Method and system for measuring magnetic field of inertial measurement system

Abstract

The invention discloses a method for measuring a magnetic field of an inertial measurement system, which realizes accurate reconstruction of a global magnetic field vector field by planning a discrete point test layout, synchronously acquiring multi-dimensional data and constructing a magnetic field vector fitting model. The method comprises the steps of test area division and discrete point layout, multi-source data synchronous acquisition, data preprocessing and abnormal value elimination, global magnetic field vector field fitting modeling, and fitting result verification and optimization. According to the method, a discrete point layering layout strategy is adopted, a multivariable fitting algorithm is combined, magnetic field space change characteristics can be effectively captured, global magnetic field vector field fitting errors are controlled, and the method is suitable for inertial measurement system magnetic field performance detection and optimization design and has the advantages of being high in measurement speed, high in representation precision and high in practicability.

Description

Technical Field

[0001] This invention relates to the field of inertial navigation and magnetic field testing technology, and to a method and system for measuring the magnetic field of an inertial measurement system. Background Technology

[0002] The measurement accuracy of an inertial measurement system (INS) directly determines the navigation and control performance of a carrier. In INS, magnetic field interference is one of the key factors affecting the output accuracy of core components such as gyroscopes and accelerometers. Especially in fields with high navigation and control requirements, accurate characterization of the magnetic field distribution can effectively improve the layout of the magnetic field source and sensitive devices, as well as the application of magnetic shielding technology in the system, thereby enhancing the performance of the INS.

[0003] Traditional methods for measuring the magnetic field of inertial measurement systems mainly include single-point measurement and uniform grid traversal. Single-point measurement can only acquire magnetic field data at individual locations, failing to reflect the global magnetic field distribution. While uniform grid traversal can cover a certain area, it suffers from redundant test point layouts, low testing efficiency, and fails to consider the non-uniformity of magnetic field spatial variations, resulting in insufficient global magnetic field fitting accuracy. Furthermore, existing methods often ignore the impact of environmental interference factors such as temperature drift and attitude coupling on magnetic field testing, further reducing the accuracy of the magnetic field vector field characterization and making it difficult to meet the requirements of high-precision inertial measurement systems. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and propose a method and system for measuring the magnetic field of an inertial measurement system. By using discrete point layering, multi-source data synchronous acquisition, interference compensation and high-precision fitting modeling, the global magnetic field vector can be accurately reconstructed.

[0005] To achieve the above-mentioned objectives, the present invention provides at least one of the following technical solutions: a method for measuring the magnetic field of an inertial measurement system, comprising: S1. Establish a three-dimensional rectangular test coordinate system with the geometric center of the inertial measurement system as the origin; divide the test space into: core area, transition area, and outer area; test points are evenly distributed in the core area at a set interval a, test points are distributed in the transition area at a set interval b, and test points are distributed in the outer area at a set interval c, forming a discrete point test matrix. S2, fix the three-dimensional fluxgate magnetometer on the displacement platform; move it sequentially to each test point according to the preset discrete point test matrix, and collect the magnetic field vector data (B) of each test point in the three-dimensional rectangular test coordinate system. x B y B z Simultaneously, the inertial measurement system collects corresponding attitude data and ambient temperature data using its built-in inertial instruments. S3 uses a sliding window filtering algorithm to smooth the raw data collected in S2; and uses the Grubbs criterion to identify and remove outliers in the magnetic field vector data. S4, perform global magnetic field vector field fitting modeling; S5 verifies and optimizes the fitting results of the global magnetic field vector field, and outputs the global magnetic field vector fitting model and related performance parameters.

[0006] Furthermore, the core area refers to the platform assembly of the inertial measurement system and the area 2mm outside the platform assembly's rotation area; the transition area refers to the area 2mm outside the platform assembly's rotation area to the physical outer contour of the inertial measurement system; and the outer area refers to the area 200mm outside the physical outer contour of the inertial measurement system.

[0007] Furthermore, the three-dimensional fluxgate magnetometer and displacement platform must meet the magnetic field accuracy requirements of the inertial measurement system.

[0008] Furthermore, the sampling frequency of data, attitude data, and ambient temperature data at each test point is no less than 30Hz, and 20 to 50 sets of data are continuously collected at each test point. All data are synchronized through timestamps.

[0009] Furthermore, in step S3, the sliding window size is 5 to 10 sets of data to remove high-frequency noise.

[0010] Furthermore, in S4, global magnetic field vector field fitting modeling is performed, including: Based on the preprocessed magnetic field vector data in S3, a three-dimensional spatial magnetic field vector field fitting model is constructed. A radial basis function neural network algorithm is used to fit the magnetic field components in the x, y, and z directions of the three-dimensional rectangular test coordinate system, yielding the mathematical expression for the global magnetic field vector field: B(x,y,z)=[B x (x,y,z), B y (x,y,z), B z (x,y,z)] T Where (x,y,z) are the spatial coordinates in the three-dimensional rectangular test coordinate system; The radial basis function neural network has 50 to 100 hidden nodes, uses a Gaussian activation function, optimizes grid weights using gradient descent, and terminates the iteration when the fitting error is less than 0.1% or the number of iterations reaches 200.

[0011] Furthermore, in step S5, the fitting results of the global magnetic field vector field are verified and optimized, including: Several verification points are randomly selected from the test points. The selection ratio of verification points is not less than 15% of the total test points, and they are evenly distributed in the core area, transition area, and outer area. Calculate the deviation between the measured magnetic field vector value and the predicted value of the fitted model at each verification point; If the maximum deviation between the measured magnetic field vector value and the predicted value of the fitted model at each verification point is greater than 0.5%, then additional test points will be set up in the corresponding area, and S2~S4 will be repeated for refitting. If the maximum deviation between the measured magnetic field vector values ​​at each verification point and the predicted values ​​of the fitted model is ≤0.5%, then the global magnetic field vector fitting model will be output.

[0012] An inertial measurement system magnetic field measurement system includes: a three-dimensional fluxgate magnetometer, a drive device for a displacement platform, a test track for the displacement platform, a data acquisition unit, and power supply and control equipment for the inertial measurement system; The three-dimensional fluxgate magnetometer is fixed on the displacement platform. The power supply and control equipment of the inertial measurement system supplies power to the inertial measurement system and controls the drive device of the displacement platform to drive the displacement platform to move on the test track of the displacement platform according to the preset path. The data acquisition unit synchronously collects magnetic field data, attitude data and temperature data of the inertial instruments of the inertial measurement system.

[0013] Furthermore, several test points are set on the test track of the displacement platform. The method for setting the test points is as follows: The test space is divided into a core area, a transition area, and an outer area. Test points are evenly distributed in the core area at a set interval 'a', test points are distributed in the transition area at a set interval 'b', and test points are distributed in the outer area at a set interval 'c', forming a discrete point test matrix.

[0014] Furthermore, the core area refers to the platform assembly of the inertial measurement system and the area 2mm outside the platform assembly's rotation area; the transition area refers to the area 2mm outside the platform assembly's rotation area to the physical outer contour of the inertial measurement system; and the outer area refers to the area 200mm outside the physical outer contour of the inertial measurement system.

[0015] Compared with the prior art, the present invention has the following advantages: (1) The present invention adopts a discrete point layering strategy of “dense core and sparse periphery”. Under the premise of ensuring global coverage, it focuses on improving the test accuracy of areas with drastic magnetic field changes. Compared with the traditional uniform distribution traversal method, it reduces the number of test points by more than 40%, greatly improves the measurement efficiency, and solves the technical problem that the traditional magnetic field measurement method cannot accurately characterize the global magnetic field distribution of the inertial test system.

[0016] (2) This invention simultaneously collects magnetic field data, attitude data and temperature data, establishes a dataset of disturbances under multi-factor environments, and can realize statistical analysis of magnetic field distribution and system characteristic data, providing a reference for subsequent system accuracy analysis; (3) The present invention uses the radial basis function neural network algorithm to fit the global magnetic field vector field, which can accurately capture the nonlinear spatial distribution characteristics of the magnetic field and the fitting error is less than 0.5%, which is a significant improvement in fitting accuracy compared with the traditional polynomial fitting method. (4) The present invention introduces a verification and optimization mechanism, and verifies the effectiveness of the fitting model through uniformly distributed verification points to ensure the reliability of the global magnetic field vector field characterization. It is applicable to the magnetic field measurement of high-precision inertial measurement systems such as aerospace and weapon equipment, and has broad application prospects. Attached Figure Description

[0017] Figure 1 This is a flowchart of the magnetic field measurement method of the inertial measurement system of the present invention; Figure 2 This is a schematic diagram of the layout and related equipment of the inertial measurement system test device according to an embodiment of the present invention; Figure 3 This is a cloud map showing the magnetic field distribution of the inertial measurement system according to an embodiment of the present invention. 1-Inertial measurement system, 2-Three-dimensional fluxgate magnetometer, 3-Drive device for precision displacement platform, 4-Test track for precision displacement platform, 5-Data acquisition unit, 6-Power supply and control equipment for inertial measurement system. Detailed Implementation

[0018] The features and advantages of the present invention will become clearer and more explicit from the following detailed description.

[0019] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments. Although various aspects of embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated otherwise.

[0020] An inertial measurement system magnetic field measurement system includes: a three-dimensional fluxgate magnetometer 2, a drive device 3 for a displacement platform, a test track 4 for the displacement platform, a data acquisition unit 5, and an inertial measurement system power supply and control device 6. The three-dimensional fluxgate magnetometer 2 is fixed on the displacement platform. The inertial measurement system power supply and control device 6 supplies power to the inertial measurement system 1 and controls the drive device 3 to drive the displacement platform to move along a preset path on the test track 4 of the displacement platform. The data acquisition unit 5 synchronously acquires magnetic field data, attitude data of the inertial instruments of the inertial measurement system 1, and temperature data.

[0021] Several test points are set on test track 4 of the displacement platform. The method for setting the test points is as follows: The test space is divided into a core area, a transition area, and an outer area. Test points are evenly distributed in the core area at a set interval of 20mm, test points are distributed in the transition area at a set interval of 50mm, and test points are distributed in the outer area at a set interval of 100mm, forming a discrete point test matrix.

[0022] The core area refers to the platform assembly of the inertial measurement system and the area 2mm outside the platform assembly's rotation area; the transition area refers to the area 2mm outside the platform assembly's rotation area to the physical outer contour of the inertial measurement system; the outer area refers to the area 200mm outside the physical outer contour of the inertial measurement system.

[0023] like Figure 1 As shown, a method for measuring the magnetic field of an inertial measurement system includes: (1) Test area division and discrete point layout: A three-dimensional rectangular test coordinate system is established with the geometric center of the inertial measurement system as the origin (which can be consistent with the coordinate definition of the inertial measurement system). The test space is divided into three layers: core area (2mm outside the platform component and its rotation area), transition area (2mm from the rotation area of ​​the platform component to the physical outer contour of the system), and outer area (from the physical outer contour of the system to the outer 200mm). Test points are evenly distributed in the core area at 20mm intervals, test points are distributed in the transition area at 50mm intervals, and test points are distributed in the outer area at 100mm intervals, forming a discrete point test matrix of "dense in the core and sparse in the periphery" to ensure the test coverage of areas with drastic magnetic field changes.

[0024] (2) Synchronous acquisition of multi-source data: A high-precision three-dimensional fluxgate magnetometer (test accuracy ±100μT) is fixed on a precision displacement platform (positioning accuracy ±0.03mm). The displacement platform is controlled by a host computer to move in the order of the discrete point test matrix, reaching each test point in sequence. Each test point is stopped for 0.5s, and magnetic field vector data (B) is acquired synchronously. x B y B z The attitude data and ambient temperature data of the inertial measurement system were sampled at a frequency of 50Hz. 25 sets of data were collected at each test point. All data were synchronized at the millisecond level through timestamps to ensure the spatiotemporal consistency of the data. (3) Data preprocessing and outlier removal: The original data collected was smoothed by a sliding window filtering algorithm to remove high-frequency noise; the Grubbs criterion (significance level α=0.05) was used to identify and remove outliers in the magnetic field vector data.

[0025] (4) Global magnetic field vector field fitting modeling: The test magnetic field vector data are classified according to the spatial coordinates (x, y, z) of the test points to construct a training dataset. A radial basis function neural network (with 80 hidden layer nodes and a Gaussian activation function) is used to fit and train the magnetic field components in the x, y, and z directions respectively. During the training process, the grid weights are optimized by gradient descent to minimize the fitting error. Training is stopped when the fitting error is less than 0.1% or the number of iterations reaches 200, and the mathematical expression of the global magnetic field vector field is obtained: B(x, y, z) = [B x (x,y,z), B y (x,y,z), B z (x,y,z)] T This enables accurate prediction of magnetic field vectors at any spatial location.

[0026] (5) Validation and optimization of fitting results: Randomly select 15% of the test points and distribute the validation points evenly in the core area, transition area and outer area to ensure that the distribution of validation points is relatively uniform. Calculate the absolute deviation and relative deviation between the measured value of the magnetic field vector of each validation point and the predicted value of the fitting model. If the maximum relative deviation is greater than 0.5%, add test points in the area with large deviation and repeat steps (2) to (4) to collect and fit data again. If the maximum relative deviation is ≤0.5%, the fitting model is deemed to be effective and output the global magnetic field vector field fitting model, fitting error distribution, magnetic field strength of each area and other performance parameters.

[0027] Example: This embodiment uses a spherical inertial measurement system as the measurement object. The system has a size of Φ200mm, and the global magnetic field vector field fitting error is required to be ≤0.5%. The method includes the following steps: (1) Test area division and discrete point layout: A three-dimensional test coordinate system is established with the geometric center of the inertial measurement system 1 as the origin. The core area has a diameter of 100mm and about 60 test points are laid out at 20mm intervals. The transition area has a diameter of 100mm to 200mm and about 80 test points are laid out at 50mm intervals. The outer area has a diameter of 200mm to 400mm and about 70 test points are laid out at 100mm intervals. A total of 210 test points are set out.

[0028] (2) Multi-source data synchronous acquisition: A three-dimensional fluxgate magnetometer 2 (model CH-330F) and an auxiliary measurement motor (positioning accuracy ±0.03mm) are used. The three-dimensional fluxgate magnetometer 2 is fixed on the precision displacement platform. The inertial measurement system power supply and control equipment 6 supplies power to the inertial measurement system 1 and controls the drive device 3 of the precision displacement platform to drive the platform to move. The host computer controls the platform to move along the test track 4 of the precision displacement platform according to the preset path. Each test point is stopped for 0.5s. The data acquisition device 5 synchronously acquires magnetic field data, attitude data, and temperature data at a sampling frequency of 50Hz. 25 sets of data are acquired at each measurement point, such as... Figure 2 As shown.

[0029] (3) Data preprocessing and outlier removal: Origin data processing software was used, and sliding window filtering (8 groups of windows) was used to smooth the original data. Three outlier data points were removed by the Grubbs criterion.

[0030] (4) Global magnetic field vector field fitting model: A radial basis function neural network was constructed with 80 hidden layer nodes and Gaussian activation function. The magnetic field components in the x, y, and z directions were fitted and trained respectively. After 120 training iterations, the fitting error dropped to 0.08%, and the training was stopped, resulting in the global magnetic field vector field fitting model.

[0031] (5) Validation and optimization of the fitting results: 36 validation points were selected (12 in the core area, 12 in the transition area, and 12 in the outer area). The relative error between the measured and predicted values ​​was calculated. The maximum relative error was 0.32%, which is less than 0.5%, indicating that the fitting model is effective. The global magnetic field vector field fitting distribution map is output, as shown below. Figure 3 As shown.

[0032] The present invention has been described in detail above with reference to specific embodiments and exemplary examples; however, these descriptions should not be construed as limiting the present invention. Those skilled in the art will understand that various equivalent substitutions, modifications, or improvements can be made to the technical solutions and embodiments of the present invention without departing from the spirit and scope of the invention, and all such modifications and improvements fall within the scope of the present invention. The scope of protection of the present invention is defined by the appended claims.

[0033] The contents not described in detail in this specification are common knowledge to those skilled in the art.

Claims

1. A method for measuring the magnetic field of an inertial measurement system, characterized in that, include: S1, with the geometric center of the inertial measurement system as the origin, establish a three-dimensional rectangular test coordinate system; The test space is divided into: core area, transition area, and outer area; test points are evenly distributed in the core area at a set interval 'a', test points are distributed in the transition area at a set interval 'b', and test points are distributed in the outer area at a set interval 'c', forming a discrete point test matrix. S2, fix the three-dimensional fluxgate magnetometer on the displacement platform; move it sequentially to each test point according to the preset discrete point test matrix, and collect the magnetic field vector data (B) of each test point in the three-dimensional rectangular test coordinate system. x B y B z Simultaneously, the inertial measurement system collects corresponding attitude data and ambient temperature data using its built-in inertial instruments. S3 uses a sliding window filtering algorithm to smooth the raw data collected in S2; and uses the Grubbs criterion to identify and remove outliers in the magnetic field vector data. S4, perform global magnetic field vector field fitting modeling; S5 verifies and optimizes the fitting results of the global magnetic field vector field, and outputs the global magnetic field vector fitting model and related performance parameters.

2. The method for measuring the magnetic field of an inertial measurement system according to claim 1, characterized in that, The core area refers to the platform assembly of the inertial measurement system and the area 2mm outside the platform assembly's rotation area; the transition area refers to the area 2mm outside the platform assembly's rotation area to the physical outer contour of the inertial measurement system; the outer area refers to the area 200mm outside the physical outer contour of the inertial measurement system.

3. The method for determining the magnetic field of an inertial measurement system according to claim 2, characterized in that, The three-dimensional fluxgate magnetometer and displacement platform must meet the magnetic field accuracy requirements of the inertial measurement system.

4. The method for determining the magnetic field of an inertial measurement system according to claim 3, characterized in that, The sampling frequency of data, attitude data, and ambient temperature data at each test point is no less than 30Hz. 20 to 50 sets of data are continuously collected at each test point, and all data are synchronized through timestamps.

5. The method for measuring the magnetic field of an inertial measurement system according to claim 4, characterized in that, In S3, the sliding window size is 5 to 10 sets of data to remove high-frequency noise.

6. The method for measuring the magnetic field of an inertial measurement system according to claim 5, characterized in that, In step S4, global magnetic field vector field fitting modeling is performed, including: Based on the preprocessed magnetic field vector data in S3, a three-dimensional spatial magnetic field vector field fitting model is constructed. A radial basis function neural network algorithm is used to fit the magnetic field components in the x, y, and z directions of the three-dimensional rectangular test coordinate system, yielding the mathematical expression for the global magnetic field vector field: B(x,y,z)=[B x (x,y,z), B y (x,y,z), B z (x,y,z)] T Where (x,y,z) are the spatial coordinates in the three-dimensional rectangular test coordinate system; The radial basis function neural network has 50 to 100 hidden nodes, uses a Gaussian activation function, optimizes grid weights using gradient descent, and terminates the iteration when the fitting error is less than 0.1% or the number of iterations reaches 200.

7. The method for determining the magnetic field of an inertial measurement system according to claim 6, characterized in that, In step S5, the fitting results of the global magnetic field vector field are verified and optimized, including: Several verification points are randomly selected from the test points. The selection ratio of verification points is not less than 15% of the total test points, and they are evenly distributed in the core area, transition area, and outer area. Calculate the deviation between the measured magnetic field vector value and the predicted value of the fitted model at each verification point; If the maximum deviation between the measured magnetic field vector value and the predicted value of the fitted model at each verification point is greater than 0.5%, then additional test points will be set up in the corresponding area, and S2~S4 will be repeated for refitting. If the maximum deviation between the measured magnetic field vector values ​​at each verification point and the predicted values ​​of the fitted model is ≤0.5%, then the global magnetic field vector fitting model will be output.

8. A magnetic field measurement system for an inertial measurement system, characterized in that, include: Three-dimensional fluxgate meter (2), drive device for displacement platform (3), test track for displacement platform (4), data acquisition unit (5) and power supply and control equipment for inertial measurement system (6). The three-dimensional fluxgate meter (2) is fixed on the displacement platform. The power supply and control equipment (6) of the inertial measurement system supplies power to the inertial measurement system (1) and controls the drive device (3) of the displacement platform to drive the displacement platform to move on the test track (4) of the displacement platform according to the preset path. The data acquisition device (5) synchronously collects magnetic field data, attitude data and temperature data of the inertial instruments of the inertial measurement system (1).

9. The inertial measurement system magnetic field measurement system according to claim 8, characterized in that: Several test points are set on the test track (4) of the displacement platform. The method for setting the test points is as follows: The test space is divided into a core area, a transition area, and an outer area. Test points are evenly distributed in the core area at a set interval 'a', test points are distributed in the transition area at a set interval 'b', and test points are distributed in the outer area at a set interval 'c', forming a discrete point test matrix.

10. The inertial measurement system magnetic field measurement system according to claim 9, characterized in that, The core area refers to the platform assembly of the inertial measurement system and the area 2mm outside the platform assembly's rotation area; the transition area refers to the area 2mm outside the platform assembly's rotation area to the physical outer contour of the inertial measurement system; the outer area refers to the area 200mm outside the physical outer contour of the inertial measurement system.

Images