A system for determining the dynamic attitude of a ship using high-frequency GNSS data

Abstract

This invention relates to the field of data analysis technology, specifically to a system for obtaining the dynamic attitude of a ship using high-frequency GNSS data. The system includes a data preprocessing module, a data precalculation module, a data integration module, and a data analysis module. The data preprocessing module establishes a target dataset; the data precalculation module obtains the theoretical rotation angle and actual error value of the target vector; the data integration module combines the theoretical rotation angle and actual error value to obtain an analysis array; and the data analysis module performs data analysis on the analysis array to obtain the relevant types of the theoretical rotation angle and actual error value, establishes an error calculation model, obtains the theoretical error value based on the error calculation model, and calculates the tilt angle based on the theoretical error value to reflect the ship's tilt attitude. This effectively solves the problems of poor accuracy and low frequency in traditional methods of obtaining ship attitude.

Description

Technical Field

[0001] This invention relates to the field of data analysis technology, and in particular to a system for obtaining the dynamic attitude of a ship from high-frequency GNSS data. Background Technology

[0002] Hull attitude refers to the ship's motion state during navigation, reflecting its tilt angle and rolling state, etc. Hull attitude plays an important role in navigation safety, transportation efficiency and the development of intelligent ships.

[0003] In existing technologies, tilt sensors, such as inclinometers and levels, are typically used to acquire relevant information about the hull. The hull's attitude information is obtained by processing this information. However, tilt sensors have significant drawbacks, such as slow dynamic response. When the hull's attitude changes rapidly, they cannot capture the angle change in time, resulting in hysteresis errors. Furthermore, under conditions of impact interference, the measurement data from inclinometers is prone to fluctuations or errors, resulting in poor accuracy and an inability to provide timely and effective responses to the hull's attitude. Summary of the Invention

[0004] The purpose of this invention is to solve the problems in the background art, and to propose a system for obtaining the dynamic attitude of a ship using high-frequency GNSS data.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a system for obtaining the dynamic attitude of a ship using high-frequency GNSS data, comprising:

[0006] As a further aspect of the present invention, three target points are respectively set at three different positions on the hull, and a total station is used to collect the initial position information of the three target points respectively. Initial data is established based on the three position information and uploaded to the data preprocessing module; the three target points are at the same horizontal level.

[0007] The data preprocessing module is used to preprocess the initial data to obtain a standard vector set, analyze the standard vector set, and determine the target dataset.

[0008] The method for preprocessing the initial data is as follows:

[0009] The three target vectors formed by the coordinates of the three target points are calculated by vector calculation method, and the three target vectors are combined to obtain the standard vector set.

[0010] The magnitude of the target vector is calculated according to the vector magnitude formula. The vector magnitudes are then compared, and the target vector with the largest magnitude is marked as a non-target vector. The non-target vectors are then removed from the standard vector set.

[0011] Mark the vector magnitude corresponding to the target vector as the target magnitude L, obtain the two corresponding coordinates in the target vector and mark them as target coordinates, calculate the height difference between the two target coordinates by subtracting their height values, and mark the height difference as the target height h. Construct the target dataset {L, h} based on the target magnitude L and the target height h.

[0012] The data pre-calculation module is used to obtain the theoretical rotation angle based on the target dataset and the measured rotation angle based on the existing equipment. It calculates the actual error value of the target vector based on the measured rotation angle and the theoretical rotation angle, with a one-to-one correspondence between the theoretical rotation angle and the actual error value.

[0013] The data integration module uses the theoretical rotation angle and the corresponding actual error value as the first sub-item and the second sub-item, respectively, and combines the first sub-item and the second sub-item to obtain the analysis array;

[0014] The data analysis module splits and arranges several analysis arrays to obtain a reference sequence and a characteristic sequence. It then performs data analysis on the reference sequence and the characteristic sequence to obtain the correlation type between the theoretical rotation angle and the actual error value. Based on the correlation type, it establishes an error accounting model and calculates the inclination angle of the hull based on the error accounting model.

[0015] As a further aspect of the present invention, the method for calculating the actual error value is as follows:

[0016] By training formula The actual error value SWθ of the target vector is calculated; Cθ is the measured rotation angle, and Jθ is the theoretical rotation angle, which is calculated using the arcsine function Jθ=arcsin(h / L);

[0017] By setting up several different attitudes of the ship hull, i.e. different measurement rotation angles, and repeating the above method for calculating the actual error value, several analysis arrays are obtained.

[0018] As a further aspect of the present invention, the method for constructing the error accounting model is as follows:

[0019] Several analysis arrays are numbered, and the analysis array numbers are assigned to the first and second sub-items in the analysis array. The first and second sub-items in the analysis array are split, and the reference sequence is obtained by sorting them from smallest to largest according to the first sub-item.

[0020] Obtain the number of the first sub-item in the reference sequence and use it as the reference number. Sort the reference numbers and sort the second sub-items according to the reference number sorting to obtain the feature sequence. Determine the correlation type between the first and second sub-items.

[0021] As a further aspect of the present invention, the first sequence element of the feature sequence is set as the starting element, the second sequence element of the feature sequence is set as the target element, the size relationship between the target element and the starting element is compared, the second sequence element is updated to the starting element, the third sequence element is updated to the target element, the size relationship between the target element and the starting element is compared, and the feature sequence is traversed in sequence to complete the traversal, and a comparison model is established based on the size relationship between the starting element and the target element.

[0022] The expression for the comparison model is: BX(i, i+1) is the comparison analysis value between the starting element and the target element. (i, i+1) represents the comparison relationship between the starting element and the target element. i is the element number of the feature sequence. i is a positive integer. i∈[1, N-1] and N is the total number of elements in the feature sequence.

[0023] As a further aspect of the present invention, the comparison analysis value BX(i, i+1) is monitored. If the analysis value BX(i, i+1) is the same, the traversal continues. If BX(i, i+1) changes, the sequence element with element number i+1 corresponding to the current comparison analysis value BX(i, i+1) is marked as an abnormal element until the feature sequence is traversed and the traversal result is obtained.

[0024] By judging the formula The monitoring and analysis value JX is calculated, and data analysis is performed on the monitoring and analysis value JX. If the monitoring and analysis value JX is not less than the positive correlation threshold ZY or the monitoring and analysis value JX is not greater than the negative correlation threshold FY, then a decision instruction is generated. If the monitoring and analysis instruction JX∈(FY, ZY), then it is determined that the first sub-item and the second sub-item are not related.

[0025] 7. A system for obtaining ship dynamic attitude from high-frequency GNSS data according to claim 6, characterized in that: according to the decision instruction, two sequence elements adjacent to the abnormal element are marked as decision elements; the decision elements are arranged from smallest to largest according to the element number i of the decision elements to obtain a decision sequence; it is determined whether the sequence elements in the decision sequence are linearly changing; if they are linearly changing, the correlation type of the first sub-item and the second sub-item is determined to be linearly correlated; otherwise, the first sub-item and the second sub-item are determined to be uncorrelated.

[0026] As a further aspect of the present invention, the correlation coefficient is calculated as follows:

[0027] Through the year-on-year parameter formula The year-on-year parameter between two adjacent elements in the reference sequence and two adjacent elements in the feature sequence is calculated. Where CKm and TZj represent the sequence elements of the reference sequence and the sequence elements of the feature sequence, respectively; h and j are the sequence element numbers of the reference sequence and the feature sequence, respectively; m=j, m∈[1,N-1];

[0028] Calculated using the year-on-year coefficient formula; The year-on-year coefficient TX of the first and second sub-items;

[0029] Year-on-year parameters The reference value is obtained by calculating the difference between the reference value and the year-on-year coefficient TX. The minimum reference value is then obtained and its corresponding year-on-year parameter is calculated. Mark the two adjacent elements in the reference sequence corresponding to the target year-on-year parameter and the two adjacent elements in the feature sequence as the first key element and the second key element, respectively. Calculate the mean of the two first key elements and use the result as the first standard reference item. Calculate the mean of the two second key elements and use the result as the second standard reference item.

[0030] Using the geometric formula The standard error value was calculated. FC is the first standard reference sub-item, and SC is the second standard reference sub-item.

[0031] As a further aspect of the present invention, if the correlation type is unrelated, multiple second sub-items of the same first sub-item are obtained through testing, and the average of the multiple second sub-items is calculated to obtain the average second sub-item. The first sub-item and the average second sub-item are recorded. Each time a corresponding second sub-item of the same first sub-item is obtained, the average second sub-item is updated. The average second sub-item is marked as a feature error value. e is the number of the first sub-item, e is a positive integer, and e∈[1,N-1];

[0032] Based on the first sub-item and the characteristic error value Establish a mapping function F(e), and obtain the feature error value of the first sub-item based on the mapping function F(e). ;

[0033] The expression for the error accounting model is:

[0034] LWθ is the theoretical error value;

[0035] By solving the formula The calculated inclination angle Sθ of the hull is obtained, and the inclination angle Sθ reflects the tilting attitude of the hull.

[0036] Compared with existing technologies, the advantages of this invention are as follows: a target dataset is established through a data preprocessing module; the theoretical rotation angle and actual error value of the target vector are obtained through a data precalculation module; the theoretical rotation angle and actual error value are combined to obtain an analysis array through a data integration module; the analysis array is analyzed through a data analysis module to obtain the relevant types of the theoretical rotation angle and actual error value, and an error accounting model is established; the theoretical error value is obtained according to the error accounting model; and the inclination angle is calculated based on the theoretical error value to reflect the inclination attitude of the hull. This effectively solves the problems of poor accuracy and low frequency in traditional methods of obtaining hull attitude. Attached Figure Description

[0037] Figure 1 This is a flowchart of the modules of the present invention. Detailed Implementation

[0038] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0039] Reference Figure 1 A system for obtaining the dynamic attitude of a ship using high-frequency GNSS data includes three target points set at three different locations on the ship. A total station is used to collect the initial position information of the three target points. Initial data is established based on the three position information and uploaded to the data preprocessing module. The three target points are located at the same horizontal level.

[0040] The data preprocessing module is used to preprocess the initial data to obtain a standard vector set, analyze the standard vector set, and determine the target dataset. It should be noted that the lateral spacing between the three target points is not less than 2 / 3 of the ship's width, and the longitudinal spacing is not less than 1 / 2 of the ship's length.

[0041] The method for preprocessing the initial data is as follows:

[0042] The three target point vectors formed by the coordinates of the three target points are calculated by vector calculation method, and the three target point vectors are combined to obtain a standard vector set. The initial position coordinates of the target points are obtained by acquiring satellite positioning data with frequencies from 1559MHz to 1610MHz from three high-frequency GNSS receivers. The target point vectors are calculated by vector calculation formula. Satellite positioning and vector calculation are existing technologies and will not be described in detail here.

[0043] The magnitude of the target vector is calculated using the vector magnitude formula. The vector magnitudes are then compared, and the target vector with the largest magnitude is marked as a non-target vector. The non-target vectors are then removed from the standard vector set. The remaining target vectors in the standard vector set are marked as target vectors. The calculation of the vector magnitude can be achieved using existing calculation formulas, which will not be elaborated here.

[0044] Mark the vector magnitude corresponding to the target vector as the target magnitude L, obtain the two corresponding coordinates in the target vector and mark them as target coordinates, calculate the height difference between the two target coordinates by subtracting their height values, and mark the height difference as the target height h. Construct the target dataset {L, h} based on the target magnitude L and the target height h.

[0045] The data pre-calculation module is used to obtain the theoretical rotation angle based on the target dataset and the measured rotation angle based on the existing equipment. It calculates the actual error value of the target vector based on the measured rotation angle and the theoretical rotation angle, with a one-to-one correspondence between the theoretical rotation angle and the actual error value.

[0046] The data integration module uses the theoretical rotation angle and the corresponding actual error value as the first sub-item and the second sub-item, respectively, and combines the first sub-item and the second sub-item to obtain the analysis array;

[0047] The method for calculating the actual error value is as follows:

[0048] By training formula The actual error value SWθ of the target vector is calculated; Cθ is the measured rotation angle, and Jθ is the theoretical rotation angle, which is calculated using the arcsine function Jθ=arcsin(h / L);

[0049] By setting up several different attitudes of the ship hull, i.e. different measurement rotation angles, and repeating the above method for calculating the actual error value, several analysis arrays are obtained.

[0050] The data analysis module splits and arranges several analysis arrays to obtain a reference sequence and a characteristic sequence. It then performs data analysis on the reference sequence and the characteristic sequence to obtain the correlation type between the theoretical rotation angle and the actual error value. Based on the correlation type, it establishes an error accounting model and calculates the inclination angle of the hull based on the error accounting model.

[0051] The method for constructing the error accounting model is as follows:

[0052] S101. Number several analysis arrays and assign the analysis array numbers to the first and second sub-items in the analysis array. Split the first and second sub-items in the analysis array and sort them from smallest to largest according to the first sub-item to obtain the reference sequence.

[0053] Obtain the number of the first sub-item in the reference sequence and use it as the reference number. Sort the reference numbers and sort the second sub-items according to the reference number sorting to obtain the feature sequence. Determine the correlation type between the first sub-item and the second sub-item.

[0054] S102. Set the first sequence element of the feature sequence as the starting element, and the second sequence element of the feature sequence as the target element. Compare the size relationship between the target element and the starting element. Then update the second sequence element as the starting element, update the third sequence element as the target element, and compare the size relationship between the target element and the starting element. Repeat this process to traverse the feature sequence and establish a comparison model based on the size relationship between the starting element and the target element.

[0055] The expression for the comparison model is: BX(i, i+1) is the comparison analysis value between the starting element and the target element, (i, i+1) represents the comparison relationship between the starting element and the target element, i is the element number of the feature sequence, i is a positive integer, i∈[1, N-1], and N is the total number of elements in the feature sequence; it should be noted that the total number of elements in the reference sequence is the same as the total number of elements in the feature sequence.

[0056] The comparison analysis value BX(i, i+1) is monitored. If the analysis value BX(i, i+1) is the same, the traversal continues. If BX(i, i+1) changes, the sequence element with the element number i+1 corresponding to the current comparison analysis value BX(i, i+1) is marked as an abnormal element until the feature sequence is traversed and the traversal result is obtained.

[0057] By judging the formula The monitoring and analysis value JX is calculated, and data analysis is performed on the monitoring and analysis value JX. If the monitoring and analysis value JX is not less than the positive correlation threshold ZY or the monitoring and analysis value JX is not greater than the negative correlation threshold FY, then a decision instruction is generated. If the monitoring and analysis value JX ∈ (FY, ZY), then it is determined that the first sub-item and the second sub-item are not related. The positive correlation threshold ZY and the negative correlation threshold FY are both obtained based on historical test big data. The positive correlation threshold ZY can be 0.8, and the negative correlation threshold FY can be 0.2.

[0058] According to the decision instruction, both sequence elements adjacent to the abnormal element are marked as decision elements. The decision elements are arranged in ascending order according to their element number i to obtain the decision sequence. It is determined whether the sequence elements in the decision sequence are linearly related. If they are linearly related, the correlation between the first and second sub-items is determined to be linearly related; otherwise, the first and second sub-items are determined to be unrelated. It should be noted that linear changes can be increasing or decreasing.

[0059] If the correlation type is linear, the reference ratio sequence is obtained by calculating the sequence elements of the reference sequence, and the characteristic ratio sequence is obtained by refining the sequence elements of the characteristic sequence. The correlation coefficient between the first sub-item and the second sub-item is calculated based on the reference ratio sequence and the characteristic ratio sequence.

[0060] The correlation coefficient is calculated as follows:

[0061] Through the year-on-year parameter formula The year-on-year parameter between two adjacent elements in the reference sequence and two adjacent elements in the feature sequence is calculated. Where CKm and TZj represent the sequence elements of the reference sequence and the sequence elements of the feature sequence, respectively; h and j are the sequence element numbers of the reference sequence and the feature sequence, respectively; m=j, m∈[1,N-1];

[0062] Calculated using the year-on-year coefficient formula; The year-on-year coefficient TX of the first and second sub-items;

[0063] Year-on-year parameters The reference value is obtained by calculating the difference between the reference value and the year-on-year coefficient TX. The minimum reference value is then obtained and its corresponding year-on-year parameter is calculated. Mark the two adjacent elements in the reference sequence corresponding to the target year-on-year parameter and the two adjacent elements in the feature sequence as the first key element and the second key element, respectively. Calculate the mean of the two first key elements and use the result as the first standard reference item. Calculate the mean of the two second key elements and use the result as the second standard reference item.

[0064] Using the geometric formula The standard error value was calculated. FC is the first standard reference sub-item, and SC is the second standard reference sub-item;

[0065] If the correlation type is unrelated, multiple second sub-items of the same first sub-item are obtained through testing, and the average second sub-item is calculated by averaging the multiple second sub-items. The first sub-item and the average second sub-item are recorded. Specifically, the average second sub-item is updated each time a corresponding second sub-item of the same first sub-item is obtained. The average second sub-item is then marked as the feature error value. e is the number of the first sub-item, e is a positive integer, and e∈[1,N-1];

[0066] Based on the first sub-item and the characteristic error value Establish a mapping function F(e), and obtain the feature error value of the first sub-item based on the mapping function F(e). ;

[0067] The expression for the error accounting model is:

[0068] LWθ is the theoretical error value;

[0069] By solving the formula The calculated inclination angle Sθ of the hull is obtained, and the inclination angle Sθ reflects the tilting attitude of the hull.

[0070] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A system for determining the dynamic attitude of a ship hull from high frequency GNSS data, characterized by: The three targets are arranged at three different positions of the ship body, and a total station instrument is arranged at each position to collect initial position information of the three targets, and initial data is established according to the three position information and uploaded to a data preprocessing module; The three targets are at the same height. The data preprocessing module is used for preprocessing the initial data to obtain a standard vector set, and analyzing the standard vector set to determine a target data set. The method for preprocessing the initial data is: Three target vectors are calculated by vector calculation method, and the three target vectors are combined to obtain a standard vector set. The length of the target vector is calculated according to the vector module formula, and the length of the target vector is obtained. The maximum vector length is obtained by comparing the lengths of the target vectors, and the corresponding target vector is marked as a non-target vector. The length of the target vector is marked as the target length L, and the two coordinates corresponding to the target vector are obtained and marked as target coordinates. The height difference between the two target coordinates is obtained by calculating the difference between the height values of the two target coordinates, and the height difference is marked as the target height h. The target data set {L, h} is constructed according to the target length L and the target height h.

2. The system for calculating the dynamic attitude of the ship body according to claim 1, characterized in that: The data preprocessing module is used for obtaining a theoretical rotation angle according to the target data set and a measurement rotation angle according to an existing device, calculating an actual error value of the target vector according to the measurement rotation angle and the theoretical rotation angle, and the theoretical rotation angle and the actual error value correspond one by one. By training formula The actual error value SWθ;Cθ of the target vector is calculated, the measured rotation angle is Cθ, and the theoretical rotation angle is Jθ, which is calculated by the inverse sine function Jθ=arcsin(h / L). The data integration module combines the theoretical rotation angle and the corresponding actual error value as a first sub-item and a second sub-item, respectively, and combines the first sub-item and the second sub-item to obtain an analysis array.

3. The system for calculating the dynamic attitude of the ship body according to claim 2, characterized in that: The data analysis module splits and arranges a plurality of analysis arrays to obtain a reference sequence and a feature sequence, and performs data analysis on the reference sequence and the feature sequence to obtain a correlation type of the theoretical rotation angle and the actual error value. The error accounting model is established according to the correlation type, and the inclination of the ship body is calculated according to the error accounting model. The method for calculating the actual error value is: A plurality of different postures of the ship body are set, i.e. The method for constructing the error accounting model is: The analysis arrays are numbered, and the first sub-item and the second sub-item in the analysis array are numbered. The first sub-item and the second sub-item in the analysis array are split, and the first sub-item is sorted from small to large to obtain a reference sequence. The number of the first sub-item in the reference sequence is obtained and used as a reference number, the reference number is sorted, the second sub-item is sorted according to the reference number, a feature sequence is obtained, and the correlation type of the first sub-item and the second sub-item is determined.

4. The system for calculating the dynamic attitude of the ship body according to claim 3, characterized in that: The first bit sequence element of the characteristic sequence is set as a starting element, the second bit sequence element of the characteristic sequence is set as a target element, the size relationship between the target element and the starting element is compared, the second bit sequence element is updated as the starting element, the third bit sequence element is updated as the target element, the size relationship between the target element and the starting element is compared, and the characteristic sequence is sequentially traversed, and a comparison model is established according to the size relationship between the starting element and the target element; The expression of the comparison model is: ; BX(i, i+1) is a comparison analysis value of the start element and the target element, (i, i+1) represents a comparison relationship of the start element and the target element, i is an element number of the feature sequence, i is a positive integer, i∈[1, N-1], N is a total number of elements of the feature sequence.

5. The system for calculating the dynamic attitude of the ship's hull from high-frequency GNSS data according to claim 4, characterized in that: The comparison analysis value BX(i, i+1) is monitored, if the analysis value BX(i, i+1) is the same, the traversal is continued, if the BX(i, i+1) changes, the sequence element with the element number i+1 corresponding to the current comparison analysis value BX(i, i+1) is marked as an abnormal element, until the characteristic sequence is traversed, and a traversal result is obtained; By judging formula The monitoring analysis value JX is calculated, and data analysis is performed on the monitoring analysis value JX; if the monitoring analysis value JX is not less than the positive correlation threshold ZY or the monitoring analysis value JX is not greater than the negative correlation threshold FY, a decision instruction is generated; if the monitoring analysis value JX is in the range (FY, ZY), it is judged that the first sub-item and the second sub-item are irrelevant.

6. The system for calculating the dynamic attitude of the ship's hull from high-frequency GNSS data according to claim 5, characterized in that: According to the decision instruction, the two sequence elements adjacent to the abnormal element are marked as decision elements, the decision elements are arranged from small to large according to the element number i of the decision elements to obtain a decision sequence, whether the sequence elements in the decision sequence are linearly changed is judged, if the sequence elements are linearly changed, it is judged that the correlation type of the first subterm and the second subterm is linear correlation, otherwise, it is judged that the first subterm and the second subterm are irrelevant.

7. The system for calculating the dynamic attitude of the ship hull according to claim 6, characterized in that: The calculation method of the correlation coefficient is: The same parameter formula is calculated by The same parameter between two adjacent elements in the reference sequence and two adjacent elements in the characteristic sequence is calculated ; wherein CKm and TZj represent sequence elements of the reference sequence and sequence elements of the characteristic sequence respectively, h and j are sequence element numbers of the reference sequence and sequence element numbers of the characteristic sequence respectively, m = j, m ∈ [1, N-1]; The same ratio coefficient is calculated by the same ratio coefficient formula; The same ratio coefficient TX of the first sub-item and the second sub-item; The same ratio parameter The reference value is obtained by difference calculation of the same ratio parameter and the same ratio coefficient TX, the minimum reference value is obtained, and the same ratio parameter corresponding to the minimum reference value is marked as a target same ratio parameter The two adjacent elements in the reference sequence corresponding to the target same ratio parameter are marked as a first key element and a second key element, respectively, and the two first key elements are subjected to mean value calculation and the result is taken as a first standard reference item, and the two second key elements are subjected to mean value calculation and the result is taken as a second standard reference item. By the geometric progression formula The standard error value is calculated ; FC is the first standard reference sub-item, and SC is the second standard reference sub-item.

8. The system for calculating the dynamic attitude of the ship hull according to claim 7, characterized in that: If the relevant type is no correlation, a plurality of second sub-items of the same first sub-item are obtained by testing, and the plurality of second sub-items are averaged to obtain an average second sub-item, and the first sub-item and the average second sub-item are recorded; wherein, the average second sub-item is updated every time a corresponding second sub-item of the same first sub-item is obtained; the average second sub-item is marked as a characteristic error value , e is the number of the first sub-item, e is a positive integer, e∈[1, N-1] According to the first subterm and the feature error value The mapping function F(e) is established, and the feature error value of the first subterm is obtained according to the mapping function F(e) ; The expression of the error accounting model is: ; LW0 is the theoretical error value; The inclination angle Sθ of the ship body is calculated by the equation The inclination angle Sθ of the ship body is calculated by the equation The inclination angle Sθ of the ship body is calculated by the equation

Images