A radar calibration method and system in a mixed environment

Abstract

The invention relates to the technical field of radar calibration, and discloses a radar calibration method in a mixed environment. The method comprises: collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals, performing adaptive signal separation on the filtered echo signals to identify target echo signals and clutter signals, tracking the identified target echo signals to estimate target states, calibrating target position trajectories through sequential analysis, performing error estimation based on the target position trajectories obtained through analysis and the actual target position estimation, calibrating radar antennas based on the estimated errors, and improving the accuracy of radar calibration.

Description

Technical Field

[0001] The present invention relates to the technical field of radar calibration, and in particular to a radar calibration method and system in a mixed environment. Background Art

[0002] Radar technology has a wide range of applications in military, civil, security and other fields, such as target detection and tracking, weather forecasting and meteorological monitoring, aviation navigation, etc. However, in practical applications, radar systems often need to work in complex and mixed environments, which brings great challenges to radar measurement and calibration. Traditional radar calibration methods are often based on simple environmental assumptions, ignoring the impact of various interference sources and obstacles in mixed environments on radar measurements, resulting in inaccurate and unstable calibration results. In response to this problem, the present invention proposes a radar calibration method and system in a mixed environment, comprehensively analyzes the calibration inaccuracy phenomenon caused by the increase of radar measurement errors in mixed environments, reduces errors through adaptive signal processing, and improves the accuracy of radar calibration. Summary of the invention

[0003] In view of this, the present invention provides a radar calibration method in a mixed environment, the purpose of which is to: 1) organically combine the target tracking technology of wavelet decomposition filtering based on dynamic window, adaptive signal separation and improved long short-term memory network to form a comprehensive radar target recognition and calibration scheme, thereby improving the comprehensiveness and applicability of the system; 2) by adopting the wavelet decomposition method based on dynamic window to filter the radar echo signal, compared with the traditional fixed window filtering method, it can more effectively suppress noise and clutter, and improve the extraction effect of the target signal; 3) when the echo signal after filtering is adaptively separated, the joint diagonalization technology is adopted to effectively distinguish the target echo signal from the clutter signal, improve the extraction accuracy and robustness of the target echo signal, and realize more reliable target signal recognition; 4) introduce the improved long short-term memory network as the main implementation method of sequential analysis, which is used to track and estimate the identified target echo signal and calibrate the target position trajectory, thereby improving the accuracy and robustness of target tracking.

[0004] To achieve the above object, the present invention provides a radar calibration method in a mixed environment, comprising the following steps:

[0005] S1: collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals, wherein wavelet decomposition based on dynamic windows is an implementation method of the filtering process;

[0006] S2: performing adaptive signal separation on the filtered echo signal to identify the target echo signal and the clutter signal, wherein joint diagonalization is an implementation method of the adaptive signal separation;

[0007] S3: Tracking the identified target echo signal to estimate the target state, and calibrating the target position trajectory through sequential analysis, wherein the improved long short-term memory network is an implementation method of the sequential analysis;

[0008] S4: Estimating the error based on the target position trajectory obtained by analysis and the actual target position trajectory, and calibrating the radar antenna based on the estimated error.

[0009] As a further improvement method of the present invention:

[0010] Optionally, in step S1, collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals includes:

[0011] S11: dividing the collected radar echo signal into multiple initial signal segments according to a preset window size, performing spectrum density analysis on the echo signal in each initial signal segment, and dynamically dividing and recombining each initial signal segment signal according to the power spectrum density to obtain a signal segment after adaptive window adjustment;

[0012] S12: performing wavelet transform on the echo signal in each window after adjustment, and calculating the wavelet coefficient of the echo signal in each window;

[0013] S13: Calculate the energy spectrum according to the calculated wavelet coefficients of the echo signal in each window, and delete the wavelet coefficients below the preset threshold according to the preset threshold to obtain the filtered wavelet coefficients. The energy spectrum calculation formula is:

[0014] E(f,t)=|W(f,t)| 2

[0015] in:

[0016] W(f,t) represents the wavelet coefficient of the echo signal at frequency f and time t;

[0017] S14: Reconstruct the signal according to the wavelet coefficients of the echo signal in each window obtained by calculation to obtain the filtered echo signal.

[0018] Optionally, performing spectrum density analysis on the echo signal in each initial signal segment and dynamically dividing and recombining each initial signal segment signal according to the power spectrum density to obtain a signal segment after adaptive window adjustment includes:

[0019] S111: Calculate the amplitude steady-state value and the phase steady-state value of the echo signal in the window, and perform weighted fusion on the calculated amplitude steady-state value and phase steady-state value to obtain a steady-state evaluation value of the echo signal in the window, where the amplitude steady-state value A ssIt is obtained by time averaging the echo signal x(t) in the window. The calculation formula is:

[0020]

[0021] in:

[0022] T represents the length of the echo signal in the window;

[0023] Phase steady-state value φ ss It is obtained by time averaging the phase of the echo signal x(t) within the window. The calculation formula is:

[0024]

[0025] in:

[0026] arg(x(t)) represents the phase of the echo signal;

[0027] S112: Perform steady-state judgment based on the calculated steady-state evaluation value. If the evaluation value does not exceed the specified threshold, it means that the echo signal in the current window is a steady-state signal; if it exceeds the specified threshold, it means that the echo signal in the current window is a non-steady-state signal. Iteratively divide the non-steady-state echo signal into two windows and calculate the steady-state evaluation value of the echo signal in each binary window. Iteratively execute step S112 until the echo signals in all windows of each iteration are steady-state signals.

[0028] S113: performing greedy neighbor window merging on the echo signals of all windows to obtain echo signals after merging multiple windows, wherein the echo signal after merging the windows satisfies that the steady-state evaluation value does not exceed the specified threshold and the steady-state evaluation value calculated by merging the echo signals of the adjacent windows exceeds the specified threshold, and the greedy neighbor window merging merges the neighbor windows according to the window time;

[0029] S114: The calculated echo signal after merging the multiple windows is a signal segment after adaptive window adjustment.

[0030] Optionally, in step S2, performing adaptive signal separation on the filtered echo signal to identify the target echo signal and the clutter signal includes:

[0031] S21: Calculate the autocovariance matrix according to the filtered echo signal and perform normalization operation. The autocovariance matrix calculation formula is:

[0032]

[0033] in:

[0034] C represents the autocovariance matrix of the echo signal;

[0035] X represents the echo signal matrix, with a dimension of N*n, where n represents the echo signal length and N represents the number of echo signal samples;

[0036] X T represents the transposed matrix of the filtered signal matrix;

[0037] S22: Perform eigenvalue decomposition on the autocovariance matrix to obtain eigenvectors and eigenvalues, and obtain the whitening transformation matrix by multiplying the inverse square root of the eigenvalue by the transpose of the eigenvector matrix. The calculation formula is:

[0038]

[0039] in:

[0040] D represents the eigenvalue matrix of the covariance matrix;

[0041] V represents the eigenvector matrix;

[0042] W represents the whitening transformation matrix;

[0043] S23: Jointly diagonalizing the whitening transformation matrix and the autocovariance matrix to obtain a joint diagonalization transformation matrix, wherein the objective function of the joint diagonalization transformation matrix is ​​realized by minimizing the F norm between the whitening transformation matrix and the autocovariance matrix, and the objective function calculation formula is:

[0044]

[0045] in:

[0046] ||.|| F represents the Frobenius norm:

[0047] P represents the joint diagonalization transformation matrix;

[0048] D1 and D2 represent the diagonal matrices corresponding to the whitening transformation matrix and the autocovariance matrix, respectively;

[0049] S24: Calculate the reconstructed echo signal matrix according to the calculated joint diagonal transformation matrix, and reconstruct the target echo signal according to the solved reconstructed echo signal matrix. The calculation formula for echo signal matrix reconstruction is:

[0050] C'=CP

[0051] in:

[0052] C' represents the reconstructed echo signal matrix;

[0053] The energy spectrum is calculated for each row in the echo signal matrix, and the row with the largest energy spectrum in the matrix is ​​taken as the target echo signal, and the remaining rows are clutter signals.

[0054] Optionally, in step S3, tracking the identified target echo signal to estimate the target state, and calibrating the target position trajectory through sequential analysis, includes:

[0055] A target trajectory tracking model is constructed based on an improved long short-term memory network, and a sequential analysis is performed based on the constructed model. The model takes the identified target echo signal sequence as input to infer the target trajectory, and includes an input layer, an attention layer, an LSTM layer, and an output layer. The input layer receives the target echo signal as the input of the model; the attention layer calculates the importance weights of signal values ​​at different positions for target trajectory prediction through Softmax attention; the LSTM layer is used to infer the target signal position; and the output layer is used to output the target trajectory. The specific process includes:

[0056] S31: Extract the features of the target echo signal by combining the attention mechanism to obtain the feature vector of the target echo signal. The specific process includes:

[0057] S311: Map the original echo signal to the attention score space using a fully connected layer;

[0058] S312: Perform a softmax operation on the attention score to obtain a normalized attention weight;

[0059] S313: Multiply the attention weight by the original target echo signal feature to obtain a weighted feature representation. The calculation formula is:

[0060]

[0061] in:

[0062] Feature i represents the value of the i-th feature;

[0063] α ij Represents the attention weight value of the i-th feature of the model at the j-th time step;

[0064] X j represents the value of the target echo signal at the jth time step;

[0065] S32: Combining all calculated target echo signal feature vectors into a target echo signal feature vector sequence, and calculating the target trajectory through sequential analysis, specifically including:

[0066] S321: extract information from the target echo signal feature vector at time t through the forget gate and the input gate respectively, and the calculation formula is:

[0067] f t =σ(W f ·[h t-1,x t ]+b f )

[0068] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0069] in:

[0070] x t Represents the target echo signal feature vector at time step t;

[0071] h t-1 represents the hidden state at time step t-1;

[0072] W f and b f Represents the weight matrix parameters and bias vector parameters of the forget gate;

[0073] W i and b i Represents the weight matrix parameters and bias vector parameters of the input gate;

[0074] σ(·) represents the activation function, f t and i t They represent the information extracted by the forget gate and input gate at time t respectively;

[0075] S322: Estimate the hidden state based on the extracted information. The calculation formula is:

[0076] C t =tanh(W C ·[h t-1 ,[x t ,f t ,i t ]]+b C )

[0077] in:

[0078] C t represents the hidden state estimate vector at time step t;

[0079] W C and b C Represents the hidden state estimation weight matrix parameters and bias vector parameters;

[0080] S323: Calculate the target echo signal position according to the calculated hidden state estimation vector, and splice it to form a target trajectory. The calculation formula is:

[0081] o t =σ(Wo ·[h t-1 ,C t ]+b o )

[0082] in:

[0083] W o and b o Represents the weight matrix parameters and bias vector parameters of the output layer;

[0084] o t Indicates the target echo signal position at time t.

[0085] Optionally, in step S4, performing error estimation based on the target position trajectory obtained by analysis and the actual target position trajectory, and calibrating the radar antenna based on the estimated error, comprises:

[0086] S41: Calculate the error between the target position trajectory and the actual target position trajectory as a basis for correcting the direction of the radar antenna, where the error calculation formula at time t is:

[0087]

[0088] in:

[0089] e t represents the error at time t;

[0090] o t Indicates the target echo signal position at time t;

[0091] represents the actual position of the target at time t;

[0092] S42: According to the error estimation value, the radar antenna is corrected, wherein the correction includes azimuth correction and elevation correction, wherein the azimuth correction is performed on the azimuth angle of the radar antenna, and the elevation correction is performed on the elevation angle of the radar antenna;

[0093] The azimuth correction calculation formula is:

[0094]

[0095] in:

[0096] α k represents the azimuth after the kth round of correction;

[0097] δ1 represents the azimuth adjustment step coefficient;

[0098] Represents the gradient of the error objective function;

[0099] The pitch correction calculation formula is:

[0100]

[0101] in:

[0102] β k represents the pitch angle after the kth round of correction;

[0103] δ2 represents the pitch angle adjustment step coefficient;

[0104] Represents the gradient of the objective function according to the error.

[0105] In order to solve the above problems, the present invention provides a radar calibration system in a mixed environment, the system comprising:

[0106] The signal acquisition module is used to collect radar echo signals and filter the collected echo signals to obtain filtered echo signals, and perform adaptive signal separation on the filtered echo signals to identify target echo signals and clutter signals;

[0107] The target position trajectory calibration module is used to track the identified target echo signal to estimate the target state and calibrate the target position trajectory through sequential analysis;

[0108] The radar calibration module is used to estimate the error based on the analyzed target position trajectory and the actual target position trajectory, and calibrate the radar antenna based on the estimated error.

[0109] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:

[0110] A memory storing at least one instruction;

[0111] Communication interface, enabling electronic equipment to communicate; and

[0112] The processor executes the instructions stored in the memory to implement the radar calibration method in a mixed environment as described above.

[0113] In order to solve the above problem, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is executed by a processor in an electronic device to implement the radar calibration method in a mixed environment described above.

[0114] Compared with the prior art, the present invention proposes a radar calibration method in a mixed environment, which has the following advantages:

[0115] First, this scheme proposes a comprehensive radar target recognition and calibration method, which organically combines the target tracking technology of wavelet decomposition filtering based on dynamic window, adaptive signal separation and improved long short-term memory network to form a comprehensive radar target recognition and calibration scheme, which improves the comprehensiveness and applicability of the system. In addition, by using the wavelet decomposition method based on dynamic window to filter the radar echo signal, it can more effectively suppress noise and clutter and improve the extraction effect of target signal compared with the traditional fixed window filtering method.

[0116] At the same time, when performing adaptive signal separation on the filtered echo signal, this scheme uses the joint diagonalization technology to effectively distinguish the target echo signal from the clutter signal, improve the extraction accuracy and robustness of the target echo signal, and achieve more reliable target signal recognition. It also introduces an improved long short-term memory network as an implementation method of sequential analysis, which is used to track and estimate the identified target echo signal and calibrate the target position trajectory, thereby improving the accuracy and robustness of target tracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0117] Figure 1 A schematic diagram of a flow chart of a radar calibration method in a mixed environment provided by an embodiment of the present invention;

[0118] Figure 2 A functional module diagram of a radar calibration system in a mixed environment provided by an embodiment of the present invention;

[0119] Figure 3 A schematic diagram of the structure of an electronic device for implementing a radar calibration method in a cluttered environment provided by an embodiment of the present invention.

[0120] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0121] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0122] The embodiment of the present application provides a radar calibration method in a mixed environment. The execution subject of the radar calibration method in a mixed environment includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided in the embodiment of the present application. In other words, the radar calibration method in a mixed environment can be executed by software or hardware installed in a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0123] Embodiment 1:

[0124] S1: collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals, wherein wavelet decomposition based on dynamic windows is an implementation method of the filtering process.

[0125] In the step S1, collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals include:

[0126] S11: dividing the collected radar echo signal into multiple initial signal segments according to a preset window size, performing spectrum density analysis on the echo signal in each initial signal segment, and dynamically dividing and recombining each initial signal segment signal according to the power spectrum density to obtain a signal segment after adaptive window adjustment;

[0127] S12: performing wavelet transform on the echo signal in each window after adjustment, and calculating the wavelet coefficient of the echo signal in each window;

[0128] S13: Calculate the energy spectrum according to the calculated wavelet coefficients of the echo signal in each window, and delete the wavelet coefficients below the preset threshold according to the preset threshold to obtain the filtered wavelet coefficients. The energy spectrum calculation formula is:

[0129] E(f,t)=|W(f,t)| 2

[0130] in:

[0131] W(f,t) represents the wavelet coefficient of the echo signal at frequency f and time t;

[0132] S14: Reconstruct the signal according to the wavelet coefficients of the echo signal in each window obtained by calculation to obtain the filtered echo signal.

[0133] The step of performing spectrum density analysis on the echo signal in each initial signal segment and dynamically dividing and recombining each initial signal segment signal according to the power spectrum density to obtain a signal segment after adaptive window adjustment includes:

[0134] S111: Calculate the amplitude steady-state value and the phase steady-state value of the echo signal in the window, and perform weighted fusion on the calculated amplitude steady-state value and phase steady-state value to obtain a steady-state evaluation value of the echo signal in the window, where the amplitude steady-state value A ss It is obtained by time averaging the echo signal x(t) in the window. The calculation formula is:

[0135]

[0136] in:

[0137] T represents the length of the echo signal in the window;

[0138] Phase steady-state value φ ss It is obtained by time averaging the phase of the echo signal x(t) within the window. The calculation formula is:

[0139]

[0140] in:

[0141] arg(x(t)) represents the phase of the echo signal;

[0142] S112: Perform steady-state judgment based on the calculated steady-state evaluation value. If the evaluation value does not exceed the specified threshold, it means that the echo signal in the current window is a steady-state signal; if it exceeds the specified threshold, it means that the echo signal in the current window is a non-steady-state signal. Iteratively divide the non-steady-state echo signal into two windows and calculate the steady-state evaluation value of the echo signal in each binary window. Iteratively execute step S112 until the echo signals in all windows of each iteration are steady-state signals.

[0143] S113: performing greedy neighbor window merging on the echo signals of all windows to obtain echo signals after merging multiple windows, wherein the echo signal after merging the windows satisfies that the steady-state evaluation value does not exceed the specified threshold and the steady-state evaluation value calculated by merging the echo signals of the adjacent windows exceeds the specified threshold, and the greedy neighbor window merging merges the neighbor windows according to the window time;

[0144] S114: The calculated echo signal after merging the multiple windows is a signal segment after adaptive window adjustment.

[0145] S2: performing adaptive signal separation on the filtered echo signal to identify the target echo signal and the clutter signal, wherein joint diagonalization is an implementation method of the adaptive signal separation.

[0146] In the step S2, adaptive signal separation is performed on the filtered echo signal to identify the target echo signal and the clutter signal, including:

[0147] S21: Calculate the autocovariance matrix according to the filtered echo signal and perform normalization operation. The autocovariance matrix calculation formula is:

[0148]

[0149] in:

[0150] C represents the autocovariance matrix of the echo signal;

[0151] X represents the echo signal matrix, with a dimension of N*n, where n represents the echo signal length and N represents the number of echo signal samples;

[0152] X T represents the transposed matrix of the filtered signal matrix;

[0153] S22: Perform eigenvalue decomposition on the autocovariance matrix to obtain eigenvectors and eigenvalues, and obtain the whitening transformation matrix by multiplying the inverse square root of the eigenvalue by the transpose of the eigenvector matrix. The calculation formula is:

[0154]

[0155] in:

[0156] D represents the eigenvalue matrix of the covariance matrix;

[0157] V represents the eigenvector matrix;

[0158] W represents the whitening transformation matrix;

[0159] S23: Jointly diagonalizing the whitening transformation matrix and the autocovariance matrix to obtain a joint diagonalization transformation matrix, wherein the objective function of the joint diagonalization transformation matrix is ​​realized by minimizing the F norm between the whitening transformation matrix and the autocovariance matrix, and the objective function calculation formula is:

[0160]

[0161] in:

[0162] ||.|| F represents the Frobenius norm:

[0163] P represents the joint diagonalization transformation matrix;

[0164] D1 and D2 represent the diagonal matrices corresponding to the whitening transformation matrix and the autocovariance matrix, respectively;

[0165] S24: Calculate the reconstructed echo signal matrix according to the calculated joint diagonal transformation matrix, and reconstruct the target echo signal according to the solved reconstructed echo signal matrix. The calculation formula for echo signal matrix reconstruction is:

[0166] C'=CP

[0167] in:

[0168] C' represents the reconstructed echo signal matrix;

[0169] The energy spectrum is calculated for each row in the echo signal matrix, and the row with the largest energy spectrum in the matrix is ​​taken as the target echo signal, and the remaining rows are clutter signals.

[0170] S3: Tracking the identified target echo signal to estimate the target state, and calibrating the target position trajectory through sequential analysis, wherein the improved long short-term memory network is an implementation method of the sequential analysis.

[0171] In the step S3, the target echo signal obtained by identification is tracked to estimate the target state, and the target position trajectory is calibrated by sequential analysis, including:

[0172] A target trajectory tracking model is constructed based on an improved long short-term memory network, and a sequential analysis is performed based on the constructed model. The model takes the identified target echo signal sequence as input to infer the target trajectory, and includes an input layer, an attention layer, an LSTM layer, and an output layer. The input layer receives the target echo signal as the input of the model; the attention layer calculates the importance weights of signal values ​​at different positions for target trajectory prediction through Softmax attention; the LSTM layer is used to infer the target signal position; and the output layer is used to output the target trajectory. The specific process includes:

[0173] S31: Extract the features of the target echo signal by combining the attention mechanism to obtain the feature vector of the target echo signal. The specific process includes:

[0174] S311: Map the original echo signal to the attention score space using a fully connected layer;

[0175] S312: Perform a softmax operation on the attention score to obtain a normalized attention weight;

[0176] S313: Multiply the attention weight by the original target echo signal feature to obtain a weighted feature representation. The calculation formula is:

[0177]

[0178] in:

[0179] Feature i represents the value of the i-th feature;

[0180] α ij Represents the attention weight value of the i-th feature of the model at the j-th time step;

[0181] X j represents the value of the target echo signal at the jth time step;

[0182] S32: Combining all calculated target echo signal feature vectors into a target echo signal feature vector sequence, and calculating the target trajectory through sequential analysis, specifically including:

[0183] S321: extract information from the target echo signal feature vector at time t through the forget gate and the input gate respectively, and the calculation formula is:

[0184] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0185] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0186] in:

[0187] x t Represents the target echo signal feature vector at time step t;

[0188] h t-1 represents the hidden state at time step t-1;

[0189] W f and b f Represents the weight matrix parameters and bias vector parameters of the forget gate;

[0190] W i and b i Represents the weight matrix parameters and bias vector parameters of the input gate;

[0191] σ(·) represents the activation function, f t and i t They represent the information extracted by the forget gate and input gate at time t respectively;

[0192] S322: Estimate the hidden state based on the extracted information. The calculation formula is:

[0193] C t =tanh(W C ·[h t-1 ,[x t ,f t ,i t ]]+b C )

[0194] in:

[0195] C t represents the hidden state estimate vector at time step t;

[0196] W C and b C Represents the hidden state estimation weight matrix parameters and bias vector parameters;

[0197] S323: Calculate the target echo signal position according to the calculated hidden state estimation vector, and splice it to form a target trajectory. The calculation formula is:

[0198] o t =σ(W o ·[h t-1 ,C t ]+b o )

[0199] in:

[0200] W o and b o Represents the weight matrix parameters and bias vector parameters of the output layer;

[0201] o t Indicates the target echo signal position at time t.

[0202] S4: Estimating the error based on the target position trajectory obtained by analysis and the actual target position trajectory, and calibrating the radar antenna based on the estimated error.

[0203] In the step S4, error estimation is performed based on the target position trajectory obtained by analysis and the actual target position trajectory, and the radar antenna is calibrated based on the estimated error, including:

[0204] S41: Calculate the error between the target position trajectory and the actual target position trajectory as a basis for correcting the direction of the radar antenna, where the error calculation formula at time t is:

[0205]

[0206] in:

[0207] e t represents the error at time t;

[0208] o t Indicates the target echo signal position at time t;

[0209] represents the actual position of the target at time t;

[0210] S42: According to the error estimation value, the radar antenna is corrected, wherein the correction includes azimuth correction and elevation correction, wherein the azimuth correction is performed on the azimuth angle of the radar antenna, and the elevation correction is performed on the elevation angle of the radar antenna;

[0211] The azimuth correction calculation formula is:

[0212]

[0213] in:

[0214] α k represents the azimuth after the kth round of correction;

[0215] δ1 represents the azimuth adjustment step coefficient;

[0216] Represents the gradient of the error objective function;

[0217] The pitch correction calculation formula is:

[0218]

[0219] in:

[0220] β k represents the pitch angle after the kth round of correction;

[0221] δ2 represents the pitch angle adjustment step coefficient;

[0222] Represents the gradient of the objective function according to the error.

[0223] Embodiment 2:

[0224] like Figure 2 , which is a functional module diagram of a radar calibration system in a cluttered environment provided by an embodiment of the present invention, which can implement the radar calibration method in a cluttered environment in Example 1.

[0225] The radar calibration system 100 in a cluttered environment of the present invention can be installed in an electronic device. According to the functions to be implemented, the radar calibration system in a cluttered environment can include a patrol signal acquisition module 101, a calibration target position trajectory module 102, and a radar calibration module 103. The module of the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0226] The signal acquisition module 101 is used to acquire radar echo signals and filter the acquired echo signals to obtain filtered echo signals, and perform adaptive signal separation on the filtered echo signals to identify target echo signals and clutter signals;

[0227] The target position trajectory calibration module 102 is used to track the identified target echo signal to estimate the target state and calibrate the target position trajectory through sequential analysis;

[0228] The radar calibration module 103 is used to estimate the error based on the target position trajectory obtained by analysis and the actual target position trajectory, and calibrate the radar antenna based on the estimated error. Figure 1 The radar calibration method in a mixed environment is the same technical means as described in , and can produce the same technical effects, so I will not go into details here.

[0229] Embodiment 3:

[0230] like Figure 3 FIG. 1 is a schematic diagram of the structure of an electronic device for implementing a radar calibration method in a cluttered environment provided by an embodiment of the present invention.

[0231] The electronic device 1 may include a processor 10 , a memory 11 , a communication interface 13 and a bus, and may also include a computer program stored in the memory 11 and executable on the processor 10 , such as a program 12 .

[0232] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a mobile hard disk of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 1. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 may be used not only to store application software and various types of data installed in the electronic device 1, such as the code of the program 12, etc., but also to temporarily store data that has been output or is to be output.

[0233] The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips, etc. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules stored in the memory 11 (such as the program 12 for implementing the radar calibration method in a mixed environment), and calls the data stored in the memory 11 to execute various functions of the electronic device 1 and process data.

[0234] The communication interface 13 may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices, and to achieve connection and communication between internal components of the electronic device.

[0235] The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0236] Figure 3 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0237] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, etc. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0238] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.

[0239] The program 12 stored in the memory 11 in the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve:

[0240] Collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals;

[0241] Adaptively separate the filtered echo signal to identify the target echo signal and clutter signal;

[0242] Track the identified target echo signal to estimate the target state, and calibrate the target position trajectory through sequential analysis;

[0243] The error is estimated based on the target position trajectory obtained by analysis and the actual target position trajectory, and the radar antenna is calibrated based on the estimated error.

[0244] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figures 1 to 3 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0245] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. And the terms "including", "comprising" or any other variants thereof in this article are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "including a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0246] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A radar calibration method in a mixed environment, characterized in that: The method comprises: S1: collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals, wherein wavelet decomposition based on dynamic windows is an implementation method of the filtering process; Collecting radar echo signals and filtering the collected echo signals to obtain filtered echo signals, including: S11: dividing the collected radar echo signal into multiple initial signal segments according to a preset window size, performing spectrum density analysis on the echo signal in each initial signal segment, and dynamically dividing and recombining each initial signal segment signal according to the power spectrum density to obtain a signal segment after adaptive window adjustment; S12: performing wavelet transform on the echo signal in each window after adjustment, and calculating the wavelet coefficient of the echo signal in each window; S13: Calculate the energy spectrum according to the calculated wavelet coefficients of the echo signal in each window, and delete the wavelet coefficients below the preset threshold according to the preset threshold to obtain the filtered wavelet coefficients. The energy spectrum calculation formula is: E(f,t)=|W(f,t)| 2 in: W(f,t) represents the wavelet coefficient of the echo signal at frequency f and time t; S14: performing signal reconstruction according to the wavelet coefficients of the echo signal in each window obtained by calculation after filtering to obtain a filtered echo signal; Performing spectrum density analysis on the echo signal in each initial signal segment and dynamically segmenting and recombining each initial signal segment according to the power spectrum density to obtain a signal segment after adaptive window adjustment, including: S111: Calculate the amplitude steady-state value and the phase steady-state value of the echo signal in the window, and perform weighted fusion on the calculated amplitude steady-state value and phase steady-state value to obtain a steady-state evaluation value of the echo signal in the window, where the amplitude steady-state value A ss It is obtained by time averaging the echo signal x(t) in the window. The calculation formula is: in: T represents the length of the echo signal in the window; Phase steady-state value φ ss It is obtained by time averaging the phase of the echo signal x(t) within the window. The calculation formula is: in: arg(x(t)) represents the phase of the echo signal; S112: Perform steady-state judgment based on the calculated steady-state evaluation value. If the evaluation value does not exceed the specified threshold, it means that the echo signal in the current window is a steady-state signal; if it exceeds the specified threshold, it means that the echo signal in the current window is a non-steady-state signal. Iteratively divide the non-steady-state echo signal into two windows and calculate the steady-state evaluation value of the echo signal in each binary window. Iteratively execute step S112 until the echo signals in all windows of each iteration are steady-state signals. S113: performing greedy neighbor window merging on the echo signals of all windows to obtain echo signals after merging multiple windows, wherein the echo signal after merging the windows satisfies that the steady-state evaluation value does not exceed the specified threshold and the steady-state evaluation value calculated by merging the echo signals of the adjacent windows exceeds the specified threshold, and the greedy neighbor window merging merges the neighbor windows according to the window time; S114: The calculated echo signal after merging the multiple windows is a signal segment after adaptive window adjustment; S2: performing adaptive signal separation on the filtered echo signal to identify the target echo signal and the clutter signal, wherein joint diagonalization is an implementation method of the adaptive signal separation; Adaptively separate the filtered echo signal to identify the target echo signal and clutter signal, including: S21: Calculate the autocovariance matrix according to the filtered echo signal and perform normalization operation. The autocovariance matrix calculation formula is: in: C represents the autocovariance matrix of the echo signal; X represents the echo signal matrix, with a dimension of N*n, where n represents the echo signal length and N represents the number of echo signal samples; X T represents the transposed matrix of the filtered signal matrix; S22: Perform eigenvalue decomposition on the autocovariance matrix to obtain eigenvectors and eigenvalues, and obtain the whitening transformation matrix by multiplying the inverse square root of the eigenvalue by the transpose of the eigenvector matrix. The calculation formula is: in: D represents the eigenvalue matrix of the covariance matrix; V represents the eigenvector matrix; W represents the whitening transformation matrix; S23: Jointly diagonalizing the whitening transformation matrix and the autocovariance matrix to obtain a joint diagonalization transformation matrix, wherein the objective function of the joint diagonalization transformation matrix is ​​realized by minimizing the F norm between the whitening transformation matrix and the autocovariance matrix, and the objective function calculation formula is: in: ||.|| F represents the Frobenius norm: P represents the joint diagonalization transformation matrix; D1 and D2 represent the diagonal matrices corresponding to the whitening transformation matrix and the autocovariance matrix, respectively; S24: Calculate the reconstructed echo signal matrix according to the calculated joint diagonal transformation matrix, and reconstruct the target echo signal according to the solved reconstructed echo signal matrix. The calculation formula for echo signal matrix reconstruction is: C′=CP in: C' represents the reconstructed echo signal matrix; Calculate the energy spectrum of each row in the echo signal matrix, and take the row with the largest energy spectrum in the matrix as the target echo signal, and the remaining rows as clutter signals; S3: Tracking the identified target echo signal to estimate the target state, and calibrating the target position trajectory through sequential analysis, wherein the improved long short-term memory network is an implementation method of the sequential analysis; S4: Estimating the error based on the target position trajectory obtained by analysis and the actual target position trajectory, and calibrating the radar antenna based on the estimated error.

2. The radar calibration method in a mixed environment as claimed in claim 1, characterized in that: In the step S3, the target echo signal obtained by identification is tracked to estimate the target state, and the target position trajectory is calibrated by sequential analysis, including: A target trajectory tracking model is constructed based on an improved long short-term memory network, and a sequential analysis is performed based on the constructed model. The model takes the identified target echo signal sequence as input to infer the target trajectory, and includes an input layer, an attention layer, an LSTM layer, and an output layer. The input layer receives the target echo signal as the input of the model; the attention layer calculates the importance weights of signal values ​​at different positions for target trajectory prediction through Softmax attention; the LSTM layer is used to infer the target signal position; and the output layer is used to output the target trajectory. The specific process includes: S31: Extract the features of the target echo signal by combining the attention mechanism to obtain the feature vector of the target echo signal. The specific process includes: S311: Map the original echo signal to the attention score space using a fully connected layer; S312: Perform a softmax operation on the attention score to obtain a normalized attention weight; S313: Multiply the attention weight by the original target echo signal feature to obtain a weighted feature representation. The calculation formula is: in: Feature i represents the value of the i-th feature; α ij Represents the attention weight value of the i-th feature of the model at the j-th time step; X j represents the value of the target echo signal at the jth time step; S32: Combining all calculated target echo signal feature vectors into a target echo signal feature vector sequence, and calculating the target trajectory through sequential analysis, specifically including: S321: extract information from the target echo signal feature vector at time t through the forget gate and the input gate respectively, and the calculation formula is: f t =σ(W f ·[h t-1 ,x t ]+b f ) i t =σ(W i ·[h t-1 ,x t ]+b i ) in: x t Represents the target echo signal feature vector at time step t; h t-1 represents the hidden state at time step t-1; W f and b f Represents the weight matrix parameters and bias vector parameters of the forget gate; W i and b i Represents the weight matrix parameters and bias vector parameters of the input gate; σ(·) represents the activation function, f t and i t They represent the information extracted by the forget gate and the input gate at time t respectively; S322: Estimate the hidden state based on the extracted information. The calculation formula is: C t =tanh(W C ·[h t-1 ,[x t ,f t ,i t ]]+b C ) in: C t represents the hidden state estimate vector at time step t; W C and b C Represents the hidden state estimation weight matrix parameters and bias vector parameters; S323: Calculate the target echo signal position according to the calculated hidden state estimation vector, and splice it to form a target trajectory. The calculation formula is: the t =σ(W o ·[h t-1 ,C t ]+b o ) in: W o and b o Represents the weight matrix parameters and bias vector parameters of the output layer; o t Indicates the target echo signal position at time t.

3. The radar calibration method in a mixed environment as claimed in claim 1, characterized in that: In the step S4, error estimation is performed based on the target position trajectory obtained by analysis and the actual target position trajectory, and the radar antenna is calibrated based on the estimated error, including: S41: Calculate the error between the target position trajectory and the actual target position trajectory as a basis for correcting the direction of the radar antenna, where the error calculation formula at time t is: in: e t represents the error at time t; o t Indicates the target echo signal position at time t; represents the actual position of the target at time t; S42: According to the error estimation value, the radar antenna is corrected, wherein the correction includes azimuth correction and elevation correction, wherein the azimuth correction is performed on the azimuth angle of the radar antenna, and the elevation correction is performed on the elevation angle of the radar antenna; The azimuth correction calculation formula is: in: α k represents the azimuth after the kth round of correction; δ1 represents the azimuth adjustment step coefficient; Represents the gradient of the error objective function; The pitch correction calculation formula is: in: β k represents the pitch angle after the kth round of correction; δ2 represents the pitch angle adjustment step coefficient; Represents the gradient of the objective function according to the error.

4. A radar calibration system in a mixed environment, characterized in that: The system comprises: The signal acquisition module is used to collect radar echo signals and filter the collected echo signals to obtain filtered echo signals, and perform adaptive signal separation on the filtered echo signals to identify target echo signals and clutter signals; The target position trajectory calibration module is used to track the identified target echo signal to estimate the target state and calibrate the target position trajectory through sequential analysis; A radar calibration module is used to estimate the error based on the target position trajectory obtained by analysis and the actual target position trajectory, and calibrate the radar antenna based on the estimated error to implement a radar calibration method in a mixed environment as described in any one of claims 1 to 3.

Images