A detection and positioning method based on a micro-magnetic sensor array combination

Abstract

The application provides a detection and positioning method based on a micro-magnetic sensor array combination, comprising the following steps: underwater detection is performed through a micro-magnetic sensor array to obtain detection data; a preprocessing unit is used for noise reduction processing on the detection data transmitted by the micro-magnetic sensor array; a sensor calibration unit is used for analyzing the detection data transmitted by the micro-magnetic sensor array to establish a unified orthogonal coordinate system; the detection data transmitted by the sensor calibration unit is combined with autonomous learning AI to dynamically learn the monitoring environment and the target scene mode, a magnetic anomaly target detection algorithm of adaptive minimum entropy filtering is combined with the detection data to calculate the magnetic anomaly targets with low signal-to-noise ratio, and positioning data is obtained. The application improves the signal-to-noise ratio, reduces the errors caused by the non-orthogonal and non-parallel of the three axes of the sensor, and reduces the interference caused by environmental noise, geomagnetic influence and accidental errors.

Description

Technical Field

[0001] This invention mainly relates to the technical field of underwater detection and positioning, specifically to a detection and positioning method based on a combination of micro magnetic sensor arrays. Background Technology

[0002] In recent years, underwater detection technology has been needed in civilian fields such as seabed topography exploration, large-area ocean monitoring, and underwater information acquisition, as well as in military fields such as detecting enemy ships, submarines, and aircraft carriers.

[0003] According to patent application CN201911099314.0, a method for underwater target detection and positioning based on an electric field electrode array is described. In this method, the electric field electrodes are arranged in an arc-shaped array. A detection array group is formed by combining electric field electrodes that meet certain conditions. An electric field positioning algorithm is used to calculate the underwater target positioning parameters for different detection array groups. Finally, the positioning parameters of multiple groups are fused to obtain the accurate three-directional coordinate position and electric dipole moment of the underwater target. This underwater detection and positioning method is suitable for underwater safety surveillance, exhibiting good algorithm adaptability and high positioning accuracy.

[0004] While the aforementioned underwater detection and positioning methods are applicable to the field of underwater safety surveillance, have good algorithm adaptability, and high positioning accuracy, traditional underwater detection and positioning methods have low filtering capabilities and are difficult to effectively reduce the impact of the environment, thus affecting the detection and positioning accuracy. Summary of the Invention

[0005] This invention provides a detection and positioning method based on a combination of micro-magnetic sensor arrays to solve the technical problems mentioned in the background.

[0006] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0007] A detection and localization method based on a combination of micro-magnetic sensor arrays includes the following steps:

[0008] Step 1: Underwater detection is performed using a micro-magnetic sensor array to obtain detection data;

[0009] Step 2: The detection data transmitted by the micro magnetic sensor array is received through the preprocessing unit, and the detection data transmitted by the micro magnetic sensor array is subjected to noise reduction processing through the preprocessing unit.

[0010] Step 3: Receive the detection data transmitted by the preprocessing unit through the sensor calibration unit, and analyze the detection data transmitted by the micro-magnetic sensor array through the sensor calibration unit to establish a unified orthogonal coordinate system;

[0011] Step 4: The AI ​​learns dynamically and deeply by combining the detection data transmitted by the sensor calibration unit to learn the amplitude and fluctuation characteristics of the environmental field signal in the time and frequency domains, and obtains a class of characteristic parameters for the specific target type being monitored.

[0012] Step 5: The detection data transmitted by the autonomous learning AI is received by the identification and positioning unit, and the detection data is combined with the detection data to calculate the positioning data for magnetic anomaly targets with low signal-to-noise ratio.

[0013] Furthermore, in step one, the preprocessing unit includes:

[0014] A filtering and noise reduction module is used to perform noise reduction processing on the detection data transmitted by the micro magnetic sensor array;

[0015] The data preprocessing module is used to identify abnormal data in the noise-reduced detection data.

[0016] Furthermore, in step three, the sensor calibration unit includes:

[0017] The three-axis orthogonal calibration module is used to receive the detection data processed by the data preprocessing module, and also to establish a unified orthogonal coordinate system for the received detection data in order to perform orthogonal calibration.

[0018] The three-axis parallel calibration module is used to establish a unified parallel coordinate system for the orthogonally calibrated detection data in order to perform parallel calibration.

[0019] Furthermore, in step five, after receiving the detection data processed by the autonomous learning AI, the identification and positioning unit uses the following formula to obtain the entropy filter output value. When the entropy filter output value drops below the threshold, the magnetic anomaly target is obtained.

[0020] (1) The detection data has the following normal probability density function:

[0021]

[0022] Where x i σ is the discrete value of the probe data, u is the average value of the probe data, and σ is the discrete value of the probe data. 2 It is the variance of the probe data;

[0023] (2) The average value of the probe data is calculated using the following formula:

[0024]

[0025] Where x i is the discrete value of the probe data, and u is the average value of the probe data;

[0026] (3) Detection data, its discrete value x i The probability of a noisy sample is:

[0027]

[0028] Where x i These are discrete values ​​of the probe data, using a quantization step size Δx. i The analog-to-digital converter, p(x i Approximately:

[0029] (4) The entropy filter calculates the entropy of a sample of a certain length L as follows:

[0030]

[0031] Where x i It is the discrete value of the probe data.

[0032] Furthermore, in step five, after receiving the detection data processed by the autonomous learning AI, the identification and positioning unit constructs multiple single-axis and three-axis sensor arrays, and uses the Euler deconvolution method to combine multi-modal feature matching of magnetic field components, total field and first-order and second-order magnetic gradient tensors to accurately locate the magnetic dipole source by characterizing and describing the magnetic target signal features.

[0033] Furthermore, in step five, the identification and positioning unit uses the following formula to obtain the magnetic induction intensity value in order to accurately locate the magnetic dipole source.

[0034]

[0035] Where m is the dipole magnetic moment, and the unit is Am. 2 r and r0 are the position vectors of the observation point and the magnetic target, respectively.

[0036] Furthermore, in step five, the identification and positioning unit includes:

[0037] The target recognition alarm module is used to receive the detection data processed by the self-learning AI, and calculate the magnetic anomaly target with low signal-to-noise ratio by combining the detection data with the detection data through the minimum entropy filtering magnetic anomaly target detection algorithm, obtain the entropy filter output value, and compare the entropy filter output value with the threshold stored in the storage module. If the entropy filter output value drops below the threshold, the target is magnetically abnormal.

[0038] The data transmission module is used to send alarm information to the monitoring module after the target identification alarm module does not detect the target magnetic anomaly, and also to send a confirmation end information to the monitoring module after the target identification alarm module does not detect the target magnetic anomaly.

[0039] Furthermore, the identification and positioning unit also includes:

[0040] The target localization module is used to receive the detection data processed by the self-learning AI, construct multiple single-axis and three-axis sensor arrays, and use the Euler deconvolution method to combine multi-modal feature matching of magnetic field components, total field and first-order and second-order magnetic gradient tensors to accurately locate the magnetic dipole source by characterizing and describing the magnetic target signal features, obtain the localization information, and send the localization information to the monitoring module.

[0041] Furthermore, the identification and positioning unit also includes:

[0042] The monitoring module is used to display alarm information and confirm the end of the monitoring process.

[0043] Furthermore, the identification and positioning unit also includes:

[0044] The storage module is used to store thresholds that can be read by the target recognition alarm module.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0046] Firstly, this invention improves the signal-to-noise ratio, thereby increasing the accuracy of target detection and localization for weak magnetic field signals.

[0047] Secondly, this invention reduces the errors caused by the non-orthogonality and non-parallelism of the sensor's three axes, thereby improving the accuracy of target detection and positioning.

[0048] Third, this invention reduces the interference caused by singularity problems, environmental noise, geomagnetic influences and random errors when the magnetic moment of a single magnetic gradiometer is perpendicular to the observation direction, thus achieving accurate positioning of the magnetic dipole source.

[0049] Fourth, the adaptive minimum entropy filtering magnetic anomaly target detection algorithm used in this invention has a significant advantage over the orthogonal basis function method in detecting magnetic anomaly targets with low signal-to-noise ratio.

[0050] The present invention will be explained in detail below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the present invention;

[0052] Figure 2 This is a flowchart of the present invention.

[0053] In the diagram: 10. Micro-magnetic sensor array; 20. Preprocessing unit; 21. Filtering and noise reduction module; 22. Data preprocessing module; 30. Sensor calibration unit; 31. Three-axis orthogonal calibration module; 32. Three-axis parallel calibration module; 40. Autonomous learning AI; 50. Recognition and positioning module; 51. Target recognition alarm module; 52. Data transmission module; 53. Storage module; 54. Monitoring module; 55. Target positioning unit. Detailed Implementation

[0054] To facilitate understanding of the present invention, a more comprehensive description of the present invention will be given below with reference to the accompanying drawings, which illustrate several embodiments of the present invention. However, the present invention can be implemented in different forms and is not limited to the embodiments described in the text. Rather, these embodiments are provided to make the disclosure of the present invention more thorough and complete.

[0055] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly on the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0056] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly associated with those skilled in the art to which this invention pertains. The terminology used herein in the specification of this invention is for the purpose of describing particular embodiments and is not intended to limit the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0057] For an example, please refer to the appendix. Figure 1-2 A detection and localization method based on a combination of micro magnetic sensor arrays includes the following steps:

[0058] Step 1: Underwater detection is performed using the micro-magnetic sensor array 10 to obtain detection data;

[0059] Step 2: The detection data transmitted by the micro magnetic sensor array 10 is received by the preprocessing unit 20, and the detection data transmitted by the micro magnetic sensor array 10 is subjected to noise reduction processing by the preprocessing unit 20.

[0060] Step 3: The sensor calibration unit 30 receives the detection data transmitted by the preprocessing unit 20 and analyzes the detection data transmitted by the micro magnetic sensor array 10 to establish a unified orthogonal coordinate system.

[0061] Step 4: The AI40 autonomously learns the detection data transmitted by the sensor calibration unit 30 to perform dynamic deep learning on the monitoring environment and target scene patterns, learns to acquire the amplitude and fluctuation characteristics of the environmental field signal in the time domain and frequency domain, and acquires a class of characteristic parameters for the specific target type being monitored.

[0062] Step 5: The identification and positioning unit 50 receives the detection data transmitted by the autonomous learning AI 40, and calculates the location data for magnetic anomaly targets with low signal-to-noise ratio by combining the detection data with the detection data using the adaptive minimum entropy filtering magnetic anomaly target detection algorithm.

[0063] Furthermore, in step one, the preprocessing unit 20 includes:

[0064] The filtering and noise reduction module 21 is used to perform noise reduction processing on the detection data transmitted by the micro magnetic sensor array 10;

[0065] Data preprocessing module 22 is used to identify abnormal data in the noise-reduced detection data;

[0066] It should be noted that noise reduction is achieved through the filtering and noise reduction module 21, and the waveband of the detection data is monitored through the data preprocessing module 22, which improves the signal-to-noise ratio and the accuracy of target detection and positioning for weak magnetic target signals.

[0067] Furthermore, in step three, the sensor calibration unit 30 includes:

[0068] The three-axis orthogonal calibration module 31 is used to receive the detection data processed by the data preprocessing module 22, and also to establish a unified orthogonal coordinate system for the received detection data in order to perform orthogonal calibration.

[0069] The three-axis parallel calibration module 32 is used to establish a unified parallel coordinate system for the orthogonally calibrated detection data in order to perform parallel calibration.

[0070] It should be noted that precise calibration of the micro-magnetic sensor array 10 and the establishment of a unified orthogonal coordinate system for the micro-magnetic sensor array 10 reduce the errors caused by the non-orthogonality and non-parallelism of the three axes of the micro-magnetic sensor array 10, thereby improving the accuracy of target detection and positioning.

[0071] Furthermore, in step five, after receiving the detection data processed by the self-learning AI40, the identification and positioning unit 50 uses the following formula to obtain the entropy filter output value. When the entropy filter output value drops below the threshold, the magnetic anomaly target is obtained.

[0072] (1) The detection data has the following normal probability density function:

[0073]

[0074] Where x i σ is the discrete value of the probe data, u is the average value of the probe data, and σ is the discrete value of the probe data. 2 It is the variance of the probe data;

[0075] (2) The average value of the probe data is calculated using the following formula:

[0076]

[0077] Where x i is the discrete value of the probe data, and u is the average value of the probe data;

[0078] (2) Detection data, its discrete value x i The probability of a noisy sample is:

[0079]

[0080] Where x i These are discrete values ​​of the probe data, using a quantization step size Δx. i The analog-to-digital converter, p(x i Approximately:

[0081] (4) The entropy filter calculates the entropy of a sample of a certain length L as follows:

[0082]

[0083] Where x i It is the discrete value of the probe data.

[0084] Furthermore, in step five, after receiving the detection data processed by the self-learning AI40, the identification and positioning unit 50 constructs multiple single-axis and three-axis sensor arrays, and uses the Euler deconvolution method to combine multi-modal feature matching of magnetic field components, total field and first-order and second-order magnetic gradient tensors. By characterizing and describing the characteristics of the magnetic target signal, the magnetic dipole source is accurately located, overcoming the singularity problem of a single magnetic gradiometer when the magnetic moment of the target location is perpendicular to the observation direction, as well as interference caused by environmental noise, geomagnetic influence and random errors.

[0085] Furthermore, in step five, the identification and positioning unit 50 uses the following formula to obtain the magnetic induction intensity value in order to accurately locate the magnetic dipole source.

[0086]

[0087] Where m is the dipole magnetic moment, and the unit is Am. 2Let r and r0 be the position vectors of the observation point and the magnetic target, respectively. The above equation is Euler's negative cubic homogeneous equation, i.e., f(kr) = k -3 f(r) therefore satisfies the Euler equation: Right now This is the single-point localization equation of the gradient magnetic tensor;

[0088] It should be noted that the Euler deconvolution method is used when a ferromagnetic object is much larger than its diameter at the test point and is uniformly magnetized. The magnetic field characteristics it produces on the ground are similar to those of a magnetic dipole or a sphere.

[0089] Furthermore, in step five, the identification and positioning unit 50 includes:

[0090] The target recognition alarm module 51 is used to receive the detection data processed by the self-learning AI40, and calculate the magnetic anomaly target with low signal-to-noise ratio by combining the detection data with the detection data through the minimum entropy filtering magnetic anomaly target detection algorithm, obtain the entropy filter output value, and compare the entropy filter output value with the threshold stored in the storage module 53. If the entropy filter output value drops below the threshold, the target is magnetically abnormal.

[0091] The data sending module 52 is used to send alarm information to the monitoring module 54 after the target identification alarm module 51 does not detect the target magnetic anomaly, and is also used to send a confirmation end information to the monitoring module 54 after the target identification alarm module 51 does not detect the target magnetic anomaly.

[0092] Furthermore, the identification and positioning unit 50 also includes:

[0093] The target localization module 55 is used to receive the detection data processed by the self-learning AI40, construct multiple single-axis and three-axis sensor arrays, and use the Euler deconvolution method to combine multi-modal feature matching of magnetic field components, total field and first-order and second-order magnetic gradient tensors to accurately locate the magnetic dipole source by characterizing and describing the magnetic target signal features, obtain localization information, and send the localization information to the monitoring module 54.

[0094] Furthermore, the identification and positioning unit 50 also includes:

[0095] The monitoring module 54 is used to display alarm information and confirm the end of the monitoring process.

[0096] Furthermore, the storage module 53 is used to store threshold values ​​that are read by the target identification alarm module 51.

[0097] The specific operation method of this invention is as follows:

[0098] First, underwater detection is performed using the micro-magnetic sensor array 10 to obtain detection data. The detection data transmitted by the micro-magnetic sensor array 10 is received by the preprocessing unit 20, and the detection data transmitted by the micro-magnetic sensor array 10 is subjected to noise reduction processing by the preprocessing unit 20.

[0099] The sensor calibration unit 30 receives the detection data transmitted by the preprocessing unit 20 and analyzes the detection data transmitted by the micro magnetic sensor array 10 to establish a unified orthogonal coordinate system.

[0100] The AI40 autonomous learning system, combined with the detection data transmitted by the sensor calibration unit 30, performs dynamic deep learning on the monitoring environment and target scene patterns. It learns to acquire the amplitude and fluctuation characteristics of the environmental field signal in the time and frequency domains, and acquires a class of characteristic parameters for the specific target type being monitored.

[0101] The identification and positioning unit 50 receives the detection data transmitted by the autonomous learning AI 40, and calculates the location data for magnetic anomaly targets with low signal-to-noise ratio by combining the detection data with the detection data using an adaptive minimum entropy filtering magnetic anomaly target detection algorithm.

[0102] The present invention has been described by way of example in conjunction with the accompanying drawings. Obviously, the specific implementation of the present invention is not limited to the above-described manner. Any non-substantial improvement made by adopting the inventive concept and technical solution of the present invention, or the direct application of the inventive concept and technical solution of the present invention to other occasions without modification, shall be within the protection scope of the present invention.

Claims

1. A detection and positioning method based on a combination of micro-magnetic sensor arrays, characterized in that, Includes the following steps: Step 1: Underwater detection is performed using a micro-magnetic sensor array (10) to obtain detection data; Step 2: The detection data transmitted by the micro magnetic sensor array (10) is received by the preprocessing unit (20), and the detection data transmitted by the micro magnetic sensor array (10) is subjected to noise reduction processing by the preprocessing unit (20). Step 3: The sensor calibration unit (30) receives the detection data transmitted by the preprocessing unit (20) and analyzes the detection data transmitted by the micro-magnetic sensor array (10) to establish a unified orthogonal coordinate system. Step 4: The AI ​​(40) learns autonomously and combines the detection data transmitted by the sensor calibration unit (30) to perform dynamic deep learning on the monitoring environment and target scene mode, learns to obtain the amplitude and fluctuation characteristics of the environmental field signal in the time domain and frequency domain, and obtains a class of characteristic parameters for the specific target type being monitored. Step 5: The detection data transmitted by the autonomous learning AI (40) is received by the identification and positioning unit (50), and the detection data is combined with the detection data to calculate the magnetic anomaly target with low signal-to-noise ratio by the detection algorithm of adaptive minimum entropy filtering to obtain the positioning data; In step five, after the identification and positioning unit (50) receives the detection data processed by the self-learning AI (40), the identification and positioning unit (50) uses the following formula to obtain the entropy filter output value. When the entropy filter output value drops below the threshold, the magnetic anomaly target is obtained. (1) The detection data has the following normal probability density function: Where x i σ is the discrete value of the probe data, u is the average value of the probe data, and σ is the discrete value of the probe data. 2 It is the variance of the probe data; (2) The average value of the probe data is calculated using the following formula: Where x i is the discrete value of the probe data, and u is the average value of the probe data; (3) Detection data, its discrete value x i The probability of a noisy sample is: Where x i These are discrete values ​​of the probe data, using a quantization step size Δx. i The analog-to-digital converter, p(x i Approximately: (4) The entropy filter calculates the entropy of a sample of a certain length L as follows: Where x i It is the discrete value of the probe data.

2. The detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 1, characterized in that, In step one, the preprocessing unit (20) includes: The filtering and noise reduction module (21) is used to perform noise reduction processing on the detection data transmitted by the micro magnetic sensor array (10); The data preprocessing module (22) is used to judge abnormal data in the noise-reduced detection data.

3. The detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 1, characterized in that, In step three, the sensor calibration unit (30) includes: The three-axis orthogonal calibration module (31) is used to receive the detection data processed by the data preprocessing module (22) and to establish a unified orthogonal coordinate system for the received detection data in order to perform orthogonal calibration. The three-axis parallel calibration module (32) is used to establish a unified parallel coordinate system for the orthogonally calibrated detection data in order to perform parallel calibration.

4. The detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 1, characterized in that, In step five, after receiving the detection data processed by the autonomous learning AI (40), the identification and positioning unit (50) constructs multiple single-axis and three-axis sensor arrays, and uses the Euler deconvolution method to combine the multi-modal feature matching of magnetic field components, total field and first-order and second-order magnetic gradient tensors to accurately locate the magnetic dipole source by characterizing and describing the magnetic target signal features.

5. The detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 1, characterized in that, In step five, the identification and positioning unit (50) uses the following formula to obtain the magnetic induction intensity value in order to accurately locate the magnetic dipole source. Where m is the dipole magnetic moment, and the unit is Am. 2 r and r0 are the position vectors of the observation point and the magnetic target, respectively.

6. The detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 1, characterized in that, In step five, the identification and positioning unit (50) includes: The target recognition alarm module (51) is used to receive the detection data processed by the self-learning AI (40), and calculate the magnetic anomaly target with low signal-to-noise ratio by combining the detection data with the detection algorithm of the minimum entropy filter, obtain the entropy filter output value, and compare the entropy filter output value with the threshold stored in the storage module (53). If the entropy filter output value drops below the threshold, the target is magnetically abnormal. The data sending module (52) is used to send alarm information to the monitoring module (54) after the target identification alarm module (51) does not detect the target magnetic anomaly, and is also used to send confirmation end information to the monitoring module (54) after the target identification alarm module (51) does not detect the target magnetic anomaly.

7. The detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 6, characterized in that, The identification and positioning unit (50) further includes: The target localization module (55) is used to receive the detection data processed by the autonomous learning AI (40), construct multiple single-axis and three-axis sensor arrays, and use the Euler deconvolution method to combine the multi-modal feature matching of magnetic field components, total field and first-order and second-order magnetic gradient tensors to accurately locate the magnetic dipole source by characterizing and describing the magnetic target signal features, obtain the localization information, and send the localization information to the monitoring module (54).

8. The detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 7, characterized in that, The identification and positioning unit (50) further includes: The monitoring module (54) is used to display alarm information and confirm the end information.

9. A detection and positioning method based on a combination of micro-magnetic sensor arrays according to claim 6, characterized in that, The identification and positioning unit (50) further includes: The storage module (53) is used to store thresholds that can be read by the target identification alarm module (51).

Images