A broadband millimeter wave flow velocity measurement radar device and a high-precision direction finding method

Abstract

The present invention discloses a broadband millimeter wave velocity measurement radar device and a high-precision direction-finding method, which are mainly used for high-precision measurement of river flow data. The device comprises a transmitting millimeter wave array antenna, a receiving millimeter wave array antenna, a transmitting filter, a power amplifier, a phase shifter, a transmitting mixer, a signal transmitter, a receiving filter, a receiving amplifier, an orthogonal mixer, an analog-to-digital converter, a signal processor, network transmission and control and other key components. The high-precision direction-finding method of the present invention is particularly targeted at the broadband signal of the millimeter wave array, and realizes high-precision direction of arrival (DOA) estimation. The method converts the DOA estimation problem into a sparse reconstruction problem by constructing a dictionary matrix of different frequency points and adopting sparse reconstruction technology, and solves it by an iterative method, thereby effectively improving the accuracy of azimuth estimation. The technical solution of the present invention not only improves the accuracy of velocity measurement, but also realizes all-round velocity measurement, and has broad application prospects.

Description

Technical Field

[0001] The invention relates to the technical field of radar detection and parameter estimation, in particular to a broadband millimeter wave flow velocity measurement radar device and a high-precision direction finding method. Background Art

[0002] In order to achieve high-precision river flow data, in addition to the current contact flow measurement method, the non-contact side-scan radar method has attracted attention in the industry in recent years. The side-scan radar with an operating frequency lower than 1GHz can achieve basic river surface velocity measurement, but due to the low operating frequency and narrow bandwidth, the velocity measurement accuracy is poor. It is difficult to meet the needs of high-precision flow testing, and higher-precision flow testing equipment is urgently needed. In the process of river surface velocity testing, electromagnetic waves are first emitted to hit the water surface, and then the surface velocity of the river is estimated through the echo signal of the electromagnetic wave. The key process is the direction of arrival (DOA) of the reflected electromagnetic wave. This direction information can be used to accurately obtain the position of the reflected electromagnetic wave, thereby obtaining the corresponding velocity information on the river surface at different positions, and realizing a full-scale estimation of the surface velocity of the river. In order to improve the estimation performance of the river surface velocity, it is urgent to realize a velocity sensing device based on broadband signals and a high-precision signal direction estimation method of broadband signals. Summary of the invention

[0003] In order to solve the deficiencies of the prior art, the present invention proposes a broadband millimeter-wave flow velocity measurement radar device and a high-precision direction finding method. The present invention adopts the following technical scheme: a broadband millimeter-wave flow velocity measurement radar device, including a transmitting millimeter-wave array antenna, a receiving millimeter-wave array antenna, a signal transmitter, and a signal processor. The signal transmitter is connected to the transmitting millimeter-wave array antenna through a transmitting mixer, a phase shifter, a power amplifier, and a transmitting filter in sequence; the receiving millimeter-wave array antenna is connected to the signal processor through a receiving filter, a receiving amplifier, an orthogonal mixer, and an analog-to-digital converter in sequence; the signal processor completes the control of the radar transmission signal and the processing of the received signal. Furthermore, the transmitting millimeter-wave array antenna and the receiving millimeter-wave array antenna are different antenna arrays, and the isolation between the transmitting and receiving antennas reaches more than 40dB. Furthermore, the signal processor is composed of FPGA or DSP. Furthermore, the control of the signal processor transmitting the signal realizes the control of the generation of the broadband millimeter-wave RF signal and synchronizes it with the receiving signal processing, and realizes the sampling, filtering, downsampling, storage, flow velocity inversion calculation, and signal echo direction of arrival (DOA) calculation functions of the received signal during the receiving processing.

[0004] The high-precision direction finding method using the above-mentioned broadband millimeter-wave flow velocity measurement radar device is characterized by comprising the following steps:

[0005] Step 1: Enter basic information:

[0006] Array antenna receiving signal , number of antennas , the number of incoming signals , signal bandwidth , Maximum carrier frequency ;

[0007] Step 2: Initialization:

[0008] Signal Index , discretization spatial angle , margin parameter , support set ;

[0009] Step 3: Algorithm steps:

[0010] 1. For receiving signals Perform DFT (discrete Fourier transform) to obtain the frequency domain signal ;

[0011] 2. Construct dictionary matrices for different frequency points ;

[0012] 3. Formulate the DOA estimation problem as an optimization problem

[0013] ;

[0014] 4. Repeat algorithm steps 1-3 until all incoming signals are traversed ;

[0015] 5. Determine the maximum value index ;

[0016] 6. Update support set: ;

[0017] 7. Calculate the orthogonal projection residual of the frequency domain signal ;

[0018] 8. Based on the current support set , output DOA estimation result ;

[0019] Step 4: Output DOA estimation results .

[0020] For the broadband signal of the millimeter wave array, high-precision wave direction estimation is achieved. When the spacing between the antennas is , the number of antennas is , when there are multiple incoming wave signals at the same time, assuming that the number of incoming wave signals is , No. indivual The incoming signal can be expressed as , then indivual The received signal of the antenna can be expressed as

[0021] ;

[0022] in, represents additive white Gaussian noise, Indicates The antenna receives the The time delay of the incoming signal relative to the 0th antenna.

[0023] After discrete Fourier transform in step 1 of the algorithm, The frequency domain signal of frequency points can be expressed as

[0024] ;

[0025] in, is the frequency domain representation of the noise signal, For the signal direction, the steering vector of the broadband signal is

[0026] ;

[0027] The frequency domain expression of the incoming signal is:

[0028] ;

[0029] in, is the sampling period, is the number of sampling points.

[0030] In step 2 of the algorithm, based on the received signal model, high-precision signal direction estimation based on sparse reconstruction is achieved, and different dictionary matrices are constructed for different frequency points.

[0031] ;

[0032] in, represents the discretized spatial domain angle, is the number of discretized angles. Since each frequency point signal contained in the broadband signal contains the same azimuth information, the signal azimuth estimation problem can be transformed into the following sparse reconstruction problem:

[0033] ;

[0034] in, is a sparse vector, which has non-zero values ​​only at the discretized spatial angle position corresponding to the target orientation, and the rest are zero. is the regularization coefficient.

[0035] In step 5 of the algorithm, an iterative method is used to solve the problem. The support set is updated in the iterative process as follows:

[0036] ;

[0037] in, is the received signal margin.

[0038] The beneficial technical effects achieved by adopting the above technical solution are:

[0039] The present invention transforms the DOA estimation problem into a sparse reconstruction problem by constructing dictionary matrices of different frequency points and adopting sparse reconstruction technology, and solves it by an iterative method, thereby effectively improving the accuracy of azimuth estimation.

[0040] Simulation results show that the proposed method exhibits superior DOA estimation performance under different signal-to-noise ratios, numbers of antennas, bandwidths, snapshot numbers and grid spacing conditions, especially in terms of broadband characteristics, which is significantly better than the existing algorithms.

[0041] The technical solution of the present invention not only improves the accuracy of flow velocity measurement, but also can realize all-round flow velocity measurement because the device is easy to install, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] FIG1 is a schematic diagram showing the composition of a broadband millimeter-wave flow velocity measurement radar device.

[0043] Figure 2 is a schematic diagram of the array receiving signal.

[0044] Figure 3 shows the comparison of DOA estimation performance under different SNR conditions.

[0045] Figure 4 shows the comparison of DOA estimation performance under different numbers of antennas.

[0046] Figure 5 shows the comparison of DOA estimation performance under different bandwidth conditions.

[0047] Figure 6 shows the comparison of DOA estimation performance under different snapshot numbers.

[0048] Figure 7 shows the comparison of DOA estimation performance under different grid spacing conditions.

[0049] In the figure, 101 transmits millimeter wave array antenna, 103 receives millimeter wave array antenna, 102 multiple antenna units, 201 transmits filter, 202 power amplifier, 203 phase shifter, 204 transmits mixer, 205 signal transmitter, 206 receives filter, 207 receives amplifier, 208 orthogonal mixer, 301 analog-to-digital converter, 302 signal processor, 303 network transmission and control module. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present invention are clearly and completely described in conjunction with Figures 1 to 7. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] A broadband millimeter wave velocity measurement radar device comprises a transmitting millimeter wave array antenna 101, a receiving millimeter wave array antenna 103, a signal transmitter 205, and a signal processor 302. The signal transmitter is connected to the transmitting millimeter wave array antenna 101 through a transmitting mixer 204, a phase shifter 203, a power amplifier 202, and a transmitting filter 201 in sequence; the receiving millimeter wave array antenna 103 is connected to the signal processor 302 through a receiving filter 206, a receiving amplifier 207, an orthogonal mixer 208, and an analog-to-digital converter 301 in sequence; the signal processor completes the control of the radar transmitting signal and the processing of the receiving signal. The signal processor has a built-in network transmission and control module 303.

[0052] The transmitting millimeter wave array antenna 101 is composed of a plurality of antenna units 102. The transmitting millimeter wave array antenna and the receiving millimeter wave array antenna are different antenna arrays, and the isolation between the transmitting and receiving antennas reaches more than 40 dB.

[0053] The signal processor is composed of FPGA or DSP. The control of the signal processor transmits the signal to realize the control of the generation of broadband millimeter-wave RF signals and synchronizes with the received signal processing. In the receiving process, the received signal is sampled, filtered, downsampled, stored, flow velocity inversion calculation, and signal echo direction of arrival (DOA) calculation functions are realized.

[0054] For the linear array shown in Figure 2, the spacing between antennas is , the number of antennas is , when there are multiple incoming wave signals at the same time, assuming that the number of incoming wave signals is , No. indivual The incoming signal can be expressed as The first indivual The received signal of the antenna can be expressed as

[0055] ;

[0056] in, represents additive white Gaussian noise, Indicates The antenna receives the The time delay of the incoming signal relative to the 0th antenna can be expressed as

[0057] ;

[0058] in, represents the propagation speed of electromagnetic waves, Indicates The direction of the incoming signal is DOA. Usually, the incoming signal can be expressed as ,in is the equivalent complex baseband signal,

[0059] is the carrier frequency (the highest frequency of the signal). When the received signal is narrowband, the signals received by different antennas in the array meet the condition

[0060] ;

[0061] At this time, different time-delay signals can be expressed as

[0062] ;

[0063] Therefore, the time delay of a narrowband signal can be equivalent to the phase difference of signals received by different antennas, and DOA estimation can be completed by solving this phase difference.

[0064] However, when the signal bandwidth is wide, the narrowband signal assumption is not satisfied, that is,

[0065]

[0066] It is not possible to use the phase information in the narrowband signal to estimate the DOA of the signal. When the first Moment The received signal can be expressed as

[0067] ;

[0068] definition and , and after down-conversion, the received signal can be simplified to

[0069] ;

[0070] After collecting the received signals from different antennas to form a vector, the received signal can be defined as

[0071] ;

[0072] After DFT, the frequency domain signal of the pth frequency point can be expressed as

[0073] ;

[0074] The frequency domain signals of all antennas are expressed as vectors, which can be expressed as

[0075] ;

[0076] Among them, we define the frequency domain expression of the incoming signal after DFT as

[0077] ;

[0078] Typically, for complex sampled signals, the signal bandwidth With sampling frequency Similarly, the received signal can be further simplified as

[0079] ;

[0080] The steering vector of the broadband signal is defined as

[0081] ;

[0082] In summary, the frequency domain model of the received signal in ULA , the present invention will realize the direction of the incoming wave and fully exploit the characteristics of broadband signals.

[0083] From the ULA broadband receiving signal model (9), it can be seen that broadband signals in the array will cause different phase changes between multiple antennas. If the traditional narrowband signal DOA estimation algorithm is used directly, the difference between the narrowband steering vector and the broadband steering vector will cause deviations in DOA estimation. To address this problem, this paper proposes a broadband DOA estimation method based on sparse reconstruction to improve the DOA estimation accuracy under broadband conditions. Different from the current broadband DOA algorithm based on sparse reconstruction, this method improves the DOA estimation accuracy by mining the same DOA estimation information of signals at different frequencies, that is, the DOA information at different frequencies in the broadband signal is the same.

[0084] First, different dictionary matrices can be constructed for different frequency points, where The dictionary matrix of frequency points can be expressed as

[0085] ;

[0086] in, represents the discretized spatial domain angle, is the number of discretized angles. frequency points, the DOA estimation problem can be transformed into the following sparse reconstruction problem:

[0087] ;

[0088] in, is a sparse vector, which has non-zero values ​​only at the discretized spatial angle position corresponding to the target orientation, and the rest are zero. is the regularization coefficient.

[0089] Secondly, since the DOA information of each frequency point signal contained in the broadband signal is the same, data of different frequency points can be collected to form the following optimization problem:

[0090] ;

[0091] If the existing coefficient reconstruction algorithm is directly used to solve the optimization problem, it will lead to sparse vector The support set of is different, and it is impossible to achieve an accurate DOA estimation algorithm. Therefore, this paper proposes a new sparse reconstruction algorithm. The specific algorithm is shown in Table 1, which realizes the sparse vector The same support set is solved to complete the DOA estimation of broadband signals.

[0092]

[0093]

[0094] The performance of the proposed wideband sparse DOA estimation algorithm is simulated and verified. The simulation parameters are shown in Table 2. Different wideband signals are received and the method proposed in this paper is used to estimate their angle of arrival. The simulation parameters in Table 2 are used unless otherwise specified.

[0095] Table 2 System simulation parameter settings

[0096] parameter Value Number of antennas 16 Carrier frequency 1 GHz bandwidth 200 MHz Antenna spacing Half wavelength Number of sampling points 10 Number of incoming signals 3 Monte Carlo simulation times <![CDATA[10 3 ]]> Space angle interval 0.1° Received signal SNR 5 dB

[0097] First, we simulated and compared the DOA estimation performance under different receiving signal-to-noise ratio (SNR) conditions, such as Figure 3 As shown in the figure, it can be found that when SNR ≥ -10dB, the DOA estimation performance of the proposed method is better than that of the MUSIC (Multiple Signal Classification) algorithm and the OMP (Orthogonal Matching Pursuit) algorithm. At this time, the DOA estimation performance is measured by the root mean square error (RMSE), which is defined as follows:

[0098] ;

[0099] in, represents the number of Monte Carlo simulations, Indicates The actual DOA of the signal during the simulation, Indicates The DOA estimation result of the simulation. Since 10 sampling points are used, the SNR of the received signal will be increased by 10dB in theory. Therefore, when SNR≥-5dB, the improvement of SNR brings limited improvement in DOA estimation accuracy. It can be found from the figure that when SNR≥-5dB, the DOA estimation error of the proposed method is within 1°, while the estimation error of the current algorithm is more than 5°, so the proposed method can significantly improve the DOA estimation accuracy.

[0100] Secondly, when SNR=5dB, we simulated and compared the DOA estimation performance under different numbers of antennas, as shown in Figure 4. It can be found from the figure that with the increase in the number of antennas, the proposed method can further improve the DOA estimation accuracy. When the number of antennas increases from 6 to 32, the DOA estimation error is reduced from 1.3° to 0.093°, so the DOA estimation accuracy is significantly improved. The current existing algorithms do not improve the DOA estimation accuracy. On the contrary, with the increase in the number of antennas, the DOA estimation accuracy is gradually deteriorating. This is because more antennas will increase the aperture of the antenna array, and the signal delay at the head and tail receiving antennas will be larger, which is not conducive to the reception of broadband signals. Therefore, when the number of antennas increases from 6 to 32, the DOA estimation error also deteriorates from 4.8° to 15.13°.

[0101] Next, in order to accurately describe the impact of different broadband signals on DOA estimation performance, as shown in Figure 5, we gradually increase the bandwidth of the received signal from 100MHz to 800MHz. It can be seen from the figure that for the method proposed in this paper, the increase in bandwidth does not significantly affect the accuracy of DOA estimation, and the DOA estimation error gradually deteriorates from 0.25° at 100MHz to 0.38° at 800MHz. However, the DOA estimation performance of the existing algorithms will deteriorate significantly with the increase of bandwidth. It can be seen from the figure that the DOA estimation error of the MUSIC algorithm gradually deteriorates from 2.57° of the 100MHz bandwidth signal to 24.42° at 800MHz, and the OMP algorithm gradually deteriorates from 2.52° to 27.16°. Therefore, from the simulation results, it can be found that the method proposed in this paper has better broadband characteristics.

[0102] In addition, for the DOA estimation performance under different snapshot conditions, Figure 6 shows the simulation results. It can be found that when the snapshots gradually increase from 10 to 103, the DOA estimation error of the proposed method gradually decreases from 0.27° to 0.041°. Therefore, more snapshots are beneficial to the improvement of the estimation performance of the proposed method. However, due to the insufficient broadband characteristics of the MUSIC algorithm and the OMP algorithm, even if the number of snapshots is increased, the DOA estimation performance will not be significantly improved.

[0103] Finally, since the spatial domain needs to be divided into grids for spatial spectrum estimation in the DOA estimation process, we analyzed the impact of different grid divisions on the DOA estimation performance, as shown in Figure 7. It can be found that when the grid interval gradually increases from 0.1° to 1°, the DOA estimation accuracy of the proposed method gradually deteriorates from 0.25° to 0.39°. Therefore, more detailed grid division in the DOA estimation process can significantly improve the DOA estimation accuracy. However, due to the insufficient broadband characteristics of the MUSIC algorithm and the OMP algorithm, when the grid division accuracy is improved, the DOA estimation accuracy will not be significantly improved.

[0104] The broadband millimeter-wave velocity measurement radar device and high-precision direction-finding method of the present invention achieve a wider signal bandwidth in the millimeter-wave frequency band, thereby ensuring the estimation accuracy of the velocity. In view of the broadband characteristics of millimeter waves, a high-precision direction-finding method based on broadband signals is formed, thereby improving the azimuth estimation accuracy of the incoming wave signal. Therefore, it has higher river surface velocity estimation performance, is easy to install, and can more conveniently realize all-round velocity measurement.

[0105] Of course, the above description is only a preferred embodiment of the present invention, and the present invention is not limited to the above embodiments. It should be noted that all equivalent substitutions and obvious deformation forms made by any technician familiar with the field under the guidance of this specification fall within the essential scope of this specification and should be protected by the present invention.

Claims

1. A high-precision direction finding method for a broadband millimeter-wave flow velocity measurement radar device, characterized in that: The steps include: Step 1: Enter basic information: Array antenna receiving signal ,in , number of antennas N, number of incoming signals K, signal bandwidth B, maximum carrier frequency ; Step 2: Initialization: Signal Index , discretization spatial angle ,in , the margin parameter of the pth frequency point , support set ; Step 3: Algorithm steps: S1. For receiving signals Perform DFT (discrete Fourier transform) to obtain the frequency domain signal ; S2. Construct dictionary matrices for different frequency points ; S3. Formulate the DOA (signal echo direction) estimation problem as an optimization problem ； S4. Determine the maximum value index ; S5. Update support set: ; S6. Calculate the orthogonal projection residual of the frequency domain signal ; S7. Iterate the calculation steps S4 to S6 until K signals are calculated; S8. Based on the current support set , output DOA estimation result ; Step 4: Output DOA estimation results ; in, is the number of sampling points, is the number of discretized angles, P is the number of frequency points, p is the pth frequency point, is a sparse vector, is the regularization coefficient.

2. The high-precision direction finding method of the broadband millimeter-wave flow velocity measurement radar device according to claim 1, characterized in that: For the broadband signal of the millimeter wave array, high-precision wave direction estimation is achieved. When the spacing between the antennas is , the number of antennas is , when there are multiple incoming wave signals at the same time, assuming that the number of incoming wave signals is , No. The incoming signal is expressed as ,in , then The received signal of the antenna is expressed as ； in, , represents additive white Gaussian noise, Indicates The antenna receives the The time delay of the incoming signal relative to the 0th antenna.

3. The high-precision direction finding method of the broadband millimeter-wave flow velocity measurement radar device according to claim 1, characterized in that: After the discrete Fourier transform in step S1 of the algorithm, The frequency domain signal of a frequency point is expressed as ； in, is the frequency domain representation of the noise signal, For the signal direction, the steering vector of the broadband signal is ； The frequency domain expression of the incoming signal is: ； in, is the sampling period, is the number of sampling points, c is the propagation speed of electromagnetic waves, and d is the antenna spacing.

4. The high-precision direction finding method of the broadband millimeter-wave flow velocity measurement radar device according to claim 3, characterized in that: In step S2 of the algorithm, based on the received signal model, high-precision signal direction estimation based on sparse reconstruction is realized, and different dictionary matrices are constructed for different frequency points. ； in, represents the discretized spatial domain angle, where , is the number of discretized angles. Since each frequency point signal contained in the broadband signal contains the same azimuth information, the signal azimuth estimation problem is transformed into the following sparse reconstruction problem: ； in, is a sparse vector, which has non-zero values ​​only at the discretized spatial angle position corresponding to the target orientation, and the rest are zero. is the regularization coefficient.

5. The high-precision direction finding method of the broadband millimeter-wave flow velocity measurement radar device according to claim 1, characterized in that: In step S5 of the algorithm, an iterative method is used to solve the problem. The support set is updated in the iterative process as follows: ； in, is the received signal margin.

6. A broadband millimeter wave flow velocity measurement radar device, characterized in that: The invention is used to implement the high-precision direction finding method described in any one of claims 1 to 5, comprising a transmitting millimeter wave array antenna, a receiving millimeter wave array antenna, a signal transmitter, and a signal processor, wherein the signal transmitter is connected to the transmitting millimeter wave array antenna via a transmitting mixer, a phase shifter, a power amplifier, and a transmitting filter in sequence; the receiving millimeter wave array antenna is connected to the signal processor via a receiving filter, a receiving amplifier, an orthogonal mixer, and an analog-to-digital converter in sequence; and the signal processor completes the control of radar transmitting signals and the processing of receiving signals.

7. A broadband millimeter wave flow velocity measurement radar device according to claim 6, characterized in that: The transmitting millimeter wave array antenna and the receiving millimeter wave array antenna are different antenna arrays, and the isolation between the transmitting and receiving antennas reaches more than 40dB.

8. A broadband millimeter wave flow velocity measurement radar device according to claim 6, characterized in that: The signal processor is composed of FPGA or DSP.

9. The broadband millimeter wave flow velocity measurement radar device according to claim 6, characterized in that: The signal processor controls the transmission of signals to achieve control over the generation of broadband millimeter-wave radio frequency signals and synchronizes the processing of received signals, and implements sampling, filtering, downsampling, storage, flow velocity inversion calculation, and signal echo direction of arrival (DOA) calculation functions for received signals during the receiving processing process.

Images