JP4881037B2 - Radar equipment - Google Patents

Description

  The present invention relates to a radar apparatus suitable for use in a millimeter wave radar system.

Conventionally, an FM-CW (Frequency Modulated-Continuous Wave) type radar system has been used. Such a radar system is mounted on a vehicle or the like, for example, and is used for assisting safe driving by detecting the distance or relative speed between the vehicle and surrounding vehicles.
FIG. 8 shows a configuration related to a conventional radar system and a radar apparatus in the system.

In FIG. 8, reference numeral 27 denotes a radar apparatus, and reference numeral 140 denotes a target (for example, a vehicle, a person, an obstacle, etc.) that is a detection target of the radar apparatus. Here, the number of the target objects 140 is one here, but of course, the same applies to the case where two or more target objects exist.
The radar apparatus 27 shown in FIG. 8 pays attention to its main part, for example, transmission / reception composed of a transmission antenna 29 for transmitting scanning radio waves and a reception antenna 30 for receiving reflected waves from the target 140. An antenna unit 28; scanning control means 31 for controlling scanning positions of the transmitting antenna 29 and the receiving antenna 30; a transmitting unit 32 and a receiving unit 33 for transmitting and receiving the scanning radio wave; An analog / digital converter (hereinafter referred to as an A / D converter) 34 for converting an analog signal received by the unit 33 into a digital signal, and a fast Fourier transform (FFT) on the digital signal from the A / D converter 34 : Fast Fourier Transform (Fast Fourier Transform) 35, and the digital object related to the target 140 obtained by the Fast Fourier Transform 35 It comprises a target recognition means 36 for recognizing the target 140 by the peak value in the frequency domain of the tall signal.

In the radar system having such a configuration, in the transmission antenna 29 whose scanning position is controlled by the scanning control unit 31, for example, one period of a triangular wave having a triangular wave frequency-modulated per one transmission direction is transmitted. The signal is transmitted from the unit 32 via the transmission antenna 29.
Then, the reflected signal reflected by the target 140 is received by the receiving antenna 30, amplified and frequency converted by the receiving unit 33, and further, analog signals output from the receiving unit 33 by the A / D converter 34. The beat signal is converted into a digital signal.

  Next, the output (digital signal) from the receiving unit 33 is subjected to a fast Fourier transform process by a fast Fourier transformer 35 to detect a frequency component of the reflected signal. The power spectrum values corresponding to the frequency components (hereinafter also referred to as power values) are rearranged in ascending order from the higher order (the power value is larger), and the frequency component peculiar to the target 140 and the power value corresponding thereto are obtained. Calculation processing such as the distance to the target 140 and the relative speed is performed.

In the above-described radar device 27, noise components that circulate from the transmission unit 32 to the reception unit 33, high-order reflected waves from the target 140, and the like affect the reception sensitivity of the reception unit 33, and this is solved. Therefore, a reception method of a radar apparatus that receives the received wave by subtracting the noise component at the reception unit 33 is disclosed in Patent Document 1 below.
JP 2001-183450 A

However, for example, in a radar device mounted on a vehicle or the like, it is necessary to improve the detection accuracy of the radar device 27 in order to calculate the distance or relative speed between the vehicle and surrounding vehicles with high accuracy. Specifically, improvement in the resolution of the A / D converter 34 is required.
In recent years, the resolution of the A / D converter 34 has been improved, and as a result, even a minute noise (noise) component that has not been seen in the prior art appears in the FFT calculation result.

Thus, for example, even if a target peak value as shown in FIG. 9 (a) (see, for example, reference numerals P3, P6, P8, and P10) is obtained, the actual peak value as shown in FIG. Thus, even unnecessary peak values (see, for example, symbols P1, P2, P4, P5, P7, P9, and P11) appear.
In the conventional peak detection process, the detected peak value is accumulated in a storage device (not shown) such as a memory built in the radar device 27, and the peak value exceeding the memory capacity has been accumulated in the memory. By deleting the peak values from the smaller ones, the higher one (the peak value is larger) of the plurality of peak values is stored in the memory, and these are used as the target peak values. In other words, among the plurality of peak values detected by the radar device 27, the highest peak value that can be stored in the memory is calculated as a necessary peak value (target peak value).

  However, in such a conventional method, there is a limit on the capacity of the memory due to factors such as a reduction in the circuit scale of the memory and cost reduction, so that an upper limit is also imposed on the number of detected target peak values. The originally required peak value (target peak value) becomes lower than other peak values (unnecessary peak value: a peak value appearing in the vicinity of the target peak value). That is, the target peak value that should be detected is no longer detected due to the noise component (unnecessary peak value) that appears due to the improvement in detection accuracy of the radar device 27, which makes it impossible to perform a desired target recognition process. There is.

  The present invention was devised in view of such a problem.For example, by suppressing unnecessary peak value detection in a millimeter wave radar system, a desired target peak value can be detected without expansion of a memory. An object is to provide a radar apparatus.

In order to achieve the above object, the present invention is characterized by using the following radar apparatus. That is,
(1) A radar apparatus according to the present invention includes an analog / digital converter that performs analog / digital conversion on a received wave, and a fast Fourier transformer that performs fast Fourier transform on the received wave that has been analog / digital converted by the analog / digital converter. First peak detection is performed on the power spectrum obtained by fast Fourier transform by the fast Fourier transformer, based on the difference between the peak and valley of the power spectrum and the comparison of the levels of the peaks. A first detection unit for detecting a target peak from the peaks of the power spectrum, and a target peak detected by the target peak detection unit. And a target recognition means for recognizing the target.

  (2) The radar apparatus according to the present invention includes an analog / digital converter that performs analog / digital conversion on a received wave, and a fast Fourier transform that performs a fast Fourier transform on the received wave that has been analog / digital converted by the analog / digital converter. And a second peak detection means for performing peak detection by thinning out the bits of the power spectrum with respect to the power spectrum obtained by fast Fourier transform by the fast Fourier transformer. Among the peaks of the power spectrum detected using the means, the target peak detecting means for detecting the target peak based on the data of the bits before the thinning process is performed, and the target peak detecting means Target recognition means for recognizing a target based on the target peak It is characterized by comprising.

  (3) Further, the radar apparatus of the present invention includes an analog / digital converter that performs analog / digital conversion on a received wave, and a fast Fourier transform that performs fast Fourier transform on the received wave that has been analog / digital converted by the analog / digital converter. And a power spectrum obtained by performing a fast Fourier transform with the fast Fourier transformer, a frequency difference corresponding to each of the plurality of peaks and a level between the plurality of peaks among the plurality of peaks of the power spectrum And a target peak detector for detecting a target peak from the peaks of the power spectrum using the third peak detector, and a target peak detector Target recognition based on the target peak detected by the peak detector It is characterized in that a target recognition unit which performs.

(4) Here, the target peak detecting means may detect the peak of the power spectrum by canceling the detection by the peak detecting means during the diagnosis process.
(5) Further, the radar apparatus includes an image recognition unit that recognizes an image of a target, and the target peak detection unit performs detection by the peak detection unit based on image recognition information from the image recognition unit. The peak of the power spectrum may be detected by canceling.

(6) Further, the radar apparatus of (1) further includes image recognition means for recognizing an image of a target, and the first peak detection means is configured to perform a first peak based on image recognition information from the image recognition means. You may comprise including the 1st setting value change means which changes the 1st setting value which is a setting value about the difference of the peak and trough of a power spectrum used by a detection means.
(7) The radar apparatus of (2) further includes image recognition means for recognizing an image of a target, and the second peak detection means uses the second peak based on the image recognition information from the image recognition means. You may comprise including the 2nd setting value change means to change the 2nd setting value which is a setting value about the said thinning-out process used by the detection means.

(8) Further, the radar device of (3) further includes image recognition means for recognizing an image of a target, and the third peak detection means is configured to perform a third peak based on image recognition information from the image recognition means. You may comprise including the 3rd setting value change means to change the 3rd setting value which is a setting value about the difference of the frequency used with a detection means.
(9) The radar apparatus according to the present invention includes an analog / digital converter that performs analog / digital conversion on a received wave, and a fast Fourier transform that performs fast Fourier transform on the received wave that has been analog / digital converted by the analog / digital converter. And a power spectrum obtained by performing a fast Fourier transform with the fast Fourier transformer, and performing peak detection based on a difference between a peak and a valley of the power spectrum and a level comparison between the peaks. 1 peak detection means, 2nd peak detection means for performing peak detection by thinning out the bits of the power spectrum, and a frequency difference corresponding to each of the plurality of peaks among the plurality of peaks of the power spectrum and the plurality of the plurality of peaks Based on a comparison of levels between peaks, Target peak detecting means for detecting a target peak from among the peaks of the power spectrum using any one of the third peak detecting means for performing peak detection, and the target detected by the target peak detecting means Target recognition means for recognizing a target based on the peak, wherein the target peak detection means is configured to detect the first peak detection means and the second peak detection means according to the peak state of the power spectrum. And selecting means for selecting one of the third peak detecting means.

  (10) A radar apparatus according to a technique related to the present invention includes an analog / digital converter that performs analog / digital conversion of a received wave, and a fast Fourier transform of the received wave that has been analog / digital converted by the analog / digital converter. And a power spectrum obtained by performing a fast Fourier transform using the fast Fourier transformer, and a peak based on a difference between the peak and valley of the power spectrum and a comparison of the levels of the peaks. The target peak is selected from the peaks of the power spectrum by using any one of the peak detecting means of the first peak detecting means for detecting and the second peak detecting means for detecting the peak by thinning out the bits of the power spectrum. Target peak detection means for detecting the target peak and the target peak. Target recognition means for recognizing a target based on the target peak detected by the peak detection means, wherein the target peak detection means is configured to perform the first step according to the peak state of the power spectrum. You may provide the selection means which selects either a peak detection means and a 2nd peak detection means.

  (11) A radar apparatus according to a technique related to the present invention includes an analog / digital converter that performs analog / digital conversion of a received wave, and a fast Fourier transform of the received wave that has been analog / digital converted by the analog / digital converter. And a second peak detecting means for detecting a peak by thinning out the bits of the power spectrum, and a power spectrum obtained by performing a fast Fourier transform by the fast Fourier transformer. Using any one of the peak detection means of the third peak detection means for performing peak detection based on a difference in frequency corresponding to each of the plurality of peaks and a comparison of levels of the plurality of peaks. The target from the peak of the power spectrum Target peak detecting means for detecting a peak, and target recognition means for recognizing a target based on the target peak detected by the target peak detecting means, wherein the target peak detecting means comprises the power You may provide the selection means which selects either said 2nd peak detection means and 3rd peak detection means according to the state of the peak of a spectrum.

  (12) Furthermore, a radar apparatus according to a technique related to the present invention includes an analog / digital converter that performs analog / digital conversion of a received wave, and a fast Fourier transform of the received wave that has been analog / digital converted by the analog / digital converter. And a power spectrum obtained by performing a fast Fourier transform with the fast Fourier transformer, based on a difference between a peak and a valley of the power spectrum and a comparison between levels of the plurality of peaks. First peak detecting means for performing peak detection and peak detection based on a difference in frequency corresponding to each of the plurality of peaks of the plurality of peaks of the power spectrum and a comparison of levels of the plurality of peaks. Using any of the peak detection means of the third peak detection means, A target peak detecting means for detecting a target peak from the peaks of the power spectrum, and a target recognition means for recognizing a target based on the target peak detected by the target peak detecting means, The target peak detection means may include a selection means for selecting one of the first peak detection means and the third peak detection means according to the peak state of the power spectrum.

According to the present invention, at least one of the following effects or advantages can be obtained.
(1) About the power spectrum obtained by performing the fast Fourier transform by the fast Fourier transformer, based on the difference between the peak and valley of the power spectrum and the comparison of the levels of the peaks, from among the peaks of the power spectrum Since the target peak is detected, unnecessary peak detection can be suppressed.

(2) Since the target peak can be detected from the peaks of the power spectrum by thinning out the bits of the power spectrum, unnecessary peak detection can be suppressed.
(3) Furthermore, a target peak is selected from among the peaks of the power spectrum based on a frequency difference corresponding to each of the plurality of peaks of the plurality of peaks of the power spectrum and a comparison of levels of the plurality of peaks. Can also be detected, so that unnecessary peak detection can be suppressed.

(4) Further, during the diagnosis process, the radar device cancels the detection by each of the peak detection means (first to third peak detection means) and detects a target peak from the peaks of the power spectrum. Therefore, it is possible to speed up the target peak detection process.
(5) Further, the radar apparatus includes an image recognition unit that recognizes an image of the target, cancels the detection by each of the peak detection units based on the image recognition information from the image recognition unit, and outputs the power spectrum. Since the target peak can be detected from among the peaks, the target peak detection process can be speeded up.

(6) Here, the first peak detecting unit uses the first peak detecting unit based on the image recognition information from the image recognizing unit to determine the difference between the peak and valley of the power spectrum. Since the first set value, which is the set value, can be changed, the target peak detection process can be performed efficiently.
(7) Further, the second peak detecting unit determines a second set value that is a set value for the bit thinning process used by the second peak detecting unit based on the image recognition information from the image recognizing unit. Since it can be changed, the target peak detection process can be performed efficiently.

(8) Further, a third set value, which is a set value for the difference in frequency used by the third peak detecting unit, based on the image recognition information from the image recognizing unit. Therefore, the target peak detection process can be performed efficiently.
(9) Since the target peak can be detected by selectively using any one of the first to third peak detecting means according to the peak state of the power spectrum, various Therefore, it is possible to perform the target peak detection process optimal for the actual peak measurement value, and to further improve the detection accuracy of the radar apparatus.

  (10) Further, as described above, since the detection of unnecessary peaks can be suppressed, the use area of a storage device such as a memory mounted in the radar device can be saved, thereby expanding the memory. It is possible to detect the target peak without performing.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram schematically showing an outline of a radar system using a radar apparatus as an embodiment of the present invention.
As shown in FIG. 1, this radar system is the above-described FM-CW radar system, and is configured to perform various measurements on a target 14 as a radar measurement target.

Furthermore, the radar apparatus 1 is mounted on a vehicle such as an automobile, for example, and includes a transmission / reception unit 2, a scanning control unit 5, an image recognition unit 6, a transmission unit 7, a reception unit 8, and an A / D converter 9. And a fast Fourier transformer 10, a target peak detector 11 as a target peak detector, a target object recognizer 12, and a diagnosis processor 13.
Here, the transmission / reception unit 2 includes a transmission antenna 3 and a reception antenna 4, and performs input / output control with the outside. The transmission antenna 3 is for transmitting scanning radio waves, and the reception antenna 4 is for receiving reflected waves reflected by the target 14. Since the transmission antenna 3 and the reception antenna 4 have a narrow geometric angle range in a direction in which the maximum gain can be obtained and have a sharp beam directivity, the scanning control unit 5 uses the transmission antenna 3 and the reception antenna. 4 is controlled so that the target 14 can be detected in a wide angle range.

The scanning control means 5 is for controlling the directivity directions of the transmission antenna 3 and the reception antenna 4, and can change the transmission direction of the continuous transmission wave generated by the transmission unit 7. In addition, the antenna scanning position can be controlled at the time of the diag processing by an input from the diag processing means 13 described later.
The transmission unit 7 is for generating radio waves for scanning, and the reception unit 8 receives a reflected wave from the target 14 received by the reception antenna 4 and performs amplification, frequency conversion, and the like. Is for. Here, the radio waves for scanning generated by the transmitter 7 together with the reflected waves are input to the receiver 8, and the radio waves for scanning generated when reflected by the target 14 by comparing the two. It is possible to detect a phase shift and a power change.

The A / D converter 9 is for converting an analog signal into a digital signal. The A / D converter 9 converts the analog signal input from the receiving unit 8 into a digital signal, which is then sent to the next-stage fast Fourier transformer 10. Output.
The fast Fourier transformer 10 is for performing a fast Fourier transform on the digital signal input from the A / D converter 9, thereby detecting the frequency component of the reflected wave from the target 14.

  Then, the target peak detection unit 11 uses any one of peak detection means of the first peak detection means 15, the second peak detection means 16, the third peak detection means 17 and the normal peak detection means 18 to be described later. It functions as target peak detection means for detecting a target peak from among the peaks, and detects the target peak necessary for recognizing the target 14 from the frequency component of the reflected wave detected by the fast Fourier transformer 10. It is for doing. As this detection method, as will be described later, an optimum method (peak detection means) is selected depending on the signal input from the image recognition means 6 and the diagnosis processing means 13 or the peak state obtained from the fast Fourier transformer 10. The

The diagnostic processing means 13 is for executing diagnostic processing in the radar apparatus 1. In this embodiment, the diagnostic processing means 13 is also used for controlling the operations of the scanning control means 5 and the target peak detection unit 11 by the output from the diagnostic processing means 13. used.
Here, the diagnosis process refers to a self-diagnosis process that the radar apparatus 1 has. In this diagnosis processing, for example, the scanning control means 5 is automatically controlled at a predetermined time interval when radar detection is performed, and a scanning radio wave transmitted from the transmission antenna 3 is preliminarily set to a target peak mounted on the vehicle. The peak detection is performed by reflecting on another known target, and the detection accuracy of the radar apparatus 1 is maintained by correcting the deviation between the target peak and the actual peak detection processing result. Is. As described above, the diag processing means 13 controls the scanning positions of the transmitting and receiving antennas 3 and 4 to a predetermined position at the time of diag processing at the time of diag processing, and the target peak detection unit 11 uses a normal peak described later as a method of target peak detection. The peak detection means 18 can be controlled to be selected.

The target recognition means 12 calculates the distance to the target 14 and the relative speed with the vehicle using the target peak detected by the target peak detector 11 described above.
The image recognition means 6 recognizes the target 14 as an image and generates an image signal. For example, a camera or the like is used.
Furthermore, the structure of the target peak detection part 11 is demonstrated using FIG.

As shown in FIG. 2, the target peak detector 11 in the radar apparatus 1 of the present embodiment includes a first peak detector 15, a second peak detector 16, a third peak detector 17, a normal peak detector 18, and The control unit 22 includes a selection unit 25, a CPU (Central Processing Unit) 23, and a memory 24.
The normal peak detecting means 18 simply detects the power spectrum input from the fast Fourier transformer 10 from the peak having the large power value, and sets these as the target peak value.

  The selection means 25 selects any one of a first peak detection means 15, a second peak detection means 16, a third peak detection means 17, and a normal peak detection means 18 described later as a target peak detection method. Therefore, an optimum peak detection method for detecting the target peak can be selected in accordance with the frequency component (peak state) of the reflected wave input from the fast Fourier transformer 10. For example, when the obtained power spectrum state has a large power difference between peaks (peaks) and valleys (bottoms), the first peak detector 15 is selected to perform target peak detection processing. Thus, detection of unnecessary peaks can be efficiently suppressed. In addition, for example, when it is desired to detect a peak of a certain power or more as a target peak among the obtained peaks, the second peak detection unit 16 can be selected to perform target peak detection processing. If the target peak appears periodically in the state of the obtained power spectrum, the target peak is efficiently detected by selecting the third peak detection means 17 and performing the target peak detection process. You can also.

The control unit 22 includes a CPU 23 for performing various controls according to external control signals (for example, an image signal from the image recognition unit 6 and a control signal from the diagnosis processing unit 13), and various values. And a memory 24 for storing.
The first peak detection means 15 does not obtain a target peak based on only the magnitude of the peak of the power spectrum input from the fast Fourier transformer 10, but the difference between the peak of the power spectrum and the valleys before and after the peak. And a predetermined value (first set value), and a target peak is obtained based on a level comparison between the peaks, and a first setting for changing the first set value. Value changing means 19 is provided.

  Further, the first set value changing means 19 is based on the image recognition information (image signal) from the image recognizing means 6 and the difference between the peak and valley of the power spectrum used by the first peak detecting means 15. The first set value, which is the set value for, is changed so that the target peak value detection processing result is optimized. For example, when the control unit 22 determines that the power value has a large or small variation based on the image signal from the image recognition unit 6 and the detected peak, the control unit 22 sets a plurality of first set values. The target peak value by the first peak detecting means 15 is determined by the first setting value changing means 19 actually changing the first setting value. The detection accuracy can be improved.

Here, a specific processing flow of the first peak detection means 15 will be described.
First, the CPU 23 detects, for example, a difference between adjacent power spectra in frequency, and detects a first peak value (for example, a peak appearing in the lowest frequency portion) based on a change in the sign of the difference. . Then, the difference between the peak value and the values in the valleys before and after the peak value is obtained, and when both the obtained results are larger than the first set value, the peak value is stored in the memory 24 or the like as a target peak. Is done. On the other hand, when either one of the differences between the peak value and the values in the valleys before and after the peak value is equal to or less than the first set value, the peak value is stored in the memory 24 or the like as a temporary peak.

  Similarly, the CPU 23 compares the next peak value with the temporary peak value, and if the temporary peak value is larger, the difference between the temporary peak value and the previous trough, and the temporary peak value and the next peak. When the difference from the trough after is obtained and both of the obtained results are larger than the first set value, the peak value is stored as a target peak in the memory 24 or the like and is less than or equal to the first set value. In some cases, it is stored again as a temporary peak. Further, when the temporary peak value is equal to or less than the next peak value, the difference between the next peak value and the subsequent valley and the difference between the next peak value and the valley before the temporary peak are obtained and obtained. If both the obtained results are larger than the first set value, the next peak value is stored as the target peak in the memory 24 or the like, and if it is equal to or less than the first set value, the next peak is a temporary peak. Is newly stored.

Furthermore, the operation when the above-described processing flow is applied to an example of a certain peak will be described more specifically with reference to FIGS. 5 (a) and 5 (b).
In the example of the peak shown in FIG. 5 (a), the CPU 23 first obtains B-A and B-C. Since both are larger than the first set value, B is stored in the memory 24 as the target peak. The

  Next, DC and DE are obtained. Since the value of DE is equal to or less than the first set value, the valley E is discarded and D is stored in the memory 24 as a temporary peak. Further, the CPU 23 sets the value of the valley C as the comparison target with the next peak F without setting the value of the valley E as the comparison target with the next peak F. That is, FC and FG are then obtained, and since the value of FG is equal to or less than the first set value, F is a temporary peak and the valley information G is discarded from the memory.

Here, the magnitudes of the temporary peaks D and F having at least one valley (here, valley C) in a common valley are compared, and since F> D, the temporary peak D is also discarded.
Next, HC and HI are obtained. Again, since the peak H shares the valley information C of the peak F with the peak F, the magnitudes of F and H are compared, and F> H. Therefore, H is discarded.
Then, by repeating the operation described above, for example, in the example of the peak shown in FIG. 5A, peaks B and F are detected as target peaks.

On the other hand, in the example of the peak shown in FIG. 5B, first, the CPU 23 obtains B′−A ′ and B′−C ′, both of which are larger than the first set value, so that B ′ is the target. It is stored in the memory 24 as a peak.
Next, D′−C ′ and D′−E ′ are obtained. Since the value of D′−E ′ is equal to or smaller than the first set value, D ′ is stored in the memory 24 as a temporary peak. Further, the CPU 23 sets the value of the valley C ′ as the comparison target with the next peak F ′ without setting the value of the valley E ′ as the comparison target with the next peak F ′. That is, next, F′-C ′ and F′-G ′ are obtained. Since the value of F′-G ′ is equal to or less than the first set value, F ′ is set as a temporary peak, and valley information G ′ is obtained. Discard from memory.

Here, the magnitudes of the temporary peaks D ′ and F ′ having at least one valley (here, valley C ′) in a common valley are compared, and the temporary peak F ′ is discarded because D ′> F ′. .
Next, H′-C ′ and H′-I ′ are obtained. Again, since the peak H ′ shares the valley information C ′ of the peak D ′ with the peak D ′, the relationship between D ′ and H ′ The magnitudes are compared, and since D ′> H ′, H ′ is discarded.

Then, by repeating the operation described above, for example, in the example of the peak shown in FIG. 5B, peaks B ′ and D ′ are detected as target peaks.
In this way, in the target peak value detection process by the first peak detection means 15, unnecessary peak values are detected by not detecting peaks whose power difference between peaks and valleys is smaller than the predetermined first set value. It can be suppressed.

  The second peak detecting means 16 performs a bit thinning process (power value averaging process by bit shift), which will be described later, on the bits of the power spectrum with respect to the power spectrum input from the fast Fourier transformer 10. Thus, peak detection is performed, and second setting value changing means 20 is further provided. The second set value changing means 20 is a set value for the bit thinning process used by the second peak detecting means 16 based on the image recognition information (image signal) from the image recognizing means 6. (2) The set value is changed so that the target peak value detection processing result is optimized. The second set value changing unit 20 also uses the second set value in the second peak detecting unit 16 with little variation in power value based on, for example, the image signal from the image recognizing unit 6 and the detected peak. When the determination is made, the control means 22 determines the second set value to a larger value so that the target peak detection is optimally performed, and the second set value changing means 20 actually sets the second set value. By changing the value, the target peak value detection processing speed by the second peak detector 16 can be improved.

As shown in FIG. 3, the second peak detection means 16 further includes a 32-bit memory 26, a power spectrum value cutout unit 27, a peak determination unit 28, and a peak storage buffer 29.
Here, the 32-bit memory 26 is a storage device for storing the output from the fast Fourier transformer 10 (here, the power spectrum value expressed by 16 bits and the frequency value expressed by 16 bits), Although a 32-bit memory is used here, a larger capacity storage device can be used.

  Further, the power spectrum value cutout unit 27 is for realizing the power value averaging process by the bit shift as the above-described bit thinning process. 2 Based on the information regarding the set value, the lower bits (2 bits in this case) of the output data from the 32-bit memory 26 are deleted and output to the peak determination unit 28 (bit extraction processing).

Based on the output from the power spectrum value cutout unit 27, the peak determination unit 28 determines a peak value from a larger (higher) peak value as a target peak value, and a frequency corresponding to the target peak value. The value is output to the peak storage buffer 29.
Further, the peak storage buffer 29 is based on the output from the peak determination unit 28 (frequency value corresponding to the target peak value), and all the bit data (here, the power spectrum value 16) output from the power spectrum value cutout unit 27. The target peak value is calculated from the 32-bit data comprising 16 bits and a frequency value of 16 bits, and is stored in the peak storage buffer 29.

  In other words, the second peak detecting means 16 performs predetermined processing by the CPU 23 without considering the lower data for the number of bits set by the second set value changing means 20 among the bit data representing the power spectrum value. The target peak is detected based on the upper bits of the number of bits (16-2 = 14 bits in this embodiment), the frequency at which the target peak appears is calculated, and then the power is calculated based on the frequency. A detailed target peak value is calculated from all the bit data before being cut out by the spectrum value cutout unit 27.

  In other words, by not considering the lower bits, a peak as shown in FIG. 9A is obtained, and based on this, the frequency corresponding to the target peak value (for example, symbols P3, P6, P8, P10) is obtained. Is detected. Based on these frequencies, as shown in FIG. 9 (b), target peak values (for example, codes P3, P6, P8) from the peak before the bit cutout process is performed by the power spectrum value cutout unit 27. , P10) is calculated in more detail.

  Thus, in the peak value detection by the second peak detection means 16, the frequency value corresponding to the target peak value is calculated by not considering the data of the power spectrum value for the number of bits determined by the second set value. In addition, since the target peak value is calculated from all the bit data related to the power spectrum value based on this frequency value, detection of an unnecessary peak value can be suppressed.

  Further, the third peak detecting means 17 is configured to determine, for the power spectrum input from the fast Fourier transformer 10, a frequency difference corresponding to each of the plurality of peaks of the plurality of peaks of the power spectrum, and each of the plurality of peaks. Peak detection is performed on the basis of the comparison between levels, and a third set value changing means 21 is further provided. The third set value changing means 21 is also a set value for the difference in frequency used by the third peak detecting means 17 based on the image recognition information (image signal) from the image recognizing means 6. 3 The set value is changed so that the detection processing result of the target peak value is optimized. Based on the image signal from the image recognition unit 6 and the detected peak, the target peak value detection processing by the third peak detection unit 17 is performed on the frequency difference (third set value) in the third peak detection unit 17. The result is changed so as to obtain an optimum result. For example, when the control unit 22 determines that the peak value is large in the vicinity of the same frequency band, the third setting value is determined to be the minimum value of the frequency difference between the plurality of peaks, and further, the third setting value is determined. By the value changing means 21 actually changing the third set value, it becomes possible to efficiently suppress detection of unnecessary peak values.

As shown in FIG. 4, the third peak detection means 17 further includes a 32-bit memory 30, a peak determination unit 31, and a peak storage buffer 32.
Similarly to the 32-bit memory 26, the 32-bit memory 30 stores the output from the fast Fourier transform 10 (here, the power spectrum value expressed by 16 bits and the frequency value expressed by 16 bits). Here, a 32-bit memory is used, but a larger-capacity storage device can also be used. The 32-bit memory 30 can be shared with the 32-bit memory 26.

The peak determining unit 31 determines a target peak value by, for example, steps described later, and the peak storage buffer 32 is for storing an output (target peak value) from the peak determining unit 31. The peak storage buffer 32 may be shared with the peak storage buffer 29.
Here, the operation of the third peak detecting means 17 will be specifically described with reference to FIG.

First, the peak determination unit 31 stores the peak J in the peak storage buffer 32, and then stores the peak K in the peak storage buffer 32. At this time, since the peak determination unit 31 determines that K> J, the peak value is stored in the peak storage buffer 32 in the order of K and J.
Next, similarly to the above, the peak determination unit 31 stores the peak L in the peak storage buffer 32, and the difference between the frequency of the peak L and the frequency of the peak K is equal to or less than the third set value. Since the peak determination unit 31 determines that L> K, the peak K is discarded, and the peak value is stored in the peak storage buffer 32 in the order of L and J.

Similarly, the peak determination unit 31 stores the peak M in the peak storage buffer 32. The difference between the frequency of the peak M and the frequency of the peak L is equal to or less than the third set value, and L> M Therefore, the peak M is discarded, and the peak value is stored in the peak storage buffer 32 in the order of L and J.
Further, similarly, the peak N is stored, and it is determined that the difference between the frequency of the peak N and the frequency of the peak L is larger than the third set value and further N> L. , Peak values are stored in the order of N, L, and J.

  Similarly, the peak determination unit 31 stores the peak O in the peak storage buffer 32. The difference between the peak O frequency and the peak N frequency is equal to or less than the third set value, and N> O Therefore, the peak O is discarded, and the peak value is stored in the peak storage buffer 32 in the order of N, L, and J. At this time, even if the peak determination unit 31 discards the peak O, only the frequency information of the discarded peak O may be stored in the peak storage buffer 32 as comparison data. That is, the peak comparison and the frequency comparison are performed independently, and even if a certain peak value is discarded, the frequency corresponding to the discarded peak value is used as comparison data as frequency comparison data. is there. In the above example, when the peak O is discarded, the peak value O is discarded. However, the frequency value is not discarded, and the frequency value of O is used as data for comparison with the P frequency value. However, immediately before the comparison with P, what is stored in the peak storage buffer 32 is N peak values (power values) and frequency values.

Next, the peak determination unit 31 stores the peak P. The difference between the frequency of the peak P and the frequency of the peak N (or O) is equal to or less than the third set value, and P> N Since it is determined that there is a peak, the peak N is discarded, and the peak value is stored in the peak storage buffer 32 in the order of P, L, and J.
Finally, the peak determination unit 31 stores the peak Q in the peak storage buffer 32, the difference between the frequency of the peak Q and the frequency of the peak P is greater than the third set value, and P>L> Since it is determined that Q> J, peak values are stored in the peak storage buffer 32 in the order of P, L, Q, and J.

Thereafter, by repeating the above-described operation, for example, in the example of the peak shown in FIG. 6, peaks P, L, Q, and J are detected as target peaks.
As described above, in the peak value detection by the third peak detecting means 17, only the peak value separated by the frequency difference determined by the third set value is detected as the target peak, and therefore appears in the vicinity of the target peak value. Detection of unnecessary peak values can be suppressed.

Here, the target peak detection process when the diagnosis process is performed in the radar apparatus 1 will be described.
First, for example, when the control means 22 detects that the radar apparatus 1 is performing diag processing by a control signal input from the diagnosis processing means 13, the CPU 23 selects the normal peak detection means 18 as a peak detection method. The peak detector is controlled as follows.

  The normal peak detection means 18 selects a peak having a large peak value from the power spectrum obtained by the normal peak detection process, that is, the fast Fourier transformer 10, and uses these as target peak detection. The detection is realized, and the processing is simple as compared with the first to third peak detection means described above, and the processing speed is high.

  In this way, in a situation where unnecessary peaks are difficult to detect, such as during diag processing, the target peak detection unit 11 performs the above-described peak detection means (first peak detection means 15, second peak detection means 16, and third peak detection). By canceling the detection by the means 17) and performing the target peak detection using the normal peak detection means 18, the target peak detection process can be speeded up.

  Further, in the situation where unnecessary peaks are not easily detected other than during the diagnosis process, that is, in the situation where no obstacle exists around the vehicle, the target peak is detected using the normal peak detecting means 18 having a high processing speed. Needless to say, it is desirable to perform the detection process. Therefore, the radar apparatus 1 according to the present embodiment determines whether or not the unnecessary peak is difficult to be detected based on the image recognition information of the target 14 obtained from the image recognition unit 6, and determines the unnecessary peak. Is determined to be difficult to detect, the detection by each of the peak detection means (the first peak detection means 15, the second peak detection means 16, and the third peak detection means 17) is canceled, and the normal peak The detection means 18 can perform target peak detection processing.

  Further, the control means 22 corresponds to the state of the power value obtained from the fast Fourier transformer 10, the first peak detection means 15, the second peak detection means 16, the third peak detection means 17 and the normal peak detection means 18. It is also possible to perform target peak detection processing by selecting any of the above. For example, when it is determined that there is an obstacle around the image recognition means 6 (for example, when there is another vehicle ahead), the first peak detection means 15, the second peak detection means 16, and The target peak detection process can be performed using any of the third peak detection means 17, and in other cases, the target peak detection can be performed using the normal peak detection means 18.

  Further, for example, when it is determined that there is an obstacle in the surroundings, and the peak difference between detected peaks is small and the power value is highly varied, the target peak detection process is performed by the first peak detection means 15. To do. As a result, it is possible to detect a target peak by regarding a peak whose difference from an adjacent power value is smaller than a value determined by the first set value as an unnecessary peak.

Further, for example, when the peak difference of the detected peaks is small and there is no variation in the peak differences, the target peak detection process is performed by the second peak detection means 16. This makes it possible to detect a target peak by regarding a peak with a close power value as an unnecessary peak.
Further, for example, when there are many peaks in the vicinity of the same frequency, the third peak detection means 17 performs target peak detection processing. This makes it possible to detect a target peak by regarding a peak appearing in the vicinity of the target peak value as an unnecessary peak.

Here, the operation of the target peak detection process in the present embodiment will be described using the flowchart shown in FIG.
First, when the flow is started, it is determined whether or not the diagnosis processing is performed by the diagnosis processing means 13 (step S1). If it is determined that the diagnosis processing is performed (yes route of step S1). ), The target peak detection process is executed using the normal peak detection means 18 (step S7). On the other hand, when it is determined in step S1 that the diagnosis process is not performed (no route of step S1), the control unit 22 acquires image recognition information from the image recognition unit 6 in step S2.

  Next, it is determined from this image recognition information whether or not there are obstacles around the vehicle (step S3). If it is determined that there are no obstacles around the vehicle (no route in step S3). In step S 7, the target peak detection process is executed using the normal peak detection means 18. On the other hand, when it is determined that there is an obstacle around the vehicle (yes route of step S3), the peak value obtained from the fast Fourier transformer 10 is further analyzed by the control means 22, and the adjacent frequencies are analyzed. It is determined whether or not peaks are frequently detected in the region (step S4). In addition, this determination process (referred to as a near peak determination process) is performed when, for example, the control unit 22 compares the number of peaks appearing in a certain frequency band with a predetermined threshold and the number of peaks is larger than the threshold. This is realized by determining that a peak is frequently detected in the adjacent frequency region, and by determining that the peak is not more than the threshold value.

  If it is determined in step S4 that peaks are frequently detected in the adjacent frequency region (yes route in step S4), the third set value is optimized for target peak detection based on the image recognition information. Setting (step S5), the target peak detection process is executed using the third peak detection means 17 (step S6). That is, for example, when the image recognition information is an image signal in which a peak value appears frequently in the vicinity of the same frequency band, an unnecessary peak value can be obtained by changing the third setting value to a larger value. This effectively suppresses the detection of the above. On the other hand, when it is determined that the peak is not frequently detected in the adjacent frequency region (no route of step S4), the peak value obtained from the fast Fourier transformer 10 is further analyzed by the control means 22, and the peak is obtained. It is determined whether there is a variation in the difference (step S8). In this determination process (referred to as variation determination process), for example, the control means 22 compares a predetermined threshold with the number of peaks exceeding a predetermined value for a peak difference between a plurality of peaks. If the number exceeding the predetermined value is larger, it is determined that there is a variation, and otherwise, it is realized by determining that there is no variation.

  In step S8, if it is determined that there is a variation in peak difference (yes route in step S8), the first set value is set to an optimum value for target peak detection based on the image recognition information ( Step S9), target peak detection processing is executed using the first peak detection means 15 (step S10). That is, for example, when the image recognition information is an image signal in which a large or small variation is seen in the power value, the target peak value is detected by changing the first set value to a larger value. This is to improve accuracy.

  On the other hand, if it is determined that there is no variation in the peak difference (no route in step S8), the second set value is set to an optimum value for target peak detection based on the image recognition information (step S11). The target peak detection process is executed using the peak detection means 16 (step S12). That is, for example, when the image recognition information is an image signal in which many unnecessary peaks appear, the detection processing speed of the target peak value can be improved by changing the second setting value to a larger value. It is for illustration.

  Thus, in the radar system according to the above embodiment, based on the peak value obtained by the fast Fourier transformer 10, the image recognition information obtained by the image recognition means 6, the diagnosis processing control signal from the diagnosis processing means 13, and the like, Since the optimum peak detection method is automatically selected when the target 14 is detected by the radar, the detection of unnecessary peaks can be efficiently suppressed, and therefore the detection accuracy of the target peak can be improved.

Furthermore, as described above, since the detection of unnecessary peaks can be suppressed, the use area of the storage device such as the memory 24 can be saved, and thereby the desired target peak can be saved without expanding the storage device. Can be detected.
As mentioned above, although one Embodiment of this invention was described in detail, this invention is not limited to said Embodiment, In the range which does not deviate from the summary of this invention, it can change arbitrarily and can implement.

  For example, when the target peak detection means 11 has any two of the first peak detection means 15, the second peak detection means 16, and the third peak detection means 17 and the normal peak detection means 18 as the peak detection means. However, by appropriately modifying the above-described flow, the target peak detection process can be executed by selecting the optimal target peak detection method for radar detection according to the situation at that time.

  In this case, the configuration of the target peak detection means 11 is, for example, a combination of the first peak detection means 15 and the second peak detection means 16 with the third peak detection means 17 omitted in FIG. In the case of a combination of the first peak detection means 15 and the third peak detection means 17, the second peak detection means 16 is omitted in FIG. 2, and the second peak detection means 16 and the third peak detection means 17 are obtained. In this case, the first peak detecting means 15 is deleted in FIG.

  Further, in the case of the combination of the first peak detecting means 15 and the second peak detecting means 16, the processing flow is obtained by deleting steps S4, S5 and S6 in FIG. 7, and the yes route of step S3 is the step. The process moves to S8. In the case of the combination of the first peak detecting means 15 and the third peak detecting means 17, the processing flow is the one in which steps S8, S11 and S12 are deleted in FIG. 7, and the no route of step S4 is the step. The process moves to S9. In the case of the combination of the second peak detection means 16 and the third peak detection means 17, the processing flow is the one in which steps S8, S9 and S10 are deleted in FIG. 7, and the no route of step S4 is the step. The process moves to S11.

  Further, for example, the target peak detection means 11 includes, as a peak detection means, any one of the first peak detection means 15, the second peak detection means 16, and the third peak detection means 17 and the normal peak detection means 18. Even in the case where it is included, the above-described flow can be modified as appropriate to execute the target peak detection process by selecting the optimal target peak detection method for radar detection according to the situation at that time.

  The configuration of the target peak detection means 11 in this case is, for example, in the case of a combination of the first peak detection means 15 and the normal peak detection means 18, the second peak detection means 16 and the third peak detection means 17 in FIG. In the case of the combination of the second peak detecting means 16 and the normal peak detecting means 18, the first peak detecting means 15 and the third peak detecting means 17 are deleted in FIG. In the case of the combination of the third peak detecting means 17 and the normal peak detecting means 18, the first peak detecting means 15 and the second peak detecting means 16 are omitted from FIG.

  In the case of the combination of the first peak detection means 15 and the normal peak detection means 18, the processing flow is obtained by deleting steps S4, S5, S6, S8, S11 and S12 in FIG. The yes route moves the process to step S9. In the case of the combination of the second peak detecting means 16 and the normal peak detecting means 18, the processing flow is obtained by deleting steps S4, S5, S6, S8, S9 and S10 in FIG. 7, and step S3. The yes route moves the process to step S11. In the case of the combination of the third peak detecting means 17 and the normal peak detecting means 18, the processing flow is obtained by deleting steps S4, S8, S9, S10, S11 and S12 in FIG. The yes route moves the process to step S5.

Then, in the target peak detection means 11 constituted by a combination of the above-described peak detection means, the target peak detection by the normal peak detection means 18 is performed during the diagnosis process, and the image recognition information from the image recognition means 6 is also included. On the basis of this, even when there are no obstacles around the target, the target peak detection by the normal peak detection means 18 is performed.
Further, in the target peak detecting means 11 having the above configuration, the first set value changing means 19, the second set value changing means 20, and the third set value changing means 21 are based on the image recognition information from the image recognizing means 6. In addition to being able to change the 1 setting value, the 2nd setting value, and the 3rd setting value, each setting value can be changed by the user's operation, and can also be set to a fixed value that does not depend on the image recognition information.

1 is a block diagram schematically showing a radar system according to an embodiment of the present invention. It is a block diagram which shows the structure of the target peak detection part in the said radar system. It is a block diagram which shows the structure of the 2nd peak detection means in the said radar system. It is a block diagram which shows the structure of the 3rd peak detection means in the said radar system. (A), (b) is a figure which shows an example of peak distribution. It is a figure which shows an example of peak distribution. It is a figure which shows the flow of the target peak detection process for implementing one Embodiment of this invention. It is a block diagram which shows the conventional radar system typically. (A) is a figure which shows an example of the peak detection process result at the time of detecting a peak value with low precision, (b) is a figure which shows an example of the peak detection process result at the time of detecting a peak value with high precision. is there.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Radar apparatus 2 Transmission / reception antenna part 3 Transmission antenna 4 Reception antenna 5 Scan control means 6 Image recognition means 7 Transmission part 8 Reception part 9 Analog / digital converter 10 Fast Fourier transform 11 Target peak detection part (target peak detection means)
12 target recognition means 13 diagnosis processing means 14 target 15 first peak detection means 16 second peak detection means 17 third peak detection means 18 normal peak detection means 19 first set value change means 20 second set value change means 21 Third set value changing means 22 Control means 23 CPU
24 memory 25 selection means 26, 30 32-bit memory 27 power spectrum value cutout unit 28, 31 peak determination unit 29, 32 peak storage buffer

Claims (9)

An analog-digital converter that converts the received wave from analog to digital;
A fast Fourier transformer for fast Fourier transforming the received wave analog-digital converted by the analog-digital converter;
First peak detection for performing peak detection on the power spectrum obtained by fast Fourier transform by the fast Fourier transformer based on the difference between the peak and valley of the power spectrum and the comparison of the levels of the peaks. Means for detecting a target peak from the peaks of the power spectrum using the first peak detecting means, and
A radar apparatus comprising: target recognition means for recognizing a target based on the target peak detected by the target peak detection means. An analog-digital converter that converts the received wave from analog to digital;
A fast Fourier transformer for fast Fourier transforming the received wave analog-digital converted by the analog-digital converter;
The power spectrum obtained by fast Fourier transform by the fast Fourier transformer has second peak detecting means for detecting a peak by thinning out the bits of the power spectrum, and the second peak detecting means is used. Target peak detection means for detecting a target peak based on bit data before being subjected to the thinning-out processing from among the peaks of the power spectrum detected by
A radar apparatus comprising: target recognition means for recognizing a target based on the target peak detected by the target peak detection means. An analog-digital converter that converts the received wave from analog to digital;
A fast Fourier transformer for fast Fourier transforming the received wave analog-digital converted by the analog-digital converter;
About the power spectrum obtained by performing the fast Fourier transform by the fast Fourier transformer, the frequency difference corresponding to each of the plurality of peaks among the plurality of peaks of the power spectrum and the comparison of the levels of the plurality of peaks, Third peak detection means for performing peak detection based on the target peak detection means for detecting a target peak from the peaks of the power spectrum using the third peak detection means,
A radar apparatus comprising: target recognition means for recognizing a target based on the target peak detected by the target peak detection means.   The target peak detection means is configured to detect the peak of the power spectrum by canceling the detection by the peak detection means during the diagnosis process. The radar apparatus according to claim 1. Image recognition means for recognizing the target object,
The target peak detecting means is configured to detect the peak of the power spectrum by canceling the detection by the peak detecting means based on the image recognition information from the image recognizing means. The radar apparatus according to any one of claims 1 to 3. Image recognition means for recognizing the target object,
The first peak detecting means is a first set value for a difference between a peak and a valley of the power spectrum used by the first peak detecting means based on image recognition information from the image recognizing means. 2. The radar apparatus according to claim 1, further comprising first setting value changing means for changing the setting value. Image recognition means for recognizing the target object,
A second peak detecting unit configured to change a second set value, which is a set value for the bit thinning process used by the second peak detecting unit, based on image recognition information from the image recognizing unit; The radar apparatus according to claim 2, further comprising a set value changing unit. Image recognition means for recognizing the target object,
A third peak detector for changing a third set value, which is a set value for the difference in frequency used by the third peak detector, based on image recognition information from the image recognizer; The radar apparatus according to claim 3, comprising a set value changing unit. An analog-digital converter that converts the received wave from analog to digital;
A fast Fourier transformer for fast Fourier transforming the received wave analog-digital converted by the analog-digital converter;
First peak detection for performing peak detection on the power spectrum obtained by fast Fourier transform by the fast Fourier transformer based on the difference between the peak and valley of the power spectrum and the comparison of the levels of the peaks. Means, a second peak detecting means for detecting a peak by thinning out the bits of the power spectrum, and a frequency difference corresponding to each of the plurality of peaks among the plurality of peaks of the power spectrum and the plurality of peaks Target peak detection means for detecting a target peak from among the peaks of the power spectrum using any of the peak detection means of the third peak detection means for performing peak detection based on the comparison of the levels of
A target recognition means for recognizing a target based on the target peak detected by the target peak detection means;
The target peak detection means includes a selection means for selecting any one of the first peak detection means, the second peak detection means, and the third peak detection means according to the peak state of the power spectrum. A radar device characterized by that.

Images