A high resolution imaging radar array design method

Abstract

The application relates to a high-resolution imaging radar array design method, the imaging radar has a plurality of transmitting array elements and a plurality of receiving array elements, and comprises the following steps: setting a two-dimensional region for arranging a plurality of antenna array elements; according to an unambiguous angle measurement range, setting grid intervals in the azimuth and elevation directions, and dividing the two-dimensional region into a two-dimensional gridded region; limiting an antenna array element placement region and the number of antenna array elements in the region; adopting a particle swarm optimization algorithm to perform MIMO virtualization on the position distribution of the transmitting array elements and the receiving array elements of the two-dimensional gridded region, form a virtual two-dimensional array, and preset a target maximum sidelobe value of the virtual two-dimensional array to be not greater than T0; and optimizing the positions of the transmitting and receiving array elements by taking the maximum sidelobe of an array pattern as a target function. The imaging radar designed by adopting the method has the advantages of low cost, high resolution and low sidelobe value.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of millimeter wave radar, in particular to a high-resolution imaging radar array design method. BACKGROUND

[0002] Millimeter wave radar is an indispensable key component of intelligent driving (ADAS) system. In the process of driving, through the detection of surrounding motor vehicles, non-motor vehicles, pedestrians and related road conditions, the distance, speed, direction and angle of the surrounding target objects are obtained to provide information for the ADAS system to make decisions. In particular, it has strong all-weather working ability, which is an irreplaceable advantage of laser and video sensors for light perception, so it plays an irreplaceable role in ADAS systems and future unmanned driving.

[0003] The current vehicle-mounted radar has the problem of low angle resolution capability, and the imaging capability cannot meet the needs of intelligent driving, such as the two times of Tesla colliding with the front passing vehicles in Florida. This makes the current high-level automatic driving system still configured with laser radar, but the laser radar has the problems of insufficient environmental adaptability and difficulty in meeting the vehicle requirements, which limits the development and application of intelligent driving technology. The L3 and above level automatic driving system puts forward the requirements of two-dimensional high angle resolution capability of azimuth and pitch, and the point cloud target output capability of laser radar.

[0004] 4D millimeter wave imaging radar can be applied to vehicle-end intelligent driving and road-end intelligent transportation fields. The main function is to reconstruct the environment and target three-dimensional geometric information and motion information in real time with high precision, establish the relative spatial position relationship and relative motion speed between the radar placement point and the target, and construct the vehicle running scene to provide rich road scene information for the vehicle-end and road-end, support high-level automatic driving and high-level intelligent road construction.

[0005] Currently, the industry focuses on increasing the number of transmitting and receiving antennas while considering angle resolution and cost, and has made some achievements, but the effect is not satisfactory. For example, the invention CN 112924938 A mentions a 12-transmitting 16-receiving millimeter wave 4D imaging radar microstrip antenna array, which is a special two-dimensional transmitting and receiving array for four pieces of radio frequency front end cascade. It has the problems of low angle resolution, high sidelobe, and difficult engineering implementation. SUMMARY

[0006] The high-resolution imaging radar array design method proposed by the present application preliminarily avoids the problem of crossing lines by limiting the transmitting and receiving channel area, and maximizes the virtual array aperture by setting fixed-position antenna elements, thereby ensuring maximum angle resolution. It has the advantages of high angle resolution, low sidelobe, easy engineering implementation, etc.

[0007] The high-resolution imaging radar array design method provided by this invention, wherein the imaging radar has multiple transmitting elements and multiple receiving elements, includes the following steps:

[0008] S1. For an imaging radar with N transmitting elements and M receiving elements, a two-dimensional region Y×Z is set up for arranging these antenna elements.

[0009] S2. Set the grid intervals for azimuth and elevation directions according to the unambiguous angle measurement range, and divide the two-dimensional region into a two-dimensional gridded region. The grid points are the locations where transmitting and receiving array elements can be placed.

[0010] S3. Set constraints for the two-dimensional gridded area. Based on the fan-out characteristics of the transceiver channel of the integrated transceiver RF front-end chip, limit the placement area of ​​the transmitting and receiving array elements and the number of transmitting and receiving array elements in the area. The position and number of transmitting or receiving array elements in this two-dimensional gridded area need to take into account the crossover problem of the integrated transceiver RF wiring.

[0011] S4. A particle swarm optimization algorithm is used to search the two-dimensional gridded region. The position distribution of the searched transmitting and receiving array elements is virtualized using MIMO (multiple input multiple output) to form a virtual two-dimensional array. The highest sidelobe value of the virtual two-dimensional array target is preset to be no greater than T0, and N and M are positive integers. At the same time, in order to maximize the aperture of the virtual two-dimensional array, the receiving and transmitting array elements are fixedly placed at specific positions.

[0012] S5. Calculate the highest sidelobe value T of the virtual two-dimensional array pattern. Continuously optimize the positions of the transmitting and receiving antennas until the calculated highest sidelobe value T is less than the preset target highest sidelobe value T0, and output the layout of the virtual two-dimensional array at this time.

[0013] Preferably, in S2, based on the unambiguous angle measurement range ±θ and Set the grid spacing for azimuth and elevation to dy and dz, respectively, where dy = λ / 2 / sin(θ) and dz = λ / 2 / sin(φ); and λ is the wavelength, θ is the azimuth angle. The pitch angle.

[0014] Preferably, in step S5, if T is greater than T0, the particle swarm optimization algorithm in step S4 is executed for further search iteration; otherwise, the layout of the virtual two-dimensional array is output.

[0015] Preferably, the particle swarm optimization algorithm is initialized with a swarm of random particles, and then finds the optimal solution through iteration; in each iteration, it updates itself by tracking the individual optimal and the global optimal; after finding these two optimal values, the particles update their velocity and position using the following formula;

[0016] v i =v i +c1×rand()×(pbest i -x i )+c2×rand()×(gbest i -x i )

[0017] x i =x i-1 +v i

[0018] where i = 1, 2, …, N is the particle number, v i is the particle velocity; rand() is a random number between (0, 1); x i is the current position of the particle; c1, c2 are learning factors, pbest i is the individual optimum, gbest i is the global optimum.

[0019] Preferably, the particle decides the next movement by its own experience and the best experience in the companions, and the optimization speed of the particle satisfies the following formula:

[0020] v i = ω × v i +c1×rand()×(pbest i -x i )+c2×rand()×(gbest i -x i )

[0021] where ω is an inertia factor, and the value thereof is non-negative; when the value thereof is large, the global optimization ability is strong and the local optimization ability is weak; and when the value thereof is small, the global optimization ability is weak and the local optimization ability is strong.

[0022] Preferably, the inertia factor ω can be linearly changed in the search process of the particle swarm optimization algorithm.

[0023] Preferably, the left and right ends of the virtual two-dimensional array are respectively fixedly provided with a transmitting array element and a receiving array element to ensure that the two-dimensional column azimuth aperture of the virtual array is maximized, and the elevation distance of the fixed antenna element is maximized.

[0024] The particle swarm optimization algorithm is used for iterative search of a two-dimensional gridding area, and the MIMO equivalent virtual array of the optimal solution (a possible transceiving antenna array element position distribution) searched is calculated, and the MIMO virtual mode can be any one of TDM (time division multiplexing) MIMO virtual, FDM (frequency division multiplexing) MIMO virtual, CDM (code division multiplexing) MIMO virtual, DDM (Doppler diversity multiplexing) MIMO virtual, etc., to form a virtual two-dimensional transceiving array.

[0025] Further, the virtual two-dimensional transceiving array formed by an array with N transmitting array elements and M receiving array elements has N*M virtual transceiving array elements in total, and the azimuth and elevation positions of the i*j virtual array are respectively:

[0026]

[0027] wherein i=1...N, j=1...M, are the azimuth and elevation positions of the i-th transmitting array element, are the azimuth and elevation positions of the j-th receiving array element.

[0028] When the beam pointing is [0°, 0°] (azimuth and elevation), the antenna array pattern corresponding to the angle (θ0, φ0) (azimuth and elevation) is F(θ0, φ0)=w H a(θ0, φ0);

[0029] wherein a(θ0, φ0) is a steering vector of the virtual array, w H =a H (0, 0) is an array weight vector.

[0030]

[0031] The size relationship between the highest sidelobe value T0 of the antenna array pattern of the virtual array and T is judged, and if T>T0, the search iteration process is repeated until T≤T0, and a two-dimensional imaging radar antenna array design with low sidelobe and high resolution is obtained.

[0032] The present application has the following beneficial effects:

[0033] By the above method, the imaging radar designed by the present application realizes higher two-dimensional angle resolution without increasing hardware cost, and has the advantages of low sidelobe, easy engineering implementation, etc. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 The flowchart of the method of the present application is shown in the figure;

[0035] Figure 2A transceiving channel arrangement mode of an integrated transceiving radio frequency front-end chip is provided in an embodiment of the present application.

[0036] Figure 3 An antenna array region segmentation mode and corresponding transceiving antenna number constraint are provided in an embodiment of the present application.

[0037] Figure 4 A high-resolution imaging radar antenna array layout is provided in an embodiment of the present application.

[0038] Figure 5 A virtual transceiving antenna array graph after MIMO virtualization of a high-resolution imaging radar antenna array is provided in an embodiment of the present application.

[0039] Figure 6 An azimuth angle resolution distribution graph of a high-resolution imaging radar antenna array is provided in an embodiment of the present application.

[0040] Figure 7 A pitch angle resolution distribution graph of a high-resolution imaging radar antenna array is provided in an embodiment of the present application.

[0041] Figure 8 A peak sidelobe ratio distribution graph of a high-resolution imaging radar antenna array is provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] The high-resolution imaging radar array design method provided in the present application is further described in detail below in combination with the accompanying drawings and specific embodiments. The advantages and features of the present application will be more apparent according to the following description.

[0043] Preferably, the imaging radar has a plurality of transmitting array elements and a plurality of receiving array elements, and the imaging radar array design method comprises the following steps:

[0044] S1, for an imaging radar having N transmitting array elements and M receiving array elements, a two-dimensional region YxZ is set for arranging the transmitting array elements and the receiving array elements, N and M are positive integers. Referring to Figure 1 Taking a radar system with 12 transmitting array elements and 16 receiving array elements as an example, the constraint array arrangement two-dimensional region is 106x95mm.

[0045] S2, according to the non-ambiguous angle measurement range, the azimuth and pitch directions ±θ and ±φ are set with grid intervals dy and dz, and the two-dimensional region is divided into a two-dimensional grid region according to the grid intervals dy and dz, and the grid points of the two-dimensional grid region are position points where the transmitting array elements and the receiving array elements can be placed;

[0046] Specifically, dy = λ / 2 / sin(θ), dz = λ / 2 / sin(φ), where λ is the wavelength, θ is the azimuth angle, ​The pitch angle.

[0047] Taking an azimuth-free angular measurement range of ±90° and an elevation-free angular measurement range of ±32° as an example, the grid intervals are set to dy=0.5λ=1.947mm and dz=0.944λ=3.675mm respectively, and the two-dimensional region is divided into 55×26 grid points.

[0048] S3. Set constraints for the two-dimensional gridded region. Since different integrated transceiver RF front-end chips have different transceiver channel distributions, their fan-out patterns vary, typically with 90°, 180°, or 270° distributions. Based on the fan-out characteristics of the integrated transceiver RF front-end chip's transceiver channels, limit the placement area of ​​the transmitting and receiving array elements, as well as the number of transmitting or receiving array elements within each area. Furthermore, the position and number of transmitting or receiving array elements in this two-dimensional gridded region must avoid RF wiring intersections. RF wiring refers to the microstrip lines used to connect the integrated transceiver RF front-end chip to the transmitting and receiving array elements.

[0049] Specifically, taking the AWR2243P, a mainstream integrated RF front-end chip with three transmit and four receive channels, as an example, its transmit and receive channel fan-out is as follows: Figure 2 As shown, Rx represents the receive channel, and Tx represents the transmit channel. Four integrated RF front-end chips with three transmit and four receive channels are cascaded. Taking advantage of the fan-out characteristics of the receive and transmit channels of the integrated transceiver RF front-end chips, 12 transmit elements and 16 receive elements are placed alternately. The azimuth and elevation dimensions of both the receive and transmit element regions are equal to the array size to achieve the maximum array aperture. The 55×26 gridded two-dimensional region (array surface) is divided into 7 regions, as shown... Figure 3 As shown, region 1 constrains 4 receiving array elements, with an azimuth of 0 ≤ y < 12dy and an elevation of 16dz ≤ z ≤ 25dz; region 2 constrains 6 transmitting array elements, with an azimuth of 12dy ≤ y < 42dy and an elevation of 16dz ≤ z ≤ 25dz; region 3 constrains 4 receiving array elements, with an azimuth of 42dy ≤ y ≤ 54dy and an elevation of 16dz ≤ z ≤ 25dz; and region 4 constrains 3 transmitting array elements, with an azimuth of 0 ≤ y < 12dy and an elevation of 16dz ≤ z ≤ 25dz. Region 5 is constrained to have 8 receiving array elements, with an azimuth of 12dy≤y<42dy and an elevation of 0≤z<9dz; Region 6 is constrained to have 3 transmitting array elements, with an azimuth of 42dy≤y≤54dy and an elevation of 0≤z<9dz; Region 7 is for placing integrated transceiver RF front-end chips, but no transmitting or receiving array elements are placed there, with an azimuth of 0≤y≤54dy and an elevation of 9dz≤z<16dz.

[0050] S4, search the two-dimensional grid area by using the particle swarm optimization algorithm, perform MIMO (multiple input multiple output) virtualization on the position distribution of the transmitting array elements and the receiving array elements obtained in the search to form a virtual two-dimensional array; and preset a peak sidelobe value T0 of the virtual two-dimensional array.

[0051] Taking a frequency of 77GHz radar system as an example, the wavelength λ is 3.893mm, and the preset peak sidelobe ratio of the virtual two-dimensional array is not greater than -10dB, i.e., T0=-10.

[0052] Preferably, the MIMO radar virtual two-dimensional array is the sum of the transmitting array aperture and the receiving array aperture, so that fixed transmitting array elements and receiving array elements are arranged at both ends of the grid to ensure that the azimuth aperture of the virtual two-dimensional array is maximum, and the elevation distance of the transmitting array elements and the receiving array elements is maximized.

[0053] According to the constraint condition set above, a 55*26 grid point space is searched to obtain a kind of transmitting and receiving antenna layout as shown in Figure 4 The maximum actual aperture in the azimuth direction of the antenna is 54dy, and the maximum actual aperture in the elevation direction is 25dz.

[0054] S5, calculate the highest sidelobe value T of the virtual two-dimensional array pattern, and optimize the positions of the transmitting array elements and the receiving array elements by taking the highest sidelobe of the virtual array pattern as the objective function, and continuously optimize the positions of the transmitting array elements and the receiving array elements until the calculated highest sidelobe value T is less than the preset target highest sidelobe value T0.

[0055] If T is greater than T0, continue to perform S4 particle swarm optimization algorithm search iteration, otherwise output the layout of the virtual two-dimensional array.

[0056] Preferably, the particle swarm optimization algorithm is initialized as a group of random particles (random solutions), and then the optimal solution is found through iteration. In each iteration, the particle updates itself by tracking two "extreme values" (individual optimal, global optimal); when the two optimal values are found, the particle updates its speed and position by the following formula.

[0057] v i =v i +c1×rand()×(pbest i -x i )+c2×rand()×(gbest i -x i )

[0058] x i =x i-1 +v i

[0059] where i = 1, 2, …, N is the particle index, v i is the velocity of the particle; rand() is a random number between (0, 1); x i is the current position of the particle; c1, c2 are learning factors, pbest i is the individual optimum, gbest i is the global optimum. The particle decides the next movement by its own experience and the best experience of the companions, and the particle further optimizes the velocity to satisfy the following formula:

[0060] v i = ω × v i + c1 × rand() × (pbest i - x i ) + c2 × rand() × (gbest i - x i )

[0061] where ω is an inertia factor, which is non-negative, and a larger value has a stronger global optimization ability and a weaker local optimization ability, and a smaller value has a weaker global optimization ability and a stronger local optimization ability. The inertia factor ω can be linearly changed in the search process of the particle swarm optimization algorithm, and the dynamic inertia factor ω can obtain better optimization results than a fixed value.

[0062] The above particle swarm optimization algorithm is used to iteratively search a two-dimensional grid region, and the optimal solution (a possible position distribution of the transceiving antenna array element) obtained by the search is used to calculate the MIMO equivalent virtual array. The MIMO virtual method can be any one of TDM (time division multiplexing) MIMO virtual, FDM (frequency division multiplexing) MIMO virtual, CDM (code division multiplexing) MIMO virtual, DDM (Doppler diversity multiplexing) MIMO virtual, etc., to form a virtual two-dimensional array.

[0063] Further, an array with N transmitting array elements and M receiving array elements forms a virtual two-dimensional array with N·M virtual transceiving array elements, and the azimuth and elevation positions of the i·j virtual array are:

[0064]

[0065] where i = 1…N, j = 1…M, are the azimuth and elevation positions of the i-th transmitting array element, respectively, are the azimuth and elevation positions of the j-th receiving array element, respectively.

[0066] When the beam pointing is [0°, 0°] (azimuth, elevation), the antenna array pattern corresponding to the angle (θ0, φ0) (azimuth, elevation) is F(θ0, φ0) = w Ha(θ0,φ0);

[0067] Where a(θ0,φ0) is the steering vector of the virtual array, w H =a H (0,0) is the array weight vector;

[0068]

[0069] Determine the relationship between the highest sidelobe value T0 and T in the antenna array pattern of the virtual array; if T>T0, repeat the search iteration process until T≤T0, to obtain a low-sidelobe, high-resolution two-dimensional imaging radar antenna array design.

[0070] The array arrangement of the transmitting and receiving elements after MIMO virtualization is as follows: Figure 5 As shown, the maximum virtual aperture of the virtual two-dimensional array is 108dy in the azimuth direction and 50dz in the elevation direction. The virtualized antenna array aperture is 210.276×183.75mm.

[0071] Specifically, a simulation target is set, with its azimuth and elevation angles traversing the ranges [-60°, 60°] and [-20°, 20°] respectively at 1° intervals. Digital beamforming is used for angle estimation, and the array's angle measurement performance is analyzed. For example... Figure 6 The optimal azimuth angle resolution shown is 1.2°; Figure 7 The optimal pitch angle resolution shown is 1.18°; Figure 8 The peak sidelobes of the two-dimensional spatial spectrum are below -10dB. Compared with existing publicly available technical solutions, the high-resolution imaging radar array design method of this invention significantly improves the azimuth and elevation angle resolutions of the virtual two-dimensional array while ensuring low peak sidelobes.

[0072] Through the above methods, the millimeter-wave imaging radar of this invention achieves higher two-dimensional angular resolution compared with the prior art without increasing hardware costs.

[0073] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A high-resolution imaging radar array design method, characterized in that, The imaging radar has multiple transmitting elements and multiple receiving elements, and includes the following steps: S1. Set up a two-dimensional area that can accommodate multiple antenna array elements; S2. Set the grid intervals for azimuth and elevation directions according to the unambiguous angle measurement range to divide the two-dimensional region into a two-dimensional gridded region; S3. Set two-dimensional gridded area constraints, and limit the placement area of ​​the transmitting array element and the number of transmitting array elements in the area according to the fan-out characteristics of the transmitting and receiving channel of the integrated transceiver RF front-end chip. S4. Using the particle swarm optimization algorithm, the two-dimensional gridded region is searched, and the position distribution of the searched transmitting and receiving array elements is virtualized by MIMO to form a virtual two-dimensional array. And it is preset that the highest sidelobe value of the virtual two-dimensional array target is no greater than T0; S5. Calculate the highest sidelobe value T of the virtual two-dimensional array pattern, continuously optimize the antenna element positions until the calculated highest sidelobe value T is less than the preset target highest sidelobe value T0, and output the layout of the virtual two-dimensional array at this time.

2. The high-resolution imaging radar array design method as described in claim 1, characterized in that, The S2 statement, based on the unambiguous angle measurement range [-θ, +θ] and Set the grid intervals for azimuth and elevation, with intervals of dy and dz respectively, where dy = λ / 2 / sin(θ) and dz = λ / 2 / sin(φ); Where λ is the wavelength and θ is the azimuth angle. The pitch angle.

3. The high-resolution imaging radar array design method as described in claim 1, characterized in that, In step S5, if T is greater than T0, the particle swarm optimization algorithm in step S4 is returned to perform the search iteration; otherwise, the layout of the virtual two-dimensional array is output.

4. The high-resolution imaging radar array design method as described in claim 3, characterized in that, The particle swarm optimization algorithm is initialized with a swarm of random particles, and then finds the optimal solution through iteration. In each iteration, the particle updates itself by tracking the individual optimal and the global optimal. Once these two optimal values ​​are found, the particle updates its velocity and position using the following formula. v i =v i +c1×rand()×(pbest i -x i )+c2×rand()×(gbest i -x i ) x i =x i-1 +v i Where i = 1, 2, ..., N are particle numbers, v i It represents the particle's velocity; rand() is a random number between (0, 1); x i It represents the particle's current position; c1 and c2 are learning factors, and pbest i It is the individual optimal, gbest i It is the global optimum.

5. The high-resolution imaging radar array design method as described in claim 4, characterized in that, The particle determines its next move based on its own experience and the best experience among its peers. The optimal velocity of the particle satisfies the following equation: v i =ω×v i +c1×rand()×(pbest i -x i )+c2×rand()×(gbest i -x i In this equation, ω is the inertia factor. Its value is non-negative. A larger value indicates a strong global optimization ability and a weak local optimization ability, while a smaller value indicates a weak global optimization ability and a strong local optimization ability.

6. The high-resolution imaging radar array design method as described in claim 5, characterized in that, The inertia factor ω can change linearly during the particle swarm optimization algorithm search process.

7. The high-resolution imaging radar array design method as described in claim 1, characterized in that, The left and right ends of the virtual two-dimensional array are fixedly equipped with transmitting and receiving array elements to ensure the maximum azimuth aperture of the virtual two-dimensional array, while maximizing the elevation distance between the transmitting and receiving array elements.

Images