WO2021131601A1

WO2021131601A1 - Radar device

Info

Publication number: WO2021131601A1
Application number: PCT/JP2020/045180
Authority: WO
Inventors: 黒田　浩司; 武田　茂樹; 梅比良　正弘
Original assignee: 日立Astemo株式会社
Priority date: 2019-12-25
Filing date: 2020-12-04
Publication date: 2021-07-01
Also published as: JPWO2021131601A1; DE112020005328T5

Abstract

The present invention improves the azimuth detection accuracy of a radar device. A radar device 1 that uses a virtual expansion antenna system performs, by a DSP 108, a calibration process on first reception information based on a reception signal received through an existing reception antenna 110a, 110b, and performs virtual expansion process for virtually expanding the reception antenna 110a, 110b with respect to the first reception information on which the calibration process has been performed, to convert the first reception information into second reception information based on the virtual expansion antenna system. Then, the radar device estimates the incoming direction of the reception signal on the basis of the second reception information.

Description

Radar device

The present invention relates to a radar device.

Conventionally, a radar device that detects obstacles around the vehicle has been known for use in automatic driving of vehicles and driving support systems. Since such a radar device needs to be mounted on a vehicle, miniaturization is required.

The technique described in Patent Document 1 is known for miniaturization of radar devices. In Patent Document 1, a technique called MIMO (Multiple Input Multiple Output) is used to reduce the influence of mutual coupling between transmitting antennas in a radar device that transmits and receives simultaneously with a plurality of antennas, and the direction in which the target exists is determined. Techniques for improving detection accuracy are disclosed.

Japanese Unexamined Patent Publication No. 2019-128235

In the radar device, in addition to the above-mentioned MIMO method, other methods such as SIMO (Single Input Multiple Output) method in which radio waves transmitted from one transmitting antenna are received by a plurality of receiving antennas are also used. However, since the technique described in Patent Document 1 cannot be applied to these radar devices, it is difficult to improve the directional detection accuracy.

The radar device according to the present invention uses a virtual extended antenna method, and calibrates the first received information based on the received signal received by an existing receiving antenna, and the calibrated process is performed. By performing a virtual expansion process for virtually expanding the reception antenna with respect to the first reception information, the first reception information is converted into the second reception information based on the virtual expansion antenna method, and the first reception information is converted into the second reception information. The arrival direction of the received signal is estimated based on the received information of 2.

According to the present invention, the directional detection accuracy of the radar device can be improved.

It is a figure which shows the structure of the radar apparatus which concerns on one Embodiment of this invention. It is a figure explaining the virtual expansion array antenna formed by the virtual expansion processing. It is a figure explaining the operation principle of a virtual expansion array antenna. It is a flowchart of the calibration process and the virtual extension process in the radar apparatus which concerns on 1st Embodiment of this invention. It is a figure explaining the effect of the calibration process which concerns on 1st Embodiment of this invention. It is a figure explaining the outline of the space averaging method. It is a flowchart of the calibration process and the virtual extension process in the radar apparatus which concerns on 2nd Embodiment of this invention. It is a figure explaining the effect of the calibration process which concerns on the 2nd Embodiment of this invention.

(First Embodiment)
FIG. 1 is a diagram showing a configuration of a radar device according to an embodiment of the present invention. The radar device 1 shown in FIG. 1 is used, for example, mounted on an automobile, and includes a waveform generator 101, a voltage controlled oscillator 102, an amplifier 103, a

low noise amplifier

104a, 104b, a

mixer

105a, 105b, and a low-pass filter. It includes 106a, 106b,

AD converters

107a, 107b, a digital signal processor (DSP) 108, a transmitting antenna 109, and receiving

antennas

110a, 110b.

The waveform generator 101 generates a voltage waveform in which the voltage continuously changes at a predetermined cycle under the control of the DSP 108, and outputs the voltage waveform to the voltage controlled oscillator 102. The voltage controlled oscillator 102 generates a transmission signal having an oscillation frequency controlled according to the voltage waveform input from the waveform generator 101, and outputs the transmission signal to the amplifier 103 and the

mixers

105a and 105b. The amplifier 103 amplifies the transmission signal input from the voltage controlled oscillator 102 and outputs it to the transmission antenna 109. The transmission antenna 109 emits the transmission signal input from the amplifier 103 into space. As a result, the FMCW signal in which the continuous wave is frequency-modulated is transmitted from the radar device.

The

receiving antennas

110a and 110b are arranged at predetermined intervals, and the received signal reflected by the object is received and output to the

low noise amplifiers

104a and 104b, respectively. The

low noise amplifiers

104a and 104b amplify the received signals input from the

receiving antennas

110a and 110b, respectively, and output them to the

mixers

105a and 105b. The

mixers

105a and 105b are composed of multipliers, and the frequency difference between these signals is obtained by multiplying the transmission signal input from the voltage controlled oscillator 102 and the reception signal input from the

low noise amplifiers

104a and 104b. Beat signals corresponding to the above are generated and output to the low-

pass filters

106a and 106b, respectively.

The low-

pass filters

106a and 106b take out the low-frequency components of the beat signal input from the

mixers

105a and 105b, respectively, and output them to the

AD converters

107a and 107b. The

AD converters

107a and 107b generate a digital value of the beat signal by converting the beat signal input from the low-

pass filters

106a and 106b into a digital signal at predetermined sampling cycles, and output the beat signal to the DSP 108.

The DSP 108 is a processor that performs predetermined arithmetic processing by executing a program stored in advance, and has a Fourier transform unit 108a, a virtual expansion processing unit 108b, a peak detection unit 108c, and a direction estimation unit 108d as its functions. It should be noted that these functions may be realized by using a logic circuit such as FPGA (Field Programmable Gate Array) or other hardware instead of the DSP.

The Fourier transform unit 108a performs a fast Fourier transform (FFT) on the digital values of the beat signals obtained by the

AD converters

107a and 107b, respectively, to obtain information representing the frequency component of each beat signal.

The virtual expansion processing unit 108b performs virtual expansion processing for virtually expanding the number of receiving antennas with respect to the information of the frequency component of each beat signal input from the Fourier transform unit 108a. By this virtual expansion processing, expansion antennas corresponding to the

reception antennas

110a and 110b are virtually generated, and the frequency information of the beat signal corresponding to the reception information of the

reception antennas

110a and 110b and the reception information of each expansion antenna correspond to each other. The frequency information of the beat signal to be output is output from the virtual expansion processing unit 108b.

In the following description, the reception information before the virtual expansion processing by the virtual expansion processing unit 108b, that is, the information on the frequency component of the beat signal based on the reception signals actually received by the

reception antennas

110a and 110b is referred to as "first reception information". It is called. Further, the reception information after the virtual expansion processing by the virtual expansion processing unit 108b, that is, the frequency component information of the beat signal based on the reception signals received by the

reception antennas

110a and 110b, and the reception signal virtually received by each expansion antenna Together with the information on the frequency component of the beat signal based on it, it is referred to as "second reception information". In other words, the virtual expansion processing unit 108b converts the first received information into the second received information by performing the virtual expansion processing. The details of the virtual extension processing will be described later.

The peak detection unit 108c detects a peak that exceeds a preset threshold value based on the second reception information input from the virtual expansion processing unit 108b. Then, based on the detected peak, the frequency of the beat signal corresponding to the distance to the object is obtained, and the distance to the object and the relative velocity are calculated.

The direction estimation unit 108d estimates the arrival direction for each peak detected by the peak detection unit 108c based on the second received information input from the virtual expansion processing unit 108b. As a result, the transmission signal is reflected by the object, the arrival direction of the received signal received by the

receiving antennas

110a and 110b is estimated, and the direction of the object with respect to the radar device 1 is obtained.

The radar device 1 described above generates, for example, a voltage waveform of a triangular wave or a sawtooth wave by a waveform generator 101, and outputs this to a voltage controlled oscillator 102 to transmit a transmission signal in which a continuous wave is frequency-modulated. The reflected wave whose transmitted signal is reflected by the object is input to the

mixers

105a and 105b as a received signal after a delay time proportional to the distance d to the object. Therefore, a beat signal having a frequency proportional to the delay time can be obtained.

Note that FIG. 1 shows a configuration example of the radar device 1 in the case where the transmission signal transmitted from one transmission antenna 109 is received by the two

reception antennas

110a and 110b after being reflected by the object. , The number of transmit or receive antennas is not limited to this. Any number of receiving antennas as long as one transmitting antenna is provided with a plurality of receiving antennas for receiving each of the received signals generated by the transmission signal transmitted from the transmitting antenna being reflected by the object. The radar device 1 can be configured with. Here, as the number of receiving antennas is increased, the resolution when the direction estimation unit 108d estimates the arrival direction of the received signal can be improved, but there is a demerit that the mounting space of the radar device 1 increases. Therefore, it is preferable to determine the optimum number of receiving antennas from the relationship between the required resolution of the radar device 1 and the mounting space.

Next, the virtual expansion processing executed by the virtual expansion processing unit 108b will be described with reference to FIGS. 2 and 3. In FIGS. 2 and 3, the number of receiving antennas is different from that in FIG. 1 in order to explain the virtual expansion process in an easy-to-understand manner. Specifically, two receiving

antennas

110a and 110b are shown in FIG. 1, but in FIGS. 2 and 3, the virtual expansion process will be described by taking the case where the radar device 1 has five receiving antennas 21 to 25 as an example. .. However, as described above, the radar device 1 can use an arbitrary number of receiving antennas, and even if the number of receiving antennas is different, the virtual expansion process can be performed by the same method.

FIG. 2 is a diagram illustrating a virtual expansion array antenna formed by the virtual expansion process. For example, as shown in FIG. 2A, it is assumed that the actual array antenna is formed by arranging the five receiving antennas 21 to 25 in a row in the radar device 1. In this case, in the virtual expansion process, by performing an operation using the transformation matrix T, the receiving antennas 21 to 25 are arranged in a line at positions different from those of the receiving antennas 21 to 25 as shown in FIG. 2 (b). Virtual array antennas 31 to 35 arranged side by side can be virtually set to form a virtual array antenna. Then, the receiving antennas 21 to 25 and the virtual antennas 31 to 35 are combined to form a virtual extended array antenna as shown in FIG. 2 (c).

When the vector of the received signal received by the real array antennas formed by the receiving antennas 21 to 25 _{is expressed as x to r} , the vector of the received signal received by the virtual array antennas formed by the virtual antennas 31 to 35 is It is calculated by multiplying the received signal vectors x to _{r by the transformation matrix T.} That is, the received signal vectors x to _T of the virtual extended array antenna shown in FIG. 2C are represented by the following equation (1).

FIG. 3 is a diagram for explaining the operating principle of the virtual extended array antenna. As shown in FIG. 3, the positions of the receiving antennas 21 to 25 forming the real array antenna _{are represented by d r1} to d _r5 , respectively, and the positions of the virtual antennas 31 to 35 forming the virtual array antenna are _designated _{as d v1} to d v5. Represent each. In this case, the steering vector of the real array antenna and the steering vector of the virtual array antenna with respect to the plane wave incident from _{the θ L direction are represented by the following equations (2) and (3), respectively.}

_{Therefore, the steering vector a T} (θ) of the virtual extended array antenna shown in FIG. 3 is represented by the following equation (4).

Assuming that the angle range of the received signal arriving at the real array antenna and the virtual array antenna is the range of θ _L to θ _R , the steering vector of the real array antenna and the steering of the virtual array antenna are steered by dividing this angle range at Δθ intervals. The vector can be queued as shown in the following equations (5) and (6), respectively.

Here, between the steering vector of the real array antenna represented by the equation (5) and the steering vector of the virtual array antenna represented by the equation (6), the following equation is used using the transformation matrix T described above. The relationship (7) holds. That is, the transformation matrix T is a transformation matrix for estimating the reception signal of the virtual array antenna by the plane wave in _{the range of θ L} to θ _{R for each Δθ.}

The transformation matrix T is obtained by the general inverse matrix represented by the following equation (8).

In the radar device 1 of the present embodiment, the direction estimation unit 108d uses the received signal vector and the steering vector of the virtual extended array antennas represented by the above equations (1) and (4), respectively, in the direction of arrival of the received signal. That is, the direction of the object is estimated. As a result, even when the number of elements of the receiving antenna is small, the opening of the receiving antenna can be virtually expanded to improve the resolution when estimating the arrival direction of the received signal.

By the way, the received signal obtained by the radar device 1 contains errors due to various error factors. Therefore, a deviation of an error occurs from the theoretical value that should be originally obtained, which leads to a decrease in resolution when estimating the arrival direction of the received signal. Therefore, in the radar device 1 of the present embodiment, the virtual expansion processing unit 108b performs a calibration process for removing an error from the first received information prior to the execution of the virtual expansion processing. This improves the resolution of estimating the arrival direction of the received signal.

In the present embodiment, an error matrix indicating an error between the received signal and the theoretical value calculated based on the calibration measurement data experimentally acquired in advance is stored in the DSP 108. In the above calibration process, the virtual expansion processing unit 108b does not include an error in the received signal vector represented by the first received information obtained from the receiving antenna by using the inverse matrix of the error matrix stored in the DSP 108 in advance. Convert to a received signal vector. Then, by performing virtual expansion processing on the received signal vector after this conversion, the first received information is converted into the second received information. In addition, instead of the error matrix, the inverse matrix of the error matrix may be stored in the DSP 108, and the calibration process may be performed using this.

The direction estimation unit 108d estimates the arrival direction of the received signal by using the calibrated second reception information output from the virtual expansion processing unit 108b. Therefore, the resolution when the direction estimation unit 108d estimates the arrival direction of the received signal is improved, and as a result, the direction detection accuracy of the radar device 1 can be improved.

FIG. 4 is a flowchart of the calibration process and the virtual expansion process in the radar device 1 according to the first embodiment of the present invention. In the flowchart of FIG. 4, the processes of steps S10 and S20 are preprocesses performed in advance in an experimental environment using the receiving

antennas

110a and 110b, and the processes of steps S110 to S150 radar the receiving

antennas

110a and 110b. This is a process performed by the DSP 108 when a received signal from an object is actually received while mounted on the device 1.

In step S10, the steering vector including the error is experimentally acquired by measuring the received signal waveforms when the experimental radio waves are transmitted from a plurality of directions to the receiving

antennas

110a and 110b, respectively. In step S20, an error matrix G ~ indicating an error from the theoretical value is calculated based on the steering vector acquired in step S10.

_{In step S110, the virtual expansion processing unit 108b acquires the reception signal vectors x to r} of the radar device 1 received by the

reception antennas

110a and 110b, respectively. _{Here, the received signal vectors x to r} are acquired by acquiring the first received information output from the Fourier transform unit 108a to the virtual expansion processing unit 108b.

In step S120, the virtual expansion processing unit 108b performs calibration processing _{on the received signal vectors x to r acquired in step S110.} Here, by multiplying an inverse matrix G ~ ^-1 of the error matrix G ~ calculated in step S20 before processing on the received signal vector x ~ _r, removing the error component from the received signal vector x ~ _r. As a result, the received signal vector x _r including no error can be calculated.

In step S130, the virtual expansion processing unit 108b calculates the transformation matrix T defined by the above equation (8) for the angle range _{from θ L} to θ _R.

In step S140, the virtual expansion processing unit 108b performs virtual expansion processing using the transformation matrix T calculated in step S130 on _{the received signal vector x r calculated by the calibration process in step S120.} Here, the reception signal vector x by the virtual expansion array antenna that has been calibrated by using the following equation (9) in which the above equation (1) is _{applied to the reception signal vector x r and the above equation (4).} _{Calculate T} and the steering vector a _T (θ).
As a result, the first received information from which the error component has been removed by the calibration process can be converted into the second received information after the virtual expansion process.

In step S150, the peak detection unit 108c and the direction estimation unit 108d estimate the arrival direction of the received signal using _{the received signal vector x T} and the steering vector a _{T (θ) calculated in step S140 after the virtual expansion processing.} To do. Here, as described above, the peak detection unit 108c detects the peak, and the direction estimation unit 108d estimates the arrival direction of the received signal for each detected peak. After executing the process of step S150, the flowchart of FIG. 4 ends.

FIG. 5 is a diagram illustrating the effect of the calibration process according to the first embodiment of the present invention. In FIG. 5, the waveform shown by reference numeral 51 shows an example of a spatial spectrum when the received signal is subjected to virtual expansion processing without performing calibration processing, and the waveform shown by reference numeral 52 is obtained by performing calibration processing on the received signal. An example of the spatial spectrum when the virtual expansion processing is performed after the processing is shown. Comparing these spatial spectra, the spatial spectrum 51 has two peaks integrated and the boundary is not clear, whereas the spatial spectrum 52 has a clear boundary between the two peaks.
Therefore, it can be seen that the calibration process improves the resolution of estimating the arrival direction of the received signal. The

spatial spectra

51 and 52 show examples of spatial spectra obtained from received signals using a well-known algorithm called MUSIC (Multiple Signal Classification).

According to the first embodiment of the present invention described above, the following effects are exhibited.

(1) The radar device 1 using the virtual extended antenna method performs calibration processing on the first reception information based on the reception signals received by the existing

reception antennas

110a and 110b (step S120), and the calibration processing is performed. By performing a virtual expansion process for virtually expanding the

reception antennas

110a and 110b with respect to the first reception information (step S140), the first reception information becomes the second reception information based on the virtual expansion antenna method. Convert. Then, the arrival direction of the received signal is estimated based on the second received information (step S150). Since this is done, the directional detection accuracy of the radar device can be improved.

(2) In the calibration process of step S120, the radar device 1 removes the error from the first received information by using ^{the inverse matrix G ~ -1} of the error matrix G ~ indicating the error between the received signal and the theoretical value. Perform the calculation of. Since this is done, the error can be reliably removed from the first received information including the error, and the resolution at the time of estimating the arrival direction of the received signal can be improved.

(3) the radar apparatus 1, using the inverse matrix G ~ ^-1 which is the inverse matrix G ~ ^-1 or preset calculated from the set error matrix G ~ advance, perform calculations of the calibration process of step S120 To do. Since this is done, it is possible to obtain an inverse matrix capable of reliably removing an error from the first received information based on the calibration measurement data experimentally acquired in advance.

(4) The radar device 1 receives one transmitting antenna 109 and a plurality of receiving antennas 110a, each of which receives a receiving signal generated by reflecting a transmitting signal transmitted from one transmitting antenna 109 by an object. It is provided with 110b. Therefore, in the radar device 1, it is possible to apply the invention method capable of realizing a small size and high directional resolution.

(Second embodiment)
Next, the radar device according to the second embodiment of the present invention will be described. In this embodiment, an example in which the spatial averaging method is further applied to the radar device 1 described in the first embodiment will be described.

FIG. 6 is a diagram for explaining the outline of the spatial averaging method. As shown in FIG. 6, in the spatial averaging method, the receiving antennas 21 to 25 and the virtual antennas 31 to 35 described in the first embodiment are divided into two groups by shifting the overlapping range to divide the receiving antennas 21 into two groups. A first sub-array composed of 24 to 24 and virtual antennas 31 to 35 and a second sub-array composed of receiving antennas 21 to 25 and virtual antennas 31 to 34 are set. Then, the correlation matrices of the first sub-array and the second sub-array are calculated respectively, and a new correlation matrix is obtained by taking the average of these, and the arrival direction of the received signal is estimated. Thereby, the estimation accuracy of the arrival direction of the received signal can be improved. Since the spatial averaging method itself is a well-known technique, the details thereof will be omitted.

FIG. 7 is a flowchart of the calibration process and the virtual expansion process in the radar device 1 according to the second embodiment of the present invention. In the flowchart of FIG. 7, the parts that perform the same processing as the flowchart of FIG. 4 described in the first embodiment are assigned the same step numbers as those of FIG.

In steps S10 and S20, the same processing as in FIG. 4 is performed as preprocessing.

In steps S110 to S140, the virtual expansion processing unit 108b performs the same processing as in FIG. 4, respectively.

In step S145, the virtual expansion processing unit 108b obtains the second reception information obtained by the virtual expansion processing in step S140, that is, the received signals after the virtual expansion processing represented by the above equations (9) and (4), respectively. Spatial averaging processing based on the spatial averaging method is performed on the vector x _T and the steering vector a _{T (θ).}

In step S150, the peak detection unit 108c and the direction estimation unit 108d use the second reception information after the spatial averaging process calculated in step S145 to determine the arrival direction of the received signal as in the first embodiment. presume. After executing the process of step S150, the flowchart of FIG. 7 ends.

FIG. 8 is a diagram illustrating the effect of the calibration process according to the second embodiment of the present invention. In FIG. 8, the waveform shown by reference numeral 81 shows an example of a spatial spectrum when the received signal is subjected to virtual expansion processing and spatial averaging processing without performing calibration processing, and the waveform shown by reference numeral 82 is a received signal. An example of the spatial spectrum when the virtual expansion processing and the spatial averaging processing are performed after the calibration processing is performed is shown. Comparing these spatial spectra, similar to the

spatial spectra

51 and 52 of FIG. 5 described in the first embodiment, in the spatial spectrum 81, the two peaks are integrated and the boundary is not clear, whereas the space In spectrum 82, the boundary between the two peaks is clear. Therefore, it can be seen that the calibration process improves the resolution of estimating the arrival direction of the received signal. The

spatial spectra

81 and 82 show examples of spatial spectra obtained from received signals using a well-known algorithm called MUSIC.

According to the second embodiment of the present invention described above, the radar device 1 relates to the second received information obtained by performing the virtual expansion process on the first received information that has been calibrated. Spatial averaging processing is performed (step S145), and the arrival direction of the received signal is estimated based on the second received information after the spatial averaging processing (step S150). Since this is done, the estimation accuracy of the arrival direction of the received signal can be further improved.

The embodiments and various modifications described above are merely examples, and the present invention is not limited to these contents as long as the features of the invention are not impaired. Moreover, although various embodiments and modifications have been described above, the present invention is not limited to these contents. Other aspects conceivable within the scope of the technical idea of the present invention are also included within the scope of the present invention.

1 Radar device 101 Waveform generator 102 Voltage controlled oscillator 103

Amplifier

104a, 104b

Low noise amplifier

105a,

105b Mixer

106a, 106b Low-

pass filter

107a, 107b AD converter 108 Digital signal processor (DSP)
108a Fourier transform unit 108b Virtual expansion processing unit 108c Peak detection unit 108d Direction estimation unit 109

Transmitting antenna

110a, 110b Receiving antenna

Claims

A radar device that uses a virtual extended antenna system.
The first received information based on the received signal received by the existing receiving antenna is calibrated and processed.
By performing a virtual expansion process that virtually expands the receiving antenna with respect to the first received information that has undergone the calibration process, the first received information is subjected to a second reception information based on the virtual expansion antenna method. Convert to received information and
A radar device that estimates the arrival direction of the received signal based on the second received information.
In the radar device according to claim 1,
In the calibration process, a radar device that performs a calculation for removing the error from the first received information by using an inverse matrix of an error matrix indicating an error between the received signal and a theoretical value.
In the radar device according to claim 2,
A radar device including an arithmetic processing unit that performs the calculation using the inverse matrix calculated from the preset error matrix or the preset inverse matrix.
In the radar device according to claim 1,
A radar device that performs spatial averaging processing on the second received information and estimates the arrival direction of the received signal based on the second received information after the spatial averaging processing.
In the radar device according to claim 1,
One transmitting antenna and
A radar device including a plurality of receiving antennas each receiving the received signal generated by reflecting a transmitted signal transmitted from the one transmitting antenna by an object.