A Conformal Phased Array Radar Target Detection Method

Abstract

This invention discloses a target detection method for conformal phased array radar, belonging to the field of radar target detection technology. The invention first constructs a compensation quantity based on the array element curvature tensor to initially compensate for the array element echo; then, the array elements are projected onto a reference plane to compensate for phase changes, obtaining a virtual plane equivalent echo; after standardization, the equivalent echo is multiplied by a curvature weight and a curvature-enhanced energy spectrum is generated through a two-dimensional Fourier transform; the target and annular background window are extracted, and the energy and curvature similarity are calculated to obtain a two-dimensional saliency; the energy saliency is enhanced by the curvature saliency to obtain the target's comprehensive discriminative power, and the presence of a target is determined when the discriminative power is greater than a threshold. This invention, through curvature compensation and two-dimensional fusion discrimination, offsets the phase distortion and energy dispersion caused by array curvature, significantly improving the detection accuracy in complex scenes.

Description

Technical Field

[0001] This invention relates to the field of radar target detection technology, and specifically to a method for target detection using a conformal phased array radar. Background Technology

[0002] With the development of modern radar technology, phased array radars are widely used for target detection in complex environments. Traditional phased array radars typically perform signal processing based on planar arrays, and their target detection methods rely on the coherent superposition of array element echoes and two-dimensional Fourier transform to obtain the energy spectrum. Common target detection methods in existing technologies include matched filtering, constant false alarm rate detection (CFAR), and peak detection based on the energy spectrum. These methods can extract target energy characteristics relatively accurately in planar array environments. However, for non-planar conformal phased array radars, because the array elements are distributed on curved surfaces, the geometry between the elements introduces additional phase distortion. Directly using traditional methods for target detection often leads to energy spectrum distortion, thereby reducing the accuracy of target detection.

[0003] The problem with existing technologies is that, for conformal phased array radars, the curvature effect of the array elements is not effectively compensated, resulting in low target detection accuracy. Summary of the Invention

[0004] To address the aforementioned shortcomings in the existing technology, the present invention provides a conformal phased array radar target detection method that solves the problem of low target detection accuracy in the existing technology.

[0005] To achieve the above-mentioned objectives, the technical solution adopted by this invention is: a conformal phased array radar target detection method, comprising the following steps:

[0006] S1. Based on the curvature tensor of each array element on the conformal phased array, construct the curvature compensation quantity, and compensate the corresponding array element echo to obtain the array element compensated echo.

[0007] S2. Select a reference plane and project each array element onto the reference plane to obtain a virtual plane;

[0008] S3. Find the phase change before and after projection, compensate the array element echo again, and obtain the equivalent echo of each array element in the virtual plane.

[0009] S4. Standardize the virtual plane, multiply each equivalent echo on the plane by the curvature weight, and then perform a two-dimensional Fourier transform to obtain the curvature-enhanced energy spectrum.

[0010] S5. Extract the target window and the annular background window for curvature-enhanced energy spectrum extraction;

[0011] S6. Obtain the energy and curvature similarity of the target window and the circular background window, and obtain the target energy salience and the target curvature salience;

[0012] S7. The target energy salience is enhanced by using the target curvature salience to obtain the target comprehensive discrimination. When the target comprehensive discrimination at the detection position is greater than the threshold, there is a target at the corresponding detection position.

[0013] Furthermore, S1 includes the following steps:

[0014] S11. For each element of the conformal phased array, first obtain its original echo;

[0015] S12. Based on the target's azimuth angle, determine the azimuth projection direction of the target's direction onto the array surface;

[0016] S13. Then, by combining the curvature tensor and the direction compensation weight at the location of the array element, the curvature compensation amount corresponding to the array element is calculated.

[0017] S14. Perform phase compensation on the original echo of the corresponding array element according to the curvature compensation amount to obtain the array element compensated echo.

[0018] Furthermore, the process of obtaining the directional compensation weights includes:

[0019] A1. On a conformal phased array, take the array element to be compensated as the center and select multiple array elements that are spatially adjacent to it to form an array set.

[0020] A2. For the incident direction to be estimated, calculate the complex phase matching results between the original echoes of all array elements in the array set and the theoretical phase in that direction;

[0021] A3. Sum the matching results belonging to the matrix set;

[0022] A4. Traverse the incident directions and select the incident direction that maximizes the summation result as the coarsely estimated incident angle;

[0023] A5. Based on the roughly estimated angle of incidence, find the corresponding compensation weight by looking up a table.

[0024] Furthermore, S2 includes the following sub-steps:

[0025] S21. Select a reference plane and obtain the unit normal vector of the reference plane;

[0026] S22. Calculate the vector difference of each array element relative to the reference point on the reference plane, and obtain the displacement of the array element along the reference plane by the inner product operation of the unit normal vector.

[0027] S23. Based on the displacement of the array element along the unit normal vector of the reference plane, project the array element onto the reference plane to obtain the virtual coordinates of the array element.

[0028] S24. Extract the two-dimensional coordinates of the virtual coordinates of each array element, and arrange the array elements according to the two-dimensional coordinates to form a virtual plane.

[0029] Furthermore, S3 includes the following sub-steps:

[0030] S31. Based on the coordinates of the array elements before projection, obtain the uncompensated geometric phase of each array element before projection.

[0031] S32. Calculate the geometric phase after projection based on the virtual coordinates of the array elements after projection.

[0032] S33. Subtract the uncompensated geometric phase before projection from the geometric phase after projection to obtain the phase change of the array element.

[0033] S34. The phase change of the array elements is compensated to the array element compensation echo to obtain the equivalent echo.

[0034] Furthermore, S4 includes the following sub-steps:

[0035] S41. Map the coordinates of the virtual plane to the orthogonal basis of the reference plane to obtain the normalized virtual plane;

[0036] S42. Obtain the curvature weights based on the curvature values ​​corresponding to the curvature tensor of each array element.

[0037] S43. Multiply the equivalent echo of each array element by the corresponding curvature weight, and perform complex domain multiplication to obtain the energy-enhanced echo.

[0038] S44. Arrange the energy-enhanced echoes according to the coordinates of the standardized virtual plane, and perform a two-dimensional Fourier transform to obtain the direction-distance two-dimensional spectrum.

[0039] S45. Extract energy at each position of the direction-distance two-dimensional spectrum to obtain the curvature-enhanced energy spectrum.

[0040] Furthermore, the formula for curvature weight in S42 is:

[0041] ;

[0042] in, For the first The curvature weights of each array element, where exp is an exponential function. For the first The curvature value of each array element. This is the adjustment coefficient;

[0043] The formula for obtaining the energy-enhanced echo in S43 is: ,in, For the first The energy-enhanced echo of each array element For the first The curvature weights of each array element For the first The equivalent echo of each array element.

[0044] Furthermore, S5 includes the following steps:

[0045] S51. In the curvature-enhanced energy spectrum, select an r×r target window centered on the location to be detected, where r is the side length of the target window;

[0046] S52. Construct an R×R window centered on the same detection location. Remove the target window from the R×R window to obtain a ring-shaped background window, where R is the side length of the ring-shaped background window and R > r.

[0047] Furthermore, S6 includes the following sub-steps:

[0048] S61. Take the average value of all energies in the target window to obtain the target window energy, and take the average value of all energies in the annular background window to obtain the background window energy.

[0049] S62. Calculate the curvature similarity for each position in the target window, and take the average of all curvature similarities to obtain the target window similarity.

[0050] S63. Calculate the curvature similarity for each position in the annular background window, and take the average of all curvature similarities to obtain the background window similarity.

[0051] S64. Based on the ratio of target window energy to background window energy, obtain the target energy salience.

[0052] S65. The target curvature salience is obtained based on the ratio of the target window similarity to the background window similarity.

[0053] Furthermore, the formula for the overall discriminant of the target is:

[0054] ;

[0055] in, The overall discrimination of the target at the location to be detected. The target energy salience at the location to be detected. To achieve the desired curvature enhancement, This is the proportionality coefficient.

[0056] The beneficial effects of this invention are as follows:

[0057] 1. This invention constructs a compensation amount based on the curvature tensor of each array element on a conformal phased array and compensates for the array element echo, which can effectively reduce the phase distortion caused by the curvature of the array elements and improve the accuracy of the array element echo.

[0058] 2. This invention projects each array element onto a reference plane to obtain a virtual plane, realizing the equivalent transformation from a non-planar array to a planar array. Then, by calculating the phase change before and after projection, the array element compensation echo is corrected again to obtain the virtual plane equivalent echo, further correcting the phase error caused by projection and improving the coherence of the target signal.

[0059] 3. This invention standardizes and weights the virtual plane equivalent echo, and then performs a two-dimensional Fourier transform to obtain a curvature-enhanced energy spectrum, which makes the target signal more prominent in the energy spectrum and enhances the target energy identification.

[0060] 4. This invention extracts the target window and the annular background window, clearly distinguishing the target region from the background region. By calculating the energy and curvature similarity between the target window and the background window, it obtains the target energy salience and the target curvature salience, making the target features more prominent in complex backgrounds and enhancing the reliability of target detection. The target curvature salience is used to enhance the target energy salience, resulting in a comprehensive target discriminative power. When the target location exceeds a threshold, the target is accurately identified, achieving high-precision and high-reliability target detection, thus solving the problem of low detection accuracy in traditional methods. Attached Figure Description

[0061] Figure 1 This is a flowchart of a conformal phased array radar target detection method. Detailed Implementation

[0062] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0063] like Figure 1 As shown, a conformal phased array radar target detection method includes the following steps:

[0064] S1. Based on the curvature tensor of each array element on the conformal phased array, construct the curvature compensation quantity, and compensate the corresponding array element echo to obtain the array element compensated echo.

[0065] S2. Select a reference plane and project each array element onto the reference plane to obtain a virtual plane;

[0066] S3. Find the phase change before and after projection, compensate the array element echo again, and obtain the equivalent echo of each array element in the virtual plane.

[0067] S4. Standardize the virtual plane, multiply each equivalent echo on the plane by the curvature weight, and then perform a two-dimensional Fourier transform to obtain the curvature-enhanced energy spectrum.

[0068] S5. Extract the target window and the annular background window for curvature-enhanced energy spectrum extraction;

[0069] S6. Obtain the energy and curvature similarity of the target window and the circular background window, and obtain the target energy salience and the target curvature salience;

[0070] S7. The target energy salience is enhanced by using the target curvature salience to obtain the target comprehensive discrimination. When the target comprehensive discrimination at the detection position is greater than the threshold, there is a target at the corresponding detection position.

[0071] The process of obtaining the curvature tensor of each array element in this invention includes: obtaining the three-dimensional actual installation coordinates of each array element in the conformal phased array. For each array element, perform a local second-order surface fitting z=f(x,y) and take the second-order partial derivative to obtain the curvature tensor of each array element as follows:

[0072] ;

[0073] in, Let be the curvature tensor of the m-th array element. Let m be the coordinates of the m-th array element. Let x be the x-coordinate of the m-th element. Let be the y-coordinate of the m-th array element. Let f(x,y) be the z-coordinate of the m-th array element, and f(x,y) be the second-order surface fitting function for the x and y coordinates.

[0074] In this embodiment, S1 includes the following steps:

[0075] S11. For each element of the conformal phased array, first obtain its original echo;

[0076] S12. Based on the target's azimuth angle, determine the azimuth projection direction of the target's direction onto the array surface;

[0077] S13. Then, by combining the curvature tensor and the direction compensation weight at the location of the array element, the curvature compensation amount corresponding to the array element is calculated.

[0078] S14. Perform phase compensation on the original echo of the corresponding array element according to the curvature compensation amount to obtain the array element compensated echo.

[0079] The formula for array element compensation echo is:

[0080] ;

[0081] ;

[0082] in, For the compensation echo of the m-th array element, The original echo of the m-th array element. It is a natural constant. For both virtual and real units, , The azimuth projection of the target direction onto the array face. Let be the curvature tensor of the m-th array element. For transpose, It is the azimuth angle. The pitch angle, To compensate for the weight, Let be the curvature compensation amount for the m-th array element.

[0083] Due to their curved surface layout, conformal arrays exhibit additional phase errors in the target echo due to curvature differences among different array elements, directly affecting the accuracy of subsequent processing such as Fourier transform. This invention calculates the phase compensation amount using "curvature tensor + directional projection," accurately offsetting the phase deviation of the curved array elements.

[0084] In this embodiment, the process of obtaining the orientation compensation weight includes:

[0085] A1. On a conformal phased array, take the array element to be compensated as the center and select multiple array elements that are spatially adjacent to it to form an array set.

[0086] A2. For the incident direction to be estimated, calculate the complex phase matching results between the original echoes of all array elements in the array set and the theoretical phase in that direction;

[0087] A3. Sum the matching results belonging to the matrix set;

[0088] A4. Traverse the incident directions and select the incident direction that maximizes the summation result as the coarsely estimated incident angle;

[0089] A5. Based on the roughly estimated angle of incidence, find the corresponding compensation weight by looking up a table.

[0090] The formula for obtaining the directional compensation weight is as follows:

[0091] ;

[0092] in, This is a rough estimate of the angle of incidence. To select the independent variable that maximizes the result, It is a natural constant. For both virtual and real units, For theoretical phase, Let || be the original echo of the m-th array element, and || be the modulus. It is the azimuth angle. The pitch angle, This is a rough estimate of the pitch angle in the angle of incidence. This is the azimuth angle in the roughly estimated angle of incidence. For array sets.

[0093] In this embodiment, a table is used to establish a correspondence between the roughly estimated incident angle and the compensation weight, which is pre-stored in the database, and each compensation weight is an empirical value.

[0094] This invention reduces the impact of surface noise by using the echo correlation of local array elements to form an array set through "adjacent array elements". At the same time, by traversing the direction using "complex phase matching + maximum energy criterion", it can accurately lock the incident direction that best matches the echo phase, making the coarsely estimated incident angle closer to the real target direction, and making it easier to obtain higher precision compensation weights.

[0095] In this embodiment, S2 includes the following sub-steps:

[0096] S21. Select a reference plane and obtain the unit normal vector of the reference plane;

[0097] S22. Calculate the vector difference of each array element relative to a reference point on the reference plane, and obtain the displacement of the array element along the reference plane by performing an inner product operation with the unit normal vector:

[0098] ;

[0099] in, For the first The displacement of each array element along the unit normal vector of the reference plane. The unit normal vector of the reference plane. For the first conformal phased array The coordinates of each array element Choose any reference point on the reference plane. This is a transpose operation;

[0100] S23. Based on the displacement of the array element along the unit normal vector of the reference plane, project the array element onto the reference plane to obtain the virtual coordinates of the array element:

[0101] ;

[0102] in, For the first The virtual coordinates of each array element satisfy... , The distance from the reference plane to the origin of the global coordinate system;

[0103] S24. Extract the virtual coordinates of each array element. Two-dimensional coordinates The array elements are arranged in two-dimensional coordinates to form a virtual plane.

[0104] In this embodiment, , , and All are 3×1 column vectors.

[0105] The reference plane is a manually selected reference plane used to project the curved elements of the conformal phased array onto this plane to obtain the virtual coordinates of the elements, thereby forming a virtual plane. In this embodiment, the reference plane is the tangent plane of the "area of ​​elements to be processed" on the carrier surface—by fitting the coordinates of the elements in this area, a plane that best fits the local surface is obtained.

[0106] The problem with conformal arrays is that curved surface layouts cannot directly utilize mature planar array signal processing algorithms (such as 2D Fourier transform and planar beamforming). This invention converts curved array elements into virtual planar coordinates through "reference plane projection," enabling conformal arrays to be applied to 2D Fourier transforms while retaining the advantage of conformal arrays "fitting the shape of the carrier."

[0107] In this embodiment, S3 includes the following sub-steps:

[0108] S31. Based on the coordinates of the array elements before projection, obtain the uncompensated geometric phase of each array element before projection.

[0109] S32. Calculate the geometric phase after projection based on the virtual coordinates of the array elements after projection.

[0110] S33. Subtract the uncompensated geometric phase before projection from the geometric phase after projection to obtain the phase change of the array element.

[0111] S34. Compensate the phase change of the array elements to the array element compensation echo to obtain the equivalent echo: .

[0112] No. The geometric phase (relative to the reference point) of each element before projection is: ,in, For the first Uncompensated geometric phase before projection of each array element For radar operating wavelength, The target direction is the unit vector. For the first conformal phased array The coordinates of each array element; the formula for calculating the geometric phase after projection is: , For the first The geometric phase after projection of each array element.

[0113] When a conformal array is projected from a curved surface onto a virtual plane, the change in the spatial position of the array elements introduces an additional geometric phase difference (phase inconsistency before and after projection). This invention precisely cancels out the phase interference caused by projection through the process of "calculating the geometric phase difference before and after projection → compensating for this difference," ensuring that the phase of the virtual plane array elements is completely aligned with the phase of the "ideal plane array," thus solving the core problem of "phase distortion after projection" and guaranteeing the phase consistency of the virtual array.

[0114] In this embodiment, S4 includes the following sub-steps:

[0115] S41. Map the coordinates of the virtual plane to the orthogonal basis of the reference plane to obtain the normalized virtual plane;

[0116] S42. Obtain the curvature weights based on the curvature values ​​corresponding to the curvature tensor of each array element.

[0117] S43. Multiply the equivalent echo of each array element by the corresponding curvature weight, and perform complex domain multiplication to obtain the energy-enhanced echo.

[0118] S44. Arrange the energy-enhanced echoes according to the coordinates of the standardized virtual plane, and perform a two-dimensional Fourier transform to obtain the direction-distance two-dimensional spectrum.

[0119] S45. Extract energy at each position of the direction-distance two-dimensional spectrum to obtain the curvature-enhanced energy spectrum.

[0120] The curvature tensor corresponds to the curvature value, that is, the curvature tensor corresponds to the curvature scalar. The curvature scalar is a scalar feature extracted from the curvature tensor and is used to quantify the overall curvature of a point on a surface, such as Gaussian curvature, mean curvature, etc.

[0121] In this embodiment, the formula for curvature weight in S42 is:

[0122] ;

[0123] in, For the first The curvature weights of each array element, where exp is an exponential function. For the first The curvature of each array element is worth a normalized value. For adjustment coefficients, The empirical value is 0.3;

[0124] The formula for obtaining the energy-enhanced echo in S43 is: ,in, For the first The energy-enhanced echo of each array element For the first The curvature weights of each array element For the first The equivalent echo of each array element.

[0125] In conformal phased arrays, the array elements are distributed on curved surfaces, and their geometry deviates from that of traditional planar arrays. This causes changes in the equivalent position, phase response, and spatial sampling relationship of the array elements, resulting in problems such as main lobe broadening, side lobe elevation, and even energy leakage in the direction-range two-dimensional spectrum, which directly reduces the concentration of target energy and detection performance. To address this inherent problem, this invention first constructs a unified, standardized virtual plane by mapping virtual plane coordinates to an orthogonal basis of the reference plane (S41), enabling curved array elements to have a consistent spatial reference in two-dimensional spectral processing and eliminating spatial sampling inconsistencies caused by the curved surface layout. Subsequently, curvature weights are generated based on the curvature values ​​of the array elements (S42) to adaptively compensate for the amplitude. Then, by multiplying the equivalent echo with the complex domain of the curvature weights (S43), differentiated energy enhancement is performed on the array element echoes, allowing the energy of the array element signal corresponding to the target to be amplified in a targeted manner, thus alleviating the target energy diffusion problem caused by the curvature of the array surface from an energy perspective. Finally, the echoes are arranged according to the standardized virtual plane and a two-dimensional Fourier transform is performed (S44) to obtain the direction-range two-dimensional spectrum and extract the curvature-enhanced energy spectrum (S45), effectively improving the concentration of target energy in the two-dimensional spectrum while suppressing surface clutter caused by curvature disorder, significantly improving the target detection performance of the conformal phased array.

[0126] In this embodiment, the energy in S45 is the square of the complex amplitude of each spectral point in the direction-distance two-dimensional spectrum.

[0127] In this embodiment, S5 includes the following steps:

[0128] S51. In the curvature-enhanced energy spectrum, select an r×r target window centered on the location to be detected, where r is the side length of the target window;

[0129] S52. Construct an R×R window centered on the same detection location. Remove the target window from the R×R window to obtain a ring-shaped background window. R is the side length of the ring-shaped background window, R > r. Usually, R is 7 and r is 7.

[0130] In this embodiment, S6 includes the following sub-steps:

[0131] S61. Take the average value of all energies in the target window to obtain the target window energy, and take the average value of all energies in the annular background window to obtain the background window energy.

[0132] S62. Calculate the curvature similarity for each position in the target window, and take the average of all curvature similarities to obtain the target window similarity.

[0133] S63. Calculate the curvature similarity for each position in the annular background window, and take the average of all curvature similarities to obtain the background window similarity.

[0134] S64. Based on the ratio of target window energy to background window energy, obtain the target energy salience.

[0135] S65. The target curvature salience is obtained based on the ratio of the target window similarity to the background window similarity.

[0136] This scheme uses a dual-window design of "local target window (r×r) + annular background window (R×R)" to focus on the local target features at the detection location with a small window, and to obtain the statistical characteristics of the surrounding background with an annular window. The difference is then quantified by "target / background energy ratio and curvature similarity ratio", which improves the contrast of target features.

[0137] In S64 and S65, when taking the ratio, a minimum value is added to the background window energy and background window similarity in the denominator to prevent the denominator from being 0.

[0138] The formula for calculating curvature similarity is: ,in, For curvature similarity, For position The curvature tensor, Let sim be the target curvature a priori template, and sim be the similarity function. High similarity can only be obtained when the curvature tensor of the location to be detected matches the target template, thus filtering out "high-energy but mismatched curvature clutter" at the geometric feature level.

[0139] The target curvature prior template is a pre-stored "array element curvature tensor template corresponding to the real target echo".

[0140] In this embodiment, the formula for obtaining the target comprehensive discriminant is:

[0141] ;

[0142] in, The overall discrimination of the target at the location to be detected. The target energy salience at the location to be detected. To achieve the desired curvature enhancement, This is the proportionality coefficient.

[0143] In this embodiment, the threshold for the overall discrimination of the target at the location to be detected is set based on experiments or experience.

[0144] This solution deeply integrates "target energy salience" and "target curvature salience" through the formula of target comprehensive discrimination: energy salience ensures the "signal strength basis" of the target; curvature salience strengthens the "geometric feature uniqueness" of the target, making the discrimination more consistent with the characteristics and laws of real targets.

[0145] This invention constructs a compensation amount based on the curvature tensor of each array element on a conformal phased array and compensates for the array element echo, which can effectively reduce the phase distortion caused by the curvature of the array elements and improve the accuracy of the array element echo.

[0146] This invention projects each array element onto a reference plane to obtain a virtual plane, realizing the equivalent transformation from a non-planar array to a planar array. Then, by calculating the phase change before and after projection, the array element compensation echo is corrected again to obtain the virtual plane equivalent echo, further correcting the phase error caused by projection and improving the coherence of the target signal.

[0147] This invention standardizes and weights the virtual plane equivalent echo, and then performs a two-dimensional Fourier transform to obtain a curvature-enhanced energy spectrum, which makes the target signal more prominent in the energy spectrum and enhances the target energy identification.

[0148] This invention extracts a target window and a ring-shaped background window to clearly distinguish the target region from the background region. By calculating the energy and curvature similarity between the target window and the background window, it obtains the target energy salience and the target curvature salience, making the target features more prominent in complex backgrounds and enhancing the reliability of target detection. The target curvature salience is used to enhance the target energy salience, resulting in a comprehensive target discriminative power. This accurately identifies the target when the detection position exceeds a threshold, achieving high-precision and high-reliability target detection and solving the problem of low detection accuracy in traditional methods.

[0149] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A target detection method for conformal phased array radar, characterized in that, Includes the following steps: S1. Based on the curvature tensor of each array element on the conformal phased array, construct the curvature compensation quantity, and compensate the corresponding array element echo to obtain the array element compensated echo. S2. Select a reference plane and project each array element onto the reference plane to obtain a virtual plane; S3. Find the phase change before and after projection, compensate the array element echo again, and obtain the equivalent echo of each array element in the virtual plane. S4. Standardize the virtual plane, multiply each equivalent echo on the plane by the curvature weight, and then perform a two-dimensional Fourier transform to obtain the curvature-enhanced energy spectrum. The formula for curvature weight is: ； in, For the first The curvature weights of each array element, where exp is an exponential function. For the first The curvature value of each array element. This is the adjustment coefficient; S5. Extract the target window and the annular background window for curvature-enhanced energy spectrum extraction; S6. Obtain the energy and curvature similarity of the target window and the circular background window, and obtain the target energy salience and the target curvature salience; S7. The target energy salience is enhanced by using the target curvature salience to obtain the target comprehensive discrimination. When the target comprehensive discrimination at the detection position is greater than the threshold, the target exists at the corresponding detection position.

2. The conformal phased array radar target detection method according to claim 1, characterized in that, S1 includes the following steps: S11. For each element of the conformal phased array, first obtain its original echo; S12. Based on the target's azimuth angle, determine the azimuth projection direction of the target's direction onto the array surface; S13. Then, by combining the curvature tensor and the direction compensation weight at the location of the array element, the curvature compensation amount corresponding to the array element is calculated. S14. Perform phase compensation on the original echo of the corresponding array element according to the curvature compensation amount to obtain the array element compensated echo.

3. The conformal phased array radar target detection method according to claim 2, characterized in that, The process of obtaining the directional compensation weight includes: A1. On a conformal phased array, take the array element to be compensated as the center and select multiple array elements that are spatially adjacent to it to form an array set. A2. For the incident direction to be estimated, calculate the complex phase matching results between the original echoes of all array elements in the array set and the theoretical phase in that direction; A3. Sum the matching results belonging to the matrix set; A4. Traverse the incident directions and select the incident direction that maximizes the summation result as the coarsely estimated incident angle; A5. Based on the roughly estimated angle of incidence, find the corresponding compensation weight by looking up a table.

4. The conformal phased array radar target detection method according to claim 1, characterized in that, S2 includes the following steps: S21. Select a reference plane and obtain the unit normal vector of the reference plane; S22. Calculate the vector difference of each array element relative to the reference point on the reference plane, and obtain the displacement of the array element along the reference plane by the inner product operation of the unit normal vector. S23. Based on the displacement of the array element along the unit normal vector of the reference plane, project the array element onto the reference plane to obtain the virtual coordinates of the array element. S24. Extract the two-dimensional coordinates of the virtual coordinates of each array element, and arrange the array elements according to the two-dimensional coordinates to form a virtual plane.

5. The conformal phased array radar target detection method according to claim 1, characterized in that, S3 includes the following steps: S31. Based on the coordinates of the array elements before projection, obtain the uncompensated geometric phase of each array element before projection. S32. Calculate the geometric phase after projection based on the virtual coordinates of the array elements after projection. S33. Subtract the uncompensated geometric phase before projection from the geometric phase after projection to obtain the phase change of the array element. S34. The phase change of the array elements is compensated to the array element compensation echo to obtain the equivalent echo.

6. The conformal phased array radar target detection method according to claim 1, characterized in that, S4 includes the following sub-steps: S41. Map the coordinates of the virtual plane to the orthogonal basis of the reference plane to obtain the normalized virtual plane; S42. Obtain the curvature weights based on the curvature values ​​corresponding to the curvature tensor of each array element. S43. Multiply the equivalent echo of each array element by the corresponding curvature weight, and perform complex domain multiplication to obtain the energy-enhanced echo. S44. Arrange the energy-enhanced echoes according to the coordinates of the standardized virtual plane, and perform a two-dimensional Fourier transform to obtain the direction-distance two-dimensional spectrum. S45. Extract energy at each position of the direction-distance two-dimensional spectrum to obtain the curvature-enhanced energy spectrum.

7. The conformal phased array radar target detection method according to claim 6, characterized in that, The formula for obtaining the energy-enhanced echo in S43 is as follows: ,in, For the first The energy-enhanced echo of each array element For the first The curvature weights of each array element For the first The equivalent echo of each array element.

8. The conformal phased array radar target detection method according to claim 1, characterized in that, S5 includes the following steps: S51. In the curvature-enhanced energy spectrum, select an r×r target window centered on the location to be detected, where r is the side length of the target window; S52. Construct an R×R window centered on the same detection location. Remove the target window from the R×R window to obtain a ring-shaped background window, where R is the side length of the ring-shaped background window and R > r.

9. The conformal phased array radar target detection method according to claim 1, characterized in that, S6 includes the following sub-steps: S61. Take the average value of all energies in the target window to obtain the target window energy, and take the average value of all energies in the annular background window to obtain the background window energy. S62. Calculate the curvature similarity for each position in the target window, and take the average of all curvature similarities to obtain the target window similarity. S63. Calculate the curvature similarity for each position in the annular background window, and take the average of all curvature similarities to obtain the background window similarity. S64. Based on the ratio of target window energy to background window energy, obtain the target energy salience. S65. The target curvature salience is obtained based on the ratio of the target window similarity to the background window similarity.

10. The conformal phased array radar target detection method according to claim 1, characterized in that, The formula for obtaining the target comprehensive discriminant is: ； in, The overall discrimination of the target at the location to be detected. The target energy salience at the location to be detected. To achieve the desired curvature enhancement, This is the proportionality coefficient.

Images