Ship dynamic ranging positioning method

Abstract

The application provides a ship dynamic ranging positioning method, and the specific steps include: S1, data processing and filtering of ship peripheral measuring points, initial data measured by the measuring points is pretreated by using the elimination method; S2, ship spherical center fitting virtual sample generation, a virtual sample spherical center radius and coordinates of the ship peripheral measuring points are simulated by using a Gaussian distribution virtual sample generation algorithm; S3, locking and dynamic tracking of the ship positioning point, a target ship is tracked, multi-point and multi-position dynamic data are obtained, a complete local coordinate system of the ship is established, the accurate position of the ship positioning point in the local coordinate system of the ship is determined, and multi-posture space calibration of the ship is realized, the application realizes global tracking and pose positioning calibration of the ship by a space base station, solves the problem that the application is limited due to the complex shape of the measured piece or the material characteristics of the measured surface in ship measurement, and meets the demand of ship ranging positioning, and has wide application prospects in the military and civil fields.

Description

Technical Field

[0001] This invention relates to the field of ship positioning technology, and specifically to a ship dynamic ranging and positioning method. Background Technology

[0002] Distance and positioning is fundamental to ensuring safe navigation of ships and has broad application prospects in various maritime activities, especially in ship berthing and departure from docks, navigation in densely populated areas, towing, collision avoidance, escort convoy navigation, and replenishment of naval vessels.

[0003] Traditional ship ranging and calibration methods mainly employ laser ranging and GPS relative positioning, but both have significant drawbacks: laser ranging requires multi-point manual intervention, has low automation, is difficult to manufacture, and suffers from the complex structure of large ship components, significant self-obstruction, numerous parts requiring inspection and analysis with varying measurement requirements, and the potential for cumulative errors when stitching together multi-view measurement data. Traditional mechanical contact measurement methods for ships cannot meet performance and dimensional requirements, thus failing to achieve quantitative measurement. Therefore, it is necessary to design a dynamic ship ranging and positioning method. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a dynamic ship ranging and positioning method. Based on the Gaussian distribution virtual sample generation method, the center radius and coordinates of the ship's measurement points are simulated. The center coordinates are continuously tracked and locked through the overall least squares spherical fitting algorithm, realizing global tracking and pose positioning calibration of the ship's hull by an external space base station. For complex assembled components on the hull, three-dimensional digital measurement of the hull's circumference is performed, quantitatively estimating data such as assembly errors, component displacements, and structural deformations. This solves the problem that ship measurement is limited by the complex shape of the measured parts or the material characteristics of the measured surfaces, achieving highly automated and high-precision ship ranging and positioning.

[0005] This invention provides a dynamic ranging and positioning method for ships, the specific steps of which include: S1, data processing and filtering of measurement points on the outer perimeter of the ship, selecting the outer perimeter of the ship as the measurement point, and using a rejection method to preprocess the initial data measured by the laser sensor at the measurement point to remove spikes and abnormal abrupt changes in data; this includes coarse reference sphere center calculation and data filtering, the specific sub-steps of which are:

[0006] S11. Based on the preliminary fitting of the coarse reference sphere center according to the measurement points around the ship, divide all the initial data into multiple local data groups, fit their sphere centers respectively, calculate the weight of each sphere center, and the sphere center with the highest weight value is the preliminary fitting coarse reference sphere center.

[0007] S12. Filtering algorithm threshold calculation: Calculate the distance between the coarse reference sphere center and the data points in all local data groups as the sphere center radius sample, and calculate the standard deviation of the distance of the ship sphere center radius sample as the threshold.

[0008] S13. Eliminate jump points. Based on the distance standard deviation, determine whether the distance between the data points of each local data group and the center of the ball is within the specified range, and eliminate jump points based on this.

[0009] S14. Determine whether the data filtering is finished. If some data points are removed during filtering, use the filtered data points to refit the sphere center as the coarse reference sphere center, and execute steps S12 and S13 again. If no data points are removed during filtering, the filtering is finished, and the filtered coarse reference sphere center is obtained.

[0010] S2. Virtual Sample Generation for Ship Center Fitting: A Gaussian distribution virtual sample generation algorithm is used to simulate the virtual sample center radius and coordinates of the ship's outer perimeter measurement points. The specific sub-steps are as follows:

[0011] S21. Input the spherical information of the ship to be fitted, and allocate variables to store the center information data of the sphere;

[0012] S22. Allocate space to store ship spherical sample data and Gaussian noise radius;

[0013] S23. Load the ship's spherical center radius noise, and calculate the ship's spherical coordinates by performing the loading of radius noise and polar coordinate values ​​once according to the spherical center radius;

[0014] S24. Calculate the radius and coordinates of the center of the sphere of the generated virtual ship sample;

[0015] S3. Locking and dynamic tracking of ship positioning points: Track the target ship, obtain multi-point and multi-position dynamic data within the specified actions and time limits, establish a complete local coordinate system of the ship, determine the precise position of the ship positioning point in the local coordinate system inside the ship, and realize the multi-attitude spatial calibration of the ship.

[0016] S31. Fit the coordinates of the ship's center of sphere and the radius of the sphere according to the ship's spherical equation;

[0017] S32. Linearize the random elements in the observation vector according to the ship's spherical equation, that is, at the coordinates (x, y) of the spherical point. 0 ,y 0 ,z 0 Linearize at a, discard the second-order minimum term, linearize the parameter estimate, and obtain the radius R of the ship's center, that is, at a 0 b 0 c 0 r 0 Linearization, similarly discarding the second-order minima, yields:

[0018] (2x 0 -2a 0 )Δx+(2y 0 -2b 0 )Δy+(2z 0 -2c 0 )Δz=(2x 0 -2a 0 )Δa+(2y 0 -2b 0 )Δb+(2z 0 -2c 0 )Δc+2r 0 Δr-((x 0 ) 2 +(y 0 ) 2 +(z 0 ) 2 +(a 0 ) 2 +(b 0 ) 2 +(c 0 ) 2 -(r 0 ) 2 -2a 0 x 0 -2b 0 y 0 -2c 0 z 0 )

[0019] Simplifying the equations sequentially, we obtain the radius R of the ship's center of gravity:

[0020]

[0021] S33. Dynamic tracking of the ship's center coordinates: Given that the calibration attitude and measurement attitude coordinate systems have the same scale scaling ratio in all directions, the transformation relationship between the two coordinate systems is expressed by rotation parameters, translation parameters, and scaling parameters. Given the coordinates of three points in the calibration attitude and measurement attitude coordinate systems, dynamic tracking of the ship's center coordinates is achieved.

[0022] Preferably, the specific sub-steps in step S11 are as follows:

[0023] S111. Import the data to be filtered into the function and define a temporary variable temp_Point to store it;

[0024] S112. Coarse reference sphere center grouping traversal: Based on the input variable sco, divide all data into numPoint data groups containing sco data.

[0025] S113. Use the sphere center fitting algorithm to fit the coarse reference sphere center;

[0026] S114. Coarse reference center threshold evaluation: Calculate the distance of each center to all data points and subtract them pairwise. If the result is within the coarse reference center threshold ser, the weight of that center is increased by 1. After all centers have been calculated, select the center with the highest weight as the coarse reference center and return it to determine the optimal reference solution for the coarse reference center.

[0027] Preferably, step S2 further includes:

[0028] S25. Set the coordinates of the center of the sphere to be fitted and the radius, and give the expected value and variance required to generate random noise, so as to generate random noise that conforms to a Gaussian distribution;

[0029] S26. The generated noise is applied to the set sphere center radius, and the sphere center radius is added to the Gaussian noise to produce a sphere center radius that conforms to a Gaussian distribution;

[0030] S27. Obtain the spherical coordinates using a polar coordinate to 3D coordinate transformation algorithm, specifically:

[0031] x=(r+radiusNoise)×sinα×cosβ

[0032] y=(r+radiusNoise)×sinα×sinβ

[0033] z = (r + radiusNoise) × cosα

[0034] In the formula, r is the radius of the center of the sphere to be fitted, radiusNoise is the random error applied to the radius, α is the pitch angle, and β is the azimuth angle;

[0035] The range of α is (0, π), and the range of β is (0, 2π). α and β are assigned values ​​using random numbers, specifically:

[0036] α=π×rand1

[0037] β=2π×rand2

[0038] In the formula, rand1 and rand2 are random numbers in (0,1) to obtain the spherical coordinates of the ship.

[0039] Preferably, in step S23, polar coordinates are used to describe the ship's sphere center coordinates, and the ship's sphere center coordinates are:

[0040] x = r × cosα × cosβ

[0041] y = r × cosα × sinβ

[0042] z = r × sinα

[0043] In the formula, r is the input radius, α is the pitch angle, and β is the azimuth angle;

[0044] When loading noise onto the ship's sphere radius, the noise is distributed using a Gaussian distribution achieved through the Box-Muller algorithm, specifically as follows:

[0045]

[0046]

[0047] Z~N(0,1)

[0048] Where U1 and U2 are uniformly distributed random numbers, and Z is a random number that follows a uniform distribution;

[0049] The specific expressions for generating a given mean and variance are:

[0050] Y = u + (Z × sigma)

[0051] In the formula, u and sigma are manually defined expectation and variance. The expectation and variance are set to (0,1) and (0,4), respectively. The generated random bias is imported into Matlab to observe whether the random noise distribution curve conforms to a normal distribution.

[0052] Preferably, the spherical equation of the ship in step S31 is:

[0053] (x i -x0) 2 +(y i -y0) 2 +(z i -z0) 2 =R 2

[0054] In the formula, (a, b, c) are the coordinates of the center of the sphere, r is the radius of the sphere, and (x, y, z) are the coordinates of a point on the surface of the sphere;

[0055] Expanding the spherical equation of the ship, where a, b, c, r 2 -a 2 -b 2 -c 2 Treating this as a parameter to be estimated, when n spherical points are observed, then:

[0056]

[0057] In the formula, A is the coefficient matrix, Y is the observation vector, and X is the parameter to be estimated;

[0058] The parameter X to be estimated is solved using total least squares, specifically:

[0059]

[0060] Preferably, step S33, which involves the dynamic tracking of the ship's sphere center coordinates, comprises the following sub-steps:

[0061] S331. The ship coordinate transformation model under various attitudes is as follows:

[0062]

[0063] Where λ is the scaling parameter, R is the rotation matrix, Δ is the translation vector, and A and B are the coordinates in the calibration attitude and measurement attitude coordinate systems, respectively;

[0064] The scaling parameter λ is:

[0065]

[0066] Among them, |P1P2| A Let |P1P2| be the distance between points P1 and P2 in the calibration attitude coordinate system; B Let |P1P2| be the distance between points P1 and P2 in coordinate system B; A Let |P3P2| be the distance between points P3 and P2 in coordinate system A; B Let |P1| be the distance between points P3 and P2 in coordinate system B; A Let |P1P3| be the distance between points P1 and P3 in coordinate system A; B Let P1 and P3 be the distance between points P1 and P3 in coordinate system B.

[0067] The rotation matrix R is a 3x3 orthogonal matrix. The rotation matrix R is constructed using the antisymmetric matrix S:

[0068]

[0069] and:

[0070]

[0071] Where I is the identity matrix, and R has only three variables a, b, and c. Solving for a, b, and c determines the rotation matrix R.

[0072] S332. Substituting points P1 and P2 into the equation and subtracting them to eliminate ΔX, ΔY, and ΔZ, we get:

[0073]

[0074] Among them, P 1AX Let P1 be the X-axis coordinate of point P1 in coordinate system A; P 2AX Let P2 be the X-axis coordinate of point P2 in coordinate system A; P 1AY Let P1 be the Y-axis coordinate of point P1 in coordinate system A; P 2AYLet P2 be the Y-axis coordinate of point P2 in coordinate system A; P 1BX Let P1 be the X-axis coordinate of point P1 in coordinate system B; P 2BX Let P2 be the X-axis coordinate of point P2 in coordinate system B; P 1BY Let P1 be the Y-axis coordinate of point P1 in coordinate system B; P 2BY Let P2 be the Y-axis coordinate of point P2 in coordinate system B; P 1BZ Let P1 be the Z-axis coordinate of point P1 in coordinate system B; P 2BZ Let P2 be the Z-axis coordinate of point P2 in coordinate system B; P 1AZ Let P1 be the Y-axis coordinate of point P1 in coordinate system A; P 2AY Let P2 be the Y-axis coordinate of point P2 in coordinate system A;

[0075] S333, the coordinates of the ship's center of gravity in each attitude are:

[0076]

[0077] Among them, X A12 For P 1AX With P 2AX The distance between them; X B12 For P 1BX With P 2BX The distance between them; X B13 For P 1BX With P 3BX The distance between them; Z A12 For P 1AZ With P 2AZ The distance between them; Z B12 For P 1BZ With P 2BZ The distance between them; Z B13 For P 1BZ With P 3BZ The distance between them; Y A12 For P 1AY With P 2AY The distance between them; Y B12 For P 1BY With P 2BY The distance between them; Y B13 For P 1BY With P 3BY The distance between them;

[0078] Substituting a, b, and c into the equation yields the rotation matrix R. Substituting the coordinates of any point into the ship coordinate transformation model expression yields the translation vector Δ, and the coordinates of the ship's center of gravity in the calibrated attitude coordinate system A are obtained.

[0079] Compared with the prior art, the present invention has the following advantages:

[0080] This invention relates to a dynamic ship ranging and positioning method. Based on the Gaussian distribution virtual sample generation method using the Box-Muller algorithm, the method simulates the sphere center radius and coordinates of the ship's measurement points. By continuously tracking and locking the sphere center coordinates through the overall least squares spherical fitting algorithm, it achieves global tracking and pose positioning calibration of the ship's hull by an external space base station. This solves the problem that the application of ship measurement is limited by the complex shape of the measured part or the material characteristics of the measured surface. It can meet the needs of ship ranging and positioning and has broad application prospects in multiple fields. Attached Figure Description

[0081] Figure 1 This is a schematic diagram of the overall structure of the ship dynamic ranging and positioning method of the present invention;

[0082] Figure 2 This is a flowchart of the data processing and filtering of the ship's outer perimeter measurement points in this invention;

[0083] Figure 3 This is a flowchart of the elimination method in the data processing and filtering of ship perimeter measurement points in this invention;

[0084] Figure 4 This is a flowchart of the virtual sample generation process for center-of-sphere fitting in this invention;

[0085] Figure 5 This is a flowchart of the virtual sample generation algorithm in this invention;

[0086] Figure 6 This is a schematic diagram of coordinate transformation in spherical coordinates in this invention. Detailed Implementation

[0087] To fully describe the technical content, structural features, objectives, and effects of this invention, a detailed description will be provided below in conjunction with the accompanying drawings.

[0088] The present invention provides a ship dynamic ranging and positioning method, such as... Figure 1 As shown, the specific steps include: S1, data processing and filtering of measurement points on the outer perimeter of the ship, selecting the outer perimeter of the ship as the measurement point, and using the elimination method to preprocess the initial data measured by the laser sensor for the measurement point to remove peaks and abnormal abrupt changes in data;

[0089] like Figure 2 and Figure 3 As shown, step S1, data processing and filtering of the ship's outer perimeter measurement points, includes coarse reference sphere center calculation and data filtering. The specific sub-steps are as follows:

[0090] S11. Based on the preliminary fitting of the coarse reference sphere center using the measurement points around the ship's perimeter, divide all initial data into multiple local data groups, fit the sphere center for each group, calculate the weight of each sphere center, and the sphere center with the highest weight value is the preliminary fitted coarse reference sphere center; the specific sub-steps in step S11 are as follows:

[0091] S111. Import the data to be filtered into the function and define a temporary variable temp_Point to store it;

[0092] S112. Coarse reference sphere center grouping traversal: Based on the input variable sco, divide all data into numPoint data groups containing sco data.

[0093] S113. Use the sphere center fitting algorithm to fit the coarse reference sphere center;

[0094] S114. Coarse reference center threshold evaluation: Calculate the distance of each center to all data points and subtract them pairwise. If the result is within the coarse reference center threshold ser, the weight of that center is increased by 1. After all centers have been calculated, select the center with the highest weight as the coarse reference center and return it to determine the optimal reference solution for the coarse reference center.

[0095] S12. Filtering algorithm threshold calculation: Calculate the distance between the coarse reference sphere center and the data points in all local data groups as the sphere center radius sample, and calculate the standard deviation of the distance of the ship sphere center radius sample as the threshold.

[0096] S13. Eliminate jump points. Based on the distance standard deviation, determine whether the distance between the data points of each local data group and the center of the ball is within the specified range, and eliminate jump points based on this.

[0097] S14. Determine whether the data filtering is finished. If some data points are removed during filtering, use the filtered data points to refit the sphere center as the coarse reference sphere center, and execute steps S12 and S13 again. If no data points are removed during filtering, the filtering is finished, and the filtered coarse reference sphere center is obtained.

[0098] like Figure 4 As shown, S2, virtual sample generation for ship sphere center fitting, uses a Gaussian distribution virtual sample generation algorithm based on Box-Muller to simulate the virtual sample sphere center radius and coordinates of the ship's outer perimeter measurement points. The specific sub-steps are as follows:

[0099] S21. Input the spherical information of the ship to be fitted, and allocate variables to store the center information data of the sphere;

[0100] S22. Allocate space to store ship spherical sample data and Gaussian noise radius;

[0101] S23. Apply noise to the ship's spherical center radius. Based on the spherical center radius, perform the application of radius noise and polar coordinate values ​​once each, and calculate the ship's spherical coordinates. In step S23, polar coordinates are used to describe the ship's spherical center coordinates, which are:

[0102] x = r × cosα × cosβ

[0103] y = r × cosα × sinβ

[0104] z = r × sinα

[0105] In the formula, r is the input radius, α is the pitch angle, and β is the azimuth angle;

[0106] When loading noise onto the ship's sphere radius, the noise is distributed using a Gaussian distribution achieved through the Box-Muller algorithm, specifically as follows:

[0107]

[0108]

[0109] Z~N(0,1)

[0110] Where U1 and U2 are uniformly distributed random numbers, and Z is a random number that follows a uniform distribution;

[0111] The specific expressions for generating a given mean and variance are:

[0112] Y = u + (Z × sigma)

[0113] In the formula, u and sigma are manually defined expectation and variance. The expectation and variance are set to (0,1) and (0,4), respectively. The generated random bias is imported into Matlab to observe whether the random noise distribution curve conforms to a normal distribution.

[0114] S24. Calculate the radius and coordinates of the center of the sphere of the generated virtual ship sample.

[0115] like Figure 5 As shown, step S2 further includes:

[0116] S25. Set the coordinates of the center of the sphere to be fitted and the radius, and give the expected value and variance required to generate random noise, so as to generate random noise that conforms to a Gaussian distribution;

[0117] S26. The generated noise is applied to the set sphere center radius, and the sphere center radius is added to the Gaussian noise to generate a sphere center radius that conforms to a Gaussian distribution;

[0118] S27. The spherical coordinates are obtained using a polar coordinate to 3D coordinate transformation algorithm, specifically:

[0119] x=(r+radiusNoise)×sinα×cosβ

[0120] y=(r+radiusNoise)×sinα×sinβ

[0121] z = (r + radiusNoise) × cosα

[0122] In the formula, r is the radius of the center of the sphere to be fitted, radiusNoise is the random error applied to the radius, α is the pitch angle, and β is the azimuth angle;

[0123] The range of α is (0, π), and the range of β is (0, 2π). α and β are assigned values ​​using random numbers, specifically:

[0124] α=π×rand1

[0125] β=2π×rand2

[0126] In the formula, rand1 and rand2 are random numbers in (0,1) to obtain the spherical coordinates of the ship.

[0127] S3. Locking and dynamic tracking of ship positioning points: Track the target ship, obtain multi-point and multi-position dynamic data within the specified actions and time limits, establish a complete local coordinate system for the ship, determine the precise position of the ship's positioning point in the local coordinate system within the ship's hull, and realize the multi-attitude spatial calibration of the ship.

[0128] S31. Fit the coordinates of the ship's center of sphere and the radius of the sphere according to the ship's spherical equation;

[0129] The spherical equation of the ship in step S31 is:

[0130] (x i -x0) 2 +(y i -y0) 2 +(z i -z0) 2 =R 2

[0131] In the formula, (a, b, c) are the coordinates of the center of the sphere, r is the radius of the sphere, and (x, y, z) are the coordinates of a point on the surface of the sphere;

[0132] Expanding the spherical equation of the ship, where a, b, c, r 2 -a 2 -b 2 -c 2 Treating this as a parameter to be estimated, when n spherical points are observed, then:

[0133]

[0134] In the formula, A is the coefficient matrix, Y is the observation vector, and X is the parameter to be estimated;

[0135] The parameter X to be estimated is solved using total least squares, specifically:

[0136]

[0137] S32. Linearize the random elements in the observation vector according to the ship's spherical equation, that is, at the coordinates (x, y) of the spherical point. 0 ,y 0 ,z 0 Linearize at a, discard the second-order minimum term, linearize the parameter estimate, and obtain the radius R of the ship's center, that is, at a 0 b 0 c 0 r 0 Linearization, similarly discarding the second-order minima, yields:

[0138] (2x 0 -2a 0 )Δx+(2y 0 -2b 0 )Δy+(2z 0 -2c 0 )Δz=(2x 0 -2a 0 )Δa+(2y 0 -2b 0 )Δb+(2z 0 -2c 0 )Δc+2r 0 Δr-((x 0 ) 2 +(y 0 ) 2 +(z 0 ) 2 +(a 0 ) 2 +(b 0 ) 2 +(c 0 ) 2 -(r 0 ) 2 -2a 0 x 0 -2b 0 y 0 -2c 0 z 0 )

[0139] Simplifying the equations sequentially, we obtain the radius R of the ship's center of gravity:

[0140]

[0141] S33. Dynamic tracking of the ship's center coordinates: Given that the calibration attitude and measurement attitude coordinate systems have the same scale scaling ratio in all directions, the transformation relationship between the two coordinate systems is expressed by rotation parameters, translation parameters, and scaling parameters. Given the coordinates of three points in the calibration attitude and measurement attitude coordinate systems, dynamic tracking of the ship's center coordinates is achieved.

[0142] Step S33 involves the dynamic tracking of the ship's sphere center coordinates, and the specific sub-steps are as follows:

[0143] S331. The ship coordinate transformation model under various attitudes is as follows:

[0144]

[0145] Where λ is the scaling parameter, R is the rotation matrix, Δ is the translation vector, and A and B are the coordinates in the calibration attitude and measurement attitude coordinate systems, respectively.

[0146] The scaling parameter λ is:

[0147]

[0148] Among them, |P1P2| A The distance between points P1 and P2 in the calibration attitude coordinate system;

[0149] The rotation matrix R is a 3x3 orthogonal matrix. The rotation matrix R is constructed using the antisymmetric matrix S:

[0150]

[0151] and:

[0152]

[0153] Here, I is the identity matrix, and R has only three variables: a, b, and c. Solving for a, b, and c determines the rotation matrix R.

[0154] S332. Substituting points P1 and P2 into the equation and subtracting them to eliminate ΔX, ΔY, and ΔZ, we get:

[0155]

[0156] Here P 1AX Let P1 be the X-axis coordinate of point P1 in coordinate system A, which is a known quantity. Simplified to:

[0157] X A12 =P 1AX -P 2AX

[0158]

[0159] Combining the above equations, we get:

[0160]

[0161] Substituting S, we get:

[0162]

[0163]

[0164] S333, the coordinates of the ship's center of gravity in each attitude are:

[0165]

[0166] Substituting a, b, and c into the equation yields the rotation matrix R. Substituting the coordinates of any point into the ship coordinate transformation model expression yields the translation vector Δ, and the coordinates of the ship's center of gravity in the calibrated attitude coordinate system A are obtained.

[0167] This invention relates to a dynamic ship ranging and positioning method. Based on a Gaussian distribution virtual sample generation method, it simulates the sphere center radius and coordinates of the ship's measurement points. Through a total least squares spherical fitting algorithm, it continuously tracks and locks the sphere center coordinates, enabling external space base stations to perform global tracking and pose positioning calibration of the ship's hull. For complex assembled components on the hull, it performs three-dimensional digital measurement of the hull's perimeter, quantitatively estimating data such as assembly errors, component displacements, and structural deformations. This solves the problem that ship measurement is limited by the complex shape of the measured parts or the material characteristics of the measured surfaces, achieving highly automated, high-precision, and efficient ship ranging and positioning.

[0168] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for dynamic ranging and positioning of ships, characterized in that, The specific steps include: S1. Data processing and filtering of measurement points on the outer perimeter of the ship: The outer perimeter of the ship is selected as the measurement points. The initial data obtained by the laser sensor for the measurement points is preprocessed using a rejection method to remove spikes and anomalous data changes. This includes coarse reference sphere center calculation and data filtering. The specific sub-steps are as follows: S11. Based on the preliminary fitting of the coarse reference sphere center according to the measurement points around the ship, divide all the initial data into multiple local data groups, fit their sphere centers respectively, calculate the weight of each sphere center, and the sphere center with the highest weight value is the preliminary fitting coarse reference sphere center. S12. Filtering algorithm threshold calculation: Calculate the distance between the coarse reference sphere center and the data points in all local data groups as the sphere center radius sample, and calculate the standard deviation of the distance of the ship sphere center radius sample as the threshold. S13. Eliminate jump points. Based on the distance standard deviation, determine whether the distance between the data points of each local data group and the center of the ball is within the specified range, and eliminate jump points based on this. S14. Determine whether the data filtering is finished. If some data points are removed during filtering, use the filtered data points to refit the sphere center as the coarse reference sphere center, and execute steps S12 and S13 again. If no data points are removed during filtering, the filtering is finished, and the filtered coarse reference sphere center is obtained. S2. Virtual Sample Generation for Ship Center Fitting: A Gaussian distribution virtual sample generation algorithm is used to simulate the virtual sample center radius and coordinates of the measurement points around the ship's perimeter. The specific sub-steps are as follows: S21. Input the spherical information of the ship to be fitted, and allocate variables to store the center information data of the sphere; S22. Allocate space to store ship spherical sample data and Gaussian noise radius; S23. Load the ship's spherical center radius noise, and calculate the ship's spherical coordinates by performing the loading of radius noise and polar coordinate values ​​once according to the ship's spherical center radius; S24. Calculate the radius and coordinates of the center of the sphere of the generated virtual ship sample; S3. Locking and dynamic tracking of ship positioning points: Track the target ship, obtain multi-point and multi-position dynamic data, establish a complete local coordinate system of the ship, determine the position of the ship positioning point in the local coordinate system inside the ship, and realize the multi-attitude spatial calibration of the ship. S31. Fit the coordinates of the ship's center of sphere and the radius of the sphere according to the ship's spherical equation; S32. Linearize the random elements in the observation vector according to the ship's spherical equation, that is, at the coordinates (x, y) of the spherical point. 0 , y 0 ,z 0 Linearize at a, discard the second-order minimum term, and linearize the parameter estimates to obtain the fitted estimate R of the ship's sphere center radius, i.e., at a 0 b 0 c 0 r 0 Linearization, similarly discarding the second-order minima, yields: ； By simplifying the equations sequentially, a fitted estimate of the radius of the ship's sphere is obtained. : ； S33. Dynamic tracking of the ship's center coordinates: Given that the calibration attitude and measurement attitude coordinate systems have the same scaling ratio in all directions, the transformation relationship between the two coordinate systems is expressed by rotation parameters, translation parameters, and scaling parameters. Given the coordinates of three points in the calibration attitude and measurement attitude coordinate systems, dynamic tracking of the ship's center coordinates is achieved. The specific sub-steps for dynamically tracking the ship's spherical center coordinates are as follows: S331. The ship coordinate transformation model under various attitudes is as follows: ； Where λ is the scaling parameter. Let be the rotation matrix, Δ be the translation vector, and A and B be the coordinates in the calibration attitude and measurement attitude coordinate systems, respectively. The scaling parameter λ is: ； Among them, |P1P2| A Let |P1P2| be the distance between points P1 and P2 in the calibration attitude coordinate system; B Let |P1P2| be the distance between points P1 and P2 in coordinate system B; A Let |P3P2| be the distance between points P3 and P2 in coordinate system A; B Let |P1| be the distance between points P3 and P2 in coordinate system B; A Let |P1P3| be the distance between points P1 and P3 in coordinate system A; B Let P1 and P3 be the distance between points P1 and P3 in coordinate system B. Rotation matrix It is a 3x3 orthogonal matrix. A rotation matrix is ​​constructed using the antisymmetric matrix S. : ； and: ； Where I is the identity matrix, With only three variables, a, b, and c, solving for a, b, and c determines the rotation matrix. ; S332. Substituting points P1 and P2 into the equation and subtracting them to eliminate ΔX, ΔY, and ΔZ, we get: ； Among them, P 1AX Let P1 be the X-axis coordinate of point P1 in coordinate system A; P 2AX Let P2 be the X-axis coordinate of point P2 in coordinate system A; P 1AY Let P1 be the Y-axis coordinate of point P1 in coordinate system A; P 2AY Let P2 be the Y-axis coordinate of point P2 in coordinate system A; P 1BX Let P1 be the X-axis coordinate of point P1 in coordinate system B; P 2BX Let P2 be the X-axis coordinate of point P2 in coordinate system B; P 1BY Let P1 be the Y-axis coordinate of point P1 in coordinate system B; P 2BY Let P2 be the Y-axis coordinate of point P2 in coordinate system B; P 1BZ Let P1 be the Z-axis coordinate of point P1 in coordinate system B; P 2BZ Let P2 be the Z-axis coordinate of point P2 in coordinate system B; P 1AZ Let P1 be the Z-axis coordinate of point P1 in coordinate system A; P 2AY Let P2 be the Y-axis coordinate of point P2 in coordinate system A; S333, the coordinates of the ship's center of gravity in each attitude are: ； Among them, X A12 For P 1AX With P 2AX The distance between them; X B12 For P 1BX With P 2BX The distance between them; X B13 For P 1BX With P 3BX The distance between them; Z A12 For P 1AZ With P 2AZ The distance between them; Z B12 For P 1BZ With P 2BZ The distance between them; Z B13 For P 1BZ With P 3BZ The distance between them; Y A12 For P 1AY With P 2AY The distance between them; Y B12 For P 1BY With P 2BY The distance between them; Y B13 For P 1BY With P 3BY The distance between them; Substituting a, b, and c into the equation yields the rotation matrix. Substituting the coordinates of any point into the ship coordinate transformation model expression yields the translation vector Δ, and the coordinates of the ship's center of gravity in the calibrated attitude coordinate system A are obtained.

2. The ship dynamic ranging and positioning method according to claim 1, characterized in that, The specific sub-steps in step S11 are as follows: S111. Import the data to be filtered into the function and define a temporary variable temp_Point to store it; S112. Coarse reference sphere center grouping traversal: Based on the input variable sco, divide all data into numPoint data groups containing sco data. S113. Use the sphere center fitting algorithm to fit the coarse reference sphere center; S114. Coarse reference sphere center threshold evaluation: Calculate the distance of each sphere center to all data points and subtract them pairwise. If the result is within the range of the coarse reference sphere center threshold ser, then the weight of the sphere center is increased by 1. After all sphere centers have been calculated, the sphere center with the highest weight is selected as the coarse reference sphere center and returned, thereby determining the optimal reference solution of the coarse reference sphere center.

3. The ship dynamic ranging and positioning method according to claim 1, characterized in that, Step S2 further includes: S25. Set the coordinates of the center of the sphere to be fitted and the radius, and give the expected value and variance required to generate random noise, so as to generate random noise that conforms to a Gaussian distribution; S26. The generated noise is applied to the set sphere center radius, and the sphere center radius is added to the Gaussian noise to generate a sphere center radius that conforms to a Gaussian distribution; S27. The spherical coordinates are obtained using a polar coordinate to 3D coordinate transformation algorithm, specifically: ； In the formula, r is the radius of the sphere, radiusNoise is the random error applied to the radius, α is the pitch angle, and β is the azimuth angle; The range of α is (0, π), and the range of β is (0, 2π). α and β are assigned values ​​using random numbers, specifically: ； In the formula, rand1 and rand2 are random numbers in (0,1) to obtain the spherical coordinates of the ship.

4. The ship dynamic ranging and positioning method according to claim 1, characterized in that, In step S23, polar coordinates are used to describe the coordinates of the ship's sphere center. The coordinates of the ship's sphere center are: ； In the formula, r is the radius of the sphere, α is the pitch angle, and β is the azimuth angle; When loading noise onto the ship's sphere radius, the noise is distributed using a Gaussian distribution achieved through the Box-Muller algorithm, specifically as follows: ； Where U1 and U2 are uniformly distributed random numbers, and Z is a random number that follows a standard normal distribution; The specific expressions for generating a given mean and variance are: ； In the formula, u and sigma are manually defined expectation and variance. The expectation and variance are located at (0,1) and (0,4) respectively. The generated random bias is imported into Matlab to observe whether the random noise distribution curve conforms to a normal distribution.

5. The ship dynamic ranging and positioning method according to claim 1, characterized in that, The spherical equation for the ship in step S31 is: ； In the formula, (a, b, c) are the coordinates of the sphere's center, and r is the radius of the sphere. , , () represents the coordinates of a point on the sphere; Expanding the spherical equation of the ship, where a, b, c, r 2 -a 2 -b 2 -c 2 Treating this as a parameter to be estimated, when n spherical points are observed, then: ； In the formula, A is the coefficient matrix, Y is the observation vector, and X is the parameter to be estimated; The parameter X to be estimated is solved using total least squares, specifically: 。

Images