JP4984705B2 - Radar - Google Patents

Description

  The present invention relates to an FM-CW radar used for preventing collision of automobiles, and more particularly to a radar using an array antenna.

  An FM-CW radar calculates a relative distance and a relative speed to a target based on a beat signal obtained by transmitting and receiving a frequency-modulated continuous wave.

  An array antenna in which a plurality of element antennas are arranged may be used as a transmission antenna or a reception antenna of such a radar. In a radar using an array antenna, a received signal for each element antenna is obtained by transmitting a continuous wave or receiving an incoming wave via each element antenna, and estimates the arrival direction of the incoming wave by the arrival direction estimation method.

In the arrival direction estimation method, the output of each element antenna is multiplied by an appropriate complex weight. In other words, by appropriately manipulating the amplitude and phase of the received signal of each element antenna, the directivity of the array antenna is given to an arbitrary directivity. In this way, the received signal strength with each azimuth within the detection range as the directional azimuth is detected, and the directional azimuth at which the received signal strength peaks among the directional azimuths is estimated as the arrival azimuth of the incoming wave. Further, the arrival direction of the incoming wave may be estimated with the NULL point of the directivity characteristic directed to an arbitrary direction.
Various methods such as Beamformer method, Capon method and MUSIC method are known as arrival direction estimation methods. Each arrival direction estimation method has its advantages and disadvantages.For example, the Beamformer method has the advantage of being able to estimate the arrival direction by reducing the amount of calculation. There are disadvantages that cannot be estimated. The directivity of the Capon method is better in terms of angular resolution than the directivity of the Beamformer method, but inferior in computational complexity. The directivity of the MUSIC method is superior in terms of angular resolution compared to the directivity of the Beamformer method, but inferior in computational complexity. As described above, in the arrival direction estimation method, the angular resolution and the calculation amount have a contradictory relationship.

  Also, in the arrival direction estimation method, the angular resolution and the calculation amount are different depending on the scanning pitch for switching the directivity, and if the scanning pitch is fine, the angle resolution is excellent, but the total calculation amount increases. Conversely, if the scanning pitch is made rough, the overall calculation amount can be suppressed, but the angle resolution is inferior.

  As described above, in the conventional FM-CW radar, it is difficult to achieve both high angular resolution and low calculation amount in estimating the arrival direction using the arrival direction estimation method. Therefore, a radar has been proposed that achieves both high-precision estimation of the arrival direction using the arrival direction estimation method and suppression of the calculation amount.

  In the radar of Patent Document 1, first, the distance to the target is detected based on the beat signal, and the spectrum peak frequency is equal to or higher than the boundary frequency to be determined as a long distance, or the boundary to be determined as a short distance. Detect if the frequency is below. Next, the arrival azimuth is estimated using the algorithm of the arrival azimuth estimation method that differs depending on the distance at which the target is observed depending on whether the target is a long distance or a short distance.

  This radar uses an arrival direction estimation algorithm that can reduce the amount of calculation when estimating the arrival direction for short distances, and uses an arrival direction estimation algorithm with high angular resolution for estimation of arrival directions for long distances. The aim was to achieve both resolution and reduced computational complexity.

  Similarly, the radar of Patent Document 2 detects the distance to the target based on the beat signal, and then estimates the arrival direction using a different scanning pitch for each distance at which the target is observed.

  This radar uses a coarse scanning pitch that can reduce the amount of calculation when estimating the arrival direction for a short distance, and uses a fine scanning pitch that can increase the angular resolution when estimating the direction of arrival for a long distance. It was intended to achieve a balance with the suppression of the amount.

  In addition, the radar of Patent Document 3 first estimates the arrival direction using the first arrival direction estimation method having a low angular resolution, and then surrounds the arrival direction estimated using the first arrival direction estimation method. The arrival direction is estimated with high accuracy by using the second arrival direction estimation method having high angular resolution with respect to the direction.

This radar estimates the arrival direction only by the first arrival direction estimation method having a directivity characteristic with low angular resolution for the direction where the target does not exist, and the first direction for the direction where the target exists. In addition to the estimation of the arrival direction by the arrival direction estimation method, the arrival direction is estimated by the second arrival direction estimation method having the directivity characteristics with high angular resolution. In this way, the estimation accuracy of the arrival direction of the target is improved, and the calculation amount for the direction where the target does not exist is suppressed.
Japanese Patent No. 3489114 Japanese Patent No. 3500609 JP 2003-139849 A

  However, in the radars disclosed in Patent Document 1 and Patent Document 2, since the received signal strength is calculated in all directions with respect to a long-distance target, calculation of estimation of the arrival direction for the long-distance target is performed. The quantity was very easy to grow. In addition, since the arrival direction estimation with a low angular resolution is performed in all directions on a short-distance target, only the arrival direction can be estimated with low accuracy.

  Further, in the radar disclosed in Patent Document 3, it is necessary to perform the first direction-of-arrival estimation method having directivity with low angular resolution for all directions, and the first and first directions with respect to the direction where the target exists. It was necessary to estimate the arrival direction using each of the two arrival direction estimation methods. Therefore, there is a limit to the suppression of the calculation amount of the first arrival direction estimation method.

  Therefore, an object of the present invention is to provide a radar that can suppress the calculation amount for arrival direction estimation more accurately than the conventional one while accurately estimating the direction in which the target exists.

(1) The radar according to the present invention calculates the received signal strength of an arbitrary directional direction with a predetermined directivity characteristic based on the received signal for each element antenna obtained by transmission or reception via each element antenna of the array antenna. In a radar that estimates the arrival direction of an incoming wave based on the received signal strength for each directional direction obtained by switching the directional direction at a predetermined scanning pitch, reception at the switched directional direction is performed each time the directional direction is switched. Based on the signal strength and the received signal strength at the previous directional azimuth, the peak detecting means for detecting that the peak azimuth of the received signal strength is approaching, and the peak azimuth of the received signal strength is approached by the peak detecting means And a resolution changing means for changing the subsequent angular resolution higher.

  With this configuration, when switching the azimuth direction, it is detected that the directional azimuth (peak azimuth) at which the received signal intensity reaches a peak, and the angular resolution thereafter is further improved, so that the arrival direction can be estimated with high accuracy. While performing, the calculation amount of estimation for all directions is suppressed.

(2) Further, the resolution changing means of the present invention changes the angular resolution high by reducing the scanning pitch. With this configuration, the accuracy of arrival direction estimation is increased.

(3) In addition, the resolution changing means of the present invention changes the angular resolution high by changing the directivity characteristics. For example, by switching each algorithm such as Beamformer method, Capon method, and MUSIC method, the directional characteristics are changed, and the accuracy of estimating the direction of arrival is improved.

(4) Further, in the peak detection means of the present invention, the increase width of the received signal strength before and after the switching of the directivity direction is larger than the increase width of the received signal strength during the previous switching of the directivity direction. Based on the first condition of changing, it is detected that the peak direction of the received signal intensity approaches.

  According to the first condition, the peak initial state in which the increase width of the received signal intensity gradually increases is detected to detect the approach of the peak direction. Therefore, if the approach of the peak direction is detected under this first condition and the angular resolution is increased, the missed peaks are reduced.

(5) Further, the peak detection means of the present invention is based on the second condition that the increase width of the received signal strength before and after the switching of the directivity changes beyond the threshold value. It detects that the peak direction is approaching.

  Based on this second condition, it is detected that the peak direction is approaching based on the steepness of the peak. In this case, it is possible to reduce false detection of noise or the like in which the received signal intensity gradually changes.

  Further, since the threshold value is used, it is possible to detect that the peak direction is approaching with a simple calculation, and it is possible to effectively suppress the calculation amount of this calculation.

(6) Further, in the peak detection means of the present invention, the increase width of the received signal strength before and after the switching of the directivity direction is smaller than the increase width of the received signal strength during the previous switching of the directivity direction. Based on the third condition of changing, it is detected that the peak direction of the received signal intensity approaches.

  According to the third condition, the proximity of the peak direction is detected by detecting the vicinity of the peak where the increase width of the received signal intensity gradually decreases. Therefore, if it is detected near the top of the peak under this third condition and the angular resolution is increased, the angular resolution at a portion other than the top of the peak can be suppressed, and the number of processing with high angular resolution is reduced. The total amount of calculation can be reduced.

(7) Further, in the peak detection means of the present invention, the increase width of the received signal strength before and after switching of the directivity direction first satisfies the first condition, and then satisfies the third condition. By doing so, it is detected that the peak direction of the received signal intensity approaches.

  First, by using the first condition, the approach of the peak direction can be detected and the peak direction can be captured. Further, by using the third condition, it is possible to reliably detect the approach of the peak near the top of the peak.

(8) Further, the peak detection means of the present invention includes a passage detection means for detecting that the peak direction of the received signal intensity is far, and the resolution changing means is that the peak direction is far from the peak detection means. After detection, the angular resolution is changed to a low value. This passage detection means detects that the peak direction of the received signal strength is far away, and lowers the angular resolution thereafter to calculate arrival direction estimation for all directions. The amount can be suppressed.

(9) In addition, the peak detection unit of the present invention detects that the peak direction is approaching using any of the first to third conditions, and then the peak direction is long using the passage detection unit. Based on the fourth condition that, after that, the decrease width of the received signal strength before and after the switching of the directional orientation is smaller than the threshold value, or the received signal strength increases before and after the switching of the directional orientation. Then, the approach of the next peak direction is detected.

  If the decrease in received signal strength after passing the peak is gradual, or if the received signal strength increases despite passing, the next peak may be approaching. Therefore, in this configuration, when the received signal intensity does not decrease or increases after the peak passage, the angular resolution is increased again. Thereby, for example, even when a plurality of peaks are overlapped, each of a plurality of peak directions can be detected.

  The radar according to the present invention can suppress the amount of calculation for estimating the arrival direction while estimating the arrival direction of the incoming wave with sufficient accuracy.

A radar according to a first embodiment of the present invention will be described with reference to FIGS.
FIG. 1 is a block diagram illustrating a configuration of a main part of a radar 50 according to the present embodiment.
The radar 50 according to this embodiment includes a reception antenna 1, a transmission array antenna 10, a switch circuit 3, a voltage controlled oscillator (VCO) 4, a branch circuit 5, an LNA 6, a mixer 7, an IF amplifier 8, and a signal processing circuit 9. .

  The array antenna 10 radiates an electromagnetic wave as a transmission signal toward a predetermined detection area including the front direction of the antenna. For example, if this radar is mounted on an automobile and forward detection is performed, the transmission signal is emitted to a detection area having a predetermined width in front of the automobile. Here, a continuous wave whose frequency changes in a triangular shape with time is used as the transmission signal.

  This array antenna 10 is for transmission in which element antennas 2A to 2E having the same directivity pattern are arranged in order on a straight line, and the element antennas 2A to 2E are arranged at equal intervals so that the front directions coincide with each other. It is a thing. The array antenna 10 is realized by, for example, a microstrip antenna connected in parallel or a waveguide slot array antenna connected in parallel. Specifically, the case of the microstrip antenna will be described. A plurality of patch antennas arrayed at equal intervals in a two-dimensional manner on a dielectric substrate are connected by a microstrip line in units of rows, and the microstrip line antennas in units of rows are connected. Arrange at equal intervals in the row direction. Each microstrip line antenna corresponds to each of the element antennas 2 </ b> A to 2 </ b> E, and each microstrip line antenna is connected to the transmitter via the switch circuit 3.

  When the detection operation is started, the signal processing circuit 9 gives the VCO 4 a modulation voltage whose voltage value fluctuates in a triangular waveform at a predetermined timing. The VCO 4 to which the modulation voltage is applied is within a predetermined frequency range from the timing. For example, in the 76 GHz band, a continuous wave whose frequency varies like a triangular wave is generated. The signal processing circuit 9 provides the switch circuit 3 with an antenna switching signal synchronized with the timing.

  The branch circuit 5 gives the continuous wave output from the VCO 4 to the switch circuit 3 and a part thereof to the mixer 7 as a local signal. The switch circuit 3 switches at the timing according to the switching signal from the signal processing circuit 9 to select the element antennas 2A to 2E, and applies the continuous wave output from the VCO 4 to the selected element antenna. The element antennas 2A to 2E of the array antenna 10 are sequentially selected by the switch circuit 3, and the selected element antenna radiates (transmits) a continuous wave generated by the VCO 4 to an external detection region.

  The continuous wave transmitted from the array antenna 10 is reflected by a target within the detection range and received by the receiving antenna 1. In the following description, a reception signal transmitted through each element antenna and received by the reception antenna is referred to as a reception signal of each channel. The reception antenna 1 outputs the reception signal of each channel to the LNA 6. The LNA 6 amplifies the received signal of each input channel and outputs the amplified signal to the mixer 7. The mixer 7 mixes the local signal input from the branch circuit 5 with the received signal input from the LNA 6. The IF amplifier 8 amplifies the reception signal (IF beat signal) of each channel mixed with the local signal and outputs the amplified signal to the signal processing circuit 9.

  The signal processing circuit 9 samples the received signal of each input channel, performs a Fourier transform on the received signal strength of any channel, and detects from this apparatus using a known FM-CW method calculation. The distance to the target reflecting the signal and the relative speed between the present apparatus and the target reflecting the detection signal are calculated.

  In addition, in the radar 50 of the present embodiment, in addition to the above-described calculation processing of the relative speed and relative position of the FM-CW method, processing for estimating the arrival direction is performed. The signal processing circuit 9 calculates the spectrum intensity of a specific directivity for the received signal of each channel. As the arrival direction estimation method, the known Beamformer method algorithm or the known Capon method algorithm is used, and the spectrum intensity of each direction is calculated while changing the direction direction. Then, the approach of the peak direction is detected, and the algorithm of the arrival direction estimation method and the scanning pitch are switched.

  Hereinafter, this process will be described with reference to FIG. In this description, it is assumed that the azimuth angle (θ) at which spectrum intensity is detected first is −20 °, the algorithm is the Beamformer algorithm, and the scanning pitch Δθ is 2 °.

(S11) A weight vector of a specific azimuth angle θ is calculated by a selected algorithm (Beamformer method or Capon method), and the received signal of each channel is multiplied by the weight vector and weighted. That is, the amplitude and phase of the received signal of each channel are converted, and the received signal strength is calculated when the azimuth angle θ is the directional direction. Then, the received signal strength is normalized to calculate the spectrum strength S _i .

(S12, S13) Every time the spectrum intensity with the azimuth angle θ as the directing direction is calculated by the Beamformer method algorithm, the comparison is made between the spectral intensity of the directional direction previously calculated by the Beamformer method algorithm and the spectral intensity of the latest directional direction. A peak approach detection process for detecting that the peak direction is approaching is performed. In the case of the radar apparatus of the present embodiment, specifically, a condition that the spectrum intensity increase is smaller than before (third condition) and a condition that the spectrum intensity increase is larger than the threshold (second condition). Are used to determine that the peak direction is approaching.

The third condition that the increase width of the spectrum intensity is smaller than before (the slope is positive and the slope of the slope is negative) is for detecting that the peak intensity is near the top of the peak. The second condition that the width is larger than the threshold is for determining that a peak accompanied by a sharp change in spectrum intensity is approaching. By using both of these conditions, erroneous detection due to noise or the like is prevented, and the peak direction is reliably detected. This peak approach detection process will be described later.
(S15, S17) If the peak direction is approaching, the algorithm used for subsequent weight vector acquisition is changed from the Beamformer method algorithm to the Capon method algorithm, and the scanning pitch is made fine (Δθ = 0.1 °) and the Capon method is used. The spectrum intensity is calculated by the algorithm.

(S15, S18) On the other hand, if the peak azimuth is not approaching, the weight vector is obtained with the Beamformer method algorithm and the scanning pitch is coarse, and the subsequent spectral intensity is calculated.

(S12, S14) When the spectrum intensity with the azimuth angle θ as the directing direction is calculated by the Capon algorithm instead of the Beamformer algorithm, the spectrum intensity of the directional azimuth previously calculated by the Capon algorithm and the latest It is detected that the peak azimuth is far from comparison with the spectrum intensity of the directional azimuth. In the case of the radar apparatus of this embodiment, specifically, the attenuation amount of the spectrum intensity after passing through the peak direction is measured, and the peak direction is far when the predetermined attenuation amount (for example, −3 dB) is obtained. Is detected. This peak passage detection process will be described later.

  (S16, S18) If the peak azimuth is far, the algorithm used for acquiring the weight vector is changed from the Capon algorithm to the Beamformer algorithm, the scanning pitch is coarsened (Δθ = 2 °), and the subsequent spectral intensity Perform the calculation.

  (S16, S17) On the other hand, if the peak azimuth is not far, the subsequent spectrum intensity calculation is performed with the algorithm used for acquiring the weight vector being the Capon algorithm and the scanning pitch being fine.

  By such processing, the azimuth angle θ is switched at the scanning pitch Δθ, the arrival azimuth estimation algorithm is also switched, and the spectrum intensity calculation is repeated to calculate the angular distribution of the spectrum intensity with the arbitrary azimuth angle as the directional direction.

  Next, the peak approach detection process in which the signal processing circuit 9 detects that the peak direction of the spectrum intensity is approaching will be described with reference to FIG.

(S21) First, from the difference between the spectrum intensity S _i-1 previously calculated by the Beamformer method and the latest spectrum intensity S _i , the slope ΔS _i of the spectrum intensity between the switching of the latest directivity is calculated.

(S22) Further, from the difference between the spectrum intensity S _i-1 previously calculated by the Beamformer method and the spectrum intensity S _i-2 calculated last time, the slope ΔS _{i− 1} is calculated.

(S23, S24, S25) Further, if the latest slope ΔS _i and the previous slope ΔS _i-1 are both positive, that is, increase, the second slope ΔS _i is greater than the predetermined threshold value α. And the third condition that the latest inclination ΔS _i is smaller than the previous inclination ΔS _i−1, it is determined that the peak direction is approaching. Otherwise, it is determined that the peak direction is not approaching.

  Through such processing, the radar according to the present embodiment detects that the peak azimuth is approaching, and switches between the arrival azimuth estimation algorithm and the scanning pitch using the approach of the peak azimuth as a trigger.

  In addition to the third condition in which the increase width of the spectrum intensity is reduced as in the present embodiment, the fact that the increase width of the spectrum intensity changes beyond a predetermined threshold is added to the second condition, so that noise or the like Since the influence can be eliminated, it can be determined more reliably that the peak direction approaches. However, in this case, when a plurality of peak directions are close to each other, one peak direction may be missed. Therefore, it is possible to make a comprehensive determination that one of the two conditions described above should be satisfied, and to reliably obtain the respective peak directions when a plurality of peaks are close to each other.

  Note that it is not necessary to use both the third condition that the spectrum intensity increase width is smaller than the previous condition and the second condition that the spectrum intensity increase width is larger than the threshold value. That is, it may be determined that the peak azimuth is approaching only by comparing the spectrum intensity increase width, or it may be determined that the peak azimuth is approaching only when the spectrum intensity increase width is larger than a predetermined threshold. .

  Further, the approach of the peak direction may be detected based on the first condition for detecting that the increase in the spectrum intensity is larger than before (the inclination and the inclination of the inclination are positive). This first condition is for detecting the initial peak.

  Each of the above three conditions may be used alone, or a plurality of conditions may be used in combination. When used alone, the approach of the peak can be detected by a simple calculation, so that the program capacity can be suppressed in the case of software processing, and the circuit configuration can be simplified in the case of hardware processing. On the other hand, when used in combination, the approach of the peak can be detected more accurately, and the amount of calculation can be reduced.

  Although it is impossible to determine the peak approach using the first condition and the third condition at the same time, first, the peak is detected using the first condition for detecting that the spectrum intensity increase is larger than before. Detection may be performed, and thereafter, the third condition may be switched to detect that the increase in spectrum intensity is smaller than before. Then, it becomes possible to reliably detect the approach of the peak direction near the top of the peak. Further, it may be detected that the peak direction is approaching based on the second condition first, and then that the next peak direction is approaching based on the third condition.

  Next, a peak passage detection process for detecting that the peak direction of the spectrum intensity is far will be described with reference to FIG.

(S31, S32) In the signal processing circuit 9, the spectrum intensity is decreased from the previous time, and the current spectrum intensity S _i is a fixed ratio (for example, 3 dB) as compared with the spectrum intensity S _peak of the immediately preceding peak direction. If it has decreased above, it is determined that the peak is far away, and if it is not lower than a certain rate (for example, 3 dB or more), it is determined that the peak is not far away.

  Through such processing, the radar according to the present embodiment detects that the peak direction is far away, and switches between the algorithm and the scanning pitch by using the far peak as a trigger.

As described above, the radar 50 according to the present embodiment accurately detects that the peak direction of the spectrum intensity approaches during electronic scanning, switches the arrival direction estimation algorithm to a more accurate one, and narrows the scanning pitch. Therefore, it is possible to improve the estimation accuracy of the arrival direction with respect to the direction in which the target exists. Even in this case, it is possible to suppress the amount of calculation for estimating the arrival direction for all directions.
Each of the above processes can be realized by hardware processing or software processing.

Next, a second embodiment of the present invention will be described.
The radar of the present embodiment has the same configuration as the radar shown in the first embodiment, and the arrival direction estimation process performed by the signal processing circuit is different.

  Specifically, in the detection of peak approach in the peak approach detection process, the third condition that the increase width of the received signal intensity is smaller than the previous increase width, and the decrease width of the spectrum intensity after passing the peak are The peak is detected using a fourth condition that is smaller than the threshold value or that the spectrum intensity increases after passing the peak.

  FIG. 4 shows a flow of the peak approach detection process in the present embodiment.

(S41) First, the slope ΔS _i of the latest spectrum intensity is calculated from the difference between the spectrum intensity S _i-1 calculated last time and the latest spectrum intensity S _i by the Beamformer method.

(S42) The slope ΔS _i-1 of the previous spectrum intensity is calculated from the difference between the spectrum intensity S _i-1 calculated last time by the Beamformer method and the spectrum intensity S _i-2 calculated last time.

(S43, S44, S45) When the previous peak passage detection process is not performed, both the latest slope ΔS _i and the previous slope ΔS _i-1 are positive, and the latest slope ΔS _i is If the third condition of being smaller than the slope ΔS _i-1 is cleared, it is determined that the peak direction is approaching. Otherwise, it is determined that the peak direction is not approaching.

(S46) Also, if you are performing a previous peak passage detection processing, that is, when the switching to the Beamformer method was performed from Capon method, the spectrum intensity S _i of the latest Beamformer method, the previous Capon method Compared to the spectrum intensity _Sj last detected by the Beamformer method before switching, the fourth condition that the decrease width is smaller than the threshold (α) or the spectrum intensity increases is cleared. A), it is determined that the next peak direction is approaching. Otherwise, it is determined that the peak direction is not approaching.

(S47) When the peak azimuth is approaching, the spectrum intensity S _j is updated for use in later determination.

  Through such processing, the radar according to the present embodiment detects that the peak azimuth is approaching, and switches between the arrival azimuth estimation algorithm and the scanning pitch using the approach of the peak azimuth as a trigger.

  Here, FIG. 5 shows an example of an experimental result for the radar of this embodiment.

  FIG. 5 is a diagram showing an angular spectrum distribution in an experimental example, where the horizontal axis represents the angle and the vertical axis represents the spectrum intensity. In this experimental example, the initial directivity angle is −20 °, the initial algorithm is the Beamformer algorithm, and the scanning pitch is 2 °. Further, the algorithm for switching by detecting the peak direction is the Capon algorithm, and the scanning pitch is 0.1 °. In addition, the arrival directions of incoming waves in the experimental environment are set to −5 °, 0 °, and 10 °.

  The radar in the experiment calculated the spectrum intensity while switching the azimuth at a scanning pitch of 2 ° by the Beamformer algorithm in order from the angle range of −20 ° to −6 °.

  After calculating the spectrum intensity at an azimuth angle of -6 °, it detects that the latest increase width is smaller than the previous increase width and satisfies the third condition, so that the peak direction approaches, and the algorithm is switched to the Capon algorithm. The scanning pitch was 0.1 °. Next, the spectrum intensity was calculated while switching the azimuth angle θ at the scanning pitch of 0.1 ° by the Capon algorithm in order from the angle range of −5.9 ° to −4.5 °.

  After calculating the spectrum intensity at an azimuth angle of -4.5 °, it is detected that the peak direction is far from the spectrum intensity, which is 3 dB lower than the spectrum intensity at the peak direction, and the algorithm is switched to the Beamformer algorithm, and the scanning pitch Was set to 2 °. Next, the spectrum intensity in the angle range of −2.5 ° was calculated by the Beamformer method algorithm.

  As a result of calculating the spectrum intensity by Beamformer method at azimuth angle -2.5 °, it was detected last by Beamformer method before switching to Capon method, even though it passed the peak detected by the previous Capon method. Compared with the spectrum intensity (here, the spectrum intensity at the azimuth angle of −6 °), the current spectrum intensity by the Beamformer method (azimuth angle of −2.5 °) is large, so it is determined that the fourth condition is satisfied. Then, it was detected that the next peak direction was approaching, the algorithm was switched again to the Capon method algorithm, and the scanning pitch was set to 0.1 °. Next, the spectrum intensity was calculated while switching the azimuth angle θ at a scanning pitch of 0.1 ° by the Capon algorithm in order from an angle range of −2.4 ° to 0.5 °.

  After calculating the spectrum intensity at an azimuth angle of 0.5 °, it was detected that the peak azimuth was far from 3 dB lower than the spectrum intensity of the peak azimuth, and the algorithm was switched to the Beamformer algorithm, and the scanning pitch was set to 2 °. . Next, the spectrum intensity was calculated while switching the azimuth angle θ at a scanning pitch of 2 ° by the Beamformer algorithm in order from the angle range of 2.5 ° to 8.5 °.

  After calculating the spectrum intensity at an azimuth angle of 8.5 °, it detects that the next peak direction is approaching because the latest increase width is smaller than the previous increase width and satisfies the third condition, and the algorithm is the Capon algorithm The scanning pitch was set to 0.1 °. Next, the spectrum intensity was calculated from the angle range 8.6 ° to 10.5 ° while switching the azimuth angle θ at a scanning pitch of 0.1 ° by the Capon algorithm.

  After calculating the spectrum intensity at an azimuth angle of 10.5 °, it was detected that the peak azimuth was far from 3 dB lower than the spectrum intensity of the peak azimuth, the algorithm was switched to the Beamformer algorithm, and the scanning pitch was set to 2 °. . Then, the spectrum intensity was calculated while switching the azimuth angle θ at the scanning pitch of 2 ° by the Beamformer algorithm in order from the angle range of 12.5 ° to 18.5 °.

  As shown in this experimental example, in the radar according to this embodiment, the arrival direction of the incoming wave can be estimated with high accuracy, and both the number of processes of the Capon algorithm and the number of processes at a fine scanning pitch can be suppressed. It is.

In the above embodiment, the configuration in which the peak azimuth is detected and the scan pitch and the arrival azimuth estimation algorithm are both switched has been described. However, only the scan pitch may be switched.
In the present embodiment, the fourth condition is to compare the results of the calculation of the spectrum intensity by the Beamformer method. However, the scaling of the spectrum intensity by the Beamformer method and the spectrum intensity by the Capon method is aligned, The calculation result of the spectrum intensity by the Beamformer method may be compared with the calculation result of the spectrum intensity by the Capon method.

  Moreover, although the array antenna of this embodiment is an element antenna arrange | positioned at equal intervals, you may use the array antenna in which the element antenna was arrange | positioned at unequal intervals besides that. Further, an array antenna may be used as a reception antenna instead of a transmission antenna.

  Next, a third embodiment of the present invention will be described with reference to FIGS.

  The radar according to the present embodiment has substantially the same configuration as the radar according to the first embodiment, but the processing contents of the arrival direction estimation process, the peak approach detection process, and the peak passage detection process performed by the signal processing circuit are different.

  In the arrival direction estimation processing of the present embodiment shown in FIG. 6, when it is detected that the peak direction is approaching, the scanning pitch is switched. The algorithm is fixed by the Capon method.

  Here, an explanation will be made assuming that the azimuth angle (θ) for detecting the spectrum intensity first is −20 ° and the scanning pitch Δθ is 1 °.

(S51) First, a weight vector of a specific azimuth angle θ is calculated by the Capon algorithm, and the received signal of each channel is multiplied and weighted, that is, the amplitude and phase of the received signal of each channel are converted. Then, the received signal strength is calculated when the azimuth angle θ is the directional direction. Then, the received signal strength is normalized to calculate the spectrum strength S _i .

(S52) Each time the spectrum intensity with each azimuth as the directional azimuth is calculated, it is detected that the peak azimuth is approaching from the comparison between the previously calculated directional azimuth spectrum intensity and the latest directional azimuth spectrum intensity. In the case of the radar apparatus according to the present embodiment, specifically, a condition that the spectrum intensity increase range is larger than before (first condition), and a condition that the spectrum intensity increase range is larger than a predetermined threshold (second condition). To determine whether the peak direction is approaching.

  The first condition that the increase in the spectrum intensity is larger than before is for detecting that the peak is in the initial state, and the second condition that the increase in the spectrum intensity is larger than the predetermined threshold is steep. This is for determining that a peak accompanying a change in spectrum intensity is approaching. By using both of these conditions, erroneous detection due to noise or the like is prevented, and the peak direction is reliably detected.

(S53, S54) If the peak azimuth is approaching, it is next detected that the peak azimuth is far from the comparison between the previously calculated directional azimuth spectrum intensity and the latest directional azimuth spectrum intensity. In the case of the radar apparatus of the present embodiment, specifically, every time the peak direction is passed, it is detected that the peak direction is far. This peak passage detection process will be described later.

  (S55, S57) If the peak azimuth is far, the scanning pitch is coarse (Δθ = 1 °), and the subsequent spectrum intensity is calculated.

  (S55, S56) If the peak azimuth is not far, the subsequent spectrum intensity is calculated while keeping the scanning pitch fine.

  By such processing, the calculation of the spectrum intensity is repeated by switching the azimuth angle θ with the scanning pitch Δθ, and the angular distribution of the spectrum intensity with the arbitrary azimuth angle as the directional direction is calculated.

  Next, the peak approach detection process in which the signal processing circuit 9 detects that the peak direction of the spectrum intensity is approaching will be described with reference to FIG.

(S61) First, the slope ΔS _i of the latest spectrum intensity is calculated from the difference between the spectrum intensity S _i-1 calculated last time and the latest spectrum intensity S _i .

(S62) The slope ΔS _i-1 of the previous spectrum intensity is calculated from the difference between the spectrum intensity S _i-1 calculated last time and the spectrum intensity S _i-2 calculated last time.

(S63, S64) Further, the first condition and the directivity that both the latest slope ΔS _i and the previous slope ΔS _i-1 are positive and the latest slope ΔS _i is larger than the previous slope ΔS _i-1. If the second condition that the increase width of the received signal intensity before and after the azimuth switching changes exceeding the threshold (α) is cleared, it is determined that the peak azimuth is approaching. Otherwise, it is determined that the peak direction is not approaching.

  Through such processing, the radar according to the present embodiment detects that the peak direction is approaching, and switches the scanning pitch using the approach of the peak direction as a trigger.

  A peak passage detection process for detecting that the peak direction of the spectrum intensity is far will be described with reference to FIG.

(S71) First, the signal processing circuit 9 checks whether or not the spectrum intensity has decreased, and if not, determines that the peak is not far away.

(S72) If the spectrum intensity has decreased, then the spectrum intensity S _peak of the immediately previous peak direction is compared with the current spectrum intensity S _i to check whether it has decreased, and if it has decreased, the peak It is determined that the object is far away, and if it is not lowered, it is determined that the peak is not distant.

  By such processing, the radar according to the present embodiment detects that the peak direction is far and switches the scanning pitch using the far peak as a trigger.

  As described above, the radar 50 according to the present embodiment detects that the peak azimuth is approaching, and finer the scanning pitch by approaching the peak azimuth, so that the arrival azimuth estimation accuracy with respect to the azimuth where the target exists is In addition, it is possible to suppress the calculation amount of the estimation of the arrival direction for all directions.

  Here, based on FIG. 8, the above process will be described using an example of an experimental result obtained by acquiring the spectrum intensity using the radar of the present embodiment.

  FIG. 8 is a diagram showing the angular spectrum distribution in the experimental example, where the horizontal axis represents the directivity angle and the vertical axis represents the spectrum intensity. In this experimental example, the direction of arrival is estimated by the Capon algorithm. The initial directivity angle is set to -20 ° and the scanning pitch is set to 1 °. In addition, the scanning pitch to be switched by detecting the peak direction is set to 0.1 °. In addition, the arrival directions of incoming waves in the experimental environment are set to −5 °, 0 °, and 10 °.

  First, the spectrum intensity was calculated while switching the azimuth angle in order from the angle range of −20 ° to −7 ° at a scanning pitch of 1 °.

  After calculating the spectrum intensity at the azimuth angle of -7 °, the latest increase width is larger than the previous increase width and satisfies the first condition, and the latest increase width is larger than the threshold value and satisfies the second condition. It was detected that the peak direction was approaching, and the scanning pitch was set to 0.1 °. The spectrum intensity was calculated while switching the azimuth angle θ in order from the angle range −6.9 ° to −4.9 ° at a scanning pitch of 0.1 °.

  After calculating the spectrum intensity at an azimuth angle of -4.9 °, it was detected that the peak azimuth was far and the scan pitch was set to 1 °. Next, the spectrum intensity was calculated while switching the azimuth angle θ at an angle range of −3.9 ° to −0.9 ° at a scanning pitch of 1 °.

  After calculating the spectrum intensity at the azimuth angle of −0.9 °, the latest increase width becomes larger than the previous increase width and satisfies the first condition, and the latest increase width becomes larger than the threshold and satisfies the second condition. Therefore, it was detected that the peak direction was approaching, and the scanning pitch was set to 0.1 °. The spectrum intensity was calculated while switching the azimuth angle θ at an angle range of −1.0 ° to 0.1 ° in order at a scanning pitch of 0.1 °.

  After calculating the spectrum intensity at an azimuth angle of 0.1 °, it was detected that the peak azimuth was far and the scanning pitch was set to 1 °. Next, the spectrum intensity was calculated while switching the azimuth angle θ in order from the angle range 1.1 ° to 9.1 ° at a scanning pitch of 1 °.

  After calculating the spectrum intensity at the azimuth angle of 9.1 °, the latest increase is larger than the previous increase and satisfies the first condition, and the latest increase is greater than the threshold and satisfies the second condition. From this, it was detected that the peak direction was approaching, and the scanning pitch was set to 0.1 °. The spectrum intensity was calculated while switching the azimuth angle θ in order from the angle range 9.2 ° to 10.1 ° at a scanning pitch of 0.1 °.

  After calculating the spectrum intensity at the azimuth angle of 10.1 °, it was detected that the peak azimuth was far away, and the scanning pitch was set to 1 °. Then, the spectrum intensity was calculated while switching the azimuth angle θ in order from the angle range 11.1 ° to 19.1 ° at a scanning pitch of 1 °.

  As shown in this experimental example, in the radar according to the present embodiment, the arrival direction of the incoming wave can be estimated with high accuracy and the number of processes at a fine scanning pitch can be effectively suppressed.

  In addition to detecting that the peak azimuth is approaching by the above processing, as described above, it is detected that the increase in spectrum intensity is smaller than the increase in spectrum intensity during switching to the previous directional direction. Alternatively, it may be determined that the peak direction is approaching. In that case, even if it is simple calculation, it can detect that a peak direction approaches more correctly.

  In this embodiment, since the arrival direction estimation method (Capon method) with excellent angular resolution is used, the calculation amount increases, but the measurement can be performed with high accuracy. If the arrival direction estimation method (Beamformer method) with inferior angular resolution is used, the circuit scale can be suppressed when implemented by hardware processing, and the program capacity can be reduced when implemented by software processing.

  The scan pitch to be switched is preferably set in accordance with the angular resolution of the arrival transition estimation method to be used so that there is no omission of detection of the peak direction. In this case, a relatively fine scanning pitch may be used when a high-resolution arrival direction estimation method is used, and a relatively coarse scanning pitch may be used when a low-resolution arrival direction estimation method is used. With regard to the combination of the type of arrival direction estimation method and the scanning pitch, it is preferable to adjust the setting so that sufficient angular resolution and reduction of calculation amount can be achieved while preventing detection of peak direction.

It is a figure explaining the structure of the radar of 1st Embodiment. It is a processing flow which estimates the arrival direction in the radar of the embodiment. It is a processing flow which detects the peak direction in the radar of the embodiment. It is a processing flow which detects the peak direction in the radar of 2nd Embodiment. It is a figure explaining the experimental result of the radar of the embodiment. It is a processing flow which estimates the arrival direction in the radar of 3rd Embodiment. It is a processing flow which detects the peak direction in the radar of the embodiment. It is a figure explaining the experimental result of the radar of the embodiment.

Explanation of symbols

1-receiving antenna 2-element antenna 3-switch circuit 4-VCO
5-branch circuit 6-LNA
7-mixer 8-IF amplifier 9-signal processing circuit 10-array antenna 50-radar

Claims (9)

Based on the received signal for each element antenna obtained by transmission or reception via each element antenna of the array antenna, the received signal strength of an arbitrary directional direction is calculated with a predetermined directional characteristic, and the directional direction is set to a predetermined scanning pitch. In the radar that estimates the arrival direction of the incoming wave based on the received signal strength for each directivity direction obtained by switching at
Peak detection means for detecting that the peak azimuth of the received signal strength is approached based on the received signal strength in the switched directional azimuth and the received signal strength in the previous directional azimuth each time the directional azimuth is switched,
A radar comprising: resolution changing means for changing the angle resolution thereafter when the peak detecting means detects that the peak direction of the received signal intensity approaches.   The radar according to claim 1, wherein the resolution changing unit changes the angular resolution higher by changing the scanning pitch to a finer one.   The radar according to claim 1, wherein the resolution changing unit changes the angular resolution higher by changing the directivity characteristics.   The peak detection means has a first condition that an increase width of the received signal strength before and after the switching of the directivity changes more than an increase width of the received signal strength during the previous switching of the directivity. The radar according to claim 1, wherein the radar detects that the peak direction of the received signal strength is approaching based on the radar signal.   The peak detecting means detects that the peak direction of the received signal strength approaches based on a second condition that the increase width of the received signal strength before and after the switching of the directivity direction changes beyond a threshold value. The radar according to any one of claims 1 to 4. The peak detection means has a third condition that an increase width of the received signal strength before and after the switching of the directivity direction changes smaller than an increase width of the received signal strength during the previous switching of the directivity direction. The radar according to claim 1 , wherein the radar detects that the peak direction of the received signal strength is approaching based on the radar signal. In the peak detection means, the increase width of the received signal strength before and after the switching of the directivity direction first satisfies the first condition, and then the increase of the received signal strength during the previous switching of the directivity direction 5. The radar according to claim 4, wherein when the third condition of changing smaller than the width is satisfied, it is detected that the peak direction of the received signal intensity approaches. The peak detection means comprises passage detection means for detecting that the peak direction of the received signal strength is far away,
The radar according to any one of claims 4 to 7 , wherein the resolution changing means changes the angular resolution low after detecting that the peak direction is far away by the peak detecting means.   The peak detection means detects that the peak azimuth is approaching using any of the first to third conditions, then detects that the peak azimuth is far away using the passage detection means, and then The next peak direction approaches on the basis of the fourth condition that the decrease width of the received signal strength before and after the switching of the directional direction is smaller than the threshold or the received signal strength increases before and after the switching of the directional direction. The radar according to claim 8 which detects this.

Images