JP6287674B2 - Delay time estimation device and height measurement device - Google Patents

Description

  The present invention relates to a delay time estimation device that estimates the delay time of an incoming wave and a height measurement device that measures the height of a target.

  The height measurement method that uses radio waves to measure the target height includes a method that uses the measured angle values of the array antennas arranged in a vertical plane, and a method that uses multiple receiving sensors arranged in a wide range to perform positioning. Is widely used. However, when height measurement is performed with an airborne radar, it is not possible to secure an antenna mounting space with sufficient accuracy to withstand positioning and height measurement due to mounting problems.

  In response to this problem, the following Non-Patent Document 1 uses a delay time of a direct reflected wave and an indirect reflected wave that returns to the aircraft via the sea surface etc. A height measurement method that solves the problem of sexuality is introduced. The height measurement method using the arrival delay time cannot measure the height unless the arrival delay times of both the direct reflected wave and the indirect reflected wave can be estimated. However, as the distance between the aircraft and the target increases, the arrival delay time difference between the direct reflection wave and the indirect reflection wave becomes shorter, and it becomes impossible to separate the direct reflection wave and the indirect reflection wave having a short time difference with a pulse width that is time resolution. Problems arise.

For this problem, as shown in Non-Patent Document 2 below, the range data of the incoming wave is represented as a vector having a plurality of frequency components in the frequency domain, and the range data of the incoming wave in the frequency domain is represented by There is a method for improving the time resolution of the arrival delay time by applying a super resolution method such as MUSIC (MUltiple SIgnal Classification). When the time resolution of the arrival delay time is improved by applying this super-resolution method, the problem is that the actual direct reflected wave and the indirect reflected wave have a high correlation. For example, if the super-resolution method is applied as it is based on the assumption that multiple incoming waves are uncorrelated, the actual direct reflected wave and the indirect reflected wave have high correlation, so the arrival delay time obtained by the super-resolution method The resolution and estimation accuracy of are significantly degraded. In order to avoid this degradation, moving average processing of the correlation matrix of the received data is performed when applying the super-resolution method, and the correlation between multiple incoming waves is suppressed directly in the correlation matrix used in the super-resolution method. There is a method for estimating arrival delay times of reflected waves and indirect reflected waves.

M.W.Long, Early Warning System Concepts, Chap. 8, Scitech, 2004. Nobuyoshi Kikuma, Adaptive signal processing with array antenna, pp. 274-277, Science and Technology Publishers.

With the conventional height measurement method that uses the arrival delay time to measure the target height, the arrival delay time between the direct reflection wave and the indirect reflection wave and the arrival delay time difference between the direct reflection wave and the indirect reflection wave can be grasped with high accuracy. It is important to realize a delay time estimation device. When applying the super-resolution method using moving average processing in this delay time estimation device, the moving average processing in the frequency domain is performed by reducing the vector size of the range data in the frequency domain, and the correlation between the incoming waves Repress. However, if the vector size of the range data in the frequency domain is made smaller than necessary in order to increase the moving average processing, the time resolution and estimation accuracy for estimating the incoming wave are degraded. Therefore, it becomes a problem to set the moving average processing count to an appropriate value.
The present invention has been made in order to solve the above-described problems. By performing delay time estimation using a super-resolution method in which an appropriate number of moving averages is set, arrival delay times of direct reflected waves and indirect reflected waves can be reduced. The purpose is to estimate the delay time with high accuracy. Another object of the present invention is to perform highly accurate height measurement using the obtained highly accurate arrival delay time difference.

A delay time estimation apparatus according to the present invention estimates a delay time with respect to a reference time of the first and second arrival waves by a super-resolution method using a waveform including the first and second arrival waves. In order to calculate a correlation matrix used in the super-resolution method, a plurality of calculated values for each of a plurality of different subsets including a part of a plurality of frequency components obtained by Fourier transforming the waveform Average number calculation for determining the number of times of averaging for averaging the correlation matrix based on a parameter calculated based on a delay time spectrum indicating the magnitude of a known signal included in the waveform with respect to the delay time An average processing unit that calculates a correlation matrix used in the super-resolution method by performing the averaging process based on the averaging number determined by the average number calculation unit, A super-resolution delay time estimation unit that estimates a delay time with respect to a reference time of the first and second arrival waves by performing a super-resolution method using the correlation matrix calculated by the average processing unit. It is characterized by that.

  According to the present invention, the arrival delay times of the direct reflection wave and the indirect reflection wave are conventionally set by appropriately setting the moving average number of times performed before performing the delay time estimation by the super-resolution method according to the state of the received waveform. It can be estimated with higher accuracy than the technology.

The block diagram of the height measuring apparatus by Embodiment 1, 2 of this invention. The block diagram of the delay time estimation part 1-3 of the height measuring apparatus by Embodiment 1 of this invention. The image figure of the moving average process by Embodiment 1 of this invention. It is a calculation example explaining the system which uses the spectrum shape in the moving average frequency calculation part 2-4 in Embodiment 1 of this invention. A geometry for performing height measurement in the height measurement processing unit 1-4 according to the first embodiment of the present invention. The block diagram of the delay time estimation part 1-3 of the height measuring apparatus by Embodiment 2 of this invention.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a height measuring device 1 which is a signal processing device according to an embodiment of the present invention. In the following drawings, the same reference numerals indicate the same or corresponding parts.

In this embodiment, this height measurement device is installed in an aircraft / aircraft radar that monitors the front, the Doppler frequency difference between the direct reflected wave and the indirect reflected wave from the target is extremely small, and the direct reflected wave after coherent integration This is an embodiment for estimating a target altitude when indirect reflected waves exist in the same Doppler bin. It is assumed that the aircraft has a means for grasping its altitude. Although not shown, the radar has a transmission unit that transmits a radio wave arbitrarily modulated by itself.

In FIG. 1, a target detection unit 1-2 detects a target from received data having a waveform including a direct reflection wave and an indirect reflection wave acquired by the reception antenna 1-1, and performs pulse compression, unnecessary wave suppression processing, coherent integration, and the like. By processing, a target distance estimating means for improving a target signal-to-noise ratio and a signal-to-unwanted wave ratio on a two-dimensional map of the distance of the received signal and the Doppler frequency is provided. Range data which is a Q × 1 complex data vector obtained by unnecessary wave suppression and coherent integration by the target distance estimation means is output to the delay time estimation unit 1-3. Here, Q is the number of processing range bins used for target detection.

  The delay time estimation unit 1-3 uses the Q × 1 complex data vector after the coherent integration output from the target detection unit 1-2 and the target distance information to delay the target direct reflection wave and the indirect reflection wave. Has means for estimating. Here, the delay time represents a delay time with respect to a reference time, which is a reference time, but the reference time may be any time. As an example, the reference time may be a time at which the radar transmits radio waves from a transmission unit (not shown). In this case, the delay time represents the time from when the radio wave is transmitted until it is received. Hereinafter, a description will be given assuming this example.

  The height measurement processing unit 1-4 estimates the target altitude from the own device altitude information output from the own device altimeter 1-5 and the delay time of the direct reflected wave and the indirect reflected wave output from the delay time estimating unit 1-3. Have means to do.

In the example of FIG. 1, each of the target detection unit 1-2, the delay time estimation unit 1-3, and the height measurement processing unit 1-4 that are components of the height measurement device 1 excluding the own altimeter 1-5 is dedicated. It is assumed that it is configured by hardware (for example, a semiconductor integrated circuit on which a CPU is mounted, or a one-chip microcomputer), but each component of the height measuring device 1 is configured by a computer. Also good. In this case, a program describing the processing contents of the target detection unit 1-2 and delay time estimation unit 1-3 height measurement processing unit 1-4 is stored in the memory of the computer, and the CPU of the computer stores the program in the memory. What is necessary is just to run the program currently carried out.

  FIG. 2 is a configuration diagram of the delay time estimation unit 1-3 according to the first embodiment of the present invention. The super-resolution estimation process application determination unit 2-1 determines whether or not to perform the super-resolution delay time estimation process from the range data used by the target detection unit 1-2 for target detection and information on the target position in the range data. Means to do. An example of how to determine whether super-resolution delay time estimation processing is applied will be described in detail later. Whether the interval between the direct reflected wave and the indirect reflected wave is wider than the time resolution determined by the pulse width, etc. from the delay time spectrum, etc. Can be considered as an index. The reason for including this determination unit is that the delay time estimation by the super-resolution method has a high calculation load and avoids deterioration of estimation accuracy due to deterioration of the signal-to-noise ratio by the moving average process described later.

また、超分解能推定処理適用判定部２−１が超分解能法による遅延時間推定を行うと判断した場合、レンジビン抽出部２−２から超分解能遅延時間推定部２−６までの処理が行われる。 When the super-resolution estimation process application determination unit 2-1 determines that the delay time estimation by the super-resolution method is not performed, the super-resolution estimation process application determination unit 2-1 determines the range data output from the target detection unit 1-2 and The arrival delay times of the direct reflected wave and the indirect reflected wave are calculated from the target position information and output to the height measurement processing unit 1-4. The arrival delay time τ can be calculated by dividing the target distance R indicated by the target range bin by the speed of light c as shown below.

When the super resolution estimation process application determination unit 2-1 determines that the delay time estimation is performed by the super resolution method, the processes from the range bin extraction unit 2-2 to the super resolution delay time estimation unit 2-6 are performed.

The range bin extraction unit 2-2 has means for extracting data from the integrated range data output from the target detection unit 1-2 and target range information for reducing the calculation load. The extraction range of the range bin may be arbitrarily determined according to the system, but the extraction range affects the number of degrees of freedom when performing the delay time estimation, and also has a balance with the degree of freedom consumed by the super-resolution method and moving average processing. In view of this, several times the time resolution is required.

  The range FFT processing unit 2-3 constitutes means for converting the range data extracted from the range bin extraction unit 2-2 into a frequency domain. For the conversion of the range data into the frequency domain, discrete Fourier transform, fast Fourier transform, etc. can be applied. Extraction range data may be padded with zeros in order to perform fast Fourier transform or improve frequency resolution. The frequency domain range data x may be subjected to filtering in accordance with the signal bandwidth in order to improve the signal-to-noise ratio in the time domain, and band limitation may be performed.

A moving average number calculation unit 2-4, which is an average number calculation unit, performs a movement performed by a moving average processing unit 2-5, which is an average processing unit in the subsequent stage, from the frequency domain range data output from the range FFT processing unit 2-3. A means for determining the average number of times is provided. That is, the moving average processing unit 2-5 averages the correlation matrix R _xx calculated from the frequency domain range data x in the frequency direction. If the number of moving averages is too small, the correlation between the direct reflected wave and the indirect reflected wave is calculated. Many of the super-resolution methods such as MUSIC do not work correctly. On the other hand, in order to increase the number of moving averages in the frequency direction, if the size of the range data x in the frequency domain where the correlation matrix _Rxx is calculated is reduced, the frequency width handled by the range data x becomes narrower and can be identified. Time resolution is degraded, and the signal-to-noise ratio obtained from the correlation output between the frequency domain range data x and the delay time steering vector used for identification is also degraded. As a result, if the number of moving averages is increased too much in the frequency direction, the delay time estimation accuracy deteriorates.

ここで、

であり、Mは移動平均回数、Nは移動平均後に超分解能を適用するアレーサイズ、a(τ)は時刻τに対するN×１遅延時間ステアリングベクトル、^Hは複素共役転置の演算子である。なお、超分解能法を適用するアレーサイズはアレー自由度と呼んでも構わない。 In consideration of this point, it is desirable that the moving average number calculation unit 2-4 determines an appropriate number of moving averages. In the present embodiment, the moving average number calculation unit 2-4 determines the moving average number that is the average number from the shape of the delay time spectrum before application of the super-resolution method. The delay time spectrum P (τ) with respect to the time τ can be expressed as the following equation (1).

here,

Where M is the number of moving averages, N is an array size to which super-resolution is applied after moving average, a (τ) is an N × 1 delay time steering vector with respect to time τ, and ^H is an operator of complex conjugate transposition. Note that the array size to which the super-resolution method is applied may be called array flexibility.

と表すこともでき、

は電力が規格化された後の周波数領域のレンジデータxと遅延時間ステアリングベクトルa(τ)との相関を表している。また、Ｐ（τ）はその相関を周波数領域でＭ回移動平均した結果を示している。図３に移動平均を行う際の相関行列Ｒ_xxを示す。 Here, the meaning of the delay time spectrum P (τ) will be described. Delay time spectrum P (τ) is

Can also be expressed as

Represents the correlation between the frequency domain range data x and the delay time steering vector a (τ) after the power is normalized. P (τ) indicates the result of moving average of the correlation M times in the frequency domain. FIG. 3 shows a correlation matrix R _xx when performing the moving average.

  That is, the delay time spectrum P (τ) indicates a value obtained by averaging the correlation between the range data x obtained from the received signal and the vector a (τ) determined by the delay time in the frequency domain, and the delay time spectrum P ( τ) can be said to be a function indicating the presence position of the signal as a correlation, and can be interpreted as a function indicating the magnitude of the known signal included in the waveform including the direct reflected wave and the indirect reflected wave with respect to the delay time. However, the position where the waveform of the delay time spectrum P (τ) peaks does not indicate the exact position of the signal, and in particular when the difference in arrival delay time between the direct reflected wave and the indirect reflected wave is small, P (τ) There arises a situation in which the position where the waveform becomes a peak is different from the exact position of the signal. This is because the phases of the direct reflection wave and the indirect reflection wave interfere with each other, and the details will be described below.

  FIG. 4 is a calculation example of the delay time spectrum when the arrival delay time difference between the direct reflected wave and the indirect reflected wave is incident at half the time resolution. The delay time spectrum P (τ) can be broadly divided into two states: a case where the maximum peak is broken and a case where the peak is divided depending on the phase relationship during reception of the direct reflection wave and the indirect reflection wave. This is a phenomenon that occurs because the correlation between the two waves is high and the way of interference changes depending on the phase relationship of the incoming waves at the time of reception. The case where the maximum peak breaks means that the overlapping part of the time waveform of the direct reflection wave and the time waveform of the indirect reflection wave cancel each other, and as seen from the delay time estimation algorithm, two waves that can be separated by time Is interpreted as follows. Such cancellation of the incoming wave on the spectrum is caused by the high correlation of the incoming wave, and does not occur when the incoming wave is uncorrelated and a sufficient snapshot is obtained. These peaks appear when the incoming waves cancel each other, and are different from the actual arrival delay time. Therefore, in order to estimate the accurate delay time, a moving average process for suppressing the correlation between two incoming waves in the correlation matrix used in the super-resolution method is necessary as a step before applying the super-resolution method. It becomes. The moving average process is a signal process that uses a part of the frequency component of the incoming wave as a snapshot for the average process. Since the frequency component of the incoming wave is determined by the signal bandwidth, increasing the snapshot used for the moving average can suppress the correlation while narrowing the signal bandwidth. Since the resolution and accuracy of the delay time generally strongly influence the bandwidth of the incoming wave, an unnecessarily moving average degrades the resolution and accuracy. The number of times required for this moving average varies depending on the signal-to-noise ratio (SNR) of the incoming wave. The correlation value of incoming waves that can be separated by the super-resolution method becomes smaller as the SNR becomes smaller. Therefore, it is desirable to suppress the correlation of incoming waves more at low SNR and to separate incoming waves by the super-resolution method than at high SNR. When the peak is broken, since the SNR is lower than when the peak is synthesized, it is necessary to further suppress the correlation of the incoming wave separation. Therefore, the number of moving averages when the peak is broken increases as compared with the case where the peak is synthesized. Since there is a correlation between incoming waves and the peak is broken, it can be determined that the correlation can be suppressed if the peak is broken. Therefore, the optimum number of moving averages can be evaluated by the null depth calculated between the maximum peak and the second peak in the delay time spectrum.

ここで、SP₂(M)はM回移動平均した相関行列を用いて遅延時間推定を行った時の遅延時間サーチ範囲内における二番目に大きいピークの値、SP_NULL(M)はM回移動平均した相関行列を用いて遅延時間推定を行った時の遅延時間サーチ範囲内における最大ピークと二番目に大きいピークとの間に形成されたヌルの値である。すなわち、この評価式は遅延時間スペクトラムの２番目に大きいピークの値と、１番目と２番目のピークの間にある谷の最低値との差を表している。また、この評価式は直接反射波と間接反射波を含む波形から算出されるパラメータであり、このパラメータを用いて移動平均回数計算部２−４は適切な移動平均回数Mを決定する。 An evaluation formula for selecting the moving average number using the null depth between peaks can be expressed as follows, for example.

Where SP ₂ (M) is the second largest peak value in the delay time search range when the delay time is estimated using the correlation matrix averaged M times, and SP _NULL (M) is moved M times This is a null value formed between the maximum peak and the second largest peak in the delay time search range when delay time estimation is performed using an averaged correlation matrix. In other words, this evaluation formula represents the difference between the value of the second largest peak in the delay time spectrum and the lowest value of the valley between the first and second peaks. The evaluation formula is a parameter calculated from a waveform including a direct reflected wave and an indirect reflected wave, and the moving average number calculation unit 2-4 determines an appropriate moving average number M using this parameter.

の範囲でサーチを行うとする。ここで、τ_maxはスペクトラム最大ピークに対応する到来遅延時間、T₀はパルス幅等の時間分解能である。これは、伝搬距離差の関係から直接反射波が間接反射波より先にレンジビンに表れることと、直接反射波の受信レベルの方が海面などを反射して入射してくる間接反射波に比べて一般的に受信レベルが高いという性質を利用することで、τ_maxを直接反射波の到来遅延時間と仮定したサーチ範囲である。このように設定することで、時間分解能で分離できない場合の到来遅延時間をカバーできる。 In order to calculate SP ₂ (M) and SP _NULL (M), the search range by the delay time steering vector may be arbitrarily determined, but here, in order not to misrecognize the side lobe as an incoming wave peak,

Suppose that the search is performed in the range of Here, τ _max is an arrival delay time corresponding to the maximum spectrum peak, and T ₀ is a time resolution such as a pulse width. This is because the direct reflected wave appears in the range bin earlier than the indirect reflected wave due to the propagation distance difference, and the reception level of the direct reflected wave is reflected by the sea surface and enters the indirect reflected wave. In general, the search range assumes that τ _max is the arrival delay time of the directly reflected wave by utilizing the property that the reception level is high. By setting in this way, it is possible to cover the arrival delay time when separation is not possible with time resolution.

  With such an evaluation, the null depth between peaks can be detected. However, if there is only one peak in the search range, the super-resolution method can be applied with a small number of moving averages. Therefore, the moving average number may be two or three times. In other cases, the moving average number is selected such that the number of peaks is 1 or the null depth between peaks is minimized.

に置き換える。ここで、

であり、Jはベクトルの要素を逆順にする変換行列である。この拡張を行うことで、移動平均一回当たりの相関抑圧効果を改善させることができる。 Equation (2) is a basic moving average, but may be extended to the following forward-backward averaging process in order to further improve the correlation suppression effect. Specifically, in Formula (1), R _xx is

Replace with here,

J is a transformation matrix that reverses the elements of the vector. By performing this extension, it is possible to improve the correlation suppression effect per moving average.

When the moving average number calculation unit 2-4 outputs the moving average number M, the moving average processing unit 2-5 outputs the moving average corresponding to the moving average number M to the range data output from the range FFT processing unit 2-3. Applying to the correlation matrix R _{xx of} x, the correlation matrix R _xx (the line above R) after moving average is output to the super-resolution delay time estimation processing unit 2-6. In order to clarify the processing order, the moving average number calculation unit 2-4 and the moving average processing unit 2-5 are separated. These processing units are integrated and moved in the moving average number calculation unit 2-4. The correlation matrix R _xx (line on R) used for determining the average number of times may be output as it is.

In the configuration according to the first embodiment, since the delay time is estimated for the range data after the coherent integration, the number of snapshots is one. Therefore, when applying the super-resolution method, a moving average process is indispensable in order to reduce the correlation between the direct reflected wave and the indirect reflected wave in the correlation matrix _Rxx .

ここで、E_Nは次式から算出される。

ここで、E_SとE_Nはそれぞれ移動平均後相関行列Ｒ_xx（Ｒの上に線）の信号固有値に対応する固有ベクトルを並べた行列と雑音固有値に対応する固有ベクトルを並べた行列、Λ_SとΛ_Nはそれぞれ信号固有値が対角成分に並んだ行列と雑音固有値が対角成分に並んだ行列である。 The super-resolution delay time estimation processing unit 2-6 performs delay time estimation by the super-resolution method on the correlation matrix R _xx (the line above R) after the moving average output from the moving average processing unit 2-5. When estimating the delay time by applying MUSIC which is one of the super-resolution methods, the delay time spectrum P _MUSIC (τ) is expressed as follows.

Here, E _N is calculated from the following equation.

Here, E _S and E _N are respectively a matrix in which eigenvectors corresponding to the signal eigenvalues of the correlation matrix R _xx (line on R) after moving average are arranged, and a matrix in which eigenvectors corresponding to the noise eigenvalues are arranged, Λ _S and Λ _N is a matrix in which signal eigenvalues are arranged in diagonal components and a matrix in which noise eigenvalues are arranged in diagonal components.

The moving average, the number of signal eigenvalues become two direct reflection wave and the indirect reflected waves, array freedom -2 number of noise eigenvectors, that is, the number of columns E _N. Therefore, the degree of freedom of the array necessary for estimating the delay time is required to be at least 4 including the moving average process.

  The search range in the search calculation of the expression (3) based on MUSIC does not necessarily have to be performed for the entire range extracted by the range bin extraction unit 2-2, and may be performed only in the vicinity of the target range bin.

After the search is completed and the delay time spectrum P _MUSIC (τ) applying MUSIC is obtained, peak detection processing is performed, and the arrival delay times of the direct reflected wave and indirect reflected wave are calculated from the obtained peak position, and the height measurement processing is performed. Output to part 1-4.

と表され、最小ノルム法の場合、遅延時間スペクトラムP_{min_norm}(τ)は

と表される。 Although the calculation of the delay time spectrum by MUSIC has been introduced here, the spectrum may be calculated by the Capon method or the minimum norm method. In the case of the Capon method, the delay time spectrum P _Capon (τ) is

In the case of the minimum norm method, the delay time spectrum P _{min_norm} (τ) is

It is expressed.

  The height measurement processing unit 1-4 obtains the target altitude from the arrival delay times of the direct reflected wave and the indirect reflected wave output from the delay time estimating unit 1-3 and the altitude information output from the own altimeter 1-5. A means for estimating is configured. The own device altitude calculating means by the own device altimeter 1-5 may be arbitrary. Examples include the aircraft's altitude obtained from its own coordinate calculation results by the GPS receiver, and the aircraft's altitude calculated from the arrival delay time until the radio wave irradiated directly below the aircraft by the radio altimeter returns to the aircraft. The altitude calculated from the atmospheric pressure by the barometric altimeter.

The target height estimation method in the height measurement processing unit 1-4 is arbitrary, but here, a method is used in which the height is measured based on the geometric relationship between the target altitude, the distance to the target, and the target height. FIG. 5 is a diagram showing the geometric relationship between the aircraft and the target. The symbols in the figure are as follows.

The long distance target height measurement cannot use the surface of the earth as a plane, and therefore uses a geometry reflecting the roundness of the earth as shown in FIG. The height measurement is performed by searching the target altitude for the distance to the target determined from the arrival delay time of the direct reflected wave and the altitude obtained from the own altimeter 1-5, and calculating the indirect reflected wave obtained by calculation. The altitude at which the difference between the arrival delay time and the actual arrival delay time is minimized is output as the target altitude.

 As described above, Embodiment 1 discloses a delay time estimation apparatus that estimates a delay time with respect to a reference time of a direct reflected wave and an indirect reflected wave by a super-resolution method using a waveform including a direct reflected wave and an indirect reflected wave. In order to calculate the correlation matrix used in the super-resolution method, the delay time estimation device applies to each of a plurality of different subsets including a part of a plurality of frequency components obtained by Fourier transform of the waveform. A moving average number calculation unit 2-4 that determines an average number of times for performing an averaging process for averaging a plurality of calculated correlation matrices based on a parameter calculated from the waveform, and a moving average number calculation unit 2-4 The moving average processing unit 2-5 that calculates the correlation matrix used in the super-resolution method by performing the averaging process based on the number of averaging determined in (5) and the moving average processing unit 2-5 It includes by performing the super-resolution method using the correlation matrix, a super-resolution delay time estimation unit 2-6 that estimates the delay time with respect to the reference time of the direct reflection wave and the indirect reflected waves, the. With such a configuration, it is possible to appropriately set the number of moving averages performed before performing the delay time estimation by the super-resolution method according to the reception state of the waveform, and the arrival delay times of the direct reflected wave and the indirect reflected wave can be set as compared with the prior art. It can be estimated with high accuracy.

Further, in the delay time estimation apparatus shown in the first embodiment, the parameter calculated from the waveform including the direct reflected wave and the indirect reflected wave is the delay indicating the magnitude of the known signal included in the waveform with respect to the delay time. It has the structure calculated based on a time spectrum. With such a configuration, it is possible to estimate the delay time in which the direct reflection wave and the indirect reflection wave exist with a rough accuracy using the delay time spectrum related to the delay time in which the direct reflection wave and the indirect reflection wave exist, and the delay time. It is possible to appropriately set the number of moving averages based on the state of the delay time spectrum near the time.

  In the delay time estimation apparatus shown in the first embodiment, the parameters calculated from the waveform including the direct reflected wave and the indirect reflected wave are set to the second largest peak value of the delay time spectrum, and the first and second peaks. It is the difference with the lowest value of the valley between. By such setting, a parameter having a high correlation with the phase relationship between the direct reflected wave and the indirect reflected wave included in the delay time spectrum can be set. As a result, the number of moving averages can be set appropriately according to the phase relationship between the direct reflected wave and the indirect reflected wave, and the arrival delay times of the direct reflected wave and the indirect reflected wave can be estimated with higher accuracy than the conventional technology. can do.

  The delay time estimation apparatus shown in the first embodiment is a super resolution estimation unit that determines whether to perform delay time estimation of the first and second arrival waves by the super resolution method based on the delay time spectrum. A processing application determination unit 2-1, and when the super-resolution estimation processing application determination unit 2-1 determines that the delay time is estimated by the super-resolution method, the delay times of the direct reflected wave and the indirect reflected wave by the super-resolution method are determined. Make an estimate. With this configuration, the super-resolution method is applied in a situation where it is desirable to perform the super-resolution method. The operation scale can be reduced.

  In addition, the delay time estimation apparatus shown in the first embodiment includes a transmission unit that transmits a radio wave toward a target at a reference time, and the directly reflected wave that is directly reflected by the target and the radio wave is the target. After being reflected, a waveform including an indirect reflected wave reflected at a position other than the target is received. With such a configuration, the delay time estimation device can measure the delay time from the time when the radio wave is transmitted toward the target until the direct reflected wave and the indirect reflected wave are received, and grasp the distance from the target. Can do.

  Further, the height measuring device 1 shown in the first embodiment includes a delay time estimation device including an own device altimeter 1-5, which is an altitude measuring unit that measures the altitude of the own device, and a transmission unit that has already been described, and its delay time. Height measurement processing that is a target height estimation unit that estimates the target altitude based on the delay times of the direct reflection wave and the indirect reflection wave estimated by the estimation device and the altitude of the own device measured by the own device altimeter 1-5 It has the structure provided with the part 1-4. With such a configuration, it is possible to estimate the arrival delay times of the direct reflected wave and the indirect reflected wave with higher accuracy than in the prior art, and to estimate the target altitude with high accuracy using the estimation result.

  In Embodiment 1 and the following embodiments, the direct reflected wave may be referred to as a first incoming wave, and the indirect reflected wave may be referred to as a second incoming wave.

Embodiment 2
The first embodiment shows processing for one pulse hit, while the second embodiment shows processing using received signals for a plurality of pulse hits.

  In the case of using a plurality of pulse hits, if the height measuring device of the present embodiment is mounted on an aircraft / aircraft radar that monitors the front, the difference in Doppler frequency between the direct reflected wave and the indirect reflected wave from the target is slightly different. The present embodiment is an example in which target altitude estimation is performed when a direct reflected wave and an indirect reflected wave after coherent integration exist in the same Doppler bin in such an environment. It is assumed that the aircraft has a means for grasping the altitude of the aircraft. Although not shown, the radar has a transmission unit that transmits a radio wave arbitrarily modulated by itself.

  In the second embodiment, similarly to the first embodiment, target detection is performed by suppressing unnecessary waves, coherent integration, and the like. The configuration of the height measuring device in the second embodiment is the same as that in the first embodiment, but the difference from the first embodiment is that the output from the target detection unit 1-2 to the delay time estimation unit 1-3 is the target position. It is not the range data after integration of information and Q × 1, but the target position information and the range hit matrix of Q × Ns. Here, Ns is the number of pulse hits. In addition, with the change, the processing and output performed in each processing block are different from those in the first embodiment.

  FIG. 6 is a configuration diagram of the delay time estimation unit 1-3 in the present embodiment. If the Doppler frequency is expected to be slightly different between multiple hits, the correlation between the direct and indirect reflected waves is different because the phase relationship between the direct and indirect reflected waves is different for each hit. It becomes easy to do. As a result, the number of moving averages can be reduced as compared with the case where only one hit is used.

  Unlike the first embodiment, the range bin extraction unit 3-2 of the delay time estimation unit 1-3 has an input of a Q × Ns range hit matrix, so that the output from the range bin extraction unit 3-2 is L ×. Ns matrix. Here, L is the number of extraction range bins.

  The range FFT processing unit 3-3 constitutes means for converting the range direction of the L × Ns range hit matrix output from the range bin extracting unit 3-2 into the frequency domain by discrete Fourier transform or the like. The L × Ns range hit data obtained by converting the range direction to the frequency domain is output to the moving average number calculation unit 3-4.

上記、受信データ相関行列が得られた後、移動平均回数計算部３−４は、実施の形態１の移動平均回数計算部２−４と同様の手順で移動平均回数を算出する。 The reception data correlation matrix R _xx used for calculating the moving average number based on the spectrum shape performed by the moving average number calculation unit 3-4 is expressed as follows because the range data X is an L × Ns matrix.

After the reception data correlation matrix is obtained, the moving average number calculation unit 3-4 calculates the moving average number in the same procedure as the moving average number calculation unit 2-4 of the first embodiment.

The moving average processing unit 3-5 applies the moving average of the number of times output from the moving average number calculation unit 3-4 to the reception data correlation matrix _Rxx and outputs the result to the super-resolution delay time estimation unit 3-6.

  Since the processing after super-resolution delay time estimation unit 3-6 in the second embodiment is the same as that in the first embodiment, it is omitted.

As described above, in the second embodiment, the range data X in the matrix format is handled using the reception signals for a plurality of pulse hits. Further, by calculating the reception data correlation matrix R _xx from the matrix format range data X, the same processing as in the first embodiment can be performed thereafter.
In addition, when the Doppler frequencies are expected to be slightly different, the correlation is not complete, so that the number of moving averages can be reduced compared to the case of using data after coherent integration.

 As described above, in the second embodiment, the transmission unit that transmits radio waves transmits the radio waves a plurality of times, and the moving average number calculation unit 3-4 that is an average number calculation unit includes a plurality of radio waves that are transmitted a plurality of times. The number of moving averages is determined using a waveform including a direct reflected wave and an indirect reflected wave. With such a configuration, it is possible to reduce the number of moving averages compared to a case where the transmission unit transmits radio waves once. Also, if the number of moving averages can be reduced, the super-resolution method can be applied using range data with a large vector size in the frequency domain, and the time resolution and estimation accuracy for arriving waves can be improved. Is possible.

1: Height measuring device, 1-1: Reception antenna, 1-2: Target detection unit, 1-3: Delay time estimation unit, 1-4: Height measurement processing unit, 1-5: Own altimeter, 2-1 : Super-resolution estimation processing application determination unit, 2-2: Range bin extraction unit, 2-3: Range FFT processing unit, 2-4: Moving average number calculation unit, 2-5: Moving average processing unit, 2-6: Super Resolution delay time estimation unit, 3-1: Super-resolution estimation processing application determination unit, 3-2: Range bin extraction unit, 3-3: Range FFT processing unit, 3-4: Moving average number calculation unit, 3-5: Movement Average processing unit, 3-6: Super-resolution delay time estimation unit

Claims (6)

A delay time estimation device that estimates a delay time with respect to a reference time of the first and second incoming waves by a super-resolution method using a waveform including the first and second incoming waves,
In order to calculate a correlation matrix used in the super-resolution method, an average of a plurality of correlation matrices calculated for each of a plurality of different subsets including a part of a plurality of frequency components obtained by Fourier transforming the waveform An average number of times for performing the averaging process to determine, based on a parameter calculated based on a delay time spectrum indicating the magnitude of the known signal included in the waveform with respect to the delay time, and
An average processing unit that calculates a correlation matrix used in the super-resolution method by performing the averaging process based on the averaging number determined by the average number calculation unit;
A super-resolution delay time estimation unit that estimates a delay time with respect to a reference time of the first and second arrival waves by performing a super-resolution method using the correlation matrix calculated by the average processing unit;
A delay time estimation device comprising: The parameter calculated from the waveform is a difference between a second largest peak value of the delay time spectrum and a lowest valley value between the first and second peaks. 2. The delay time estimation apparatus according to 1. A determination unit that determines whether or not to perform delay time estimation of the first and second arrival waves by a super-resolution method based on the delay time spectrum;
When the determination unit determines that performing the delay time estimation by super-resolution method, according to claim 1 or claim and performing the said first and second incoming wave delay time estimation by the super-resolution method 3. The delay time measuring device according to 2. A transmitter that transmits radio waves toward the target at the reference time;
The first incoming wave is a directly reflected wave that directly arrives when the radio wave is reflected by the target;
After the second incoming wave which the radio wave is reflected by the target, according to any one of claims 1 to 3, characterized in that the indirect wave reflected at a position other than the target Delay time estimation device. The transmitter transmits radio waves multiple times,
Claim 4 wherein the average number calculator, characterized in that determining the averaging count using a waveform, comprising a plurality of said first and second incoming wave corresponding to the radio waves transmitted the plurality of times The delay time estimation apparatus according to 1. An altitude measurement unit that measures the altitude of the device itself;
The delay time estimation apparatus according to claim 4 ,
A target altitude estimation unit that estimates the altitude of the target based on the delay times of the direct reflected wave and the indirect reflected wave estimated by the delay time estimation device, and the altitude of the own device measured by the altitude measurement unit; ,
A height measuring device comprising:

Images