of the sensor pair of the sensor and the target,
The asynchronous positioning means performs a convergence calculation using the distance difference between the sensor and the target in the first and second sensor pairs simulated in the observation at the same time by the prediction processing of the tracking filter unit as an observation value, The motion specification estimation apparatus according to claim 7, wherein the motion specification of a target at a time is estimated. Positioning is performed on the assumption that the observed value of the distance difference between the sensors of each sensor pair obtained from the observation results at different times of the plurality of sensor pairs is observed at the same time, and the target at the time 8. The provisional positioning / initial value calculating means for setting the value of the motion specifications of the target as the initial value of the target motion specifications at the reference time used in the convergence calculation of the asynchronous positioning means. Motion specification estimation device. The detection means detects an observation value of a distance difference between the target of the sensor pair from the observation result at the same time by an arbitrary sensor pair among the plurality of sensors of the reception means,
The tracking filter unit performs a tracking process on a distance difference between the sensor of the first sensor pair detected from the observation result of the first sensor pair among the plurality of sensors and the first sensor pair. When the distance difference between the second sensor pair and the target between the sensors of the second sensor pair is detected from the observation result at the same time by the second sensor pair different from the above, the observation by the second sensor pair from the tracking processing result Predicting a distance difference between a target between the sensors of the first sensor pair at a time,
The asynchronous positioning means performs a convergence calculation using the distance difference between the sensor and the target in the first and second sensor pairs simulated in the observation at the same time by the prediction processing of the tracking filter unit as an observation value, The motion specification estimation apparatus according to claim 10 , wherein the motion specification of a target at a time is estimated. The provisional positioning / initial value calculation means assumes that the observed value of the distance difference between the target of each sensor pair obtained from the observation results of the plurality of sensor pairs at different times is observed at the same time. Te performs positioning, the target of motion various values in the time, according to claim 11, wherein the set as the initial value of the target exercise specifications at the reference time to be used for convergence calculation asynchronous positioning means Motion specification estimation device. The detection means detects an observation value of a distance difference between the target of the sensor pair from the observation result at the same time by an arbitrary sensor pair among the plurality of sensors of the reception means,
The tracking filter unit performs a tracking process on a distance difference between the sensor of the first sensor pair detected from the observation result of the first sensor pair among the plurality of sensors and the first sensor pair. When the distance difference between the second sensor pair and the target between the sensors of the second sensor pair is detected from the observation result at the same time by the second sensor pair different from the above, the observation by the second sensor pair from the tracking processing result Predicting a distance difference between a target between the sensors of the first sensor pair at a time,
The temporary positioning / initial value calculating means performs a convergence calculation using a distance difference between the target of the first and second sensor pairs simulated by the prediction process of the tracking filter unit as an observation value. 14. The motion parameters according to claim 13, wherein the target motion parameters at the estimated reference time are set as initial values of the target motion parameters at the reference time used for the convergence calculation of the asynchronous positioning means. Original estimation device. A convergence calculation counting unit that counts the number of convergence calculations up to a predetermined limit number of times,
Asynchronous positioning means, the count by the convergence calculation counting unit reaches the limit number of times, end the convergence calculation from claim 1, characterized in that the output motion specifications estimate at that time of claim 14 The motion specification estimation apparatus according to any one of the above.

Images