JP2009014405A - In-vehicle radar apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an in-vehicle radar apparatus whose measurement error is made small by improving a time response property, even in the case that a state of a target significantly varies in a short time. <P>SOLUTION: The in-vehicle radar apparatus employs two processing sections of one FFT. A first FFT processing section executes a fast Fourier transform (FFT) for a plurality, M, of beat signals. A second FFT processing section executes a fast Fourier transform (FFT) for a plurality, L, of beat signals, wherein L is smaller than M. When the amount of temporal variation in a peak signal is small, a distance, an angle and a relative velocity about the target are calculated by using a peak signal obtained from a frequency analysis performed by the first FFT processing section which has an excellent S/N property. When the amount of temporal variation in the peak signal is large, the distance, the angle and the relative velocity about the target are calculated by using a peak signal obtained from a frequency analysis performed by the second FFT processing section which has an excellent response property. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an on-vehicle radar device suitable for being mounted on a vehicle such as an automobile, and more particularly to a signal processing method in the on-vehicle radar device.

  A vehicle-mounted radar is mounted on a vehicle as a sensor for detecting a target existing around the vehicle. Laser radar and millimeter wave radar are known as on-vehicle radars. The millimeter wave radar is an all-weather sensor that can stably detect a target even in a rainy or foggy state. The millimeter wave radar transmits a radio wave in a millimeter wave frequency band from a transmission antenna, and receives a reflected wave from a target such as a vehicle or a pedestrian around the vehicle. By extracting and analyzing the beat signal of the received wave with respect to the transmitted wave, the distance between the vehicle and the target, the relative speed of the target, and the like are detected.

  Several modulation schemes such as an FMCW (Frequency Modulated Continuous Wave) scheme, an FMICW (Frequency Modulated Interrupted Continuous Wave) scheme, and a two-frequency CW (Continuous Wave) scheme have been proposed as modulation schemes for transmitting radio waves.

  In the two-frequency CW system, two relatively close frequencies are periodically switched and transmitted, and the distance to the target, the relative speed, and the like are detected using the modulation degree of these received waves. The two-frequency CW method has an advantage that a necessary frequency band is narrow and only two transmission frequencies are required, so that a circuit configuration such as an oscillator can be simplified.

  In a conventional radar apparatus, frequency analysis such as FFT (Fast Fourier Transform) is performed in order to calculate a target distance, a relative speed, and the like from a beat signal. For example, in the example disclosed in Japanese Patent Application Laid-Open No. 2002-341919, when the reflected wave of the target is included in the beat signal, FFT processing with high frequency measurement accuracy is performed, and the reflected wave of the target is included in the beat signal. If not, FFT processing with low frequency measurement accuracy is executed to reduce the processing load.

Japanese Patent Laid-Open No. 2002-341019

  The target may accelerate or decelerate or turn right or left. In addition, multipath interference and radio wave interference may occur. That is, the target state, for example, the distance, relative speed, azimuth angle, or received signal strength may change significantly within a short time.

  In the conventional radar apparatus, when FFT processing with high frequency measurement accuracy is used, a long observation time is required. Therefore, the measured value is an average value for a long observation time. Therefore, the time response is not good.

  An object of the present invention is to provide an on-vehicle radar device that improves time response and has little measurement error even when the state of a target changes greatly in a short time.

  The on-vehicle radar device of the present invention includes an A / D converter that samples beat signals generated from a transmission wave and a reception wave and generates M sampling signals every predetermined time, and two FFT processing units.

  The first FFT processing unit performs frequency analysis with excellent S / N characteristics, and the second FFT processing unit performs frequency analysis with excellent responsiveness.

  The first FFT processing unit performs a fast Fourier transform (FFT) on the M beat signals. The second FFT processing unit extracts one or more consecutive L (L <M) beat signals from the M beat signals, and fast Fourier transforms the one or more extracted L beat signals. (FFT) is executed.

  When the temporal change amount of the peak signal is small, the distance, angle, and relative speed of the target are calculated using the peak signal obtained from the frequency analysis by the first FFT processing unit having excellent S / N characteristics. When the temporal change amount of the peak signal is large, the distance, angle, and relative speed of the target are calculated using the peak signal obtained from the frequency analysis by the second FFT processing unit having excellent responsiveness.

  Further, according to the present invention, when the temporal change amount of the host vehicle speed value, the yaw rate value, and the steering angle value is small, the peak signal obtained from the frequency analysis by the first FFT processing unit having excellent S / N characteristics is used. The target distance, angle, and relative speed are calculated. When the time variation of the host vehicle speed value, the yaw rate value, and the steering angle value is large, using the peak signal obtained from the frequency analysis by the second FFT processing unit with excellent responsiveness, the target distance, angle, And the relative speed is calculated.

  According to the present invention, in the in-vehicle radar device, even when the target state changes greatly in a short time, the time response can be improved and the measurement error can be reduced.

  With reference to FIG. 1, the example of the structure of the vehicle-mounted radar apparatus of this invention is demonstrated. In the on-vehicle radar device of this example, a two-frequency CW method is used as a modulation method, and a monopulse method is used as an angle detection method. The two-frequency CW method is a method in which two relatively close frequencies are periodically switched and transmitted, and the distance to the target, the relative speed, and the like are detected using the degree of modulation of these received waves. The monopulse method is a method in which the receiving antenna is divided into two left / right, and the azimuth angle of the target is detected from the sum and difference of the received signals of the left and right antennas, the power ratio, and the phase difference.

  The radar apparatus of this example includes a radar antenna unit 11 and a signal processing 12. First, the configuration and operation of the radar antenna unit 11 will be described. The radar antenna unit 11 includes an oscillator 13, a transmission antenna 14, a reception left antenna 15, a reception right antenna 16, and mixers 17 and 18. The oscillator 13 generates a transmission wave and transmits it to a target in front of the radar (for example, in front of the vehicle) via the transmission antenna 14. The reception left antenna 15 and the reception right antenna 16 receive the reflected wave from the target. The frequency of the reflected wave is a value obtained by adding the Doppler frequency ΔF to the transmitted wave. The Doppler frequency ΔF is determined by the relative speed of the target with respect to the host vehicle.

  The reception left antenna 15 and the reception right antenna 16 supply received waves to the mixer units 17 and 18, respectively. The mixer units 17 and 18 mix the transmission wave and the reception wave, and output a Doppler frequency signal as a beat signal. This beat signal is supplied to the signal processing unit 12.

  Next, the configuration and operation of the signal processing unit 12 will be described. The signal processing unit 12 includes A / D converters 19 and 20 and a microcomputer (hereinafter referred to as a microcomputer) 21. The microcomputer 21 is connected to a vehicle speed sensor 22, a yaw rate sensor 23, and a steering angle sensor 24 included in the host vehicle. Each of the A / D converters 19 and 20 samples the beat signal supplied from the radar antenna unit 11 and supplies the sampling signal to the microcomputer 21.

  With reference to FIG. 2, the configuration and processing of the signal processing unit 12 of the on-vehicle radar device according to the present invention will be described in detail. The signal processing unit 12 includes a sampling data acquisition unit 1, a first FFT processing unit 2 for high S / N, a second FFT processing unit 3 for high response, a peak detection unit 4, a peak signal history unit 5, and an effective peak selection unit. 6, a distance / angle / relative speed calculation unit 7, a vehicle information acquisition unit 8, a vehicle information history unit 9, and a target tracking unit 10. These elements may be realized as hardware, but may be realized by a program executed by the microcomputer 21.

  With reference to FIG. 3, the process of the 1st FFT process part 2 and the 2nd FFT process part 3 is demonstrated. FIG. 3 shows sampling data obtained by the A / D converters 19 and 20, that is, beat signals. The radar antenna unit 11 generates 2048 points of sampling data (beat signals) every predetermined time, for example, 44 ms. The first FFT processing unit 2 performs an FFT (Fast Fourier Transform) process once on 2048 points of sampling data (beat signals). Thereby, a spectrum of one received signal strength is obtained. The second FFT processing unit 3 divides 2048 points of sampling data into four time zones to obtain four sets of 512 points of sampling data. An FFT (Fast Fourier Transform) process is executed once on 512 points of sampling data. Accordingly, a total of four received signal strength spectra are obtained for the four time zones.

  Next, the timing of sampling and FFT processing will be described. The first FFT processing unit 2 completes and outputs the FFT process for the 2048-point beat signal in at least a time shorter than the sampling time of the 2048-point beat signal. In this example, 2048 beat signals are generated every 44 ms. Accordingly, the first FFT processing unit 2 executes the FFT processing on the 2048-point beat signal and the output of the result within 44 ms. The execution of the FFT processing is started when 2048 beat signals are input. Therefore, the next 2048 beat signals are input while the FFT process and the result are being output. Thus, real-time processing is possible by executing the FFT processing and the output of the result within a time shorter than the sampling time of 2048 beat signals.

  Similarly, the second FFT processing unit 3 executes the FFT processing on the 512-point beat signal and the output of the result within 11 ms. Thereby, real-time treatment is possible.

  Hereinafter, executing the 2048-point FFT process once for the 2048-point beat signal is simply referred to as executing the 2048-point FFT process, and executing the 512-point FFT process for the 512-point beat signal four times. Is simply referred to as performing 512-point FFT processing. The 2048-point FFT process performs frequency analysis with excellent S / N characteristics, and the 512-point FFT process performs frequency analysis with excellent responsiveness, which will be described below.

  Here, the case where the A / D converters 19 and 20 transmit 2048 points of sampling data every 44 ms has been described. However, this is merely an example, and the number of sampling data is not limited to 2048 points. Therefore, it is assumed that the number of sampling data is M and is divided into N time zones. In this case, the first FFT processing unit 2 executes M point FFT processing once, and the second FFT processing unit 3 executes M / N (= L) point FFT processing N times.

  With reference to FIG.4 and FIG.5, the peak detection process by the peak detection part 4 is demonstrated. As shown in FIG. 5, the peak detection process includes a noise estimation process in step S401 and a peak detection process in step S402.

  FIG. 4 shows the spectrum of the received signal obtained by the 2048-point FFT process and the 512-point FFT process. In FIG. 4, the horizontal axis represents frequency, and the vertical axis represents received signal strength. In the noise estimation processing in step S401, noise is estimated from this received signal strength signal. As shown in FIG. 4, the result of the FFT processing is divided into a plurality of frequency domains, and an average value of signal strengths in each frequency domain is obtained. This average value is set as the noise level in each frequency domain. In the peak detection process in step S402, a peak signal corresponding to the target is extracted using this noise level. As shown in FIG. 4, the signal-to-noise ratio (S / N) is calculated for each peak signal with the noise level as a reference. Next, a peak signal having a signal-to-noise ratio (S / N) larger than a predetermined value (for example, 13 dB) is extracted.

  One peak signal represents one target. For example, when two vehicles are detected as targets, two peak signals are obtained. In the following discussion, it is assumed that two peak signals, that is, two targets are detected from the result of the FFT processing.

  In this way, the frequency of the peak signal is obtained. If the frequency of the peak signal is obtained, the amplitude and phase of the peak signal can be obtained. The frequency of the peak signal indicates the relative speed of the target with respect to the host vehicle, the phase of the peak signal indicates the distance from the host vehicle to the target and the angle of the target with respect to the host vehicle, and the amplitude or intensity of the peak signal is determined from the target. Represents the intensity of the reflected signal.

  With reference to FIGS. 6 and 7, the effective peak selection processing by the effective peak selection unit 6 will be described. First, the purpose of this effective peak selection process will be described. FIG. 6 is an example in which the frequency of the peak signal does not change with time. This is a case where the relative speed of the target with respect to the own vehicle has not changed. In this example, as shown in FIG. 3, 512-point FFT processing is executed four times and 2048-point FFT processing is executed once for 2048 points of sampling data. The results of 512-point FFT processing are shown in FIG. 6A to FIG. 6D, and the results of 2048-point FFT processing are shown in FIG. 6A shows sample times 0 to T1, FIG. 6B shows sample times T1 to T2, FIG. 3C shows sample times T2 to T3, and FIG. 6D shows sample times T3 to T3. The result of the 512-point FFT processing for the beat signal at T4 is shown. FIG. 6E shows the result of the 2048-point FFT process on the beat signal at the sample times 0 to T4.

  Since FIG. 6 assumes a case where the frequency of the peak signal does not change with time, the frequency of the peak signal does not change in FIGS. 6 (a) to 6 (d). Therefore, when the 2048-point FFT process is performed on the beat signal at the sample times 0 to T4, as shown in FIG. 6E, a peak signal with a narrow peak width appears at the same frequency.

  FIG. 7 shows an example in which the frequency of the peak signal changes with time. This is a case where the relative speed of the target with respect to the own vehicle is changing. The frequency of the peak signal changes in FIGS. 7 (a) to 7 (d). Therefore, when the 2048-point FFT process is performed on the beat signal at the sample times 0 to T4, a peak signal with a narrow peak width appears as shown in FIG.

  That is, if the 512-point FFT processing result is used, the time change of the frequency can be obtained in detail as shown in FIGS. 7A to 7D, whereas the 2048-point FFT processing result is obtained. If it is used, as shown in FIG. 7E, the time change of the frequency cannot be obtained, and the frequency cannot be specified. That is, according to the 512-point FFT processing, frequency analysis with excellent responsiveness can be performed. On the other hand, the 2048-point FFT process cannot perform frequency analysis with excellent responsiveness, but has an advantage of being able to perform frequency analysis with excellent S / N characteristics.

  This is the conventional problem. According to the present invention, when the change in the frequency of the peak signal is large, the frequency analysis is performed using the result of the 512-point FFT processing instead of the result of the 2048-point FFT processing. Thereby, the responsiveness to the frequency can be improved and the measurement accuracy can be improved. Although the case where the frequency of the peak signal changes has been described here, the same applies to the case where the phase of the peak signal changes and the amplitude of the peak signal changes.

  With reference to FIG. 8, the target tracking process by the target tracking part 10 is demonstrated. FIG. 8 shows the transition of the target distance based on the distance and angle and the target distance calculated by the relative velocity calculation unit 7. The vertical axis is distance, and the horizontal axis is time. A curve 801 shows the transition of the target distance obtained based on the distance obtained from the result of the 2048-point FFT process. Circles indicate distance data calculated by 2048-point FFT processing. A curve 802 shows the transition of the target distance obtained based on the distance obtained from the result of the 512-point FFT process. A cross indicates distance data calculated by 512-point FFT processing. As shown in the figure, it can be seen that the curve 801 obtained by the 2048-point FFT process is less influenced by noise, that is, performs frequency analysis with excellent S / N characteristics. On the other hand, the curve 802 obtained by the 512-point FFT processing performs frequency analysis with excellent responsiveness, but it is understood that the curve 802 is easily affected by noise.

  With reference to FIG. 9, a first example of processing for detecting a target by the on-vehicle radar device according to the present invention will be described.

  In step S101, the target beat signal is sampled to obtain a sampling signal. This processing is executed by the sampling data acquisition unit 1 in FIG. Here, as shown in FIG. 3, 2048 points of sampling data, that is, beat signals are acquired. In step S102, a 2048-point FFT process is executed. This process executes the first FFT processing unit 2 of FIG. In step S103, 512-point FFT processing is executed. This process is executed by the second FFT processing unit 3 in FIG. Here, since there are 2048 sampling data, as shown in FIG. 3, the 2048-point FFT processing is executed once and the 512-point FFT processing is executed four times.

  In step S104, a peak signal detection process is executed for each result of the 2048-point FFT process and the 512-point FFT process. This processing is executed by the peak detection unit 4 in FIG. In the peak signal detection process, the frequency, amplitude, and phase of the peak signal are detected. The frequency of the peak signal indicates the relative speed of the target with respect to the host vehicle, the phase of the peak signal indicates the distance from the host vehicle to the target and the angle of the target with respect to the host vehicle, and the amplitude or intensity of the peak signal is determined from the target. Represents the intensity of the reflected signal.

  In step S105, the result of the peak signal detection process is stored. This processing is executed by the peak signal history unit 5 of FIG. That is, the frequency, amplitude, and phase of the peak signal extracted in step S104 are stored in the memory. As described above, since the 2048 point 1 FFT process is executed once and the 512 point FFT process is executed 4 times, the peak signal obtained from the result of the FFT process for a total of 5 times is recorded here.

  In steps S106 to S108, the first example of the effective peak selection process according to the present invention is executed. This process is executed by the effective peak selection unit 6 of FIG.

  In step S106, the amount of change in the frequency of the peak signal is checked from the result of the 512-point FFT process. If the amount of change in frequency exceeds the predetermined threshold value α1, the process proceeds to step S110, and if not, the process proceeds to step S107. The amount of change in the frequency of the peak signal represents the amount of change in the relative speed of the target with respect to the host vehicle. If the amount of change in the relative speed of the target is large, the process proceeds to step S110. Otherwise, the process proceeds to step S107.

  In step S107, the amount of change in the phase of the peak signal is checked from the result of the 512-point FFT process. If the phase change amount exceeds the predetermined threshold value α2, the process proceeds to step S110, and if not, the process proceeds to step S108. The amount of change in the phase of the peak signal represents the amount of change in the distance and angle from the vehicle to the target. If the distance to the target and the amount of change in the angle are large, the process proceeds to step S110, otherwise the process proceeds to step S108.

  In step S108, the amount of change in the signal strength of the peak signal is checked from the result of the 512-point FFT process. If the change amount of the signal intensity exceeds the predetermined threshold value α3, the process proceeds to step S110. Otherwise, the process proceeds to step S109. The change amount of the signal intensity of the peak signal represents the change amount of the reflection intensity of the signal from the target. If the amount of change in the reflection intensity of the signal from the target is large, the process proceeds to step S110, and if not, the process proceeds to step S109.

  In step S109, distance, angle, and relative speed calculation processing is executed. This processing is executed by the distance, angle, and relative speed calculator 7 in FIG. Here, the amount of change in the relative speed of the target, the amount of change in the phase, and the amount of change in the reflection intensity of the signal from the target are all less than a predetermined threshold. Accordingly, the target distance, angle, and relative velocity are calculated using the peak signal obtained from the result of the 2048-point FFT processing. Assume that two targets are detected from the result of one 2048-point FFT process. When the target distance, angle, and relative speed are calculated for each target, two target distances, angles, and relative speeds are obtained. Here, the distance and relative speed are calculated by the two-frequency CW method, and the angle is calculated by the monopulse method. These methods are well known in the art and will not be described in further detail.

  On the other hand, also in step S110, a distance, angle, and relative speed calculation process is executed. This processing is executed by the distance, angle, and relative speed calculator 7 in FIG. Here, at least one of the change amount of the relative speed of the target, the change amount of the phase, and the change amount of the reflection intensity of the signal from the target exceeds a predetermined threshold value. Therefore, the target distance, angle, and relative speed are calculated using the peak signal obtained from the 512-point FFT processing result. Assume that two targets are detected in each of the results of four 512-point FFT processes. When the target distance, angle, and relative speed are calculated for each target, eight target distances, angles, and relative speeds are obtained.

In step S111, target tracking processing is executed. This process is executed by the target tracking unit 10 of FIG. A filtering process such as LPF (low-pass filter) is performed on the distance, angle, and relative speed obtained in steps S109 and S110. The filtering process is executed by the number of detected targets. In step S109, one distance, angle, and relative speed are obtained for two targets. In this case, the filtering process is executed only twice. In step S110, four distances, angles, and relative velocities are obtained for two targets. In this case, the filtering process is executed 8 times. Then, the calculation result of the filtering process is used as a measurement value for the target. Using the measured values, create a path that represents the transition of the relative speed, distance, and angle of the target.
In the above embodiment, the time change amount of the frequency, phase, or signal intensity of the peak signal is used as a determination index for the effective peak selection process. However, RCS (Radar Cross Section) may be used as a determination index for effective peak selection processing. The RCS can be calculated from the distance and the signal strength, but the calculation formula is widely known in the technical field, and therefore will not be described in detail here.

  With reference to FIG. 10, the 2nd example of the process which detects a target with the vehicle-mounted radar apparatus of this invention by this invention is demonstrated. In this example, the peak width is used as a determination index for the effective peak selection process. First, in step S201, 2048 points of sampling data are acquired, and in step S202, a 2048 point FFT process is executed. In step S203, a peak signal is detected.

  In steps S204 and S205, the second example of the effective peak selection process according to the present invention is executed. This process is executed by the effective peak selection unit 6 of FIG. In step S204, peak width calculation is executed.

  The peak width calculation will be described with reference to FIG. FIG. 11 shows the spectrum of the received signal obtained from the result of the FFT process in step S202. The peak width calculation will be described using this spectrum diagram. A peak width determination threshold is set based on the signal intensity of the peak signal detected in step S203. The peak width determination threshold is smaller by a predetermined value (for example, 6 dB) than the signal intensity of the peak signal. Next, the frequencies at which the skirts present on the left and right sides of the peak signal are below the peak width determination threshold are obtained, and the difference between the two frequencies is used as the peak width.

  In step S205, it is determined whether or not the peak width exceeds a predetermined value β. If the peak width exceeds the predetermined value β, it is determined that the situation as shown in FIG. 7E has occurred, and the process proceeds to step S207. If the peak width does not exceed the predetermined value β, it is determined that the situation in FIG. 6E is present, and the process proceeds to step S206.

  In step S206, one target distance, angle, and relative velocity are calculated using the peak signal detected by the 2048-point FFT process, and the process proceeds to step S209. On the other hand, in step S207, 512-point FFT processing is performed, and the process proceeds to step S208.

  In step S208, a maximum of four target distances, angles, and relative velocities are calculated using the peak signals detected by the 512-point FFT processing, and the process proceeds to step S209. In step S209, filtering such as LPF is performed on the distance, angle, and relative speed to obtain a measurement value for the target.

  With reference to FIG. 12, the 3rd example of the process which detects a target with the vehicle-mounted radar apparatus of this invention by this invention is demonstrated. In this example, FFT selection processing is performed instead of effective peak selection processing. The vehicle information is used as a determination index for the FFT selection process. First, in step S301, 2048 points of sampling data are acquired, and in step S302, information on the vehicle speed, yaw rate, and steering angle is acquired. This process is executed by the vehicle information acquisition unit 8 of FIG.

  In step S303, the vehicle speed, yaw rate, and steering angle information acquired in step S302 is recorded in the memory. This process is executed by the vehicle information history unit 9 of FIG.

  In steps S304 to S306, an example of FFT selection processing according to the present invention is executed. This process is executed by the effective peak selection unit 6 of FIG.

  In step S304, the difference between the own vehicle speed value acquired last time and the own vehicle speed value acquired this time is obtained, and this is set as the time change amount of the own vehicle speed. Then, it is determined whether this time change amount exceeds a predetermined value γ1. If the time change amount of the host vehicle speed exceeds the predetermined value γ1, it is determined that the situation as shown in FIG. 7E has occurred, and the process proceeds to step S308. If the time change amount of the host vehicle speed does not exceed the predetermined value γ1, it is determined that the situation in FIG. 6E is present, and the process proceeds to step S305.

  In step S305, the time change amount of the yaw rate is similarly calculated, and it is determined whether this time change amount exceeds a predetermined value γ2. If the time change amount of the yaw rate exceeds the predetermined value γ2, it is determined that the situation as shown in FIG. 7E has occurred, and the process proceeds to step S308. If the time change amount of the yaw rate does not exceed the predetermined value γ2, it is determined that the situation in FIG. 6E is present, and the process proceeds to step S306.

  In step S306, the time change amount of the steering angle is similarly calculated, and it is determined whether this time change amount exceeds a predetermined value γ3. If the time change amount of the steering angle exceeds the predetermined value γ3, it is determined that the situation as shown in FIG. 7E has occurred, and the process proceeds to step S308. If the time change amount of the steering angle does not exceed the predetermined value γ3, it is determined that the situation in FIG. 6E is present, and the process proceeds to step S307.

  In step S307, a 2048-point FFT process is performed, and in step S308, a 512-point FFT process is performed, and the process proceeds to step S309.

  In step S309, a peak signal is detected from the 2048-point FFT process or 512-point FFT process, and the process proceeds to step S310. In step S310, the distance, angle, and relative speed of the target are calculated using the detected peak signal, and the process proceeds to step S311. In step S311, filtering such as LPF is performed on the distance, angle, and relative speed to obtain a measurement value for the target.

  By executing the processing as described above, even when the state of the received signal with respect to the target changes greatly in a short time, the time responsiveness can be improved and a radar apparatus with little measurement error is realized.

  The example of the present invention has been described above, but the present invention is not limited to the above-described example, and various modifications can be easily made by those skilled in the art within the scope of the invention described in the claims. It will be understood.

  The present invention is applicable to an on-vehicle radar device.

It is a figure explaining the example of the structure of the vehicle-mounted radar apparatus of this invention. It is a figure explaining the structure and process of the signal processing part of the vehicle-mounted radar apparatus by this invention. It is a figure which shows the example of the sampling data sent to the signal processing part from the radar antenna part of the vehicle-mounted radar apparatus by this invention, ie, a beat signal. With reference to FIG.4 and FIG.5, the peak detection process by the peak detection part 4 is demonstrated. As shown in FIG. 5, the peak detection process includes a noise estimation process in step S401 and a peak detection process in step S402. It is a spectrum figure explaining a peak detection process. It is a figure which shows the flow of a process explaining the peak detection process in the vehicle-mounted radar apparatus by this invention. It is a figure which shows the spectrum of the received signal obtained by the FFT process in case the frequency of a peak signal is not changing temporally. It is a figure which shows the spectrum of the received signal obtained by the FFT process in case the frequency of a peak signal is changing temporally. It is a figure explaining a target tracking process. It is a figure explaining the 1st example of the process which detects a target with the vehicle-mounted radar apparatus of this invention. It is a figure explaining the 2nd example of the process which detects a target with the vehicle-mounted radar apparatus of this invention. It is a figure explaining the peak width calculation process in the vehicle-mounted radar apparatus of this invention. It is a figure explaining the 3rd example of the process which detects a target with the radar apparatus of this invention this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Sampling data acquisition part, 2 ... FFT processing part (for high S / N), 3 ... FFT processing part (for high responsiveness), 4 ... Peak detection part, 5 ... Peak signal history part, 6 ... Effective peak selection , 7 ... Distance and angle and relative speed calculation unit, 8 ... Vehicle information acquisition unit, 9 ... Vehicle information history unit, 10 ... Target tracking unit, 11 ... Radar antenna unit, 12 ... Signal processing unit, 13 ... Oscillator, 14 ... Transmission antenna, 15 ... Left antenna for reception, 16 ... Right antenna for reception, 17, 18 ... Mixer, 19, 20 ... A / D converter, 21 ... Microcomputer, 22 ... Vehicle speed sensor, 23 ... Yaw rate sensor, 24 ... Steering Angle sensor

Claims (11)

A transmitting antenna that transmits radio waves, a receiving antenna that receives reflected waves from a target, a beat generator that generates beat signals from the transmitted waves and received waves, and M beats that are sampled at predetermined times An A / D converter that generates a sampling signal, a first FFT processing unit that analyzes the frequency of the M beat signals, and L beat signals that are consecutive among the M beat signals (L <M). One or more sets are extracted, and a peak signal is detected from the result of frequency analysis performed by the second FFT processing unit that performs frequency analysis on the extracted one or more sets of L beat signals, and the first FFT processing unit and the second FFT processing unit. An on-vehicle radar device comprising: a peak detecting unit that calculates a distance, an angle, and a relative speed of a target from the peak signal; Te,
When the temporal change amount of the peak signal is less than a predetermined threshold, the target distance, angle, and relative speed calculated from the peak signal obtained from the frequency analysis by the first FFT processing unit are output, and the peak When the temporal change amount of the signal is larger than a predetermined threshold, the target distance, angle, and relative velocity calculated from the peak signal obtained from the frequency analysis by the second FFT processing unit are output. In-vehicle radar device.   2. The on-vehicle radar device according to claim 1, wherein a temporal change amount of the frequency, phase, or signal strength of the peak signal is compared with a predetermined threshold value, and a temporal change amount of the frequency, phase, or signal strength. Is less than the threshold value, the target distance, angle, and relative velocity calculated from the peak signal obtained from the frequency analysis by the first FFT processing unit are output, and the frequency, phase, or signal intensity is output. When the amount of temporal change is larger than a predetermined threshold, the vehicle distance is output, and the target distance, angle, and relative speed calculated from the peak signal obtained from the frequency analysis by the second FFT processing unit are output. Radar equipment.   2. The on-vehicle radar device according to claim 1, wherein a temporal change amount of RCS (Radar Cross Section) of the peak signal is compared with a predetermined threshold value, and the RCS is less than the predetermined threshold value. Outputs the target distance, angle, and relative speed calculated from the peak signal obtained from the frequency analysis by the first FFT processing unit, and when the RCS is larger than the predetermined threshold, the second FFT is performed. A vehicle-mounted radar device that outputs a target distance, angle, and relative velocity calculated from a peak signal obtained by frequency analysis by a processing unit.   2. The on-vehicle radar device according to claim 1, wherein when a peak width of a peak signal obtained from a result of frequency analysis by the first FFT processing unit is equal to or less than a predetermined threshold value, the frequency is obtained by frequency analysis by the first FFT processing unit. When the peak distance of the peak signal obtained from the result of frequency analysis by the first FFT processing unit is greater than a predetermined threshold, the target distance, angle, and relative speed calculated from the peak signal are output. A vehicle-mounted radar device that outputs a target distance, angle, and relative velocity calculated from a peak signal obtained by frequency analysis by a second FFT processing unit.   The on-vehicle radar device according to claim 1, wherein the peak detection unit estimates a noise level from a frequency spectrum of a received signal obtained by frequency analysis by the first FFT processing unit and the second FFT processing unit, and calculates the noise level. A vehicle-mounted radar device characterized in that a peak signal corresponding to a target is extracted.   6. The on-vehicle radar device according to claim 5, wherein the peak detection unit divides the frequency spectrum of the received signal into a plurality of frequency regions, and calculates an average value of signal intensity in each frequency region as a noise level in each frequency region. An on-vehicle radar device.   6. The on-vehicle radar device according to claim 5, wherein the peak detector calculates a signal-to-noise ratio (S / N) for each peak signal with the noise level as a reference, and the signal-to-noise ratio (S / N) extracts a peak signal having a value greater than a predetermined value.   2. The on-vehicle radar device according to claim 1, wherein the first FFT processing unit performs a fast Fourier transform (FFT) on the M beat signals within a time shorter than a time for sampling the M beat signals. And outputting a result of frequency analysis, wherein the second FFT processing unit performs a fast Fourier transform (FFT) on the L beat signals within a time shorter than a time for sampling the L beat signals, and An on-vehicle radar device that outputs a result of frequency analysis. A transmitting antenna that transmits radio waves, a receiving antenna that receives reflected waves from a target, a beat generator that generates a beat waveform from the transmitted waves and the received waves, and M beats that are sampled at predetermined times An A / D converter that generates a sampling signal, a first FFT processing unit that analyzes the frequency of the M beat signals, and L beat signals that are consecutive among the M beat signals (L <M). One or more sets are extracted, and a peak signal is detected from the result of frequency analysis performed by the second FFT processing unit that performs frequency analysis on the extracted one or more sets of L beat signals, and the first FFT processing unit and the second FFT processing unit. An on-vehicle radar device comprising: a peak detecting unit that calculates a distance, an angle, and a relative speed of a target from the peak signal; Te,
A vehicle information acquisition unit for acquiring the own vehicle speed value, yaw rate value and steering angle value of the own vehicle;
When the temporal change amount of the host vehicle speed value, the yaw rate value or the steering angle value is compared with a predetermined threshold value, and the temporal change amount of the host vehicle speed value, the yaw rate value or the steering angle value is less than the threshold value, The target distance, angle, and relative speed calculated from the peak signal obtained from the frequency analysis by the first FFT processing unit are output, and the temporal change amount of the host vehicle speed value, yaw rate value, or steering angle value is predetermined. When it is larger than the threshold value, the on-vehicle radar device outputs a target distance, an angle, and a relative speed calculated from a peak signal obtained by frequency analysis by the second FFT processing unit. A transmission process for transmitting radio waves to the target;
A receiving step of receiving a reflected wave from the target;
A beat generation step of generating a beat signal from the transmission wave and the reception wave;
An A / D converter step of sampling the beat signal and generating M sampling signals every predetermined time;
A first FFT processing step of performing frequency analysis on the M beat signals by fast Fourier transform (FFT);
A second FFT processing step of extracting one or more sets of consecutive L (L <M) beat signals from the M beat signals, and performing frequency analysis on the extracted one or more sets of L beat signals; ,
A peak detection step of detecting a peak signal from a result of frequency analysis by the first FFT processing unit and the second FFT processing unit;
In the target detection method, the calculation step of calculating the target distance, angle, and relative speed from the peak signal,
When the temporal change amount of the peak signal is less than a predetermined threshold, the target distance, angle, and relative speed calculated from the peak signal obtained from the frequency analysis by the first FFT processing unit are output, and the peak When the temporal change amount of the signal is larger than a predetermined threshold, the target distance, angle, and relative velocity calculated from the peak signal obtained from the frequency analysis by the second FFT processing unit are output. A target detection method using an on-vehicle radar device. A transmission process for transmitting radio waves to the target;
A receiving step of receiving a reflected wave from the target;
A beat generation step of generating a beat signal from the transmission wave and the reception wave;
An A / D converter step of sampling the beat signal and generating M sampling signals every predetermined time;
A first FFT processing step of performing frequency analysis on the M beat signals by fast Fourier transform (FFT);
A second FFT processing step of extracting one or more sets of consecutive L (L <M) beat signals from the M beat signals, and performing frequency analysis on the extracted one or more sets of L beat signals; ,
A peak detection step of detecting a peak signal from a result of frequency analysis by the first FFT processing unit and the second FFT processing unit;
In the target detection method, the calculation step of calculating the target distance, angle, and relative speed from the peak signal,
A vehicle information acquisition step of acquiring the own vehicle speed value, yaw rate value or steering angle value of the own vehicle;
When the temporal change amount of the host vehicle speed value, the yaw rate value or the steering angle value is compared with a predetermined threshold value, and the temporal change amount of the host vehicle speed value, the yaw rate value or the steering angle value is less than the threshold value, The target distance, angle, and relative speed calculated from the peak signal obtained from the frequency analysis by the first FFT processing unit are output, and the temporal change amount of the host vehicle speed value, yaw rate value, or steering angle value is predetermined. When the value is larger than the threshold value, the target distance, angle, and relative speed calculated from the peak signal obtained from the frequency analysis by the second FFT processing unit are output, and the target is detected by the on-vehicle radar device. Method.

Images