System error correction method and system based on ecef and q-type cluster analysis

Abstract

This invention discloses a three-coordinate system error correction method based on ECEF and Q-type clustering analysis. Addressing the problem of multi-node inability to coordinate tracking due to sensor system errors in multi-sensor target tracking systems, this invention first collects multi-source data, preprocesses the data, then fuses the multi-source data, transfers the resulting system error information to the ECEF coordinate system for unified representation and processing, performs Q-type clustering analysis on the transformed system error data, and corrects the system error based on the Q-type clustering results, thereby improving data accuracy and consistency.

Description

Technical Field

[0001] This invention belongs to the field of systematic error correction in multi-source data fusion, specifically a systematic error correction method and system based on ECEF and Q-type clustering analysis. Background Technology

[0002] Radar multi-source data fusion is a key information processing technology aimed at improving the performance and accuracy of radar systems by integrating data from different radar systems. However, in practical applications, radar systems contain systematic errors that can affect the accuracy of the data and the fusion effect. Therefore, correcting systematic errors in radar multi-source data fusion has become an important research direction.

[0003] Radar system errors are caused by a variety of factors, including hardware and environmental aspects. For example, non-ideal characteristics of the radar antenna, receiver noise, signal attenuation, and atmospheric conditions can all lead to measurement errors in the radar system. These errors negatively impact the accuracy and reliability of radar data, thus affecting the results of multi-source radar data fusion.

[0004] Therefore, further research and exploration of systematic error correction methods and technologies in radar multi-source data fusion, as well as their effects and limitations in practical applications, are of significant theoretical and practical importance. Correcting systematic errors can improve the accuracy and consistency of radar data, further promoting the development and application of radar multi-source data fusion technology. However, traditional systematic error correction methods may suffer from problems such as positioning errors and data inconsistencies, affecting the accuracy and reliability of the fusion results. Summary of the Invention

[0005] To address the above technical issues and improve the accuracy of system error estimation in multi-source data fusion, this invention proposes a system error correction method based on ECEF and Q-type clustering analysis.

[0006] The technical solution to achieve the purpose of this invention is: a systematic error correction method based on ECEF and Q-type clustering analysis, comprising:

[0007] Step 1: Collect measurement data containing systematic errors from different data sources;

[0008] Step 2: Preprocess the collected measurement data containing systematic errors;

[0009] Step 3: Merge the preprocessed data of the same type;

[0010] Step 4: Transform the fused system error into the ECEF coordinate system to achieve a unified representation of the system error;

[0011] Step 5: Perform Q-type cluster analysis on the measurement data with systematic errors after coordinate transformation from the distance dimension, azimuth dimension, and elevation dimension respectively to obtain clustering results with different data characteristics;

[0012] Step 6: Calculate the systematic error corresponding to the clustering result measurement data from the three dimensions of distance, azimuth, and elevation, respectively. Collect the data features with the smallest systematic error in the corresponding dimensions, and recalculate the systematic error corresponding to the data features with the smallest systematic error in the three dimensions of distance, azimuth, and elevation.

[0013] Preferably, the data source includes a radar system and ground measuring instruments.

[0014] Preferably, the specific methods for preprocessing the collected measurement data containing systematic errors include: data cleaning, data alignment, and correction.

[0015] Preferably, the specific method for converting the fused system error to the ECEF coordinate system is as follows:

[0016] Step 4-1: Transform the measurement data from the relative coordinate system to the local Cartesian coordinate system;

[0017] Step 4-2: Transform the target from the local Cartesian coordinate system to the ECEF coordinate system.

[0018] Preferably, the specific formula for converting measurement data from a relative coordinate system to a local Cartesian coordinate system is as follows:

[0019]

[0020] Where (r) t θ t η t (x) represents the measured value of the target. l y l , z l ) represents the local Cartesian coordinates of the target.

[0021] Preferably, the target is located in a local Cartesian coordinate system (x... l y l , z l Transform to ECEF coordinate system (x) t y t , z t The specific formula is:

[0022]

[0023] In the formula, (x s y s , z s ) represents the coordinates of the measurement node in ECEF, and R is the rotation matrix.

[0024] Preferably, the specific method for performing Q-type cluster analysis on the coordinate-transformed measurement data with systematic errors from the distance, azimuth, and elevation dimensions to obtain clustering results with different data characteristics is as follows:

[0025] Step 5-1: Divide the measurement data with systematic errors after coordinate transformation into different categories and clusters;

[0026] Step 5-2: Calculate the Euclidean similarity based on the data features, and construct a similarity matrix between the measured data based on the similarity calculation results. Then, input the matrix into the Q-type clustering algorithm to obtain the clustering results for different data features.

[0027] Preferably, the specific method for calculating the systematic error corresponding to the clustering results of different data features from the three dimensions of distance, azimuth, and elevation angle in step 6 is as follows:

[0028] Step 6-1: Construct equations in the ECEF coordinate system based on the detection results of the two measurement nodes on the same target:

[0029] χ a +R a χ al，k =χ b +R b χ bl，k

[0030] in:

[0031]

[0032] χ a and χ b These are the ECEF coordinates of two measurement nodes a and b, respectively, R a and R b These are the rotation matrices of the two measurement nodes, χ′ al，k and χ′ bl，k These are the Cartesian coordinates of the target at the two measurement nodes.

[0033] Step 6-2: Perform a first-order Taylor expansion on the equations from Step 6-1:

[0034] χ ae，k +R a J a，k ξ a =χ be，k +R b J b，k ξ b

[0035] Where χ ae，k and χ be，k These are the ECEF coordinates of the target observed from the two nodes, J.ae，k and J be，k These are the Jacobian matrices of the target calculated when the system error is 0;

[0036] Step 6-3: Express the formula in 6-2 in matrix form, and substitute it into the ECEF coordinates of the target observed by the two measurement nodes in the measurement data corresponding to the Q-type clustering results. Solve for the systematic error by using the least squares method.

[0037] Preferably, the systematic error is solved using the least squares method:

[0038] Lξ=Δχ

[0039]

[0040] Where L = [R] a J a R b J b ], ξ=[0,0,0,0,0,0] is the initial estimate of the system error without prior information, Δχ is the ECEF coordinate difference between the two nodes observing the target; L columns are full rank to ensure It is uniquely solved, where Δr a , Δθ a ,Δη a , Δr b , Δθ b ,Δη b These are the minimum variance unbiased estimates of the two-node system errors in the three-dimensional directions of distance, azimuth, and pitch.

[0041] This invention also proposes a systematic error correction system based on ECEF and Q-type clustering analysis, comprising:

[0042] The data preprocessing module is used to preprocess the collected measurement data, which includes systematic errors.

[0043] The multi-source data fusion module is used to fuse pre-processed data of the same type.

[0044] The ECEF coordinate transformation module is used to transform the fused system error into the ECEF coordinate system to achieve a unified representation of the system error.

[0045] The Q-type clustering analysis module is used to perform Q-type clustering analysis on measurement data with systematic errors after coordinate transformation from the distance dimension, azimuth dimension, and elevation dimension, respectively, to obtain clustering results with different data characteristics.

[0046] The system error correction module is used to calculate the system error corresponding to the clustering result measurement data from the three dimensions of distance, azimuth and elevation, respectively, collect the data features with the smallest system error in the corresponding dimensions, and recalculate the system error corresponding to the data features with the smallest system error in the three dimensions of distance, azimuth and elevation.

[0047] Compared with the prior art, the present invention has the following significant advantages: the present invention combines multi-source information fusion, ECEF coordinate system and Q-type cluster analysis to provide a more accurate and efficient method for correcting system errors; by fusing information from multiple data sources, the accuracy and consistency of the data are improved; through ECEF coordinate system transformation and Q-type cluster analysis, potential patterns and regularities in the data can be identified, further improving the effect of system error correction. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating a systematic error correction method based on ECEF and Q-type clustering analysis. Detailed Implementation

[0049] A systematic error correction method based on ECEF and Q-type clustering analysis is proposed. This method integrates information from multiple radar data sources, combines the ECEF coordinate system and the Q-type clustering algorithm to analyze and correct systematic errors, achieving higher accuracy and consistency compared to traditional systematic error correction methods. A schematic diagram of the processing flow is shown below. Figure 1 As shown in the example, the specific steps are as follows:

[0050] Step 1: Collect multi-source data: Collect measurement data containing systematic errors from different data sources, including radar systems, ground measuring instruments, etc., to obtain more comprehensive and diverse data, and at the same time obtain the target's true value data such as AIS, Beidou or GPS, which is only used for comparison of the calculated systematic error results;

[0051] Step 2: Data preprocessing: Preprocess the collected measurement data containing systematic errors, including data cleaning, data alignment and correction, to improve the accuracy and consistency of the data;

[0052] Step 3: Multi-source data fusion: Fusion of pre-processed data of the same type, integrating information from different data sources to extract more accurate and reliable system error information;

[0053] Step 4: ECEF coordinate system transformation: Transform the fused system errors into the ECEF coordinate system to achieve a unified representation and processing of system errors.

[0054] Step 4-1: Filter measurement data and convert it from the relative coordinate system to the local Cartesian coordinate system:

[0055]

[0056] Where (r) t θ t η t (x) is the measurement data of the target. l y l , z l ) represents the local Cartesian coordinates of the target;

[0057] Step 4-2: Move the target from the local Cartesian coordinate system (x... l y l , z l Transform to ECEF coordinate system (x) t y t , z t ):

[0058]

[0059]

[0060] Where (x) s y s , z s ) represents the coordinates of the measurement node (such as radar) in ECEF, and R is the rotation matrix.

[0061] Step 5: Q-type cluster analysis: Perform Q-type cluster analysis on the measurement data with systematic errors after coordinate transformation. This divides the data into different categories and clusters to identify potential patterns and regularities in the data. The accuracy of the systematic error calculated from data with different characteristics will vary. By observing the accuracy of the calculated systematic error under different data characteristics through Q-type cluster analysis, summarizing the data characteristics, and collecting this characteristic data for systematic error correction, the accuracy of systematic error correction can be further improved.

[0062] Step 5-1: Divide the measurement data with systematic errors after coordinate transformation into different categories and clusters, such as expanding from three dimensions: distance, azimuth, and elevation, and setting different features;

[0063] Step 5-2: Calculate the Euclidean similarity based on the characteristics of a certain type of data, and construct a similarity matrix between the measured data based on the similarity calculation results. Then, input the matrix into the Q-type clustering algorithm to obtain the clustering results for different data characteristics.

[0064] Step 6: Calculate the systematic error corresponding to the clustering results of different data features from the three dimensions of distance, azimuth, and elevation. Collect the data features with the smallest systematic error in the corresponding dimensions, and recalculate the systematic error corresponding to the data features with the smallest systematic error in the three dimensions of distance, azimuth, and elevation.

[0065] Step 6-1: Construct equations in the ECEF coordinate system based on the detection results of two measurement nodes for the same target:

[0066] χ a +R a χ al，k =χ b +R b χ bl，k (4)

[0067] in:

[0068]

[0069] χ a and χ b These are the ECEF coordinates of two measurement nodes a and b, respectively, R a and R b These are the rotation matrices of the two measurement nodes, χ′ al，k and χ′ bl，k These are the Cartesian coordinates of the target at the two measurement nodes.

[0070] Step 6-2: Assuming the system deviation is small, expand the formula in Step 6-1 using the first-order Taylor series:

[0071] χ ae，k +R a J a，k ξ a =χ be，k +R b J b，k ξ b (6)

[0072] Where χ ae，k and χ be，k These are the ECEF coordinates of the target observed from the two nodes, J. ae，k and J be，k These are the Jacobian matrices of the target calculated when the system error is 0;

[0073] Step 6-3: Express the formula in 6-2 in matrix form, substitute the ECEF coordinates of the target observed at the two measurement nodes, and solve for the systematic error using the least squares method:

[0074]

[0075] Where L = [R] a J a R b J b ], ξ=[0,0,0,0,0,0] is the initial estimate of the system error without prior information, Δχ is the ECEF coordinate difference between the two nodes observing the target, and L is full rank to ensure It can be uniquely solved, where Δr a , Δθ a ,Δη a , Δr b , Δθ b ,Δη b These are the minimum variance unbiased estimates of the two-node system errors in the three-dimensional directions of range, azimuth, and pitch.

[0076] Step 6-4: Based on the calculation results of Step 6-3, obtain the data with the minimum systematic error in the distance, azimuth, and elevation dimensions under the current measurement data, and use the method of Step 6-3 to find the systematic error corresponding to the data with the minimum systematic error in the distance, azimuth, and elevation dimensions.

[0077] A systematic error correction system based on ECEF and Q-type clustering analysis includes:

[0078] The data preprocessing module is used to preprocess the collected measurement data, which includes systematic errors.

[0079] The multi-source data fusion module is used to fuse pre-processed data of the same type.

[0080] The ECEF coordinate transformation module is used to transform the fused system error into the ECEF coordinate system to achieve a unified representation of the system error.

[0081] The Q-type clustering analysis module is used to perform Q-type clustering analysis on measurement data with systematic errors after coordinate transformation from the distance dimension, azimuth dimension, and elevation dimension, respectively, to obtain clustering results with different data characteristics.

[0082] The system error correction module is used to calculate the system error corresponding to the clustering result measurement data from the three dimensions of distance, azimuth and elevation, respectively, collect the data features with the smallest system error in the corresponding dimensions, and recalculate the system error corresponding to the data features with the smallest system error in the three dimensions of distance, azimuth and elevation.

Claims

1. A systematic error correction method based on ECEF and Q-type clustering analysis, characterized in that, include: Step 1: Collect measurement data containing systematic errors from different data sources; Step 2: Preprocess the collected measurement data containing systematic errors; Step 3: Merge the preprocessed data of the same type; Step 4: Transform the fused system error into the ECEF coordinate system to achieve a unified representation of the system error; Step 5: Perform Q-type cluster analysis on the measurement data with systematic errors after coordinate transformation from the distance dimension, azimuth dimension, and elevation dimension respectively to obtain clustering results with different data characteristics; Step 6: Calculate the systematic error corresponding to the clustering result measurement data from the three dimensions of distance, azimuth, and elevation, respectively. Collect the data features with the smallest systematic error in the corresponding dimensions, and recalculate the systematic error corresponding to the data features with the smallest systematic error in the three dimensions of distance, azimuth, and elevation.

2. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 1, characterized in that, The data sources include radar systems and ground measuring instruments.

3. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 1, characterized in that, Specific methods for preprocessing collected measurement data containing systematic errors include: data cleaning, data alignment, and correction.

4. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 1, characterized in that, The specific method for converting the fused system error into the ECEF coordinate system is as follows: Step 4-1: Transform the measurement data from the relative coordinate system to the local Cartesian coordinate system; Step 4-2: Transform the target from the local Cartesian coordinate system to the ECEF coordinate system.

5. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 4, characterized in that, The specific formula for converting measurement data from a relative coordinate system to a local Cartesian coordinate system is as follows: Where (r) t θ t η t (x) represents the measured value of the target. l y l , z l ) represents the local Cartesian coordinates of the target.

6. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 4, characterized in that, The target is in the local Cartesian coordinate system (x l y l , z l Transform to ECEF coordinate system (x) t y t , z t The specific formula is: In the formula, (x s y s , z s ) represents the coordinates of the measurement node in ECEF, and R is the rotation matrix.

7. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 1, characterized in that, The specific method for performing Q-type clustering analysis on coordinate-transformed measurement data with systematic errors in the distance, azimuth, and elevation dimensions to obtain clustering results for different data characteristics is as follows: Step 5-1: Divide the measurement data with systematic errors after coordinate transformation into different categories and clusters; Step 5-2: Calculate the Euclidean similarity based on the data features, and construct a similarity matrix between the measured data based on the similarity calculation results. Then, input the matrix into the Q-type clustering algorithm to obtain the clustering results for different data features.

8. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 1, characterized in that, Step 6 involves calculating the systematic error corresponding to the clustering results of different data features from the three dimensions of distance, azimuth, and elevation. The specific method for this step is as follows: Step 6-1: Construct equations in the ECEF coordinate system based on the detection results of the two measurement nodes on the same target: x a +R a x' al,k =x b +R b x' bl,k in: χ a and χ b These are the ECEF coordinates of two measurement nodes a and b, respectively, R a and R b These are the rotation matrices of the two measurement nodes, χ' al,k and χ' bl,k These are the Cartesian coordinates of the target at the two measurement nodes. Step 6-2: Perform a first-order Taylor expansion on the equation from Step 6-1: x ae,k +R a J a,k x a =x be,k +R b J b,k x b Where χ ae，k and χ be，k These are the ECEF coordinates of the target observed from the two nodes, J. ae，k and J be，k These are the Jacobian matrices of the target calculated when the system error is 0; Step 6-3: Express the formula in 6-2 in matrix form, and substitute it into the ECEF coordinates of the target observed by the two measurement nodes in the measurement data corresponding to the Q-type clustering results. Solve for the systematic error by using the least squares method.

9. The systematic error correction method based on ECEF and Q-type clustering analysis according to claim 8, characterized in that, Solving for systematic errors using the least squares method: Lξ=Δχ Where L = [R] a J a R b J b ], ξ=[0,0,0,0,0,0] is the initial estimate of the system error without prior information, Δχ is the ECEF coordinate difference between the two nodes observing the target; L columns are full rank to ensure It is uniquely solved, where Δr a ,Δθ a ,Δη a ,Δr b ,Δθ b ,Δη b These are the minimum variance unbiased estimates of the two-node system errors in the three-dimensional directions of distance, azimuth, and pitch.

10. A system error correction system based on ECEF and Q-type clustering analysis using the method described in any one of claims 1 to 9, characterized in that, include: The data preprocessing module is used to preprocess the collected measurement data, which includes systematic errors. The multi-source data fusion module is used to fuse pre-processed data of the same type. The ECEF coordinate transformation module is used to transform the fused system error into the ECEF coordinate system to achieve a unified representation of the system error. The Q-type clustering analysis module is used to perform Q-type clustering analysis on measurement data with systematic errors after coordinate transformation from the distance dimension, azimuth dimension, and elevation dimension, respectively, to obtain clustering results with different data characteristics. The system error correction module is used to calculate the system error corresponding to the clustering result measurement data from the three dimensions of distance, azimuth and elevation, respectively, collect the data features with the smallest system error in the corresponding dimensions, and recalculate the system error corresponding to the data features with the smallest system error in the three dimensions of distance, azimuth and elevation.

Images