Interferometer system antenna real-time calibration method based on dynamic fitting

Abstract

According to the interferometer system antenna real-time calibration method based on dynamic fitting, a phase correction table building method and a physical model parameter method thought are combined, a phase difference correction number and a matched weight table updating method are provided, a mathematical model calibration parameter fitting thought is combined, and the calibration accuracy of the interferometer system antenna is improved. Real-time parameter fitting updating is carried out on the phase correction table on line, and the thought is suitable for real-time processing of the motion platform and can effectively shorten the table building period; a judgment strategy based on azimuth pitch angle external polygon table building coverage and mathematical model parameter fitting residual conditions is designed, and the requirement for automatic calibration and use of the electromagnetic spectrum monitoring motion platform can be met.

Description

Technical Field

[0001] This invention belongs to the field of electromagnetic spectrum monitoring data processing, and relates to a real-time calibration method for interferometer-based antennas based on dynamic fitting. Background Technology

[0002] Interferometer-based electromagnetic spectrum monitoring sensors are an important positioning system in the electromagnetic field, enabling high-precision location determination of key electromagnetic spectrum monitoring targets based on direction-finding principles. Improving the positioning accuracy of monitoring targets relies heavily on the high-precision calibration of the interferometer antenna. Although the antenna position and phase correction parameters have been precisely measured during development, factors such as vibrations during sensor operation, the influence of the space thermal environment, or stress release can cause changes in the antenna parameters, necessitating real-time calibration. Traditional interferometer antenna calibration methods include the physical model parameter method and the phase correction number tabulation method. The physical model parameter method requires collecting multiple batches of monitoring data from the reference spectrum target under different incident angles. Simultaneously, it calculates the calibration parameters based on a constructed physical parameter model of the interferometer antenna calibration (model parameters typically include antenna position and phase correction coefficients). Because it requires accumulating multiple batches of data to ensure the determination of the globally optimal solution for the calibration parameters, it is suitable for post-processing of ground-based data. The phase correction table construction method can inversely calculate the theoretical phase difference based on the location of the reference spectrum target, and construct phase correction tables under different azimuth and elevation angles. This method does not rely on historical observation data and is suitable for online real-time processing. However, due to the limited number of known reference spectrum targets, the cycle for completing the correction parameter tables for different frequency bands, azimuth angles, and elevation angles is relatively long, making it difficult to improve the positioning accuracy of electromagnetic spectrum monitoring sensors in a short period of time. Summary of the Invention

[0003] This invention proposes a real-time calibration method for interferometer antennas based on dynamic fitting. It combines the phase correction table construction method and the physical model parameter method. The phase correction table is constructed online and the mathematical model calibration parameters are fitted to the phase correction table in real time. Based on the table coverage of the azimuth and elevation angle circumscribed polygon and the residual conditions of the mathematical model parameter fitting, the fitted data model calibration parameters are dynamically applied.

[0004] The technical solution for achieving this invention is as follows: a real-time calibration method for interferometer antennas based on dynamic fitting. This method combines the phase correction table construction method and the physical model parameter method, proposing an update method for the phase correction table and its corresponding weight table. Combining the mathematical model calibration parameter fitting approach, the phase correction table is updated online in real-time. The specific steps are as follows:

[0005] Step 1: Identify the reference spectrum target online and extract the associated pulses of the reference spectrum target, then proceed to Step 2.

[0006] Step 2: Calculate the phase difference correction in real time using the correlated pulse sequence of the reference spectrum target. Then calculate the phase difference correction. Matching weight table Proceed to step 3.

[0007] Step 3: Solve for the mathematical model parameters using a fitting method based on the mathematical model. ,Will As a calibration parameter, proceed to step 4.

[0008] Step 4: Based on mathematical model parameters Evaluation of the fitting residuals and phase difference corrections Availability, based on phase difference correction circumscribed polygon Evaluate mathematical model parameters The scope of application is determined, proceed to step 5.

[0009] Step 5: Adaptively use phase difference correction The phase difference was corrected, and the unambiguous direction finding was performed again to obtain the corrected incident azimuth angle of the original monopulse signal in the antenna coordinate system. and incident elevation angle .

[0010] When performing single-pulse direction finding, the incident azimuth angle of the original single-pulse signal in the antenna coordinate system is used as the basis. The incident elevation angle of the original single-pulse signal in the antenna coordinate system ; Search grid Corresponding valid judgment table The 0 or 1 attribute, when the decision table value is 1, uses a mathematical model parameter fitting method to obtain the corresponding... Modify the measured phase difference Then, the unambiguity direction finding was performed again to obtain the corrected incident azimuth angle based on the original monopulse signal in the antenna coordinate system. and incident elevation angle ; For azimuth grid numbering, Number the pitch angle grid.

[0011] Compared with the prior art, the significant advantages of this invention are:

[0012] 1) A method for updating the phase difference correction number and its corresponding weight table is proposed. Combining the mathematical model calibration parameter fitting idea, the phase correction number table is updated in real time online. This idea is suitable for online real-time processing and can effectively shorten the table building cycle of calibration parameter table.

[0013] 2) A decision strategy based on the coverage of the table of circumscribed polygons of azimuth and elevation angles and the fitting residual conditions of mathematical model parameters was designed, which can meet the needs of online real-time autonomous edge calibration of electromagnetic spectrum monitoring sensors. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the real-time calibration method in an embodiment of the present invention.

[0015] Figure 2 A schematic diagram of parameter fitting for calibration of a mathematical model based on a phase correction table.

[0016] Figure 3 A schematic diagram illustrating the applicable range of parameters for evaluating mathematical models based on the circumscribed polygon of the phase difference correction. Detailed Implementation

[0017] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0018] This invention proposes a real-time calibration method for interferometer antennas based on dynamic fitting, comprising the following steps:

[0019] Step 1: Identify reference spectrum targets online and extract the associated pulses of the reference spectrum targets.

[0020] Most existing electromagnetic spectrum monitoring sensors have the ability to receive and process electromagnetic spectrum signals simultaneously, that is, to detect, sort, and locate radiation sources from electromagnetic spectrum signals (taking pulse signals as an example). ( Represents the radiation source generated within a single processing cycle. Parameter information of (number) and location information ,in Frequency information representing the radiation source, Represents the recirculation information of the radiation source. Pulse width information representing the radiation source.

[0021] The real-time calibration process requires rapidly acquiring the pulse sequence of the radiation source. ,in Representative radiation source The number of associated pulses is considered; therefore, a processing cycle with a period of one second is designed. During the radiation source sorting and positioning process, the associated effective pulse sequence information is automatically stored, i.e., the radiation source... Associated pulse sequence .

[0022] radiation source Parameter information With reference spectrum target Parameter information Perform matching, and simultaneously target the radiation source. Location information With reference spectrum target Location information Perform matching, where Representative reference spectrum target The reference spectrum target is automatically selected based on its number. Corresponding radiation source The matching and recognition criteria are as follows:

[0023] (1);

[0024] in, Represents the frequency threshold. This represents the recycle threshold. Represents the pulse width threshold. Represents the distance threshold; Represents frequency information. Represents double-circular information, This represents pulse width information.

[0025] Using radiation sources that meet the matching identification conditions and reference spectrum target Construct a reference spectrum target and its associated pulse sequence .

[0026] Proceed to step 2.

[0027] Step 2: Calculate the phase difference correction in real time using the correlated pulse sequence of the reference spectrum target. Then calculate the phase difference correction. Matching weight table .

[0028] Step 2.1, Design a table of phase difference corrections. ,in, The incident azimuth angle of the original monopulse signal in the antenna coordinate system. The incident elevation angle of the original monopulse signal in the antenna coordinate system. Represents the minimum value in the frequency range. Represents the maximum value within the frequency range. This indicates that the corresponding number is... Frequency range grid, azimuth grid Pitch angle grid Phase difference correction under certain conditions. Typically, the incident elevation and azimuth angle grids are divided into 1° to 5°, and the incident elevation and azimuth angle grid ranges from -65° to 65° (determined by the monitoring field of view). The frequency range is divided into several hundred MHz.

[0029] Step 2.2, based on the theoretical phase difference and measured phase difference Solve for the phase difference correction. Then, the frequency range grid, azimuth grid, and elevation grid are traversed to construct a phase difference correction table.

[0030] according to The frequency information in the full-pulse descriptor word is used to obtain the phase difference correction. Frequency grid number The phase difference correction is obtained based on the pitch angle information in the full-pulse descriptor. Pitch angle grid number The phase difference correction is obtained based on the azimuth information in the full-pulse descriptor. Azimuth grid number .

[0031] Calculate the theoretical phase difference At that time, with pulse sequence The arrival time of a single pulse is used to interpolate and obtain the attitude information of the electromagnetic spectrum sensor. and the location information of the electromagnetic spectrum sensor Then the pulse relative to the incident vector of the electromagnetic spectrum sensor antenna Represented as:

[0032] (2);

[0033] in, The rotation matrix representing the rotation from the WGS84 coordinate system to the electromagnetic spectrum sensor body coordinate system is given by... Calculated This represents the rotation matrix from the electromagnetic spectrum sensor body to the antenna coordinate system.

[0034] The baseline three-dimensional vector of the antenna pair in the antenna coordinate system is The theoretical phase difference Represented as:

[0035] (3);

[0036] Through theoretical phase difference and measured phase difference The phase difference correction can be obtained by subtracting. , Indicates pulse The wavelength.

[0037] Step 2.3, calculate the phase difference correction. Matching weight table :

[0038] Since the same phase difference correction table of an actual electromagnetic spectrum sensor will be repeatedly overwritten by multiple sets of pulses during the operation of the motion platform, a corresponding weight table is designed to achieve dynamic accumulation and updating of the phase difference correction table. , and The number of cumulative pulses with the same dimension and scale, corrected for existing phase difference values. express;

[0039] During the continuous accumulation process, when a new phase difference correction is calculated... and existing phase difference corrections within the same grid To synthesize:

[0040] (4),

[0041] At the same time, update the corresponding weight table. .

[0042] Through each step in step 2, the phase difference correction can be dynamically and continuously accumulated and updated, adapting to the online dynamic real-time processing of the motion platform, and providing discrete observations for subsequent phase correction plane fitting.

[0043] Proceed to step 3.

[0044] Step 3: Solve for the mathematical model parameters using a fitting method based on the mathematical model. ,Will As calibration parameters.

[0045] For certain electromagnetic spectrum sensors on specific platforms, the number of phase correction tables that can be covered during a single power-on operation is limited. Completely filling these tables requires a large number of power-on tasks, resulting in a significant workload. The phase corrections, caused by interferometer model parameters (antenna position and phase correction coefficients, etc.), can typically be expressed with respect to the incident azimuth angle. and pitch angle Fitting to a low-order planar model of the parameter variables, a planar fitting of the correction values ​​can be performed based on the effective values ​​in the continuously accumulating phase correction value table, thus creating a second-order data model. As shown below:

[0046] (5)

[0047] in, , , Represents the higher-order terms of the planar model. , , This represents the lower-order terms of the planar model.

[0048] For the first For each frequency range, construct the least squares equation:

[0049] (6)

[0050] in:

[0051] Second-order data model ,

[0052] parameter ,

[0053] Mathematical model parameters ,

[0054] Mathematical model parameters It can be obtained through the least squares equation:

[0055] (7),

[0056] represent The weight vector of each effective observation. Indicates the first The number of effective observations within a frequency range is determined by the cumulative pulse count of the effective observation grid. Decide, Indicates to Mathematical model parameters Find the partial derivative.

[0057] If the antenna design has good phase consistency, then the phase difference correction plane can be fitted using a low-order term of the planar model. The fitted model can be simplified as follows:

[0058] (8),

[0059] Both high-order and low-order models can be used as the final online calibration model, and their mathematical model parameters can be... These are considered as online calibration coefficients. By fitting the existing observation information in the phase difference correction table, the online calibration coefficients can be quickly obtained, reducing the need for individual calibration and table filling of phase difference corrections, and significantly shortening the calibration cycle based on the traditional grid-by-grid phase correction table building method.

[0060] Proceed to step 4.

[0061] Step 4: Based on mathematical model parameters Evaluation of the fitting residuals and phase difference corrections Availability, based on phase difference correction circumscribed polygon Evaluate mathematical model parameters Scope of application.

[0062] Step 4.1, based on mathematical model parameters The phase difference correction of the fitting residual evaluation Availability, Fitting Residuals Represented as:

[0063] (9),

[0064] During the statistical process, statistics The number of grids, This represents the threshold; when the number of statistical grids is greater than 90% of the total number of grids, the fitted plane interval is considered valid.

[0065] Step 4.2, based on phase difference correction number circumscribed polygon Evaluate mathematical model parameters Scope of application.

[0066] During the operation of the motion platform, as the phase difference correction table is dynamically updated, based on the platform's motion characteristics, the grid range that can be covered by a single reference radiation source in a single mission is linearly distributed, such as... Figure 2 As shown, the phase difference correction number The process involves dynamic accumulation; through steps 1 to 3, dynamic mathematical model parameters can be obtained. Updates are being made to allow for real-time labeling of available phase difference corrections. The range, design and phase difference correction number Accompanying valid judgment form The above-mentioned valid judgment table and The dimensions and scale are the same.

[0067] Valid Judgment Form Store 0 or 1, where 0 is an invalid identifier and 1 is a valid identifier, in order to determine the validity of the decision table. Valid identifiers, i.e., mathematical model parameters Scope of application: Based on phase difference correction. circumscribed polygon To make a valid judgment table The assignment.

[0068] like Figure 3 As shown, let to These represent the six endpoints of the effective observation coverage grid after three calibration tasks, arranged in a linear pattern. to Constructing the circumscribed polygon When determining the phase difference correction number grid circumscribed polygon When inside, the grid Corresponding valid judgment table The value is assigned to 1; when judging the phase difference correction number. grid circumscribed polygon When outside, the grid Corresponding valid judgment table The value is assigned to 0, that is:

[0069] (10)

[0070] Through steps 3 and 4, the mathematical model parameters are obtained. Online dynamic fitting and updating, and automatic determination of mathematical model parameters. The scope of application and the range of effectiveness.

[0071] Step 5: Adaptively use phase difference correction The phase difference was corrected, and the unambiguous direction finding was performed again to obtain the corrected incident azimuth angle of the original monopulse signal in the antenna coordinate system. and incident elevation angle ;

[0072] When performing single-pulse direction finding, the incident azimuth angle of the original single-pulse signal in the antenna coordinate system is used as the basis. The incident elevation angle of the original single-pulse signal in the antenna coordinate system ; Search grid Corresponding valid judgment table The 0 or 1 attribute, when the decision table value is 1, uses a mathematical model parameter fitting method to obtain the corresponding... Modify the measured phase difference Then, the unambiguity direction finding was performed again to obtain the corrected incident azimuth angle based on the original monopulse signal in the antenna coordinate system. and incident elevation angle ; For azimuth grid numbering, Number the pitch angle grid.

[0073] Adaptive use of phase difference correction When performing single-pulse direction finding, the frequency grid number is selected based on the frequency of the original single-pulse signal. Based on the incident azimuth angle of the original single-pulse signal in the antenna coordinate system ( (representing the azimuth grid number), and the incident elevation angle of the original monopulse signal in the antenna coordinate system. ( (Represents pitch angle grid number), search for grid. Corresponding valid judgment table The 0 or 1 attribute, when the decision table value is 1, uses equation (5) to obtain the corresponding value. Modify the measured phase difference Afterwards, the unambiguous direction finding can be performed again to obtain the corrected incident azimuth angle based on the original monopulse signal in the antenna coordinate system. and incident elevation angle .

Claims

1. A real-time calibration method for interferometer antennas based on dynamic fitting, characterized in that: By combining the methods of establishing phase correction tables and physical model parameters, a method for updating the phase difference correction table and its corresponding weight table is proposed. Combined with the method of fitting parameters for mathematical model calibration, the phase correction table is updated online in real time through parameter fitting.

2. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 1, characterized in that, The specific steps are as follows: Step 1: Identify the reference spectrum target online and extract the associated pulses of the reference spectrum target, then proceed to Step 2; Step 2: Calculate the phase difference correction in real time using the correlated pulse sequence of the reference spectrum target. Then calculate the phase difference correction. Matching weight table Proceed to step 3; Step 3: Solve for the mathematical model parameters using a fitting method based on the mathematical model. ,Will As calibration parameters, proceed to step 4; Step 4: Based on mathematical model parameters Evaluation of the phase difference correction of the fitting residual Availability, based on phase difference correction circumscribed polygon Evaluate mathematical model parameters Scope of application, proceed to step 5; Step 5: Adaptively use phase difference correction The phase difference was corrected, and the unambiguous direction finding was performed again to obtain the corrected incident azimuth angle of the original monopulse signal in the antenna coordinate system. and incident elevation angle ; When performing single-pulse direction finding, the incident azimuth angle of the original single-pulse signal in the antenna coordinate system is used as the basis. The incident elevation angle of the original single-pulse signal in the antenna coordinate system ; Find the grid Corresponding valid judgment table The 0 or 1 attribute, when the decision table value is 1, uses a mathematical model parameter fitting method to obtain the corresponding... Modify the measured phase difference Then, the unambiguity direction finding was performed again to obtain the corrected incident azimuth angle based on the original monopulse signal in the antenna coordinate system. and incident elevation angle ; For azimuth grid numbering, The grid number is for the pitch angle.

3. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 1, characterized in that, In step 1, the reference spectrum target is identified online autonomously, and the associated pulses of the reference spectrum target are extracted, as follows: The design employs a single-second processing cycle, automatically storing associated valid pulse sequence information during the radiation source sorting and positioning process, i.e., radiation source... Associated pulse sequence , Represents the radiation source generated within a single processing cycle. The number, The number of associated pulses representing the radiation source; radiation source Parameter information With reference spectrum target Parameter information Perform matching, and simultaneously target the radiation source. Location information With reference spectrum target Location information Perform matching, where Representative reference spectrum target The reference spectrum target is automatically selected based on its number. Corresponding radiation source The matching and recognition criteria are as follows: ， in, Represents the frequency threshold. This represents the recycle threshold. Represents the pulse width threshold. Represents the distance threshold; Represents frequency information. Represents double-circular information, Represents pulse width information; Using radiation sources that meet the matching identification conditions and reference spectrum target Construct a reference spectrum target and its associated pulse sequence .

4. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 3, characterized in that, In step 2, the phase difference correction table is calculated and constructed in real time using the associated pulse sequence of the reference spectrum target, as detailed below: Step 2.1, Design a table of phase difference corrections. ,in, The incident azimuth angle of the original monopulse signal in the antenna coordinate system. The incident elevation angle of the original monopulse signal in the antenna coordinate system. Represents the minimum value in the frequency range. Represents the maximum value of the frequency range. This indicates that the corresponding number is... Phase difference correction under conditions of frequency range grid, azimuth grid, and elevation grid; Step 2.2, based on the theoretical phase difference and measured phase difference Solve for the phase difference correction. Then, the frequency range grid, azimuth grid, and elevation grid are traversed to construct a phase difference correction table; Step 2.3, calculate the phase difference correction. Matching weight table .

5. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 4, characterized in that, In step 2.2, based on the theoretical phase difference... and measured phase difference Solve for the phase difference correction. Then, a phase difference correction table is constructed, as follows: according to The frequency information in the full-pulse descriptor word is used to obtain the phase difference correction. Frequency grid number The phase difference correction is obtained based on the pitch angle information in the full-pulse descriptor. Pitch angle grid number The phase difference correction is obtained based on the azimuth information in the full-pulse descriptor. Azimuth grid number ; Calculate the theoretical phase difference At that time, with pulse sequence The arrival time of a single pulse is used to interpolate and obtain the attitude information of the electromagnetic spectrum sensor. and the location information of the electromagnetic spectrum sensor Then the pulse relative to the incident vector of the electromagnetic spectrum sensor antenna Represented as: ； in, The rotation matrix representing the rotation from the WGS84 coordinate system to the electromagnetic spectrum sensor body coordinate system is given by... Calculated The rotation matrix representing the electromagnetic spectrum sensor body to the antenna coordinate system; The baseline three-dimensional vector of the antenna pair in the antenna coordinate system is The theoretical phase difference Represented as: ； Through theoretical phase difference and measured phase difference The phase difference correction can be obtained by subtracting. , Indicates pulse The wavelength.

6. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 5, characterized in that, In step 2.3, the phase difference correction number is constructed. Matching weight table The details are as follows: Since the same phase difference correction table of an actual electromagnetic spectrum sensor will be repeatedly overwritten by multiple sets of pulses during the operation of the motion platform, a corresponding weight table is designed to achieve dynamic accumulation and updating of the phase difference correction table. , and The number of cumulative pulses with the same dimension and scale, corrected for existing phase difference values. express; During the continuous accumulation process, when a new phase difference correction is calculated... and existing phase difference corrections within the same grid To synthesize: ； At the same time, update the corresponding weight table. .

7. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 6, characterized in that, In step 3, the mathematical model parameters are solved based on the mathematical model parameter fitting method. The details are as follows: Based on the effective values ​​in the continuously accumulating phase correction table, a plane fitting of the correction values ​​is performed, resulting in a second-order data model. As shown below: ， in, , , Represents the higher-order terms of the planar model. , , Represents the lower-order terms of the planar model; For the For each frequency range, construct the least squares equation: ， in: Second-order data model The expression is as follows: ， parameter , Mathematical model parameters , Mathematical model parameters It can be obtained through the least squares equation: ， represent The weight vector of each effective observation. Indicates the first The number of effective observations within a frequency range is determined by the cumulative pulse count of the effective observation grid. Decide, Indicates to Mathematical model parameters Partial derivative operation; If the antenna design has good phase consistency, then the phase difference correction plane can be fitted using a low-order term of the planar model. The fitted model can be simplified as follows: ， Both high-order and low-order models can be used as the final online calibration model, and their mathematical model parameters can be... These are considered as online calibration coefficients; the online calibration coefficients are quickly obtained by fitting the existing observation information in the phase difference correction table.

8. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 7, characterized in that, In step 4, based on the mathematical model parameters Evaluation of the phase difference correction of the fitting residual Availability, based on phase difference correction circumscribed polygon Evaluate mathematical model parameters The scope of application is as follows: Step 4.1, based on mathematical model parameters The phase difference correction of the fitting residual evaluation Availability: Fitting residuals Represented as: ， During the statistical process, statistics The number of grids, This represents the threshold; when the number of statistically analyzed grid cells is greater than 90% of the total number of grid cells, the fitted plane interval is considered valid. Step 4.2, based on phase difference correction number circumscribed polygon Evaluate mathematical model parameters Scope of application: During the operation of the motion platform, as the phase difference correction table is dynamically updated, based on the platform's motion characteristics, a single reference radiation source for a single mission can cover a linearly distributed grid range, and the phase difference correction... It will accumulate dynamically in order to mark the available phase difference corrections in real time. The range, design and phase difference correction number Accompanying valid judgment form The above-mentioned valid judgment table and Same dimensions and scale; Valid Judgment Form Store 0 or 1, where 0 is an invalid identifier and 1 is a valid identifier, in order to determine the validity of the decision table. Valid identifiers, i.e., mathematical model parameters Scope of application: Based on phase difference correction. circumscribed polygon To make a valid judgment table The assignment.

9. The real-time calibration method for interferometer antennas based on dynamic fitting according to claim 8, characterized in that, In step 4.2, based on the phase difference correction number circumscribed polygon Evaluate mathematical model parameters The scope of application is as follows: set up to These represent the six endpoints of the effective observation coverage grid after three calibration tasks, arranged in a linear pattern. to Constructing the circumscribed polygon ; When determining the phase difference correction grid circumscribed polygon When inside, the grid Corresponding valid judgment table The value is assigned to 1; when judging the phase difference correction number. grid circumscribed polygon When outside, the grid Corresponding valid judgment table The value is assigned to 0, that is: 。

Images