JP5911719B2 - Target detection device, guidance device, and target detection method - Google Patents

Description

  Embodiments described herein relate generally to a target detection device that detects a weak target, a guidance device that uses the target detection device, and a target detection method that is used in the target detection device.

  Among the minute target detection methods for capturing a target having a small reflection cross section with a radar, there is a shift addition method (Shift-And-Add) as a method that does not have high detection performance but requires a very small amount of calculation. In the shift addition method, a measurement waveform map of each CPI (Coherent Processing Interval) corresponding to each measurement time of the radar is shifted by a different amount for each map so as to cancel a target movement assumed between the measurement times. Then, the shifted maps are overlaid and added for each point of the map to obtain an integral gain. In the shift addition method, if the target movement is known, each map having different acquisition times can be shifted by the correct amount. As a result, the targets are added at the same point to obtain an integral gain, and a target with a small amount of reflection can be detected.

  In the shift addition method, when the expected movement of the target is not clear, it is necessary to detect the target by scanning the shift direction in which the integral gain is obtained, so that the amount of calculation increases. In particular, when even the target motion model is unknown, the amount of calculation increases explosively because it is necessary to scan by assuming every motion. For this reason, the merit of the shift addition method that the calculation amount is small is lost.

  By the way, it is assumed that the radar is a pulse Doppler radar and a map is created with two axes of a range Doppler frequency. In a situation where the received power of the reflected wave reflected by the target is so small that it cannot be detected by normal processing, the integration gain is increased as much as possible, so that the target enters one Doppler frequency bin within the range of the target acceleration that is assumed. The number of pulses is increased as much as possible. By increasing the number of pulses, the Doppler frequency bin spacing becomes very small. For this reason, even if the target is within one Doppler frequency bin within the same CPI, there may be cases where the target exists in different Doppler frequency bins in different measurement times. In such a case, the expected speed for the target has a range, and the direction and amount to shift the map becomes unclear. For this reason, the target must be detected by scanning the shift direction and the shift amount in which the integral gain can be obtained, which increases the amount of calculation.

N. Baba et al., “Stellar speckle image reconstruction by the shift-and-add method”, Applied Optics Vol. 24, p. 1405, 1985

  As described above, when the movement of the target is not clear, or when the number of pulses is increased in order to obtain an integral gain, there is a problem that the calculation amount of the target detection process by the shift addition method increases. It was.

  Therefore, even if the received power of the reflected wave reflected by the target is small and the movement of the target cannot be clearly predicted, the integration effect by the shift addition method can be enjoyed without increasing the amount of calculation. It is an object of the present invention to provide a target detection device, a guidance device, and a target detection method.

  According to the embodiment, the target detection apparatus includes a reception unit, a range division unit, a peak value selection unit, a shift unit, an addition unit, and a detection unit. The receiving unit receives the reflected wave reflected by the measurement target, and creates a measurement waveform map based on the reflected wave. The range dividing unit sets the width for each of one or more predetermined axes of the axes of the measurement waveform map based on the amount of change in motion predicted for the measurement target in a preset addition period. The measurement waveform map is divided into a plurality of ranges having the width corresponding to the predetermined axial direction. The peak value selection unit detects the maximum value of power or amplitude in the divided range, and generates a resolution reduction map using the detected value as a representative value in each range. The shift unit shifts the plurality of resolution reduction maps generated by the plurality of measurements during the addition period in the direction set for each measurement by the amount set for each measurement. The adder adds the power or amplitude of the plurality of shifted resolution reduction maps to generate an addition map. A detection part detects the point estimated that the said measuring object exists based on the power or amplitude of the said addition map.

It is a figure for demonstrating the shift addition method. It is a figure which shows the function structure of the target detection apparatus which concerns on 1st Embodiment. It is a figure which shows the flowchart of the target detection operation | movement performed by the target detection apparatus of FIG. It is a figure which shows the example of a simulation of a Doppler frequency spectrum. It is a block diagram which shows the function structure of the target detection apparatus which concerns on 2nd Embodiment. It is a figure which shows an example of the range Doppler frequency map produced with the pulse Doppler radar receiver of FIG. It is a figure which shows an example of the frequency spectrum produced by the resolution reduction process of the Doppler frequency resolution reduction part of FIG. It is a figure explaining the selection method of the maximum value in the resolution reduction process shown in FIG. It is a figure which shows the resolution reduction map shifted by the shift addition part of FIG. It is an image which shows the simulation result of the addition map produced by the target detection apparatus of FIG. It is a figure explaining the resolution reduction method utilized when creating one addition map shown by FIG. It is a figure explaining the resolution reduction method utilized when creating one addition map shown by FIG. It is a figure explaining the resolution reduction method utilized when creating one addition map shown by FIG. It is a block diagram which shows the function structure of the target detection apparatus which concerns on 3rd Embodiment. FIG. 15 is a diagram for explaining how the map range is divided by the map dividing unit in FIG. 14. It is a figure which shows an example of the division | segmentation of the map by the map division part of FIG. It is a figure which shows the other example of the division | segmentation of the map by the map division part of FIG. It is a block diagram which shows the function structure of the guidance device using the target detection apparatus which concerns on 1st thru | or 3rd embodiment.

  Hereinafter, embodiments will be described with reference to the drawings.

(First embodiment)
First, the shift addition method used in this embodiment will be described. FIG. 1 is an example for explaining the shift addition method. In the upper part of FIG. 1, a map of the measurement times 1, 2, and 3 measured in time series is shown. A map refers to a one-dimensional or more measurement result displayed with at least one or more parameters as an axis among radar measurement parameters such as range, speed, and angle. In the example of FIG. 1, the horizontal axis indicates the x position, and the vertical axis indicates the y position.

  As other examples of the map, the horizontal axis indicates a range, and the vertical axis indicates an angle such as an azimuth or an elevation angle. Moreover, you may represent in three dimensions of a range, an azimuth | direction, and an elevation angle. Further, any one of the range, azimuth, elevation angle, and speed may be used. Furthermore, the output is not limited to the radar output, and may be an image output from a camera.

  The black circles in the map of FIG. 1 indicate the target, and the gray circles indicate relatively strong points that appear due to noise. In many cases, the map used for shift addition uses the output waveform as it is. However, for convenience of illustration, only measurement points with relatively strong power are displayed here.

  Since the target according to this embodiment is weak, the target and noise are measured with similar power. Therefore, it is difficult to distinguish between a target and noise only with a single map.

  Therefore, when these three maps are superimposed and added as they are, an addition map shown in the middle of FIG. 1 is obtained. The points corresponding to the goals are arranged smoothly. In order to detect this threshold value, the map is shifted and added so that the target is the same point. The results of the measurement times 1 to 3 as the addition period are shown in the lower part of FIG. Since the target is the same point, a gain is generated by addition, and the power becomes larger than the noise point, and the threshold value can be detected.

  FIG. 2 is a diagram illustrating a functional configuration of the target detection device 10 according to the first embodiment. A target detection apparatus 10 illustrated in FIG. 1 includes an antenna 11, a radar receiver 12, a resolution reduction unit 13, a shift addition unit 14, and a detection unit 15.

  The antenna 11 receives a target reflected wave reflected by a radar wave transmitted from a radar transmitter (not shown) hitting the target.

  The radar receiver 12 demodulates the target reflected wave received by the antenna 11 and generates a map based on the demodulated waveform. The radar receiver 12 outputs the generated map to the resolution reduction unit 13.

  The resolution reducing unit 13 includes a range dividing unit 131 and a peak value selecting unit 132.

  The range dividing unit 131 divides one or more predetermined dimensions among the dimensions in the map for each width corresponding to a predetermined resolution.

  The peak selection unit 132 selects the maximum value of the waveform within each range after the division. The maximum value of the waveform may be power or amplitude. Any norm that does not consider the phase may be used, and other orders may be used. The peak value selection unit 132 creates a resolution reduction map based on the selected maximum value. The resolution reduction map is obtained by reducing the resolution of the map supplied from the radar receiver 12 by using the maximum value selected by the peak value selection unit 132 as the representative value of each point in the map.

  Note that the number of dimensions in which the resolution is reduced may be any number from one dimension to all dimensions in the map. Since the resolution of each dimension in the map supplied from the radar receiver 12 is significantly smaller than the width determined by the range where the target is expected to move during the shift addition period, the shift addition is sufficient in the original map. When it is predicted that no effect can be obtained, all dimensions predicted as such may be targeted for resolution reduction. The division width for reducing the resolution is set to a width that can cover the change in the target movement expected in the addition period with almost one bin.

  For example, when reducing the Doppler frequency resolution, at the maximum expected acceleration, the width of the Doppler frequency change when the target accelerates during the addition period, and the width of the division when the Doppler frequency resolution is reduced. And it is sufficient.

  The shift addition unit 14 includes a storage unit 141, a shift unit 142, and an addition unit 143.

  The storage unit 141 stores the resolution reduction map generated by the resolution reduction unit 13 for a preset number of measurements. Here, the preset number of measurements is set based on the addition period. When the storage unit 141 stores the resolution reduction maps for the preset number of measurements, the storage unit 141 outputs the plurality of stored resolution reduction maps to the shift unit 142.

  The shift unit 142 shifts the resolution reduction map of each measurement time by a predetermined amount in a predetermined direction. As a premise in the first embodiment, it is assumed that the range of movement change in the target map is known. Here, the shift direction and the shift amount are the direction and amount in which the shift addition gain can be obtained with a value close to or near the center of the range of change in the target motion. The shift unit 142 outputs each resolution reduction map shifted by a predetermined amount in a predetermined direction to the adding unit 143.

  The addition unit 143 generates the addition map by adding the resolution reduction map supplied from the shift unit 142 in a non-coherent manner with the phase of power or the like being dropped for each same point. The addition unit 143 outputs the generated addition map to the detection unit 15.

  The detection unit 15 detects a point that is estimated to have a high probability that a target exists in the obtained addition map. In the detection process, a point where the value after addition is larger than the threshold value may be simply determined based on the threshold value. Alternatively, only the maximum point of the addition map may be detected. In addition, more advanced detection using likelihood or the like may be performed using information before and after in time.

  Next, the target detection operation of the target detection apparatus 10 configured as described above will be described. FIG. 3 is a diagram illustrating a flowchart of the target detection operation executed by the target detection device 10.

  First, the radar receiver 12 demodulates the target reflected wave received by the antenna 11, and generates a map based on the demodulated waveform (step S31). The range dividing unit 131 divides a predetermined dimension in the map generated by the radar receiver 12 into a predetermined width (step S32). The peak selection unit 132 creates a resolution reduction map using the maximum value of the waveform in each divided range as a representative value of the range (step S33).

  Subsequently, the storage unit 141 determines whether or not a resolution reduction map for a preset number of measurements has been stored (step S34). When the resolution reduction map for the preset number of measurements is not stored (No in step S34), the target detection apparatus 10 performs the processing in steps S31 to S33 and acquires the resolution reduction map for the next measurement time. To do. When the resolution reduction maps for the preset number of measurements are stored (Yes in step S34), the storage unit 141 outputs the stored resolution reduction maps for the predetermined number of measurements to the shift unit 142 (step S35). .

  The shift unit 142 shifts the resolution reduction map for each measurement time supplied from the storage unit 141 in a predetermined direction by a predetermined amount (step S36). The adding unit 143 generates an addition map by adding the resolution reduction map of each measurement time after the shift noncoherently for each same point (step S37). The detection unit 15 detects a point in the generated addition map that is estimated to have a high probability that the target exists (step S38).

  As described above, in the first embodiment, the resolution reduction unit 13 reduces the resolution of a predetermined dimension in the map, and the shift addition unit 14 shifts the map based on the information about the approximate movement of the target, and after the shift Add maps at the same point. As a result, it is possible to detect a target with movement without increasing the amount of calculation while maintaining the effect of shift addition.

  As the resolution reduction method, a method of selecting the maximum value among the widths for reducing the resolution is used. As a result, even if the resolution is reduced, it is possible to prevent the target detection SNR (Signal to Noise Ratio) from being significantly reduced. The reason why the method for detecting the maximum value is superior to other resolution reduction methods will be described later.

  Therefore, according to the target detection apparatus 10 according to the first embodiment, the amount of calculation is increased even when the received power of the reflected wave reflected by the target is small and the movement of the target cannot be clearly predicted. The integration effect obtained by the shift addition method can be enjoyed without making it possible to detect the target.

  In the first embodiment, shift addition is performed on the norm of power, amplitude, or other dimensions. If the norm dimensions are different, the noise distribution resulting from the shift addition changes. In principle, according to the present embodiment in any dimension, it is possible to detect a target with motion without increasing the amount of calculation while maintaining the effect of shift addition. What is necessary is just to select a required dimension in consideration of the amount of calculation processing.

  In the first embodiment, the map output from the radar receiver 12 is obtained by mapping the demodulated waveform of the radar receiver 12 as it is, but the present invention is not limited to this. For example, in order to reduce the amount of calculation, the radar receiver 12 may remove a weak portion with a threshold value in advance and create a map that leaves only a bin with a strong value. The threshold at that time is preferably about the noise average value. Such removal by the threshold may be applied before the shift addition unit 14 performs shift addition on the resolution reduction map supplied from the resolution reduction unit 13.

  In the first embodiment, the resolution reducing unit 13 divides the dimension for reducing the resolution into predetermined widths. Each width at that time may overlap. In an extreme example, the resolution reduction unit 13 applies resolution reduction processing by taking a width around each point of the map before the resolution reduction, so that the interval between the points is not changed from that before the resolution reduction. May be.

  In FIG. 2 in the first embodiment, the resolution reduction unit 13, the shift addition unit 14, and the detection unit 15 are described as independent components, but the present invention is not limited to this. For example, the resolution reduction unit 13, the shift addition unit 14, and the detection unit 15 may be realized by processing of a signal processor.

(Second Embodiment)
In the second embodiment, a case will be described in which the target detection device 10 described in the first embodiment is applied to a pulse Doppler radar.

  In the pulse Doppler radar, the length of 1 CPI, the PRI (Pulse Repetition Interval) and the number of pulses are set so that sufficient gain can be obtained by Fourier transform for obtaining the Doppler frequency. The longest period in which the target speed change can be regarded as sufficiently small is the period in which the integral gain is the highest.

  If the CPI length is determined so that the integral gain is the highest, the target is contained in one frequency bin in the Doppler frequency spectrum of the range bin where the target exists at each measurement time. FIG. 4 is a diagram showing a simulation example of the Doppler frequency spectrum when 1 CPI is composed of 64 pulses. 4A shows the Doppler frequency spectrum at the measurement time 1, and FIG. 4B shows the Doppler frequency spectrum at the measurement time 20. In FIG. 4, the SNR after integration of the target reflected wave is set to a sufficiently high level of about 20 dB in order to make the drawing easier to see.

  In FIG. 4A showing the Doppler frequency spectrum of measurement time 1, there is a peak at frequency bin number 36, but in FIG. 4B showing the Doppler frequency spectrum of measurement time 20, it is at frequency bin number 39. This is because the target has an acceleration other than 0, and therefore the target reflected wave appears in different frequency bins if the measurement times are different.

  In order to shift and add such a map, it is necessary to know not only the amount of movement of the target between measurement times but also the amount of change in movement speed. In addition, it is necessary to perform scanning by many kinds of shift methods so that changes in the target moving speed can be covered.

  FIG. 5 is a block diagram illustrating a functional configuration of the target detection device 20 according to the second embodiment. A target detection apparatus 20 shown in FIG. 5 is obtained by reconfiguring the target detection apparatus 10 shown in FIG. 1 to a pulse Doppler radar and using conditions unique to the pulse Doppler radar. In addition, the same number is attached | subjected to the part which is common in the target detection apparatus 10 shown in FIG.

  The target detection device 20 includes an antenna 11, a pulse Doppler radar receiver 21, a Doppler frequency resolution reduction unit 22, a shift addition unit 23, and a threshold detection unit 24.

  The pulse Doppler radar receiver 21 demodulates the radio wave received by the antenna 11. The pulse Doppler radar receiver 21 Fourier-transforms the values of a plurality of pulses of the same CPI for each range bin to obtain a Doppler frequency spectrum. The pulse Doppler radar receiver 21 arranges Doppler frequency spectra and creates a range Doppler frequency map. FIG. 6 is a diagram illustrating an example of a range Doppler frequency map created by the pulse Doppler radar receiver 21. The pulse Doppler radar receiver 21 outputs the created range / Doppler frequency map to the Doppler frequency resolution reduction unit 22.

  The Doppler frequency resolution reduction unit 22 includes a spectrum range division unit 221 and a peak value selection unit 222, and reduces the resolution of the Doppler frequency for the received range / Doppler frequency map. In the second embodiment, the resolution of the range is not changed.

  The spectrum range dividing unit 221 divides the Doppler frequency spectrum of each range bin into widths where resolution is desired to be reduced. The range of resolution reduction is a width that can cover this by calculating the amount of change in the target Doppler frequency during the addition period from the addition period of shift addition and the maximum value of the target expected acceleration.

  The peak selection unit 222 selects the maximum value of amplitude or power in each divided range. The peak selection unit 222 creates a resolution reduction map using the maximum value selected in each range as a representative value in the range. The peak selection unit 222 outputs the created resolution reduction map to the shift addition unit 23.

  FIG. 7 is a diagram illustrating a frequency spectrum when the Doppler frequency resolution reduction unit 22 applies resolution reduction processing to the frequency spectrum illustrated in FIG. In FIG. 7, the spectrum range dividing unit 221 divides each into 10 frequency bin widths. The peak selection unit 222 selects the maximum value in the divided range. At this time, the maximum value is selected as shown in FIG. By the resolution reduction processing in the Doppler frequency resolution reduction unit 22, the fine resolution graph as shown in the upper part of FIG. 7 is converted into the coarse resolution graph as shown in the lower part.

  The shift addition unit 23 includes a spectrum range selection unit 231, a storage unit 232, a shift unit 233, and an addition unit 234.

The spectrum range selection unit 231 selects the velocity range V _r corresponding to the range of the Doppler frequency in the vicinity of the Doppler frequency where the target is supposed to exist, with respect to the range / Doppler frequency map from the Doppler frequency resolution reduction unit 22.

The range V _r will be described below.

  The Doppler frequency is proportional to the speed, ie the rate of change of the range. That is, when the map is a range Doppler frequency map, the horizontal axis and the vertical axis of the map are not independent when performing shift addition.

FIG. 6 is a schematic diagram of a range Doppler frequency map. The speed range corresponding to the Doppler frequency range on the vertical axis is V _r , and the speed corresponding to the center Doppler frequency of the range V _r is v ₀ . At this time, the maximum speed difference between the case where the target is at the upper end and the case where the target is at the lower end is V _r . Further, the difference with respect to v _{0 at} the center of the map is V _r / 2 in absolute value at the maximum at the upper end and the lower end, respectively.

The distance between the measurement times of each map and T _0. If the target is a velocity v corresponding to a Doppler frequency on the vertical axis of the map and it does not have an extreme acceleration (that is, an acceleration that allows shift addition if the resolution is reduced), In adjacent measurement rounds, the target is at a position that is offset by approximately T ₀ × v from each other. That is, it can be seen that the amount by which the map corresponding to each speed v should be shifted can be determined to be a constant value during shift addition.

If it is assumed that the target is at the center v ₀ of the width of the vertical axis of the map, the targets are located at positions shifted by T ₀ × v _{0 from} each other in adjacent measurement times. Similarly, when it is assumed that there is a target at the upper end of the map, the targets are substantially shifted from each other by T ₀ × (v ₀ + V _r / 2) between adjacent measurement times. Assuming that there is a target at the lower end of the map, the targets are located at positions that are substantially offset from each other by T ₀ × (v ₀ −V _r / 2). Similarly, assuming that the addition period of shift addition is T, at the beginning and end of the addition period, if the target is near the center of the vertical axis, it is approximately T × v ₀ , and if it is near the upper end, it is approximately T In the vicinity of × (v ₀ + V _r / 2) and the lower end, the range changes by approximately T × (v ₀ −V _r / 2). The difference in the range change amount between the target existing at the upper end and the lower end and the target existing near the center is T × V _r / 2 at maximum in absolute value. Conversely, if a map that is separated by T is shifted by T × v _0, it can be said that the difference in the range only appears at T × V _r / 2 even at the upper limit or lower limit speed at which the speed is far from v ₀ .

When the range width corresponding to the range resolution (range resolution) is _Br , the relationship between the range width _Br and _TxVr / 2 is
B _r > T × V _r / 2
In the case of satisfying, even if the entire range of v ₀ ± V _r / 2 is shifted by the shift amount corresponding to the speed of v ₀ , there is a high possibility that the target is in substantially the same range bin at least in the range direction. That is, when the range width B _r and T × V _r / 2 satisfy the above relationship, a sufficient integration effect can be expected at least in the range direction even if shift addition is performed. That is, it does not happen that a peak flows during the shift addition period and is not added.

In order to obtain a strong integration effect by shift addition, when the addition period T is desired to be long and there is no room for selection in the range width B _r , the range V _r may be reduced based on the above relational expression. The range V _r, if there is no limit to the range width B _r and addition period T is not determined only range V _r based on the above equation, the range V _r, the range width B _r and addition period What is necessary is just to set so that each parameter of T may satisfy | fill the said relational expression.

  The storage unit 232 stores the resolution reduction map in which the spectrum range is selected by the spectrum range selection unit 231 for the number of times set in advance. Here, the preset number of measurements is set based on the addition period. When the storage unit 232 stores the resolution reduction map for the preset number of measurements, the storage unit 232 outputs the stored resolution reduction map to the shift unit 233.

As shown in FIG. 9, the shift unit 233 sets the measurement number as n (n = 0, 1,...) In the resolution reduction map output from the storage unit 232 in the range direction. The map is shifted by T ₀ × n × v ₀ . The shift unit 233 outputs the resolution reduction map shifted in the range direction to the addition unit 234.

  The adding unit 234 adds the shifted resolution reduction map non-coherently with respect to the same point using power or amplitude. The adder 234 outputs the addition result to the threshold detector 24.

  The threshold detection unit 24 provides a target detection threshold and detects a point higher than that as a point where the target exists. The detection threshold may be determined based on a noise level measured in advance so that false detection (false alarm) is constant. In addition, if the strength of a target point after the shift addition is barely higher than the other strengths in the map, the average value of the heights of the points of all the maps is calculated, and how many times it is The threshold value may be determined in the form.

  Next, the simulation result about the resolution reduction process by the target detection apparatus 20 having the above configuration will be described. In this simulation, shift addition is performed on the range / Doppler frequency map created by the pulse Doppler radar receiver 21. At this time, a slight acceleration is given to the target so that the detected bin changes every moment in the Doppler frequency spectrum before the resolution reduction. In addition, 64 pulses are set to 1 CPI, and a map for 20 CPI is shift-added. Further, the width for reducing the resolution in the shift addition process is 10 frequency bins. In addition, all non-coherent addition in the shift addition process is performed with electric power. In addition, in the map, the whole was normalized by the standard deviation of the noise of the map after addition, and the level was adjusted so that the average value of the noise after normalization was 10 dB.

  FIG. 10 is an image showing the simulation result of the addition map created by the target detection apparatus 20 according to the second embodiment and the simulation result of the addition map created by other methods. FIG. 10A is a simulation result of the addition map created using the resolution reduction method according to the present embodiment, and FIG. 10B is a simulation result of the addition map created without performing the resolution reduction processing. FIGS. 10C to 10E show simulation results of the addition map created by the method shown in FIGS.

  FIG. 11 is a diagram for explaining a resolution reduction method used when creating the addition map shown in FIG. In the method shown in FIG. 11, as in the resolution reduction method according to the present embodiment, the resolution reduction width is determined, and the power of all points in the spectrum within the determined width is added to obtain a representative value. is there.

  FIG. 12 is a diagram for explaining a resolution reduction method used when creating the addition map shown in FIG. In the method shown in FIG. 12, similarly to the resolution reduction method according to the present embodiment, a resolution reduction width is determined, and spectral points within the determined width are added as complex values to obtain a representative value. . In the case of this method, when performing shift addition, it is converted into electric power and added non-coherently.

  FIG. 13 is a diagram for explaining a resolution reduction method used when creating the addition map shown in FIG. In the method shown in FIG. 13, pulses in 1 CPI are finely divided and subjected to Fourier transform, and the result is non-coherently added by power for each bin.

  According to FIG. 10, in the result of reducing the resolution shown in FIGS. 10A and 10C, the vertical axis has six frequency bins, but the resolution shown in FIG. 10B is reduced. The result of not having 64 bins. For this reason, in the result shown in FIG. 10B, the vertical axis is different from the results shown in FIGS. 10A, 10C, and 10E.

  In the result shown in FIG. 10B, the target appears slightly to the right and slightly below the center of the map. However, it can be seen that there are a plurality of frequency bins where the target exists, and the effect of the shift addition cannot be sufficiently obtained.

  On the other hand, in the results shown in FIGS. 10A, 10 </ b> C to 10 </ b> E, all the results are within a single bin. And in the result shown to Fig.10 (a) using the resolution reduction method which concerns on this embodiment, a target appears most strongly and it turns out that the effect of the method which concerns on this embodiment is high.

  As described above, in the second embodiment, the Doppler frequency resolution reduction unit 22 lowers the resolution of the Doppler frequency in the range-Doppler frequency map, and the shift addition unit 23 shifts the map based on the approximate movement information of the target. And add the shifted maps at the same point. As a result, it is possible to detect a target with movement without increasing the amount of calculation while maintaining the effect of shift addition.

  The Doppler frequency resolution in the pulse Doppler radar is determined by the number of pulses in one CPI and the PRI. When PRI and the number of pulses are determined based on the integral gain, the obtained Doppler frequency detection accuracy, that is, the moving speed accuracy is higher than necessary. Under the condition that a moderate SNR can be obtained, it is effective to reduce the Doppler frequency bin interval for clutter suppression. However, in a situation that is so weak that the target cannot be detected in the measurement result of 1 CPI, the frequency of the desired wave is unknown, and even if the bin interval is fine, there is no meaning of clutter suppression. In other words, it can be said that the situation where the shift addition is required is a state in which the speed detection accuracy is higher than necessary in order to obtain the integral gain. According to the target detection apparatus 20 according to FIG. 5, it is possible to reduce the resolution without sacrificing the integral gain due to the increase in the number of pulses.

  Further, the reason why the resolution reduction method by peak selection according to the present embodiment is more effective than the other methods is as follows. The method shown in FIG. 13 is a method in which the CPI is substantially shortened by reducing the number of pulses of Fourier transform, and detection is performed such that the Doppler frequency resolution actually decreases. The operation of lowering the resolution increases the amount of noise in the bin where the target exists, and thus has the effect of lowering the SNR originally. In the method of FIGS. 11 and 12, information of a plurality of bins is added. However, for a range including the target, a lot of noise is clearly added to the target value. For this reason, the SNR basically deteriorates. However, since the addition is performed non-coherently, a noise smoothing effect can be expected by adding the noise, and there is simply no degradation as much as the SNR obtained from the height of the waveform. However, since the target is inherently weak, a large difference does not occur between the addition result of the range including only the noise and the addition result of the range including the target, and the SNR also deteriorates depending on how noise varies.

  On the other hand, in the method according to the present embodiment, since a peak within the range is selected, noise is not added to the representative value of the width including the target. In the range including only noise, since the maximum noise value is selected, a noise value larger than the noise average value is selected. However, if the number of bins in the width is more than a certain value (3 or 4 or more), the expected value of noise selected as a representative value is smaller than when adding. When the maximum noise value is selected, if the number of bins in the range is more than a certain level, the variation when the maximum value is selected is not so large, and the same level of noise smoothing occurs as when the addition is performed. . As a result, there is no significant difference in noise dispersion from other resolution reduction methods, but in the present invention, the average value of noise is low, and the peak height due to the target appears to be relatively high, resulting in detection. Performance will increase.

  In the second embodiment, the spectrum range selection unit 231 limits the range of the Doppler frequency spectrum whose resolution has been reduced. As a result, since the shift amount in the range direction can be determined based on the speed corresponding to the center of each range, the shift amount and the shift direction are limited to only one type and the calculation amount is reduced. Detection is possible.

  Therefore, according to the target detection device 20 according to the second embodiment, even if the received power of the reflected wave reflected by the target is small and the movement of the target cannot be clearly predicted, the calculation amount is increased. The integration effect obtained by the shift addition method can be enjoyed without making it possible to detect the target.

  In the second embodiment, the map output from the pulse Doppler radar receiver 21 is obtained by mapping the demodulated waveform of the pulse Doppler radar receiver 21 as it is. However, the map is not limited to this. Absent.

  In the second embodiment, the case where the spectrum range is selected by the spectrum range selection unit 231 has been described as an example. However, the present invention is not limited to this. For example, instead of using the spectrum range selection unit 231, the shift addition unit 23 performs shift addition for the entire range of Doppler frequencies, and the threshold detection unit 24 selects a range for detecting a target based on the same criteria. Also good.

  In FIG. 5 in the second embodiment, the Doppler frequency resolution reduction unit 22, the shift addition unit 23, and the threshold detection unit 24 are described as independent components, but the present invention is not limited to this. For example, the Doppler frequency resolution reduction unit 22, the shift addition unit 23, and the threshold detection unit 24 may be realized by processing of a signal processor.

(Third embodiment)
FIG. 14 is a block diagram illustrating a functional configuration of the target detection device 30 according to the third embodiment. In addition, the same number is attached | subjected to the part which is common in the target detection apparatus 20 shown in FIG.

  The target detection device 30 shown in FIG. 14 includes an antenna 11, a pulse Doppler radar receiver 21, a Doppler frequency resolution reduction unit 22, a shift addition unit 31, threshold detection units 32-1 to 32-m (where m is a natural number), and results. An integration unit 33 is provided.

  The shift addition unit 31 includes a map division unit 311, storage units 312-1 to 312-m, shift units 313-1 to 313-m, and addition units 314-1 to 314-m.

  The map dividing unit 311 determines the velocity range Vr corresponding to the Doppler frequency range in the same manner as the spectrum range selecting unit 231 shown in FIG. 5 with respect to the range / Doppler frequency map from the Doppler frequency resolution reducing unit 22. . The map dividing unit 311 defines a plurality of maps that are limited within the determined range as divided maps by changing the center speed.

  FIG. 15 is a diagram for explaining how the map range is divided by the map dividing unit 311 according to the third embodiment. The left square is a range-Doppler frequency map after resolution reduction extended in the vertical direction. FIGS. 15A to 15C show how to define a division map.

FIG. 15A shows a method in which the Doppler frequency direction of the range-Doppler frequency map is divided into fixed ranges and each range is used as a divided map. FIG. 15B shows a method in which the Doppler frequency direction of the range-Doppler frequency map is divided into fixed ranges, and divided maps are alternately arranged so as to overlap by about half. The method shown in FIG. 15B is effective when the target may straddle the boundary of the divided map. FIG. 15C shows a method in which the Doppler frequency direction of the range-Doppler frequency map is divided into fixed ranges while sliding so as to partially overlap, and the divided range is used as a divided map. The method shown in FIG. 15C is effective when a high addition effect is expected only in the vicinity of the center of the divided map. In FIG. 15C, it is assumed that the addition period is long and the substantially effective vertical axis range _Vr is smaller than the range of the arrows.

  The map division unit 311 outputs each division map to the storage units 312-1 to 312-m. The operations of the storage units 312-1 to 312-m, the operations of the shift units 313-1 to 313-m, and the operations of the addition units 314-1 to 314-m are the same. -1, the shift unit 313-1 and the adder 314-1 will be described as an example.

  The storage unit 312-1 stores the divided map supplied from the map dividing unit 311 for the number of times set in advance. Here, the preset number of measurements is set based on the addition period. When the storage unit 312-1 stores the division maps for the preset number of measurements, the storage unit 312-1 outputs the stored division maps to the shift unit 313-1.

  The shift unit 313-1 shifts the division map output from the storage unit 312-1 in the range direction. Shift unit 313-1 outputs the division map shifted in the range direction to addition unit 314-1.

  The adder 314-1 adds the shifted divided maps non-coherently with respect to the same point using power or amplitude. Adder 314-1 outputs the addition result to threshold value detector 32-1.

  The threshold detection unit 32-1 provides a target detection threshold, and detects a higher point as a point where the target exists. The threshold detection unit 32-1 outputs the detection result to the result integration unit 33.

  The result integration unit 33 determines which frequency in which range and which range has a target, and outputs the result.

  As described above, in the third embodiment, the Doppler frequency resolution reduction unit 22 lowers the resolution of the Doppler frequency in the range-Doppler frequency map, and the shift addition unit 31 shifts the map based on the approximate movement information of the target. And add the shifted maps at the same point. As a result, it is possible to detect a target with movement without increasing the amount of calculation while maintaining the effect of shift addition.

  In the third embodiment, the map dividing unit 311 divides a map with a reduced resolution for each predetermined Doppler frequency spectrum.

  In the target detection device 20 according to the second embodiment, only the periphery of the Doppler frequency corresponding to the speed at which the target is expected to exist is extracted. However, the target approximate speed is not always known. If the type and approximate shape of the target are known, the acceleration range is limited and does not take an extremely large value, but the speed ranges from a state where it stops completely to the upper limit speed. May change.

  By dividing the reduced-resolution map for each predetermined Doppler frequency spectrum, it is possible to cope with a case where the target approximate speed is not known.

  Therefore, according to the target detection device 30 according to the third embodiment, even when the received power of the reflected wave reflected by the target is small and the movement of the target cannot be clearly predicted, the calculation amount is increased. The integration effect obtained by the shift addition method can be enjoyed without making it possible to detect the target.

In the third embodiment, the case where the map dividing unit 311 divides the resolution reduction map as illustrated in FIG. 15 is described as an example, but the present invention is not limited to this. That is, when the radar system is a high PRF in which the Doppler frequency does not have ambiguity, the division method shown in FIG. 15 is sufficient, but in the case of a low PRF in which the Doppler frequency has ambiguity, it is slightly different. It becomes a way of definition. In the low PRF, a certain map is formed by overlapping a plurality of maps having different Doppler frequency folding times. In other words, the point of the Doppler frequency spectrum did not show a single Doppler frequency, different by an integer multiple of the repetition frequency f _a, and includes a plurality of frequency components. For this reason, the speed of a point on a certain vertical axis cannot be determined as one.

  If it is known in advance which Doppler frequency folds the target speed is, the center speed of the map may be determined using the number of folds. However, when the target speed is unknown and cannot be narrowed down to one folding range, the map is divided as shown in FIG. 16, for example.

FIG. 16 is a diagram illustrating an example of map division by the map dividing unit 311 according to the third embodiment. The velocity width corresponding to the repetition frequency f _a of the Doppler frequency of the low PRF was V _a. In FIG. 16, for each of a plurality of folding times p (p is an integer, p = p _{0 to} p ₁ ) that may have a target, each folding is defined as one division map. The shift adder 31 performs shift addition by shifting the map at _a speed v ₁ + pV _a corresponding to the center frequency of each turn, with a speed corresponding to a half frequency of the repetition frequency as v ₁ . Specifically, the map division unit 311 duplicates the same map and sends it to the shift units 313-1 to 313-m via the storage units 312-1 to 312-m. The shift units 313-1 to 313-m change the shift amount corresponding to the number of times of folding p.

  The target within a certain number of loopbacks is weakly detected not only as a result of shift addition at a speed corresponding to the correct number of loopbacks but also as a result of shift addition at a speed corresponding to the number of loopbacks before and after that. There is. Since a point with the same target is detected at a coordinate point on the same map in any map, the result integration unit 33 detects a target at the same coordinate point on an adjacent map after adding several adjacent folding times Then, it is determined that the target is in the number of the most strongly detected folding. Alternatively, the result integration unit 33 determines that the target is at the center of the detected number of times of folding.

FIG. 17 is a diagram illustrating another example of map division by the map division unit 311 according to the third embodiment. In FIG. 17, the range V _r is made smaller than the velocity width V _a , and the center of the divided map range is not limited to the center of the repetition frequency, but is appropriately determined. As a result, the range of the Doppler frequency for shift addition can be made smaller than the repetition frequency of the Doppler frequency. Since the end of the spectrum of the low PRF is originally connected, the same map is placed one above the other so as to compensate for the part outside one repetition frequency. This makes it possible to detect a target at the end of the repetition frequency. As in FIGS. 15B and 15C, the range of the divided map may be overlapped by about half or more.

  Note that when the division method shown in FIG. 17 is performed, the map may be divided before the resolution is reduced, and the resolution reduction width may be defined so as to straddle the end of the repetition frequency. For example, when the number of bins of the original Doppler frequency is 128, if the resolution reduction width is 8, the resolution-reduced spectrum has 16 bins, and 128 is divisible by 8. All subsequent bins are correct representative values for the original eight bins. However, when the resolution reduction width is 10, the representative values for the original 10 bins are obtained for the first 12 bins after resolution reduction, but the 13th resolution is obtained. The bins after the reduction are representative values for eight. At high PRF, the Doppler frequency at the end of the Doppler frequency spectrum is outside the detection range in most cases, so even if such a fraction occurs, there is no problem, but at low PRF, the target is at the end of the folding range. If there is such a fraction, there is a possibility that the detection performance of the target applied to it will deteriorate. As shown in FIG. 17, when the same map is used in an up-and-down direction, a map in which all of the number of wrapping times to be inspected are overlapped is created before the resolution is reduced, and then the resolution is reduced and the map is divided. Thus, it is possible to avoid the above-mentioned problem that the fraction occurs.

  In FIG. 14 in the third embodiment, the Doppler frequency resolution reduction unit 22, the shift addition unit 31, the threshold detection units 32-1 to 32-m, and the result integration unit 33 are described as independent components. However, it is not limited to this. For example, the Doppler frequency resolution reduction unit 22, the shift addition unit 31, the threshold detection units 32-1 to 32-m, and the result integration unit 33 may be realized by processing of a signal processor.

(Other embodiments)
FIG. 18 is a block diagram illustrating a functional configuration of the guidance device 100 using the target detection device 10, 20 or 30 according to the first to third embodiments. A guidance device 100 illustrated in FIG. 18 includes a target detection device 10, 20, or 30, a target tracking device 101, and a guidance signal generation unit 102.

  The target detection device 10, 20, or 30 notifies the target tracking device 101 of the target detection result, that is, the target range and speed.

  The target tracking device 101 shares the radar receiver with the target detection device 10, 20 or 30, and uses the target range and speed notified from the target detection device 10, 20 or 30 as initial values from the output of the radar receiver. Tracking is performed by continuously detecting the target range, speed, angle, and the like. The target tracking device 101 outputs the tracking result to the guidance signal generation unit 102.

  The guidance signal generation unit 102 generates a guidance signal for guiding the flying object based on the tracking result of the range or the like supplied from the target tracking device 101. The induction signal generation unit 102 outputs the generated induction signal to a flying body drive unit (not shown) to control the drive unit.

  The target detection device 10, 20 or 30 may detect the target once and notify the detection result to the target tracking device 101, and may finish the operation. However, the target tracking device 101 starts tracking the target. May be lost on the way. Alternatively, the target detected by the target detection device 10, 20, or 30 may be a false alarm and cannot be followed. In preparation for such a case, the target detection device 10, 20 or 30 may continue the detection at all times while the target reflected wave is weak. Alternatively, when the target tracking device 101 fails to follow the target, the target detection processing may be resumed immediately upon receiving a request from the target tracking device 101. In this way, even when the target reflected wave is weak, it is possible to detect the target without significantly increasing the amount of calculation and to control the flying object to follow.

  Although several embodiments have been described, these embodiments have been presented by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and the equivalents thereof.

  DESCRIPTION OF SYMBOLS 10,20,30 ... Target detection apparatus, 11 ... Antenna, 12 ... Radar receiver, 13 ... Resolution reduction part, 131 ... Range division part, 132, 222 ... Peak value selection part, 14, 23, 31 ... Shift addition part 141, 232, 312-1 to 312-m ... storage unit, 142, 233, 313-1 to 313-m ... shift unit, 143, 234, 314-1 to 314-m ... addition unit, 15 ... detection unit , 21 ... Pulse Doppler radar receiver, 22 ... Doppler frequency resolution reduction unit, 221 ... Spectral range division unit, 231 ... Spectral range selection unit, 24, 32-1 to 32-m ... Threshold detection unit, 311 ... Map division , 33 ... Result integrating unit, 100 ... Guiding device, 101 ... Target tracking device, 102 ... Guidance signal generating unit

Claims (12)

A receiving unit that receives a reflected wave reflected by a measurement object and creates a measurement waveform map based on the reflected wave;
A width is set for each of one or more predetermined axes of the axes of the measurement waveform map based on the amount of change in motion predicted for the measurement object in a preset addition period, and the measurement waveform map A range dividing unit that divides a plurality of ranges of the width corresponding to the predetermined axial direction;
A peak value selection unit that detects a maximum value of power or amplitude in each of the divided ranges and generates a resolution reduction map in which the resolution of the measurement waveform map is reduced using the detected values as representative values in each range. When,
A shift unit that shifts a plurality of resolution reduction maps generated by a plurality of measurements during the addition period in a direction set for each measurement time by an amount set for each measurement time;
An adder for adding the power or amplitude of the plurality of shifted resolution reduction maps to generate an addition map;
A target detection apparatus comprising: a detection unit that detects a point where the measurement target is estimated to exist based on the power or amplitude of the addition map. The receiver is a pulse Doppler radar receiver,
2. The target detection apparatus according to claim 1, wherein the measurement waveform map is a two-dimensional map of a range and a Doppler frequency. The range dividing unit divides the Doppler frequency into a plurality of ranges,
The peak selection unit detects a maximum value of power or amplitude in the divided Doppler frequency range, and generates a resolution reduction map using the detected value as a representative value in each range. The target detection apparatus according to claim 2.   The shift unit reduces the plurality of resolutions in the range direction by an amount corresponding to a moving distance calculated from a moving speed corresponding to any Doppler frequency in the resolution reduction map and an elapsed time at each measurement time. The target detection apparatus according to claim 3, wherein the map is shifted. A range selection unit for selecting a range of Doppler frequency of a resolution reduction map generated by a plurality of measurements during the addition period;
The target detection apparatus according to claim 4, wherein the shift unit performs a shift process of the resolution reduction map in which the range of the Doppler frequency is selected. The range selector comprises a range resolution B _r of the resolution reduction map, the the addition period T, and a width V _r of speed corresponding to the Doppler frequency range of directions of the selected resolution reduction map, B _r> The target detection apparatus according to claim 5, wherein the target detection apparatus is set so as to satisfy a relationship of T × V _r / 2.   The said detection part selects the range of the Doppler frequency of the said addition map, and detects the point from which the power or amplitude is estimated to be larger than another point from the addition map after selection. Target detection device. Wherein the detection unit includes a range resolution B _r of the resolution reduction map, the the addition period T, and a width V _r of speed corresponding to the Doppler frequency range of directions of the selected resolution reduction map, B _r> T The target detection apparatus according to claim 7, wherein the target detection apparatus is set so as to satisfy a relationship of × V _r / 2. A plurality of resolution reduction maps generated by a plurality of measurements during the addition period, further divided in a Doppler frequency direction, further comprising a map dividing unit that generates a plurality of divided maps;
The shift unit is configured to divide the plurality of divisions in the range direction by an amount corresponding to a movement distance calculated from a movement speed corresponding to any Doppler frequency in the plurality of division maps and an elapsed time at each measurement time. The target detection apparatus according to claim 4, wherein the map is shifted. The map dividing unit calculates a range resolution B _r of the resolution reduction map, the addition period T, and a velocity width V _r corresponding to a range in the Doppler frequency direction of the divided map, such that B _r > T × V _r. The target detection apparatus according to claim 9, wherein the target detection apparatus is set so as to satisfy the relationship of / 2. The target detection apparatus according to any one of claims 1 to 10, which detects a measurement target;
A target tracking device that follows the detected measurement target and obtains information on the measurement target;
An induction device comprising: an induction signal generation unit that generates an induction signal based on the acquired information. Receive the reflected wave reflected by the measurement object, create a measurement waveform map based on the reflected wave,
A width is set for each of one or more predetermined axes of the axes of the measurement waveform map based on the amount of change of motion predicted for the measurement object in a preset addition period, and the measurement waveform map Is divided into a plurality of ranges of the width corresponding to the predetermined axial direction,
Detecting a maximum value of power or amplitude in each of the divided ranges, and generating a resolution reduction map in which the resolution of the measurement waveform map is reduced using the detected value as a representative value in each range;
Holding a plurality of resolution reduction maps acquired during the addition period;
Shifting the held plurality of resolution reduction maps in the direction set for each measurement by the amount set for each measurement,
Adding the power or amplitude of the shifted plurality of resolution reduction maps to generate an addition map;
A target detection method for detecting a point at which the measurement target is estimated to exist based on the power or amplitude of the addition map.

Images