JP6470152B2 - Radar apparatus and radar signal processing method - Google Patents

Description

  The present embodiment relates to a radar apparatus and a radar signal processing method.

  Conventional radar devices generally use CFAR (Constant False Alarm Rate) processing in many cases. However, since CFAR is a method of detecting by comparing the threshold with the point having the maximum amplitude intensity even when there are a plurality of reflection points for each target, the small target is detected by pulse compression or Doppler filter. In some cases, when the target amplitude intensity is low, the CFAR threshold is not exceeded, and detection may fail. Further, when the main lobe faces the ground surface, the sea surface, etc., there is a problem that the fixed clutter is received and the detection performance deteriorates.

Pulse compression (range compression), Ouchi, “Basics of Synthetic Aperture Radar for Remote Sensing”, Tokyo Denki University Press, pp.131-149 (2003) SAR method (Az compression), Ouchi, “Basics of Synthetic Aperture Radar for Remote Sensing”, Tokyo Denki University Press, pp.171-178 (2003) CFAR (Constant False Alarm Rate), Yoshida, “Revised Radar Technology”, IEICE, pp.87-89 (1996) (Weibull distribution, Sekine, “Radar signal processing technology”, IEICE, pp.115-118 (1991)) 2D CFAR, Guy Morris, ‘Airborne Pulsed Doppler Radar 2nd Edition’, Artech House, pp. 399-404 (1996) MUSIC, ESPRIT, Kikuma, “Adaptive Antenna Technology”, Ohm, pp.137-164 (2003)

  As described above, in the conventional radar apparatus, there is a case where a small target is not detected without exceeding the threshold when detecting a small target in CFAR (Constant False Alarm Rate) processing. Further, when the main lobe faces the ground surface, the sea surface, etc., there is a problem that the fixed clutter is received and the detection performance deteriorates.

  The present embodiment has been made in view of the above problems, and can improve the detection capability even when the target RCS (Radar Cross-Section) is small and cannot be detected by CFAR or the like under an unnecessary environment such as clutter. An object of the present invention is to provide a radar device and a radar signal processing method thereof.

  In order to solve the above-described problem, the present embodiment is a radar apparatus that transmits a pulse of N (N ≧ 1) hit, and the received signal of the transmitted pulse is compared with a PRI (Pulse Repetition Interval) axis. FFT (Fast Fourier Transform) to generate range-Doppler data, detect the target component with a predetermined threshold using the range-Doppler data, specify the approximate range of the target position from the detected target component, The range-Doppler data near the approximate range is divided into N (N ≧ 1) meshes on the range axis, and the high resolution processing is performed by focusing on the range of the representative points of the mesh. The range-Doppler data is rearranged on the range-Doppler axis, and the rearrangement result is divided into (Nh, Mh) (Nh ≧ 1, Mh ≧ 1) meshes. A histogram of width intensities is created, a feature amount for the (Nh, Mh) region is calculated from the shape of the histogram, and a target is detected using data of the (Nh, Mh) region based on the feature amount.

1 is a block diagram showing a configuration of a radar apparatus according to a first embodiment. The conceptual diagram for demonstrating the outline | summary of a synthetic aperture in the radar apparatus shown in FIG. The conceptual diagram for demonstrating the processing content of a high resolution process in the radar apparatus shown in FIG. The figure which shows a mode that the feature-value is calculated from the amplitude histogram by the mesh division | segmentation of an area | region in the radar apparatus shown in FIG. The block diagram which shows the structure of the radar apparatus which concerns on 2nd Embodiment. FIG. 6 is a diagram illustrating a state in which a Weibull-shaped factor is calculated from an amplitude histogram obtained by mesh division of a region and a target is detected based on the factor in the radar apparatus illustrated in FIG. 5. The block diagram which shows the structure of the radar apparatus which concerns on 3rd Embodiment. The conceptual diagram for demonstrating the process in the case of using two-dimensional CFAR for the target detection using a feature-value in the radar apparatus shown in FIG.

  Hereinafter, embodiments will be described with reference to the drawings. In the description of each embodiment, the same portions are denoted by the same reference numerals, and redundant description is omitted.

(First embodiment)
Hereinafter, the radar apparatus according to the first embodiment will be described with reference to FIGS. 1 to 4.

  FIG. 1 is a block diagram showing the configuration of the radar apparatus according to the first embodiment. In FIG. 1, an antenna 1 is a phased array antenna in which a plurality of antenna elements are arranged to form a large aperture array, and a transmission pulse having a specific frequency that is repeatedly supplied from a transmission / reception unit 21 of the transceiver 2 at a specific period. A signal (hereinafter referred to as a PRF (Pulse Repetition Frequency) signal) is transmitted in a designated direction and the reflected wave is received. The transceiver 2 forms a received beam in an arbitrary direction by performing phase control on the signals received by the plurality of antenna elements of the antenna 1 in the transceiver unit 21 according to instructions from the beam controller 22 and combining them. To obtain a PRF received signal. Further, the beam control unit 22 directs the transmission beam so as to sequentially scan within the search direction (AZ, EL angle range).

  A signal acquired by the transmitter / receiver 2 and transmitted / received by the antenna 1 is sent to the signal processor 3. In this signal processor 3, the received signal obtained by the transceiver 2 is converted into a digital signal by an AD (Analog-Digital) converter 31, pulse compression (Non-Patent Document 1) is performed by a pulse compressor 32, and The FFTy processing unit 33 performs FFT (Fast Fourier Transform) processing on data in the PRI (Pulse Repetition Interval) in the Y-axis direction (flight direction), and the CFAR detection unit 34 performs detection processing by CFAR (Non-patent Document 3). Here, the order of pulse compression and FFTy processing may be reversed. In the case of a low RCS (Radar Cross Section) target, the CFAR threshold is reduced and the detection probability is increased. After the detection processing by the CFAR, the target direction extraction unit 35 extracts a direction in which a target may exist. The beam controller 22 directs the beam in this direction, and a chirp signal having a wide frequency band is transmitted in order to observe the resolution with a high range axis. Here, the orientation extracted by the target direction extraction unit 35 is divided into mesh-like representative points by the mesh division unit 36.

  Moreover, in the signal processor 3, in order to observe with high resolution on the Doppler axis, the number of hits is increased and the observation time is lengthened. Further, the signal received by the transceiver 2 is converted into a digital signal by the AD converter 37, the X-axis (range axis) FFT processing is performed by the range axis FFTx converter 38, and the range reference signal is converted by the multiplier 39. Further, the multiplication unit 41 multiplies the mesh representative point by the Az reference signal, the FFTy processing unit 42 performs the FFT processing between the PRI axes, and the inverse FFTx processing unit 43 reverses the X axis (range axis). An FFT process is performed, and images for all meshes divided by the image combining unit 44 are combined (synthesized) to obtain an entire image. Note that the order of the FFTy process and the inverse FFTx process may be reversed.

  After synthesizing all meshes by image combination to obtain an entire image, the mesh dividing unit 45 divides the mesh into regions (1 to Nh, 1 to Mh) (Nh ≧ 1, Mh ≧ 1), and amplitude intensity The calculation unit 46 calculates the amplitude intensity for each mesh, and the amplitude histogram generation unit 47 generates an amplitude histogram. The mesh division unit of the range axis at this time may be the same as the mesh for image generation or may be a different size. In this amplitude histogram, when the amplitude intensity on the horizontal axis is normalized by a predetermined fixed value, the shape of the histogram changes if there is a signal having a different amplitude in the mesh compared to other meshes. Using the change in shape, the feature quantity extraction unit 48 extracts a feature quantity such as the shape of a histogram, and the target detection unit 49 detects the presence / absence of the target using the feature quantity, and distance measurement / speed measurement The unit 50 performs distance measurement / speed measurement of the target based on the distance of the representative point such as the center of the detected target mesh and the Doppler to obtain a detection output.

  Next, the outline of the synthetic aperture of the radar apparatus will be described with reference to FIG. In a radar device mounted on a flying object, a pulse is transmitted at a PRI (Pulse Repetition Interval) interval within a synthetic aperture time (1 cycle) and a reflected wave is received so that the target is always irradiated with a real aperture beam. For each pulse, data in the PRI is acquired in units of range cells. Using this acquired data, an ISAR process is performed to obtain a target ISAR image. Although FIG. 2 shows the case of a radar device mounted on a flying object, the radar device may be fixed if an ISAR image can be obtained.

In the high resolution processing of the signal processor 3, the input signal is formulated as sig (t, u) as follows. Range compression is a correlation process between an input signal and a range compression signal. The case of performing this in the frequency domain is formulated as follows.

  In order to make it on the time axis, this s may be subjected to inverse Fourier transform. However, in order to perform Az compression (Non-Patent Document 3) thereafter, the signal s is left on the (ω, u) axis. Next, a cross-range reference signal fs0 is generated.

As shown in FIG. 3, the detection position (Xn, Yn, Zn) is calculated from the distance of the representative point of each mesh divided by the range axis and the beam azimuth in the range-Doppler data of the target detection azimuth (4 ) Is a position calculated using the equation (see coordinates in FIG. 2).

Using this, the reference signal fs0 for AZ compression is given by the following equation.

The signal cs is obtained by multiplying the above-mentioned s by fs0 for AZ compression.

Using this, a signal fcs (ω, ku) is obtained by performing FFT processing on the u-axis.

The long-time integration output fp can be calculated by the inverse FFTx processing of the inverse FFTx processing unit 43 with respect to the ω-axis of fcs.

  From the output of fp (t, ku), an image of one mesh for each range can be created, and this is combined for all the meshes by the image combining unit 44 to obtain an entire image.

  Next, the mesh division unit 45 divides the mesh into regions (1 to Nh, 1 to Mh) (Nh ≧ 1, Mh ≧ 1) (FIG. 4A), and the amplitude strength unit 46 determines the amplitude for each mesh. The intensity is calculated, and an amplitude histogram is generated by the amplitude histogram generator 47 (FIG. 4B). The mesh division unit of the range axis at this time may be the same as the mesh for image generation or may be a different size. In this amplitude histogram, when the amplitude intensity on the horizontal axis is normalized by a predetermined fixed value, the shape of the histogram changes when there is a signal having a different amplitude with a larger intensity than other meshes in the mesh. Using this shape change, the feature quantity extraction unit 48 extracts a feature quantity such as the shape of a histogram (FIG. 4C), and using the feature quantity, the target detection unit 49 determines whether or not there is a target. To detect. The distance and speed of the target can be measured based on the distance of the representative point such as the center of the detected target mesh and the Doppler.

Processing after the input of the mesh division unit 45 will be described with reference to FIG. In FIG. 3, the FR data of the target detection azimuth of the high resolution processing input is dispersedly arranged in the regions 1 to N, and the target distance measurement / speed measurement is performed from the representative point range such as the center of each region and the Doppler frequency. The speed at this time can be calculated from the Doppler frequency by the following equation.

  As described above, in the present embodiment, in a coherent radar apparatus that transmits N (N ≧ 1) hit pulses, FFT (Fast Fourier Transform) processing is performed on a PRI (Pulse Repetition Interval) axis, and as necessary. Using pulse-compressed range-Doppler data (FR data) to detect by a predetermined threshold, specify the approximate range of the target position, and N (N ≧ 1) FR data near the approximate range on the range axis The FR data which is divided into meshes and is focused on the range of the representative points such as the center of the mesh and processed with high resolution is rearranged on the range-Doppler axis, and is converted into (Nh, Mh) (Nh ≧ 1, Mh). ≥1) mesh division, create amplitude intensity histogram, calculate (Nh, Mh) area feature quantity from histogram shape, and (Nh, Mh) area data based on the feature quantity The target is detected using.

  That is, according to the present embodiment, after extracting the target direction, the target is detected with high sensitivity by extracting the feature amount from the mesh-divided signal using the high-resolution transmission / reception signal on the Doppler range axis. can do.

(Second Embodiment)
In the first embodiment, a method of creating an amplitude histogram and extracting a feature amount has been described. In the present embodiment, a method using a Weibull shape factor c as a feature amount assuming a shape of an amplitude histogram as a Weibull distribution will be described. The system configuration is shown in FIG. 5 and will be described with reference to FIG. However, in FIG. 5, the same parts as those in FIG. 1 are denoted by the same reference numerals, and redundant description is omitted here. 6 (a) and 6 (b) are the same as FIGS. 4 (a) and 4 (b). In this embodiment, FIG. 6 (c) is added to improve the accuracy of target detection in FIG. 6 (d). I am trying.

  In this embodiment, a Weibull extraction unit 51 is used instead of the feature amount extraction unit 48 shown in FIG. The Weibull extraction unit 51 fits the histogram generated by the histogram generation unit 47 using the Weibull distribution and outputs the result to the target detection unit 49.

The shape of the Weibull distribution is represented by x as the amplitude of the observation value for each mesh in the region (1 to Nh, 1 to Mh) (Nh ≧ 1, Mh ≧ 1) divided by the mesh dividing unit 45 of the first embodiment. Using the shape factor c, it can be expressed by the following relational expression (see Non-Patent Document 4).

Therefore, if the relationship between x and c is tabulated in advance, <x> and <x ² > are calculated using the amplitude value x of the observation value for each mesh, and the table is cited, so that FIG. It is possible to extract c (n, m) (n = 1 to N, m = 1 to M) for each mesh shown in c). As shown in FIG. 6 (d), it is possible to detect the presence / absence of the target by arranging c on a range-Doppler axis obtained by dividing the mesh and extracting, for example, a range exceeding a predetermined threshold. Further, distance measurement / speed measurement can be performed by calculating the center of gravity including the average value and amplitude of the range-Doppler.

  As described above, in this embodiment, fitting is performed using a Weibull distribution as a feature amount for extracting the shape of the histogram. That is, after extracting the target direction, the shape factor of the Weibull distribution that fits the shape of the amplitude histogram generated as the feature value is calculated from the mesh-divided signal using the high-resolution transmission / reception signal on the range-Doppler axis. However, the target can be detected with high sensitivity by the distribution of the shape factor.

(Third embodiment)
In the second embodiment, the method of creating the amplitude histogram and extracting the feature amount by the shape factor of the Weibull distribution has been described. In this case, if clutter is large, erroneous detection may occur. In the present embodiment, a method using a two-dimensional CFAR at the time of target detection using a feature quantity such as a shape factor c will be described. The configuration of the system is shown in FIG. 7, and the processing will be described with reference to FIG. In FIG. 7, the same parts as those in FIGS. 1 and 5 are denoted by the same reference numerals, and redundant description is omitted here.

  In the present embodiment, a two-dimensional CFAR processing unit 52 is used instead of the target detection unit 49 of FIG. As shown in FIG. 8, the processing of the two-dimensional CFAR processing unit 52 (see Non-Patent Document 5) uses a feature amount such as the shape factor of the Weibull distribution arranged in the range-Doppler while reducing false detection. This is a method for detecting a target. While shifting the test cells in order on the range-Doppler axis, the threshold is determined using the average value of the surrounding reference cells, and the presence or absence of detection is determined by whether or not the value of the test cell exceeds the threshold. In the range of the reference cell, a guard cell is set according to the spread of the target shape so that the threshold becomes high due to the amplitude due to the spread of the target shape and is not detected.

  The distance measurement / speed measurement can be performed by performing the center of gravity calculation including the average value and amplitude of the detected range-Doppler cell, as in the second embodiment.

  As described above, in this embodiment, two-dimensional CFAR processing is adopted as a detection method. That is, after extracting the target direction, on the range-Doppler axis, using a high-resolution transmission / reception signal, the feature amount is extracted from the mesh-divided signal, and the false detection is reduced using the two-dimensional CFAR, The target can be detected with high sensitivity.

  In the above embodiment, the case of a radar apparatus that performs pulse compression has been described. However, the present invention can also be applied to a case where pulse compression is not performed. On the contrary, the Doppler axis may not be subjected to high resolution, and the range axis may be subjected to high resolution by pulse compression or the like.

  Further, as a technique for increasing the resolution of the range-Doppler axis, there are techniques using MUSIC (MUltiple SIgnal Classification), ESPRIT (Estimation of Signal Parameters via Rotational Invariance Techniques) processing (Non-patent Document 6), and the like.

  In addition, the present embodiment is not limited to the above-described embodiment as it is, and can be embodied by modifying the components without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

DESCRIPTION OF SYMBOLS 1 ... Antenna, 11 ... Antenna element, 2 ... Transmitter / receiver, 21 ... Transmitter / receiver, 22 ... Beam controller, 3 ... Signal processor, 31 ... AD converter, 32 ... Pulse compression part, 33 ... FFTy process part, 34 ... CFAR processing unit, 35 ... target direction extracting unit, 36 ... mesh dividing unit, 37 ... AD converting unit, 38 ... FFTx processing unit, 39 ... range reference signal multiplication unit, 41 ... AZ reference signal multiplication unit, 42 ... FFTy processing , 43 ... inverse FFT processing unit, 44 ... image combining unit, 45 ... mesh dividing unit, 46 ... amplitude intensity calculating unit, 47 ... amplitude histogram generating unit, 48 ... feature quantity extracting unit, 49 ... target detecting unit, 50 ... Ranging / speed measuring unit, 51... Weibull extracting unit, 52... 2D CFAR processing unit.

Claims (4)

In a radar device that transmits N (N ≧ 1) hit pulses,
A generator that generates range-Doppler data by performing FFT (Fast Fourier Transform) on a PRI (Pulse Repetition Interval) axis for the received signal of the transmitted pulse;
A detection unit for detecting a target component with a predetermined threshold using the range-Doppler data;
A specifying unit for specifying a rough range of the target position from the detected target component;
A dividing unit that divides range-Doppler data in the vicinity of the approximate range into N (N ≧ 1) meshes on the range axis;
A high resolution processing unit that performs high resolution processing while focusing on a range of representative points of the mesh;
A rearrangement unit that rearranges the high-resolution processed range-Doppler data on the range-Doppler axis;
A mesh division unit for dividing the rearrangement result into (Nh, Mh) (Nh ≧ 1, Mh ≧ 1) meshes;
A histogram generator for creating a histogram of amplitude intensity from the mesh division result;
A calculation unit for calculating a feature amount of (Nh, Mh) region from the shape of the histogram;
A radar apparatus comprising: a detection unit that detects a target using data of an (Nh, Mh) region based on the feature amount.   The radar apparatus according to claim 1, wherein the calculation unit fits a Weibull distribution as a feature amount for extracting the shape of the histogram.   The radar apparatus according to claim 1, wherein the detection unit performs a two-dimensional CFAR (Constant False Alarm Rate) process. Used for radar devices that transmit N (N ≧ 1) hit pulses,
The received signal of the transmitted pulse is subjected to FFT (Fast Fourier Transform) with respect to a PRI (Pulse Repetition Interval) axis to generate range-Doppler data,
A target component is detected by a predetermined threshold using the range-Doppler data,
Identify the approximate range of the target position from the detected target component,
The range-Doppler data near the approximate range is divided into N (N ≧ 1) meshes on the range axis,
Focusing on the range of the representative points of the mesh, high resolution processing,
Rearrange the high-resolution processed Range-Doppler data on the Range-Doppler axis,
The rearrangement result is divided into (Nh, Mh) (Nh ≧ 1, Mh ≧ 1) meshes,
Create a histogram of amplitude intensity from the mesh division result,
The feature amount for the (Nh, Mh) region is calculated from the shape of the histogram,
A radar signal processing method for detecting a target using (Nh, Mh) area data based on the feature amount.

Images