JP3214480B2 - FM-CW radar device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an FM-CW mounted on a vehicle or the like for detecting the distance and relative speed of a moving object.
The present invention relates to a radar device.

[0002]

2. Description of the Related Art A radar device mounted on a vehicle of this type monitors another vehicle traveling forward or backward, and
Distance and relative speed can be measured simultaneously. In particular, monitoring of a vehicle traveling ahead is to prevent following of the vehicle or a collision accident, and is expected to be very useful in the future. In addition, there is a millimeter-wave radar device that applies a millimeter-wave band radio wave to an on-vehicle radar device, and there is also an FM-CW radar system using a continuous wave of an FM modulation wave by a triangular wave as a system of the millimeter-wave radar device. . Conventionally, F
As an M-CW radar device, a continuous wave FM-modulated based on a modulation signal of a triangular wave is transmitted as a transmission wave toward an obstacle such as a car located ahead, and is reflected by the obstacle and returned. Incorporate reflected waves as received waves. Then, from the beat signal obtained by mixing the transmission wave and the reception wave at that time, for example, FFT (Fast Fouri
er Transform) to calculate a peak frequency and a peak level by performing signal processing. Thereafter, by performing recognition processing such as peak pairing and grouping, the distance to the obstacle and the relative speed can be calculated.

[0003] Such a technique is disclosed, for example, in JP-A-9-13.
It is disclosed in, for example, Japanese Patent Publication No. 3765. The technique disclosed in this publication uses the FFT method and varies the modulation signal frequency to achieve both the separation of beat signals of a plurality of targets and the improvement of relative speed calculation accuracy, and at the same time, the sampling frequency is varied. The detection frequency range is made variable to secure a detection range of a constant target distance range, and further, the beat signal frequency accuracy is improved with a minimum number of FFT points.

[0004]

However, in the conventional signal processing, since the frequency accuracy and the distance accuracy are emphasized, the processing time until the obstacle is determined becomes long, and there arises a problem that real-time processing cannot be performed. For example, in the technique disclosed in Japanese Patent Application Laid-Open No. Hei 9-133765, by increasing the number of FFT points in order to improve the frequency accuracy, the processing time required to determine an obstacle becomes longer, resulting in a lack of immediacy. Is generated.

The present invention has been made in view of such circumstances, and has as its object to calculate the number of FFT points, sampling frequency, and modulation frequency before and after two predetermined thresholds from a required maximum detection distance. The present invention is to provide an FM-CW radar device that can greatly reduce the calculation time of the FFT without lowering the frequency accuracy, and that can determine an obstacle in real time.

[0006]

In order to achieve the above object, an FM-CW radar apparatus of the present invention compares a preset threshold value of a distance with the distance of the closest obstacle, Further, by comparing the threshold value of the relative speed with the relative speed of the obstacle located at the closest distance, based on these comparison results, by changing the sampling frequency fs, the number N of FFT points, and the modulation frequency fm, it becomes necessary. It is possible to shorten the processing time until the obstacle is determined without lowering the frequency accuracy and the distance accuracy, and to prevent a rear-end collision or a collision accident of a vehicle traveling ahead. .

That is, an FM-CW radar device according to claim 1 transmits a transmission wave FM-modulated by a triangular wave toward a target, receives a reflected wave reflected from the target as a reception wave, The beat signal obtained by mixing the transmitted wave and the received wave is signal processed by the frequency analysis method,
In an FM-CW radar apparatus that calculates a distance to a target and a relative speed to the target by performing a recognition process such as peak pairing and grouping, an object located closest to a preset distance threshold and a closest distance is used. Compare the distance of the target, and further compare the threshold value of the preset relative speed and the relative speed of the closest target, based on these comparison results, based on the sampling frequency set as the initial value And the number of FFT points set as an initial value,
By changing the modulation frequency of FM modulation, the processing time until the target is determined can be shortened without lowering the required frequency accuracy and distance accuracy, and the distance to the target and the relative speed to the target can be reduced. It is characterized in that it is calculated.

In addition, the processing using the frequency analysis technique
Frequency analysis processing means for extracting spectrum data from discrete value data of the beat signal sampled at a constant time by an A / D converter; and, based on the extracted spectrum data, the peak pairing and the Target recognition means for performing processing such as grouping and calculating the distance to the target and the relative speed to the target, the distance to the target closest to the FM-CW radar device, and a preset Threshold value comparing means for comparing a sampling frequency for a beat signal in the next cycle and the number of FFT points used in the frequency analysis processing means, wherein the distance threshold value is a first threshold value. And a second value larger than the first threshold value
When the value of the distance to the target is smaller than the first threshold, the sampling frequency and the number of FFT points are reduced, and the distance between the first threshold and the target is smaller. When the value is large and the value of the distance to the target is greater than the second threshold, the FFT
It is characterized in that only the number of points is reduced.

The FM-CW radar device according to claim 2 is
2. The FM-CW radar device according to claim 1, wherein the threshold value determining unit includes a preset threshold value of the relative speed, and the sampling is performed when the value of the distance to the target is smaller than the first threshold value. 3. While reducing the frequency and the number of FFT points, comparing the threshold of the relative speed with the relative speed of the target with respect to the FM-CW radar device, the relative speed of the target is larger than the threshold of the relative speed In some cases, the FM-modulated frequency is increased.

[0010]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a diagram illustrating a basic configuration of an FM-CW radar device, and FIG. 2 is a configuration block diagram of an FM-CW radar device applied to an embodiment of the present invention. First, the configuration of the FM-CW radar device will be described with reference to FIGS. This FM-CW radar apparatus has a modulation signal generator 1 for generating a triangular modulation signal.
A voltage-controlled oscillator 2 that performs frequency modulation based on a modulation signal, a transmission antenna 3 that transmits a continuous transmission wave, and a reception antenna 4 that receives a reception wave reflected by a preceding obstacle
And a directional coupler 5 that separates a transmission wave in advance, and a mixer 6 that mixes the separated transmission wave and reception wave so that a beat signal generated by the air-fuel mixture 6 is output. Has become.

As shown in FIG. 2, the beat signal is filtered by an LPF (low-pass filter) 7 and amplified by an amplifier 8, and then is digitally converted by an A / D converter 9 to be subjected to a frequency analysis processor. It is sent to 10. The frequency analysis processor 10 converts the discrete value data of the beat signal sampled by the A / D converter 9 for a predetermined time,
A spectrum is extracted by processing using a frequency analysis method such as FFT. Then, the target recognizer 11
In, peak pairing and grouping of the extracted spectrum are performed, and the distance to the obstacle and the relative speed are calculated.

Next, the threshold value discriminating circuit 12 compares the distance to the obstacle closest to the FM-CW radar device with a preset threshold value of the distance, and performs an A / D converter for the beat signal in the next cycle. 9 and the number of FFT points of the FFT used in the frequency analysis processor 10 are changed. Further, the danger classifier 13 determines the degree of danger to the obstacle from the distance to the obstacle and the relative speed, and the alarm device 14 displays an alarm or an alarm sound to the driver according to the determined degree of danger. To warn you.

Next, the operation of the FM-CW radar apparatus according to this embodiment will be described. First, the operation principle of the A / D converter 9 in FIG. 2 will be described. When the beat signal output via the LPF 7 and the amplifier 8 is sampled by the A / D converter 9 at a certain fixed period τ, even a small change included in the original beat signal is not extracted. That is, a component having a frequency equal to or higher than the upper limit frequency determined by the sampling period τ is not included in the sampled data. This upper limit frequency is called the Nyquist frequency, which is half the sampling frequency. If it is assumed that the continuous signal contains a component higher than the Nyquist frequency, the sample value data of the harmonic component cannot be distinguished from the data for the low frequency component. In other words, in order to correctly obtain the frequency components included in the continuous signal as sample value data, two sample value data of an upper limit and a lower limit are required for a waveform for at least one cycle. Therefore, in the FM-CW radar device, by changing the sampling frequency used for the A / D conversion, the maximum detection frequency of the beat signal is changed at the same time.

Here, the operation principle of the frequency analysis processor 10 shown in FIG. 2, which is the main subject of the present invention, will be described. Although FFT is used as a frequency analysis method, first, a discrete Fourier transform DFT (Disc
rete Fourier Tarnsform) to find the frequency function from the time function. The conversion formula is g
(n), the number of data is N, and the Fourier coefficient is Gn, and is expressed by the following equations (1) and (2). However, W is a rotor.

[0015]

(Equation 1)

[0016]

(Equation 2)

At this time, when Expression (1) is represented by a matrix,
The following equation (3) is obtained.

[0018]

(Equation 3)

As is apparent from equation (3), since addition and multiplication are required N times for each row operation, N
To calculate a row, N2 multiplications are required. That is, as the number of data increases, the calculation time increases in proportion to N2. Therefore, in order to reduce the calculation time, a fast radix-2 Fourier transform that can be easily configured by an algorithm is applied. The radix-2 FFT is an algorithm in which the number of sample value data is limited to a power of two,
Although there are two methods, a time thinning method and a frequency thinning method, in both methods, since the effect of increasing the speed is the same, the time thinning method will be described here.

First, since the number N of sample value data with a radix of 2 is a multiple of two, the N sample value data g (n) in the equation (1) is converted into the following equations (4) and (5). And the even column e
(n) and an odd column h (n).

[0021]

(Equation 4)

[0022]

(Equation 5)

Therefore, from Ek and Hk, which are DFTs of e (n) and h (n) each composed of N / 2 data, Equation (1) is given as the following Equation (6). .

[0024]

(Equation 6)

Equation (6) indicates that Gk, which is a DFT for N sample value data, can be obtained from two DFTs obtained from N '= N / 2, that is, Ek and Hk. At this time, if N / 2 is further divided by 2, the number of operations can be reduced in the same manner.

That is, a description will be given with reference to FIG. 3 which is a diagram showing an FFT processing method for explaining an FFT algorithm. When N is a power of ₂ (log ₂ N), the FFT is composed of stages. That in order to obtain the Gk it is first performed N times multiplications of the rotor W ⁱ is in each stage, which is configured by N stages, multiplications are required ₂ N times Nlog. On the other hand, the addition of Ek and W ^k and Hk Similarly needed Nlog ₂ N times.

However, as is apparent from equation (2), since the actual rotor W always has values of W ⁰ = 1 and W ^{N / 2 =} −1, the multiplication of the rotor is a sum and a difference. In the end, to find Gk, (Nlog ₂ N) / 2 multiplications and
With only Nlog ₂ N additions, the final total number of operations is given by the following equation (7).

[0028]

(Equation 7)

Here, the threshold value discriminating process according to the first embodiment of the present invention will be described. FIG. 4 is a flowchart illustrating the flow of the obstacle recognition process, and FIG. 5 is a flowchart illustrating the flow of the threshold value determination process according to the first embodiment. First, a flowchart of the obstacle recognition process will be described with reference to FIG.

First, A / D conversion processing is performed by the A / D converter 9 at the sampling frequency fs set as an initial value (S1).
Thus, the FFT processing of the input sample value data is performed using the FFT point number N set as an initial value in advance (S2), and the peak frequency and peak level are obtained in the target recognizer 11. That is, the target recognizer 11 obtains the peak value of each of the upper and lower sections of the triangular wave based on the spectrum data (S
3) A peak pairing process is performed (S4). Also,
The detected peak values are subjected to grouping processing (S
5) In the obstacle determination processing, the distance and the relative speed to the obstacle are calculated (S6). Then, the threshold determination circuit 1
2, a threshold value determination process is performed (S7).

Next, the flow of the threshold value discriminating process according to the first embodiment will be described with reference to FIG. First, the threshold determination circuit 12
Then, the detected distance of the closest obstacle is compared with a threshold Rth1 set in advance to a short distance (S10), and the distance of the obstacle is smaller than the threshold Rth1, that is, the obstacle is closer to the threshold. Is determined to exist (S11, Y),
The sampling frequency fs and the number N of FFT points are reduced (S12).

Here, as is apparent from equation (7), F
When the number of FT points is reduced, the calculation time of the FFT is greatly reduced. At this time, for example, the number N of FFT points is set to 5
If the number of operations is reduced from 12 to 1/2 of 256, the number of operations is actually reduced by about 56%. At this time, FFT
At the same time as the number of points N, the sampling frequency fs is halved.
By doing so, for an obstacle that is closer to the set threshold value, the processing time until the obstacle is determined can be performed in real time without lowering the frequency accuracy. Here, the lowering of the sampling frequency fs shortens the maximum detection distance, but does not affect the obstacle that is closer to the set threshold value.

Returning to step S11, if it is determined that the closest obstacle is farther than the set threshold value Rth1 (S11, N), a threshold value Rth2 set to a long distance in advance is detected. The distance is compared with the distance of the closest obstacle, and the determination is performed again (S13). At this time, the threshold value Rth2 set in step S13 is always set to a value larger than the threshold value Rth1 set in step S10.

At this time, if it is determined that the distance of the obstacle is larger than the threshold value Rth2 again (S14, Y), the obstacle is located to a certain distance, and the necessity of the frequency accuracy is eliminated. Therefore, in order to secure only the maximum detection distance, by reducing only the number N of FFT points (S15), it is possible to shorten the processing time until the obstacle is determined and perform the processing in real time. From these results, by using the thresholds Rth1 and Rth2 indicating two different distances, by comparing obstacles existing at the shortest distance, the driver is always warned of obstacle information with high accuracy and in real time. It is possible to prevent a collision or a collision accident with an obstacle ahead.

Next, the flow of the threshold value discriminating process according to the second embodiment will be described with reference to FIG. FIG. 6 is a flowchart illustrating a flow of a threshold determination process for comparing relative speeds according to the second embodiment. The second embodiment will be described based on a formula for calculating the beat signal fb. The beat signal fb is obtained from the difference between the transmission wave and the reception wave. If there is a relative speed, the reception wave is shifted with respect to the transmission wave by the Doppler effect. At this time, the beat signal fb is given by the sum or difference of the distance component fr contained in the reflected wave and the velocity component fd due to the Doppler shift, as in the following equation (8).

[0036]

(Equation 8)

Where ΔF: frequency modulation width [Hz] fm: modulation frequency [Hz] R: distance from FM-CW radar device to obstacle [m] C: speed of light [m / sec] V: FM-CW radar Speed of obstacle relative to device
[m / sec] fo: Carrier frequency [Hz]

Here, the peak frequency of the rising section is represented by fdpku
Assuming that p is pd and the peak frequency in the descending section is fdpkdn, the distance R to the obstacle and the relative speed V are obtained by the following equations (9) to (12).

[0039]

(Equation 9)

[0040]

(Equation 10)

[0041]

[Equation 11]

[0042]

(Equation 12)

At this time, as is apparent from the equation (11), when the modulation frequency fm is changed to, for example, 2 fm, the period of the transmitted wave becomes shorter, but the maximum detection distance becomes shorter. However, by increasing the modulation frequency fm, the difference between the peak frequency in the rising section and the peak frequency in the falling section becomes small, so that peak pairing becomes easy, and the processing time until the obstacle is determined as a whole can be shortened. It becomes possible. Therefore, as shown in FIG. 6, in the second embodiment, similarly to the first embodiment, a threshold value Rth1 set in advance to a short distance and a distance to the detected closest obstacle are set. In comparison (S20), when it is determined that the distance of the obstacle is smaller than the threshold value Rth, that is, the obstacle is located closer to the threshold distance (S2).
1, Y), after changing the sampling frequency fs and the number N of FFT points to a small value (S22), again set the threshold Vth of the relative speed set in advance and the relative speed of the closest detected obstacle to the radar. And (S2
3) If it is determined that the relative speed of the obstacle is larger than the threshold value Vth (S24, Y), the peak frequency is facilitated by increasing the modulation frequency fm (S25), and the processing until the obstacle is determined. Time can be shortened.

Returning to step S21, if the closest obstacle is larger than the threshold value Rth1, (S21)
21, N), the same processing as in the first embodiment is performed. That is, the threshold value Rth2 set to a long distance in advance is compared with the detected distance of the closest obstacle, and the determination is performed again (S26). At this time, the threshold value Rth2 set in step S26 is always set to a value larger than the threshold value Rth1 set in step S20. At this time, when it is determined that the distance of the obstacle is larger than the threshold value Rth2 again (S27,
Y) Since the obstacle is located at a certain distance, the necessity of frequency accuracy is eliminated. Therefore, in order to secure only the maximum detection distance, by reducing only the number of FFT points N (S28), it is possible to shorten the processing time until the obstacle is determined and perform the processing in real time.

Therefore, similarly to the first embodiment, not only the distance threshold Rth but also the relative speed has the threshold Vth, and by comparing with the relative speed of the obstacle located at a short distance,
It is possible to warn the driver of nearby obstacle information with high accuracy and in real time, and to prevent a rear-end collision or a collision accident with a forward obstacle.

The embodiment described above is an example for explaining the present invention, and the present invention is not limited to the above embodiment, and various modifications are possible within the scope of the invention. . For example, in the above-described embodiment, the case where two levels of distance thresholds are provided has been described. However, the same effect as the present invention can be obtained by providing more thresholds and sequentially increasing the threshold values to be compared. Needless to say,

[0047]

As described above, the FM-C of the present invention
According to the W radar device, the threshold value that sets two different distances is compared with the distance to the nearest obstacle, and the sampling frequency fs or the number N of FFT points is changed based on the comparison result, thereby making the necessary. The driver can prevent the collision with the obstacle ahead or the collision accident beforehand because the processing time until the obstacle is determined can be processed in real time with high accuracy without lowering the frequency accuracy and the maximum detection frequency. Can be. Also, the relative speed is compared only when the closest obstacle is present at a position closer than the threshold set for the short distance, and based on the result, the modulation frequency is changed, so that the short distance is obtained. In addition, peak pairing of obstacles having a large relative speed can be facilitated, and the processing time until the obstacle is determined can be processed in more real time.

[Brief description of the drawings]

FIG. 1 is a diagram showing a basic configuration of an FM-CW radar device.

FIG. 2 is an FM-CW applied to an embodiment of the present invention.
FIG. 2 is a configuration block diagram of a radar device.

FIG. 3 is an FF for explaining an FFT algorithm;
It is a figure showing the T processing method.

FIG. 4 is a flowchart illustrating a flow of an obstacle recognition process.

FIG. 5 is a flowchart illustrating a flow of a threshold value determination process according to the first embodiment.

FIG. 6 is a flowchart illustrating a flow of a threshold determination process for comparing relative speeds according to the second embodiment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Modulation signal generator, 2 ... Voltage controlled oscillator, 3 ... Transmission antenna, 4 ... Receiving antenna, 5 ... Directional coupler, 6 ... Mixer, 7 ... LPF (low-pass filter), 8 ... Amplifier
9: A / D converter, 10: frequency analysis processor, 11: target recognizer, 12: threshold discriminator, 13: danger discriminator, 14: alarm

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) G01S ⁷ /00-7/42 G01S 13/00-13/95

Claims (2)

(57) [Claims] 1. A transmission wave FM-modulated by a triangular wave is transmitted toward a target, a reflected wave reflected from the target is received as a reception wave, and the transmission wave and the reception wave are mixed. In the FM-CW radar device which performs signal processing on the obtained beat signal by a frequency analysis method and performs recognition processing such as peak pairing and grouping to calculate a distance to the target and a relative speed to the target. Comparing the preset distance threshold with the distance of the closest target, further comparing the preset relative speed threshold with the relative speed of the closest target, FFT (First Fourier Transf
orm) By changing the number of points and the modulation frequency subjected to FM modulation, the processing time until the target is determined can be reduced without lowering the required frequency accuracy and distance accuracy, and the distance from the target can be reduced. and a calculates a relative velocity between the target object, by treatment with the frequency analysis technique, a / D converter
Beat signal sampled at a fixed time by the
To extract the spectrum data from the discrete value data of
The wave number analysis processing means and the peak data are extracted based on the extracted spectrum data.
Processing such as network pairing and grouping.
Calculate the distance to the target and the relative speed to the target
And a distance to the target closest to the FM-CW radar device.
The distance is compared with a preset distance threshold, and the next size
Sampling frequency for the beat signal of the
Change the number of FFT points used in frequency analysis processing means
Threshold value determining means for determining the distance , wherein the threshold value of the distance is a first threshold value and a value of the first threshold value.
When the value of the distance to the target is smaller than the first threshold,
The sampling frequency and the number of FFT points
And the value of the distance to the target is greater than the first threshold,
And the value of the distance to the target is greater than the second threshold value
If not, reduce only the number of FFT points
An FM-CW radar device characterized by the above-mentioned. 2. The method according to claim 1, wherein the threshold value determining means includes a threshold value of a preset relative speed, and when the value of the distance to the target is smaller than the first threshold value, the sampling frequency and the number of FFT points are determined. And reducing the relative speed threshold and the relative speed of the target with respect to the FM-CW radar device. When the relative speed of the target is greater than the threshold of the relative speed, FM modulation is performed. claims, characterized in that to increase the modulation frequency
Item 3. The FM-CW radar device according to item 1 .