JP6843568B2 - Radar device and arrival direction estimation method - Google Patents

Description

The disclosed embodiments relate to radar devices and methods of estimating direction of arrival.

Conventionally, a radar device having an array antenna, estimating the arrival direction of the reflected wave by analyzing the reflected wave received by the array antenna, and detecting a target based on the estimated arrival direction has been known. (See, for example, Patent Document 1).

As a technique for estimating the arrival direction of the reflected wave, the Katri-Rao product expansion array processing is performed to virtually increase the number of array elements, etc., and the resolution for the arrival direction and the number of detected reflected waves are improved. Technology has been proposed. In such a Katri-Lao product expansion array process, there is a problem that the estimation error in the arrival direction becomes large due to the correlation component between the reflected waves.

Therefore, a KR (Khatri-Rao) -EM (Expectation Maximization) -DBF (Digital Beam Forming) method has been proposed that reduces the estimation error in the arrival direction by using the EM (Expectation Maximization) algorithm method.

Takeyasu Aoyagi, Hiroki Yamada, Yoshio Yamaguchi, "A Study on DOA Estimation Error and Improvement Method Using Katri-Lao Product Expansion Array", IEICE Technical Report, January 2012, Vol.AP2011-167, No.1, pp. 119-124

However, the above-mentioned method of estimating the arrival direction has a problem that the amount of calculation required for estimating the arrival direction is large and it takes time to estimate the arrival direction.

One aspect of the embodiment is made in view of the above, and provides a radar device and an arrival direction estimation method capable of reducing the amount of calculation required for estimating the arrival direction and reducing the estimation time of the arrival direction. The purpose is to do.

The radar device according to one aspect of the embodiment includes an extended signal generation unit and an arrival direction estimation unit. The extended signal generation unit generates a correlation matrix from signals for each antenna based on reception signals output from each of a plurality of antennas included in an array antenna that receives a plurality of reflected waves from a plurality of targets, and the correlation is generated. An extended signal is generated by the non-overlapping elements of the matrix. The arrival direction estimation unit performs an estimation process of estimating the arrival direction of each reflected wave and the correlation component between the reflected waves based on the difference between the extended signal and the estimated extended signal which is an estimated value of the extended signal. Repeat while gradually reducing the search range in the direction of arrival.

According to one aspect of the embodiment, it is possible to provide a radar device and an arrival direction estimation method capable of reducing the amount of calculation required for estimating the arrival direction and reducing the estimation time of the arrival direction.

FIG. 1A is an explanatory diagram of the arrival direction estimation process according to the embodiment. FIG. 1B is a diagram showing changes in the estimation range in each arrival direction in the estimation process that is repeatedly executed until the stop condition is satisfied. FIG. 2 is a diagram showing a configuration example of a radar device according to an embodiment. FIG. 3 is a diagram showing the relationship between the transmitted wave and the reflected wave and the beat signal. FIG. 4 is a diagram showing the arrangement relationship between the virtual antenna element (extended array) and the real antenna element (real array), which are increased by the KR product expansion array processing. FIG. 5 is a diagram showing a configuration example of the side angle processing unit. FIG. 6 is a flowchart showing an example of a processing procedure executed by the angle measuring processing unit.

Hereinafter, embodiments of the radar device and the arrival direction estimation method disclosed in the present application will be described in detail with reference to the accompanying drawings. The present invention is not limited to the embodiments shown below.

[1. Arrival direction estimation process]
The radar device according to the embodiment includes an array antenna, and executes an arrival direction estimation process for estimating the arrival direction of the plurality of reflected waves by analyzing a plurality of reflected waves received by the array antenna. The arrival direction estimation process according to the embodiment is an estimation method based on maximum likelihood estimation, and is, for example, an estimation processing method by the KR (Khatri-Rao) -EM (Expectation Maximization) -DBF (Digital Beam Forming) method.

FIG. 1A is an explanatory diagram of the arrival direction estimation process according to the embodiment, and the arrival direction estimation process is executed by the radar device according to the embodiment. The radar device according to the embodiment converts the reflected wave by the target received by each receiving antenna of the array antenna into a received signal, and acquires a received signal corresponding to one or more received reflected waves (step S1).

Next, the radar device obtains a correlation matrix from the signal for each receiving antenna based on the received signal acquired in step S1 (step S2), extracts non-overlapping elements in the correlation matrix, and Katri-Rao product (Khatri-Rao product). ) An extended signal z in the extended array process (hereinafter referred to as KR product extended array process) is generated (step S3).

Further, the radar device generates an estimated extended signal z ^ which is an estimated value of the extended signal z using the arrival direction of the reflected wave and the correlation component between the reflected waves as parameters (step S4), and the extended signal z and the estimated extended signal. The difference from z ^ is calculated (step S5).

Then, the radar device updates the estimated value of the correlation component (correlation term) between the arrival direction of each reflected wave and the reflected wave based on the difference Δz between the extended signal z and the estimated extended signal z ^ (step S6). _{For example, the radar device estimates the maximum likelihood estimation value of the estimated extended signal z k} ^ for each of the direction of arrival of the reflected wave and the correlation component between the reflected waves based on the difference Δz, and the estimated extended signal z k ^ of the estimated extended signal z _k ^. Calculate the estimated value of the arrival direction that maximizes the likelihood and the estimated value of the correlation component.

Then, the radar device determines whether or not the stop condition is satisfied (step S7). Such a stop condition is, for example, a condition that the amount of change in the estimated value in the arrival direction is equal to or less than a predetermined threshold value, and a condition that the difference Δz between the extended signal z and the estimated extended signal z ^ converges within a predetermined range.

When the stop condition is not satisfied (step S7; No), the radar device repeats the processes of steps S4 to S6 (an example of the estimation process) until the stop condition is satisfied. At this time, in step S4, the radar device calculates an estimated extended signal z ^ using the estimated value of the arrival direction of each reflected wave calculated in step S6 and the estimated value of each correlation component, and the estimated extended signal z ^. Is used to execute the processes of steps S5 and S6.

On the other hand, when the stop condition is satisfied (step S7; Yes), the radar device outputs the estimated value of the arrival direction of each reflected wave updated in step S6 as a definite estimated value of the arrival direction of each reflected wave (step S8). .. In addition. The processes of steps S1 to S8 are executed every time a received signal is acquired, and a definite estimated value of the arrival direction of each reflected wave is output as an estimation result for each received signal.

As described above, in the arrival direction estimation process according to the embodiment, the estimated value of the arrival direction of each reflected wave and the estimation of each correlation component are estimated while virtually increasing the number of array elements by performing the Katri-Lao product expansion array processing. The value calculation (estimation process) is repeated until the stop condition is satisfied. This makes it possible to reduce the estimation error in the arrival direction.

Further, the radar device according to the embodiment repeats the estimation process of step S6 for calculating the estimated value of each arrival direction and the correlation component by changing the search range of each arrival direction. FIG. 1B is a diagram showing changes in the search range in each arrival direction in the estimation process of step S6, which is repeatedly executed until the stop condition is satisfied. In the example shown in FIG. 1B, an example in which two reflected waves are received by the array antenna is shown.

In the example shown in FIG. 1B, in the process of the first step S6 in which the arrival direction estimation process is started, the arrival direction of the reflected wave is set in the range of -21.8 to +21.8 [deg] (maximum search range). Calculate the estimated value of. After that, in the second process of step S6, the arrival direction of the reflected wave is estimated with the range of -14.0 to -13.0 [deg] and +12.0 to +13.0 [deg] as the search range. Further, after that, in the third process of step S6, the arrival direction of the reflected wave is estimated with the range of -13.74 to -13.14 [deg] and + 12.1 to + 12.6 [deg] as the search range. ..

Although an example of estimating the arrival direction of the reflected wave related to two targets will be described here, the content of the present embodiment can be applied to the estimation of the arrival direction of the reflected wave related to three or more targets. .. Then, in the case of the reflected waves related to the two targets, the arrival direction is estimated in the maximum search range for both the first time, but the search range is the range corresponding to the estimated angle of each reflected wave after the second time. For example, the search range of the angle of the reflected wave related to one of the two targets is in the range of -14.0 to -13.0 [deg] for the second time and -13.74 for the third time. It is in the range of ~ -13.14 [deg]. The search range for the angle of the reflected wave related to the other target is the range of +12.0 to +13.0 [deg] for the second time and the range of +12.1 to +12.6 [deg] for the third time. Is.

As described above, in the arrival direction estimation process of the radar device according to the embodiment, the estimation process of step S6 is repeated while gradually reducing the range for searching the arrival direction of the reflected wave. As a result, the amount of calculation required for estimating the arrival direction can be reduced and the estimation time for the arrival direction can be significantly reduced as compared with the case where the estimation process of step S6 is repeated without changing the range for searching the arrival direction. it can. Hereinafter, a configuration example of the radar device according to the embodiment will be specifically described.

Further, in the arrival direction estimation process of the radar device according to the embodiment, the search interval in the search range for searching the arrival direction of each reflected wave can be reduced based on the amount of change in the estimated value in the arrival direction. As a result, for example, the search interval can be increased until the estimated value in the arrival direction approaches the true value, and the amount of calculation required for estimating the arrival direction is reduced as compared with the case where the search interval is constant. However, the estimated time in the arrival direction can be reduced.

[2. Radar device configuration example]
FIG. 2 is a diagram showing a configuration example of the radar device 1 according to the embodiment. As shown in FIG. 2, the radar device 1 includes a transmission unit 10, a reception unit 20, and a signal processing unit 30. Such a radar device 1 is an in-vehicle radar device mounted on a vehicle, and is connected to a vehicle control device 2 that controls the behavior of the vehicle.

The vehicle control device 2 controls a vehicle such as a PCS (Pre-crash Safety System) or an AEB (Advanced Emergency Braking System) based on the detection result of a target by the radar device 1. The radar device 1 may be used for various purposes other than the in-vehicle radar device (for example, monitoring of an airplane or a ship).

[2.1. Transmitter 10]
The transmission unit 10 includes a signal generation unit 11, an oscillator 12, and a transmission antenna 13. The signal generation unit 11 generates a modulated signal whose voltage changes in a triangular wave shape and supplies it to the oscillator 12. The oscillator 12 generates a transmission signal ST whose frequency changes with the passage of time based on the modulated signal generated by the signal generation unit 11, and outputs the transmission signal ST to the transmission antenna 13.

The transmission antenna 13 converts the transmission signal ST from the oscillator 12 into a transmission wave TW, and outputs the transmission wave TW in front of the vehicle on which the radar device 1 is mounted (hereinafter, may be referred to as a own vehicle). .. The transmitted wave TW output by the transmitting antenna 13 is an FM-CW whose frequency fluctuates in a predetermined cycle. The transmitted wave TW transmitted from the transmitting antenna 13 to the front of the own vehicle is reflected by a target such as another vehicle and becomes a reflected wave RW.

[2.2. Receiver 20]
Receiving unit 20, a plurality of receiving antennas ₂₁ 1 to 21 ₄ to form the array antenna (hereinafter, may be collectively referred to as the receiving antenna 21), a plurality of mixers 22, and a plurality of A / D converter 23 Be prepared. The mixer 22 and the A / D converter 23 are provided for each receiving antenna 21. A plurality of receiving antennas ₂₁ 1 to 21 ₄ are arranged at equal predetermined intervals d, 4 element uniform array as array antennas by the receiving antennas ₂₁ 1 to 21 ₄ are formed.

The plurality of receiving antennas ₂₁ 1 to 21 ₄ receives the reflected wave RW from the target, according to the reflected wave RW received signal _SR 1 to SR ₄ (hereinafter, may be collectively referred to as received signal SR) is converted into. The array antenna shown in FIG. 2 is a 4-element equidistant array, but the receiving antenna 21 may be an equidistant array of 3 or less, and the receiving antenna 21 may be an equidistant array of 5 or more. There may be.

The received signal SR output from the receiving antenna 21 is amplified by an amplifier (for example, a low noise amplifier) (not shown) and then input to the mixer 22. The mixer 22 mixes a part of the transmission signal ST and the reception signal SR to remove unnecessary signal components to generate a beat signal SB, and outputs the beat signal SB to the A / D converter 23.

As a result, a beat signal SB indicating a beat frequency that is the difference between the frequency of the transmission signal ST and the frequency of the reception signal SR is generated. The beat signal SB generated by the mixer 22 is converted into a digital signal by the A / D converter 23 and then output to the signal processing unit 30.

FIG. 3 is a diagram showing the relationship between the transmitted TW wave and the reflected wave RW and the beat signal SB. As shown in the upper part of FIG. 3, the transmitted wave TW is a continuous wave whose frequency fluctuates in a predetermined period around a predetermined frequency, and the frequency of the transmitted wave TW changes linearly with time. To do. Here, the center frequency of the transmitted wave TW is f0, the displacement width of the frequency is ΔF, and the reciprocal of one cycle in which the frequency goes up and down is fm.

Since the reflected wave RW is the one reflected by the transmitted wave TW with a target, it becomes a continuous wave whose frequency fluctuates in a predetermined period around a predetermined frequency as in the transmitted wave TW, and also becomes a transmitted wave TW. On the other hand, there is a delay. The delay time τ of the reflected wave RW with respect to the transmitted wave TW is a time corresponding to the vertical distance from the own vehicle to the target. Further, in the reflected wave RW, a frequency shift of the frequency fd occurs with respect to the transmitted wave TW due to the Doppler effect according to the relative speed of the target with respect to the own vehicle.

In this way, the reflected wave RW has a frequency shift with respect to the transmitted wave TW with a delay time τ according to the vertical distance and a frequency shift according to the relative velocity with the target. Therefore, as shown in the lower figure of FIG. 3, the beat frequency of the beat signal SB includes a section in which the frequency of the transmission signal ST rises (hereinafter, referred to as an up period) and a section in which the frequency falls (hereinafter, a down section). The values will be different from those described).

The beat frequency is the frequency of the difference between the frequency of the transmitted wave TW and the frequency of the reflected wave RW. Hereinafter, the beat frequency in the up section is referred to as pup, and the beat frequency in the down section is referred to as fdn. In the lower figure of FIG. 3, the vertical axis represents frequency [kHz] and the horizontal axis represents time [msec].

[2.3. Signal processing unit 30]
As shown in FIG. 2, the signal processing unit 30 includes a transmission control unit 31, a Fourier transform unit 32, a peak extraction unit 33, an angle measurement processing unit 34, and a distance / relative velocity calculation unit 35. The signal processing unit 30 is, for example, a microcomputer including a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), an input / output port, and the like, and controls the entire radar device 1.

When the CPU of the microcomputer reads and executes the program stored in the ROM, it functions as a transmission control unit 31, a Fourier transform unit 32, a peak extraction unit 33, an angle measurement processing unit 34, and a distance / relative velocity calculation unit 35. To do. The transmission control unit 31, the Fourier transform unit 32, the peak extraction unit 33, the angle measurement processing unit 34, and the distance / relative velocity calculation unit 35 are all such as ASIC (Application Specific Integrated Circuit) and FPGA (Field Programmable Gate Array). It can also be configured in hardware.

[2.3.1. Transmission control unit 31]
The transmission control unit 31 controls the signal generation unit 11 of the transmission unit 10 to output a modulated signal whose voltage changes in a triangular wave shape from the signal generation unit 11 to the oscillator 12. As a result, the transmission signal ST whose frequency changes with the passage of time is output from the oscillator 12 to the transmission antenna 13.

[2.3.3. Fourier transform unit 32]
The Fourier transform unit 32 performs a fast Fourier transform (FFT) process on each of the beat signals SB output from each of the plurality of A / D converters 23. As a result, the Fourier transform unit 32 converts the beat signal SB output from each of the plurality of A / D converters 23 into frequency spectrum data.

The frequency spectrum converted by the Fourier transform unit 32 is output to the peak extraction unit 33. The frequency spectrum data includes signal level information for each frequency bin set at frequency intervals according to the frequency resolution of the Fourier transform unit 32.

[2.3.3. Peak extraction unit 33]
In the frequency spectrum converted by the Fourier transform unit 32, the peak extraction unit 33 sets peaks exceeding a predetermined signal level as an up section in which the frequency of the transmission signal ST rises and a down section in which the frequency of the transmission signal ST falls. Extract in each section of. Hereinafter, the frequency extracted in this way is referred to as a "peak frequency".

Since the reflected wave RW from the same target is received by the four receiving antennas 21, the extracted peak frequencies are the same between the frequency spectra of the four beat signals SB. Further, when a plurality of targets are present at different angles in the same frequency bin, the signal of one peak frequency in the frequency spectrum contains information about the plurality of targets.

Since the positions of the four receiving antennas 21 are different from each other, the phases of the reflected waves RW received by the four receiving antennas 21 are different from each other. Therefore, the phase information of the received signal SR having the same frequency bin is different for each receiving antenna 21.

[2.3.4. Angle measuring unit 34]
The angle measuring unit 34 estimates the arrival direction of each reflected wave RW based on the signals corresponding to the plurality of received signals SR output from the plurality of receiving antennas 21.

From the signal of each distance bin having a peak of a predetermined value or more specified by the peak extraction unit 33 (hereinafter, referred to as a peak signal), the angle measuring unit 34 refers to a plurality of targets existing in the same distance bin. The information is separated, and the angle of each of the plurality of targets is estimated as the arrival direction of each reflected wave RW.

The azimuth calculation unit 46 pays attention to the peak signals of the same peak frequency bin in all the frequency spectra of the four beat signals SB based on the received signal SRs of the four receiving antennas 21, and each reflection is based on the phase information of the peak signals. Estimate the direction of arrival of the wave RW. Hereinafter, the peak signals corresponding to the received signals SR _{1 to SR 4} will be referred to as peak signals x1 to x4. Further, the angle measuring unit 34 estimates the arrival direction of each reflected wave RW with respect to the peak signals x1 to x4 in the up section and the peak signals x1 to x4 in the down section, respectively.

The angle measuring unit 34 estimates each arrival direction by using, for example, the KR-EM-DBF method as an arrival direction estimation method. In such a KR-EM-DBF method, the Katri-Rao product expansion array processing is performed to virtually increase the number of array elements and the like, and the value of the correlation component (correlation term) is calculated using the EM algorithm method. By changing it, it is possible to reduce the estimation error in the arrival direction.

FIG. 5 is a diagram showing a configuration example of the angle measuring unit 34. As shown in FIG. 5, the angle measurement processing unit 34 includes an extended signal generation unit 41, a parameter initial value calculation unit 42, and an arrival direction estimation unit 43.

[2.3.4.1. Extended signal generator 41]
The extended signal generation unit 41 generates the correlation matrix Rxx from the peak signals x1 to x4 based on the received signals SR output from each of the plurality of receiving antennas 21, and generates the extended signal z by the non-overlapping elements of the correlation matrix Rxx. ..

Here, the correlation matrix Rxx can be expressed as the following equation (1). The extended signal generation unit 41 can obtain the correlation matrix Rxx by the calculation of the following equation (1). In the following equation (1), X = [x1, x2, x3, x4], E [・] represents the unsampled average, and [・] ^H represents the complex conjugate transpose.

Next, the extended signal generation unit 41 extracts non-overlapping elements from the 4 × 4 matrix in the above equation (1), and as shown in the following equation (2), the Katri-Lao product extended array process (hereinafter, hereinafter, The extended signal z in (described as KR product extended array processing) is generated.

The KR product expansion array processing is a signal processing that virtually increases the number of receiving antennas. By such KR product expansion array processing, signal processing can be performed by a virtual antenna array in which (L-1) virtual receiving antennas are increased for several L of physical receiving antennas 21.

For example, when the antenna array is a 4-element equidistant array as shown in FIG. 2, the KR product expansion array processing is performed to obtain an antenna array corresponding to a 7 (2L-1) element equidistant array as shown in FIG. Can be handled. FIG. 4 is a diagram showing the arrangement relationship between the virtual antenna element (extended array) and the real antenna element (real array), which are increased by the KR product expansion array processing.

[2.3.4.2. Parameter initial value calculation unit 42]
The parameter initial value calculation unit 42 calculates an estimated value in the arrival direction of each reflected wave RW from the peak signals x1 to x4 or the expansion signal z calculated by the expansion signal generation unit 41. The estimated value of the arrival direction is used in the arrival direction estimation unit 43 as an initial value of the calculation parameters of the estimation extension signals z ^ and z _{k ^ described later.}

For example, the parameter initial value calculation unit 42 can estimate the arrival direction of each reflected wave RW from the peak signals x1 to x4 or the extended signal z based on the DBF method. Further, the parameter initial value calculation unit 42 estimates the arrival direction of each reflected wave RW from the peak signals x1 to x4 or the extended signal z by, for example, ESPRIT (Estimation of Signal Parameters via Rotational Invariance Techniques) or MUSIC (Multiple Signal Classification). You can also do it.

[2.3.4.3. Arrival direction estimation unit 43]
The arrival direction estimation unit 43 repeats the estimation process of updating the ^{calculation parameter Θ (l)} based on the difference Δz between the extension signal z and the estimation extension signal z ^. When the stop condition is satisfied, the arrival direction estimation unit 43 outputs the latest estimated value θ _k ^ ^{(l) included in the calculation parameter Θ (l)} as a definite estimated value θ _{k in the} arrival direction. Note that " ^(l) " indicates the number of operations for estimation processing based on a set of peak signals x1 to x4 acquired at a certain timing, and "k" indicates parameters such as the arrival direction of the reflected wave and the correlation component. It is a number to distinguish.

As shown in FIG. 5, the arrival direction estimation unit 43 includes an estimation extension signal generation unit 51, a maximum likelihood estimation value calculation unit 52, a parameter estimation unit 53, a stop condition determination unit 54, and a parameter update unit 55. Be prepared.

[2.3.4.3.1.1. Estimated extended signal generation unit 51]
The estimated extended signal generation unit 51 generates an estimated extended signal z ^ estimated as an extended signal z using the estimated value of the arrival direction of each reflected wave RW and the estimated value of the correlation component between the reflected waves RW as calculation parameters.

The estimation extended signal generation unit 51 uses the parameter initial value calculation unit 42 as the initial value of the calculation parameter to estimate the arrival direction of each reflected wave RW and the initial estimation value of the correlation component between the reflected waves RW (for example,). , Fixed value) and set. Here, the estimated value of the arrival direction of each reflected wave RW calculated by the parameter initial value calculation unit 42 is "θ _k ^ ⁽⁰⁾ ", and the initial estimated value of the correlation component is "φ _k ^ ⁽⁰⁾ ". To do. In addition, " ⁽⁰⁾ " indicates that it is an initial value.

Here, for the sake of simplicity, it is assumed that _{the two reflected waves RW, the first reflected wave RW 1} and the second reflected wave RW ₂ , are received by the receiving antenna 21. Further, the initial value of the arrival direction of the first reflected wave RW _{1 is θ 1} ^ ⁽⁰⁾ , the initial value of the arrival direction of the second reflected wave RW _{2 is θ 2} ^ ^(0), and the two reflected waves RW Let the initial value of the correlation component between ₁ and RW _{2 be φ 3} ^ ⁽⁰⁾ . In this case, the initial value Θ ⁽⁰⁾ of the operation parameter is expressed by the following equation (3). Note that φ ₃ ^ ⁽⁰⁾ is, for example, zero.

The estimation extended signal generation unit 51 calculates the estimated value s ^(l) of ^{the complex amplitude using the latest calculation parameter Θ (l)} by the calculation of the following equation (4). The estimated value s ^(l) includes, for example, the _{estimated value s 1} ^(l) of the complex amplitude of _{the first reflected wave RW 1 , the estimated value s 2} ⁽⁰⁾ of the complex amplitude of the second reflected wave RW ₂ , and the estimated value s 2 (0). _{, The estimated value s 3} ^(l) of the complex amplitude of the correlation component (correlation term) between the reflected waves RW ₁ and RW ₂ is included.

For example, in the case of the first calculation, l = 0, and the estimated extended signal generation unit 51 performs the calculation of the above equation (4) ^{based on the initial value Θ (0)} of the calculation parameter, and the reflected wave and the correlation component (correlation). ^{Item) The estimated value s (0} ) of the complex amplitude is obtained.

Next, the estimated extended signal generation unit 51 generates an estimated extended signal z ^ which is an estimated value of the extended signal z based on ^{the estimated value s (l) of the complex amplitude.} For example, the estimated extended signal generation unit 51 can generate the estimated extended signal z ^ by the calculation of the following equation (5).

^{In the estimation extended signal generation unit 51, every time the calculation parameter Θ (l)} is updated by the parameter update unit 55 until the stop condition determination unit 54 determines the end of the estimation process, the updated calculation parameter Θ ^{( Based on l)} , the complex amplitude estimate s ^(l) and the estimated extension signal z ^ are repeatedly updated.

[2.3.4.3.2.2. Maximum likelihood estimation value calculation unit 52]
_{The maximum likelihood estimation value calculation unit 52 determines the maximum likelihood extension signal z k} ^ for each of the direction of arrival of the reflected wave RW and the correlation component between the reflected waves RW based on the difference between the extension signal z and the estimated extension signal z ^. Find the maximum likelihood estimate.

The maximum likelihood estimation value calculation unit 52 uses, for example, the extended signal z acquired from the extended signal generation unit 41, the estimated extended signal z ^ acquired from the estimated extended signal generation unit 51, and the estimated value s ^(l) of the complex amplitude. The maximum likelihood estimation value of the estimated extension signal zk _{^ is obtained.}

Maximum likelihood estimation value calculation unit 52, for example, the estimated extension signal z _{k ^} maximum likelihood estimates of each of the correlation components between arrival direction and the reflected wave RW calculation by the reflected wave RW of the formula (6) ( Hereinafter, the maximum likelihood estimation value may be described as _{z k ^).}

In the above equation (6), " ^(l) " indicates the number of operations of estimation processing based on a set of peak signals x1 to x4 acquired at a certain timing, and "k" is a number for distinguishing parameters. Is. Further, " _ak ^(l) " indicates a mode vector, " _sk ^(l) " is an estimated value of complex amplitude, and "β _k " is an estimated coefficient.

Here, since it is assumed that there are two reflected wave RWs, the maximum likelihood estimation value calculation unit 52 performs the maximum likelihood estimation values in the arrival directions of the two reflected waves RW by the calculations of the following equations (7) to (9). _{The maximum likelihood estimation value z 3} ^ ^(l) of the correlation component (correlation term) between z ₁ ^ ^(l) and z ₂ ^ ^(l) and the two reflected waves RW is obtained.

In the above equations (7) to (9), "a ₁ ^(l) " is the mode vector in the arrival direction of the first reflected wave RW _{1 , and "a 2} ^(l) " is the second reflected wave RW _2. The mode vector in the direction of arrival, “a ₃ ^(l) ”, is the mode vector of the correlation component (correlation term) between the _{reflected waves RW 1} and RW _2. According to the above-mentioned extended mode unit vector a (θ), for example, a ₁ ^(l) = a (θ ₁ ^ ^(l) ), a ₂ ^(l) = a (θ ₂ ^ ^(l) ), a ₃ It can be expressed as ^(l) = a (φ ₁ ^ ^(l)).

Further, in the above equations (7) to (9), "s ₁ ^(l) " is the complex amplitude of the first reflected wave RW _{1 in} the arrival direction, and "s ₂ ^(l) " is the second reflected wave RW _2. The complex amplitude in the direction of arrival, “s ₃ ^(l) ”, is the complex amplitude of the correlation component between the _{reflected waves RW 1} and RW _2. Note that "β ₁ ", "β ₂ ", and "β ₃ " are preset values, for example, 1/3.

When the two reflected waves RW are received by the receiving antenna 21, the estimated extended signal z ^ ^(l) can be expressed by the following equation (10), assuming that there is no noise. In the following formula (10), the initial values of _{s 1} ^(l) , s ₂ ^(l) , and s ₃ ^(l) _{are s 1} ⁽⁰⁾ , s ₂ ⁽⁰⁾ , and s ₃ ^(0). It is obtained by the calculation of equation (4).

^{In the maximum likelihood estimation value calculation unit 52, the estimated value s (l) of the} complex amplitude and the estimated expansion signal z ^ are updated by the estimation extension signal generation unit 51 until the stop condition determination unit 54 determines the end of the estimation process. Each time, the maximum likelihood estimation value z _k ^ is repeatedly calculated based on the updated complex amplitude estimation value s ^{(l) and the estimation extension signal z ^.}

[2.3.4.3.3. Parameter estimation unit 53]
The parameter estimation unit 53 maximizes the likelihood of the estimated extended signal z _k ^ for the arrival direction of the reflected wave RW, respectively, and estimates the arrival direction of the reflected wave RW θ _k ^ (hereinafter, the estimated value θ _k ^ of the arrival direction). ) Is calculated. Further, the parameter estimation unit 53 calculates _{an estimated value φ k} ^ of the correlation component (hereinafter, referred to as a correlation component estimated value φ _k ^) that _{maximizes the likelihood of the estimated extended signal z k} ^ for the correlation component. ..

Here, since it is assumed that there are two reflected waves RW, the parameter estimation unit 53 maximizes the likelihoods of _{the maximum likelihood estimation values z 1} ^ ^(l) and z ₂ ^ ^{(l), respectively.} Calculate the _{correlation component estimates φ 3} ^ that maximize the likelihood of the RW arrival direction estimates θ ₁ ^ and θ ₂ ^ and the maximum likelihood estimates z ₃ ^ ^{(l), respectively.}

The parameter estimation unit 53 obtains the correlation matrix _kk ^(l) of _{each estimated extension signal z k} ^ ^(l) by, for example, the calculation of the following equation (11). For example, the maximum likelihood estimation value calculation unit 52 calculates the correlation matrices c ₁ ^{(l) to} c ₃ ^(l) of _{the estimated extension signals z 1} ^ ^{(l) to} _{z 3} ^ ^(l). In the following equation (11), [・] ^H represents a complex conjugate transpose. Further, the correlation matrices c ₁ ^{(l) to} c ₃ ^(l) are extended correlation matrices based on the Katri-Lao product.

Next, the parameter estimation unit 53 uses the following equations (12) and (13) to calculate the arrival direction estimated value θ _k ^ ^{(l + 1)} _{that maximizes the likelihood of each estimated extended signal z k} ^ ^(l). Calculate the correlation component estimate φ _k ^ ^{(l + 1).}

For example, the parameter estimation unit 53 maximizes the likelihood of _{each estimated extension signal z 1} ^ ^(l) , z ₂ ^ ^(l) , and z ₃ ^ ^(l) by the calculation of the following equations (14) to (16). The estimated arrival direction θ ₁ ^ ^{(l + 1)} and θ ₂ ^ ^{(l + 1)} and the estimated correlation component φ ₃ ^ ^{(l + 1)} are calculated.

When performing the operations of the above equations (12) to (16), the parameter estimation unit 53 can gradually reduce the range (search range) of the arrival direction θ to be searched according to the number of estimation processes and the like.

For example, the parameter estimation unit 53, as shown in FIG. 1B, in the first estimation process, maximizing the estimation extension signal _{z k} ^ likelihood of the scope of -21.8~ + 21.8 [deg] as the search range Calculate the estimated arrival direction θ ₁ ^ and θ _{2 ^.} The range from -21.8 to +21.8 [deg] is the maximum search range.

Then, the parameter estimation unit 53, the estimation processing of the second and subsequent, as change amount A _k of the last computed arrival direction estimation value theta _{k ^} is small, it is also possible to reduce the time of the search range. Arrival direction estimation value theta _{k ^} of variation A _k of the last is the difference between the arrival direction estimation value of the second preceding theta _{k ^} and the previous arrival direction estimation value theta _{k ^.}

In the second estimation process, the arrival direction estimated value θ _{k ^ two times before is “θ k} ^ ⁽⁰⁾ ” calculated by the parameter initial value calculation unit 42, and the previous arrival direction estimated value θ _k ^. Is the arrival direction estimated value θ _k ^ ⁽¹⁾ calculated by the parameter estimation unit 53. Therefore, _Ak = | θ _k ^ ⁽⁰⁾ −θ _k ^ ⁽¹⁾ |.

Further, in the third estimation process, the arrival direction estimated value θ _{k ^ two times before is “θ k} ^ ⁽¹⁾ ” calculated by the parameter estimation unit 53, and the previous arrival direction estimated value θ _k ^ is It is the arrival direction estimated value θ _k ^ ⁽²⁾ calculated by the parameter estimation unit 53. Therefore, _Ak = | θ _k ^ ⁽¹⁾ −θ _k ^ ⁽²⁾ |.

Parameter estimation unit 53, the estimation processing of the second and subsequent, for example, the previous computed arrival direction estimation value theta _{k ^,} on the basis of the change amount A _k of the arrival direction estimation value θ _{k ^,} θ _{k ^} - the range of the _{Ka × a k ~θ k ^ +} Ka × a k can be a search range. "Ka" is a coefficient.

Here, as an example, Ka = 5, A ₁ = 0.1, A ₂ = 0.1, θ ₁ ^ ⁽⁰⁾ = -13.5, θ ₂ ^ ⁽⁰⁾ = +12.5. To do. In this case, in the second estimation process, the parameter estimation unit 53 calculates the _{arrival direction estimated value θ 1} ^ ⁽¹⁾ with the range of -14.0 to -13.0 [deg] as the search range, and +12. _{The estimated arrival direction θ 2} ^ ⁽¹⁾ is calculated with the range of 0 to +13.0 [deg] as the search range.

In this way, it is assumed that the result of the second estimation process is θ ₁ ^ ⁽¹⁾ = -13.44 and θ ₂ ^ ⁽¹⁾ = + 12.35. In this case, _Ak = 0.06 and _Ak = 0.05. Accordingly, the parameter estimation unit 53, the estimation processing of the third, -13.74~-13.14 [deg] range estimation extension signal _{z k} ^ DOA estimation to maximize the likelihood of a search range in value theta ₁ ⁽²⁾ ^ computes, + 12.1～ + 12.6 estimate the range of [deg] as the search range extension signal _{z k} incoming maximize the likelihood of the ^ direction estimation value θ ^₂ ^ ^₍₂₎ Is calculated.

Thus, parameter estimation unit 53, which can change the search range in accordance with the change amount A _k each time the estimation process, the method of changing the search range is not limited to such an example. For example, the parameter estimation unit 53 may change the search range for each of a plurality of times of estimation processing according to the change amount A _k.

Since the higher the search range variation A _k is large increases, even if the estimated error as large, can avoid a situation can not be estimated, also, as the search range variation A _k is small Since it becomes smaller, it is possible to prevent the estimation accuracy from deteriorating.

The parameter estimation unit 53, regardless of the direction of arrival estimate theta _{k ^} of variation A _k, or by reducing the search range every time the estimation process in stages, the search range to a plurality of times for each of the estimation process It can be made smaller in stages.

Further, the parameter estimation unit 53 has a small variation range (for example, the difference between the maximum value and the minimum value _{) of the arrival direction estimated value θ k} ^ in the past estimation process (for example, the estimation process from the 1st to the Nth time). The search range can be made smaller.

Further, the parameter estimation unit 53 _{can reduce the search range as, for example, the number of arrival direction estimated values θ k} ^ to be estimated, that is, the number of detected reflected wave RWs increases. Further, the parameter estimation unit 53 can reduce the search range as the speed of the vehicle on which the radar device 1 is mounted becomes slower, for example.

In this way, the parameter estimation unit 53 performs estimation processing while gradually reducing the range for searching the arrival direction when calculating the arrival direction estimated value θ _{k ^ that maximizes the likelihood of the estimated extension signal z k ^.} Is repeated. As a result, the amount of calculation required for estimating the arrival direction can be reduced and the estimation time for the arrival direction can be significantly reduced as compared with the case where the estimation process is repeated without changing the range for searching the arrival direction.

Further, as in the case of the search range in the arrival direction, the parameter estimation unit 53 changes the range (search range) of the correlation component to be searched by the correlation component estimated value φ _k ^ when performing the calculation (13) above. It can be made smaller step by step according to the above. For example, the parameter estimation unit 53, or reduce the search range of the correlation component in accordance with the number of the correlation as component estimation value phi _{k ^} variation A _k is small, the estimation process or to reduce the search range of the correlation components can do.

In this way, the parameter estimation unit 53 performs estimation processing while gradually reducing the range for searching for the correlation component when calculating the correlation component estimation value φ _{k ^ that maximizes the likelihood of the estimation extension signal z k ^.} Is repeated. This also makes it possible to reduce the amount of calculation required for estimating the arrival direction and reduce the estimation time in the arrival direction.

Further, the parameter estimation unit 53 can reduce the search interval in the search range in the arrival direction and the search range of the correlation component according to the amount of change in the estimated value and the like.

For example, the parameter estimation unit 53 searches the search range by setting the search interval to 0.1 in the first estimation process. As a result, the parameter estimation unit 53 changes the arrival direction θ in the range of -21.8 to +21.8 [deg] shown in FIG. 1B at intervals of 0.1 [deg], for example, in the above equation (12). ) And (13) are performed to obtain _{the arrival direction estimated value θ k} ^ and the correlation component estimated value φ _{k ^.}

In the estimation process of second or subsequent parameter estimation unit 53 obtains the arrival direction estimate theta _{k ^} of variation A _{k. When Ak} > 0.2, the parameter estimation unit 53 sets the search interval to 0.1 and searches the search range. Further, when 0.002 ≦ _Ak ≦ 0.2, the parameter estimation unit 53 sets the search interval to “ _Ak / 2” and searches the search range. _Also, if it is A k <0.002, parameter estimating unit 53 sets the search interval "0.001", to search the search range.

The parameter estimation unit 53, to reduce the search space according to the correlation component estimate phi _{k ^} DOA estimates also theta _{k ^} if similar correlation component estimate phi _{k ^} variation B _k Can be done. The amount of change B _k is, for example, B _k = φ _k ^ ^(l) −φ _k ^ ^(l-1) .

In this way, the parameter estimation unit 53 can gradually reduce the search interval in the search range in the arrival direction and the search range of the correlation component according to the amount of change in the estimated value. That is, the parameter estimation unit 53 can reduce the search interval as the amount of change in the estimated value is smaller, thereby reducing the amount of calculation required for estimating the arrival direction and reducing the estimation time in the arrival direction. it can.

The parameter estimation unit 53 reduces the search interval according to the number of estimation processes, and the smaller the variation range of the estimated values in the past estimation process (for example, the difference between the maximum value and the minimum value), the more the search interval. Can also be made smaller. Further, the parameter estimation unit 53 can reduce the search interval as the search range becomes smaller.

Parameter estimation unit 53, until the end of the estimation processing by the stop condition determining unit 54 is determined, estimated extension signal by maximum likelihood estimation value calculation unit 52 z _k ^ ^(l) is each time it is updated, is such updated Based on the estimated extension signal z _k ^ ^(l) , the arrival direction estimated value θ _k ^ ^{(l + 1)} and the correlation component estimated value φ _k ^ ^{(l + 1)} are repeatedly calculated.

[2.3.4.3.3.4. Stop condition determination unit 54]
The stop condition determination unit 54 determines whether or not the stop condition is satisfied. Stop condition, the extension signal z and the estimated extension signal z ^ change difference Δz is or condition that converges within the predetermined range, the arrival direction estimation value theta _{k ^} variation A _k and correlation component estimates phi _{k ^} and The condition is that the quantity B _k is equal to or less than a predetermined value.

For example, the stop condition determination unit 54 determines that the stop condition is satisfied when the difference between the extended signal z and the estimated extended signal z ^ is continuously within a predetermined range by a predetermined number. Further, the stop condition determination unit 54 determines that the stop condition is satisfied when the amount of change in any of _{the arrival direction estimated values θ 1} ^ and θ ₂ ^ and the correlation component estimated value φ _{3 ^ is equal to or less than a predetermined value.} ..

When the stop condition determination unit 54 determines that the stop condition is satisfied, the stop condition determination unit 54 outputs the latest arrival direction estimation values θ ₁ ^ and θ ₂ ^ calculated by the parameter estimation unit 53 as final estimation values θ ₁ and θ ₂ . On the other hand, when the stop condition determination unit 54 determines that the stop condition is not satisfied, the stop condition determination unit 54 does not output _{the definite estimated values θ 1} and θ _2.

[2.3.4.4.3.5. Parameter update unit 55]
When the parameter update unit 55 determines that the stop condition determination unit 54 does not satisfy the stop condition, the parameter update unit 55 sets the latest arrival direction estimates θ ₁ ^ ^{(l + 1)} and θ ₂ ^ ^{(l + 1) calculated by the parameter estimation unit 53.} The estimated correlation component φ ₃ ^ ^{(l + 1)} is notified to the estimated extended signal generation unit 51 as an arithmetic parameter Θ ^{(l + 1).} In this way, the parameter update unit 55 can update the calculation parameters.

The estimation extension signal generation unit 51 calculates the estimation extension signal z ^ based on the calculation parameters updated by the parameter estimation unit 53. _{The maximum likelihood estimation value calculation unit 52 calculates new estimated extension signals z 1} ^, z ₂ ^, and z ₃ ^ based on the difference between the new estimated extension signal z ^ and the extension signal z. The parameter estimation unit 53 uses the arrival direction estimates θ ₁ ^ and θ ₂ ^ that maximize the likelihood of the new estimation extension signals z ₁ ^, z ₂ ^, and z ₃ ^ and the correlation component estimation values φ ₃ ^. Calculate.

Thus, the arrival direction estimation unit 43, until the end of the estimation processing by the stop condition determining unit 54 is determined to update the calculation parameter Θ to maximize the estimated extension signal z _{k ^} likelihood.

That is, the arrival direction estimation unit 43 _{repeats the calculation of the estimated extension signals z ^ and z k} ^ while changing the calculation parameter Θ so as to maximize the likelihood _{of the estimation extension signal z k} ^, and the reflection which is the calculation parameter. Maximum likelihood estimation of the wave arrival direction and the correlation component between the reflected waves is performed. As a result, the arrival direction estimation unit 43 can estimate the arrival direction of each reflected wave RW in consideration of the correlation component between the reflected waves RW, so that the estimation error of the arrival direction of the reflected wave RW can be reduced. it can.

[2.3.5. Distance / relative velocity calculation unit 35]
The distance / relative velocity calculation unit 35 determines the peak of the up section based on the degree of agreement between the definite estimated value of the arrival direction in the up section and the definite estimated value of the arrival direction in the down section estimated by the angle measuring unit 34. And pairing is performed to associate the peaks in the down section.

In this way, the distance / relative velocity calculation unit 35 associates peaks related to the same target with each other. As a result, the distance / relative velocity calculation unit 35 derives the target data related to each of the plurality of targets existing in front of the vehicle equipped with the radar device 1. Since such target data is obtained by associating two peaks with each other, it is also called "pair data".

The distance / relative velocity calculation unit 35 is an observation value (instantaneous value) of the relative velocity and distance to each target with respect to the radar device 1 (vehicle equipped with the radar device 1) from the peak of the paired up section and down period. Is derived. For example, the distance / relative velocity calculation unit 35 derives the observed value (instantaneous value) of the relative velocity from the difference between the peak frequencies of the two peak data of the up section and the down section, which are the sources of the target data (pair data). Then, the distance can be derived from the sum of the peak frequencies of these two peak data.

Further, the distance / relative velocity calculation unit 35 is a vehicle equipped with the radar device 1 from the average value _{of the definite estimated values θ k} of the two arrival directions of the up section and the down section, which are the sources of the target data (pair data). (Or the observed value (instantaneous value) of the angle of each target with respect to (or the radar device 1) is derived. The distance / relative speed calculation unit 35 outputs target information including the distance to the detected target, the relative speed, and the angle to the vehicle control device 2.

[3. Processing by the angle measuring unit 34]
Next, an example of the flow of processing executed by the angle measuring unit 34 will be described with reference to the flowchart. FIG. 6 is a flowchart showing an example of a processing procedure executed by the angle measuring unit 34, which is a process that is repeatedly executed.

As shown in FIG. 6, the angle measuring unit 34 _{acquires peak signals x1 to x4 based on received signals SR 1 to SR 4} (step S10), and generates a correlation matrix Rxx of peak signals x1 to x4 (step). S11). The angle measuring unit 34 generates the extended signal z using the correlation matrix Rxx (step S12).

The angle measuring unit 34 estimates the arrival direction of the reflected wave based on the peak signals x1 to x4 or the extended signal z (step S13). The estimated value in step S13 is an initial value of a calculation parameter in the estimation process. The angle measuring unit 34 calculates the estimated value s ^{(l) of the complex amplitude using the latest calculation parameter Θ (l)} (step S14), ^{and based on the estimated value s (l)} of the complex amplitude, the extended signal. An estimated extended signal z ^, which is an estimated value of z, is generated (step S15).

Next, the angle measuring unit 34 calculates the difference Δz between the extended signal z and the estimated extended signal z ^ (step S16). Then, the angle measuring unit 34 obtains the maximum likelihood estimation value _{of the estimated extended signal z k} ^ for each of the arrival direction of the reflected wave RW and the correlation component between the reflected waves RW based on the difference Δz (step S17). .. The angle measuring unit 34 _{obtains the correlation matrix kk} ^(l) of the maximum likelihood estimation value of _{the estimated extension signal z k} ^ (step S18), _{and based on the correlation matrix kk} (l), the estimated extension signal z _k ^(l). The arrival direction estimation value θ _k ^ that maximizes the likelihood of ^ and the correlation component estimation value φ _k ^ are calculated and updated (step S19).

The angle measuring unit 34 determines whether or not the stop condition is satisfied (step S20), and if it determines that the stop condition is not satisfied (step S20; No), the process proceeds to step S14. On the other hand, when it is determined that the stop condition is satisfied (step S20; Yes), the angle measuring unit 34 outputs the latest estimated value θ _k ^ in the arrival direction as a definite estimated value θ _k (step S21).

In the above example, the radar device 1 detects a target using FM-CW (Frequency Modulated-Continuous Wave), but the radar device 1 is an FCM (Fast Chirp Modulation) type radar device. You may. Further, the radar device 1 may be a radar device that detects a target by using a dual frequency CW (Continuous Wave), a spread spectrum method, or a pulse method.

Further, in the above-described example, the equidistant linear array has been described as an example, but the antenna array of the radar device 1 according to the embodiment is not limited to the equidistant linear array, and even if it is an unequal interval linear array. Good.

Incidentally, the extension signal generator 41, the peak signal x1 to x4 without the reception signal _SR 1 to SR ₄ output from the receiving antenna ₂₁ 1 to 21 ₄ to generate the correlation matrix Rxx as x1 to x4, such correlation matrix The extended signal z can also be generated by the non-overlapping elements of Rxx.

As described above, the radar device 1 according to the embodiment includes an extended signal generation unit 41 and an arrival direction estimation unit 43. The extended signal generation unit 41 is a receiving antenna based on the received _{signals SR 1 to SR 4} output from each of the _{plurality of receiving antennas 21 1 to} 21 ₄ included in the array antenna that receives the plurality of reflected waves RW by the plurality of targets. The correlation matrix Rxx is generated from the peak signals x1 to x4 for each of 21 (an example of signals for each antenna), and the extended signal z is generated by the non-overlapping elements of the correlation matrix Rxx. The arrival direction estimation unit 43 estimates the arrival direction of each reflected wave RW and the correlation component between the reflected wave RWs based on the difference between the extended signal z and the estimated extended signal z ^ which is the estimated value of the extended signal z. Is repeated while gradually reducing the search range in the arrival direction θ. In this way, since the search range of the arrival direction θ is gradually reduced, the amount of calculation required for estimating the arrival direction can be reduced, and the estimation time of the arrival direction can be reduced.

Further, the arrival direction estimation unit 43 includes an estimation extension signal generation unit 51, a maximum likelihood estimation value calculation unit 52, and a parameter estimation unit 53. The estimated extended signal generation unit 51 generates the estimated extended signal z ^. The maximum likelihood estimation value calculation unit 52 uses the arrival direction of the reflected wave RW and the correlation component between the reflected waves RW as parameters, and based on the difference Δz between the extended signal z and the estimated extended signal z ^, each arrival direction and the correlation component. Calculate the maximum likelihood estimate of the estimated extension signal z _{k ^ for each of.} The parameter estimation unit 53 maximizes the likelihood of the estimated extended signal z _k ^ in each direction of arrival, and the estimated value θ _k ^ in the direction of arrival and the correlation component that maximizes the likelihood of the estimated extended signal z _k ^ in the correlation component. Calculate the estimated value φ _k ^ of. The maximum likelihood estimation value calculation unit 52 uses the estimated value θ _k ^ in the arrival direction calculated by the parameter estimation unit 53 and the estimated value φ _k ^ of the correlation component as calculation parameters, and the estimated extension signal z _k ^ in each arrival direction. Maximum likelihood estimation value of and correlation component estimation Calculate the maximum likelihood estimation value of the _{extended signal z k ^.} Thus, it is possible to perform maximum likelihood estimation estimates the arrival direction theta _{k ^} and estimated value phi _{k ^} of the correlation components, can improve arrival direction estimation accuracy.

Further, the radar device 1 according to the embodiment includes a parameter initial value calculation unit 42. _{The parameter initial value calculation unit 42 calculates an estimated value θ k} ^ in the arrival direction of the reflected wave RW based on the peak signals x1 to x4 or the extended signal z. The maximum likelihood estimation value calculation unit 52 uses the estimated value θ _k ^ in the arrival direction calculated by the parameter initial value calculation unit 42 as the initial value of the calculation parameter, and the maximum likelihood estimation value _{of the estimated extension signal z k ^ in each arrival direction.} And the maximum likelihood estimation value of the estimated extension signal z _{k ^ of the correlation component are calculated.} As a result, the amount of calculation required for estimating the arrival direction can be reduced and the estimation time in the arrival direction can be reduced as compared with the case where the initial value of the calculation parameter is set to a fixed value.

Moreover, the arrival direction estimation unit 43 to estimate theta _{k ^} of variation A _k of the arrival direction falls below a predetermined threshold value, or the difference between the extended signal z and the estimated extension signal z ^ converges to a predetermined range The estimation process is repeated until. As a result, the estimated value θ _k ^ in the direction of arrival can be accurately determined.

Moreover, the arrival direction estimation unit 43, to reduce in accordance with the search range of the arrival direction estimation value theta _{k ^} of variation A _k of DOA. As a result, the amount of calculation required for estimating the arrival direction can be reduced and the estimation time in the arrival direction can be reduced as compared with the case where the search range is not changed.

Moreover, the arrival direction estimation unit 43, the search interval for searching the arrival direction estimate theta _{k ^} in the search range of DOA, reduced in accordance with the estimated value theta _{k ^} of variation A _k of DOA. As a result, the amount of calculation required for estimating the arrival direction can be reduced, and the estimation time in the arrival direction can be reduced more effectively.

Moreover, the arrival direction estimation unit 43 as the estimated value theta _{k ^} of variation A _k of the arrival direction is reduced, to reduce the search space of the arrival directions. As a result, the amount of calculation required for estimating the arrival direction can be reduced, and the estimation time in the arrival direction can be reduced more effectively.

Further effects and variations can be easily derived by those skilled in the art. For this reason, the broader aspects of the invention are not limited to the particular details and representative embodiments expressed and described as described above. Therefore, various modifications can be made without departing from the spirit or scope of the general concept of the invention as defined by the appended claims and their equivalents.

1 Radar device 34 Angle measurement processing unit 41 Extended signal generation unit 42 Parameter initial value calculation unit 43 Arrival direction estimation unit 51 Estimated extended signal generation unit 52 Maximum likelihood estimation value calculation unit 53 Parameter estimation unit 54 Stop condition judgment unit 55 Parameter update unit

Claims (7)

A correlation matrix is generated from the signals for each antenna based on the received signals output from each of the plurality of antennas included in the array antenna that receives a plurality of reflected waves by a plurality of targets, and is extended by the non-overlapping elements of the correlation matrix. An extended signal generator that generates signals and
The estimation process for estimating the arrival direction of each reflected wave and the correlation component between the reflected waves based on the difference between the extended signal and the estimated extended signal which is an estimated value of the extended signal is performed by stepping the search range in the arrival direction. It is equipped with an arrival direction estimation unit that repeats while making it smaller .
The arrival direction estimation unit is
Radar apparatus, characterized in that you reduced in accordance with the search range to the amount of change in the estimated value of the arrival direction. The arrival direction estimation unit is
An estimated extended signal generator that generates the estimated extended signal,
With the arrival direction of the reflected wave and the correlation component between the reflected waves as parameters, the maximum likelihood of the estimated expansion signal for each of the arrival directions and the correlation component is based on the difference between the expansion signal and the estimated expansion signal. Maximum likelihood estimation value calculation unit that calculates the estimated value, and
Parameter estimation for calculating the estimated value of the coming direction that maximizes the likelihood of the estimated extended signal in each arrival direction and the estimated value of the correlation component that maximizes the likelihood of the estimated extended signal of the correlation component. With a department,
The maximum likelihood estimation value calculation unit is
Using the estimated value of the arrival direction and the estimated value of the correlation component calculated by the parameter estimation unit as the parameters, the maximum likelihood estimation value of the estimated extension signal in each arrival direction and the estimated extension signal of the correlation component. The radar device according to claim 1, wherein the maximum likelihood estimation value of is calculated. A correlation matrix is generated from the signals for each antenna based on the received signals output from each of the plurality of antennas included in the array antenna that receives a plurality of reflected waves by a plurality of targets, and is extended by the non-overlapping elements of the correlation matrix. An extended signal generator that generates signals and
The estimation process for estimating the arrival direction of each reflected wave and the correlation component between the reflected waves based on the difference between the extended signal and the estimated extended signal which is an estimated value of the extended signal is performed by stepping the search range in the arrival direction. It is equipped with an arrival direction estimation unit that repeats while making it smaller.
The arrival direction estimation unit is
The search interval for searching the estimated value in the arrival direction in the search range is reduced according to the amount of change in the estimated value in the arrival direction.
A radar device characterized by that. The arrival direction estimation unit is
An estimated extended signal generator that generates the estimated extended signal,
With the arrival direction of the reflected wave and the correlation component between the reflected waves as parameters, the maximum likelihood of the estimated expansion signal for each of the arrival directions and the correlation component is based on the difference between the expansion signal and the estimated expansion signal. Maximum likelihood estimation value calculation unit that calculates the estimated value, and
Parameter estimation for calculating the estimated value of the coming direction that maximizes the likelihood of the estimated extended signal in each arrival direction and the estimated value of the correlation component that maximizes the likelihood of the estimated extended signal of the correlation component. With a department,
The maximum likelihood estimation value calculation unit is
Using the estimated value of the arrival direction and the estimated value of the correlation component calculated by the parameter estimation unit as the parameters, the maximum likelihood estimation value of the estimated extension signal in each arrival direction and the estimated extension signal of the correlation component. Calculate the maximum likelihood estimate of
The radar device according to claim 3, wherein the radar device is characterized by the above. The arrival direction estimation unit is
The radar device according to claim 3 or 4 , wherein the search interval is reduced as the amount of change in the estimated value in the arrival direction becomes smaller. A correlation matrix is generated from the signals for each antenna based on the reception signals output from each of the plurality of antennas included in the array antenna that receives a plurality of reflected waves by a plurality of targets, and is extended by the non-overlapping elements of the correlation matrix. The first step of generating a signal and
The estimation process for estimating the arrival direction of each reflected wave and the correlation component between the reflected waves based on the difference between the extended signal and the estimated extended signal which is an estimated value of the extended signal is performed by stepping the search range in the arrival direction. only contains a second step, the carried out while repeatedly to small,
The second step is
An arrival direction estimation method, characterized in that the search range is reduced according to the amount of change in the estimated value in the arrival direction. A correlation matrix is generated from the signals for each antenna based on the reception signals output from each of the plurality of antennas included in the array antenna that receives a plurality of reflected waves by a plurality of targets, and is extended by the non-overlapping elements of the correlation matrix. The first step of generating a signal and
The estimation process for estimating the arrival direction of each reflected wave and the correlation component between the reflected waves based on the difference between the extended signal and the estimated extended signal which is an estimated value of the extended signal is performed by stepping the search range in the arrival direction. Including the second step, which is repeated while making the size smaller,
The second step is
The search interval for searching the estimated value in the arrival direction in the search range is reduced according to the amount of change in the estimated value in the arrival direction.
A method of estimating the direction of arrival, which is characterized by the fact that.

Images