Network RTK-based high-precision positioning and orientation method on water

Abstract

The application provides a water high-precision positioning and orientation method based on network RTK, comprising the following steps: acquiring original observation data of at least two GNSS antennas arranged on a water surface engineering machine and real-time difference correction data; performing data preprocessing on the original observation data; fusing the data-preprocessed original observation data with the real-time difference correction data, and obtaining first position information of each GNSS antenna based on the fused data; obtaining second position information of the water surface engineering machine based on the first position information; determining a baseline vector formed by the two GNSS antennas according to the first position information of the two GNSS antennas, and calculating a projection direction of the baseline vector in a horizontal plane to obtain a real-time heading angle. The method designs a GNSS high-precision positioning system based on network RTK technology, improves the positioning precision by arranging double antennas, realizes the orientation function, and thus meets the real-time positioning and orientation requirements of a dredger and a multi-beam scanning unmanned ship.

Description

Technical Field

[0001] This invention relates to the field of GNSS positioning technology, and more specifically to a high-precision positioning and orientation method for waterborne applications based on network RTK. Background Technology

[0002] With the rapid development of dredging projects, waterway surveying, and unmanned surface platforms, the demand for high-precision, high-reliability real-time positioning and orientation of surface vessels such as dredgers, surveying vessels, and unmanned vessels is becoming increasingly urgent. Centimeter-level accuracy in planar position and real-time heading angle is crucial for controlling the precise excavation of the cutter head, the accurate throwing of the hopper, and the automated operation of dredgers, directly impacting the quality, efficiency, and safety of engineering construction. Current positioning technologies applied to surface vessels have the following limitations:

[0003] First, traditional single-point GNSS positioning technology has an accuracy of only meters, which is far from meeting the stringent requirements of centimeter-level accuracy in dredging operations. For example, in dredging operations, a positioning error of meters will directly lead to deviations in the dredging area, resulting in resource waste or engineering accidents.

[0004] Secondly, while real-time differential RTK technology using a self-built reference station can achieve centimeter-level accuracy, it requires users to establish and maintain a fixed reference station near the work area. This approach has inherent drawbacks such as deployment of reverse coordinates, high costs, limited operating range (e.g., limited by baseline length), and unsuitability for frequently moving work scenarios. For dredgers moving in vast waters, frequent relocation and establishment of reference stations will severely reduce operational efficiency.

[0005] Furthermore, when applied to the unique construction scenario of water-based construction machinery, a series of additional challenges arise due to strong signal reflection from the water surface, obstruction from the ship's own structure, and the complex electromagnetic environment of the work area. These challenges lead to a severe deterioration in the quality of the raw observation data, specifically manifested as: severe multipath effect errors; strong reflected signals from the water surface interfering with direct signals, resulting in observations containing systematic biases that are difficult to eliminate; momentary obstruction from the ship's superstructure, bridges, and boom, as well as the ship's dynamic swaying in waves, causing interruptions in satellite signal tracking; interference from other equipment in the work area, and low signal-to-noise ratio of low-elevation satellite signals, introducing large, random errors in the observations. If these errors are not systematically suppressed and eliminated from the data source and processing flow, they will directly lead to frequent jumps in positioning results, fixation failures, a sharp decline in accuracy, and even loss of lock.

[0006] Therefore, existing technologies lack an integrated solution that can address the unique interference caused by the harsh dynamic environment of the water surface to satellite observation data, and can optimize the entire chain from data preprocessing and robust calculation to error filtering to provide stable, reliable, and directly applicable centimeter-level positioning and orientation information for engineering practice. Summary of the Invention

[0007] The purpose of this invention is to provide a high-precision positioning and orientation method for waterborne applications based on network RTK. This method designs a GNSS high-precision positioning system based on network RTK technology, improves positioning accuracy by deploying dual antennas to achieve orientation function, and optimizes measurement coordinate transformation accuracy by using Kalman filtering and Bursa seven-parameter optimization, thereby meeting the real-time positioning and orientation requirements of dredgers and multi-beam sweeping unmanned vessels.

[0008] To achieve the above objectives, the present invention proposes the following technical solution:

[0009] A high-precision positioning and orientation method for waterborne applications based on network RTK includes the following steps:

[0010] Acquire raw satellite observation data from at least two GNSS antennas mounted on the surface engineering machinery, as well as real-time differential correction data from a continuously operating reference station network;

[0011] The raw satellite observation data is preprocessed.

[0012] The preprocessed raw satellite observation data is fused with the real-time differential correction data, and the first position information of each GNSS antenna is calculated based on the fused data.

[0013] The second location information of the water surface engineering machinery is obtained based on the first location information;

[0014] Based on the first position information corresponding to two of the GNSS antennas, the baseline vector formed by the two is determined, and the projection direction of the baseline vector on the horizontal plane is calculated to obtain the real-time heading angle of the water surface engineering machinery.

[0015] As a preferred embodiment of the present invention, the raw satellite observation data includes pseudorange observations and carrier phase observations;

[0016] The data preprocessing includes detecting and repairing clock jumps in the pseudorange and carrier phase observations.

[0017] As a preferred embodiment of the present invention, the detection and repair of clock jumps in the pseudorange observations and carrier phase observations includes:

[0018] Based on the pseudorange observations and the carrier phase observations, the epoch in which the clock jump occurs is detected using the inter-epoch difference method, and the clock jump value is calculated.

[0019] The distance deviation corresponding to the clock jump value is subtracted from the epoch in which the clock jump was detected and its subsequent pseudorange and carrier phase observations.

[0020] As a preferred embodiment of the present invention, the data preprocessing further includes detecting and identifying cycle slips of the carrier phase observations.

[0021] As a preferred embodiment of the present invention, the step of detecting and identifying cycle slips of the carrier phase observations includes:

[0022] The fitting method uses the MW combined observations and the GF combined observations for cycle slip detection;

[0023] When the verification value of any combination of observations exceeds the preset threshold, it is determined that the carrier phase observation of the corresponding satellite has a cycle slip, and the carrier phase observation that has a cycle slip is marked.

[0024] As a preferred technical solution of the present invention, in the process of solving the first location information, a robust estimation algorithm is used for positioning calculation to eliminate residual gross errors.

[0025] As a preferred technical solution of the present invention, the first position information is filtered to eliminate multipath effect error.

[0026] As a preferred technical solution of the present invention, the calculated first location information is converted from the geocentric coordinate system to a designated local engineering coordinate system.

[0027] As a preferred embodiment of the present invention, the step of converting the calculated first location information from the geocentric coordinate system to a specified local engineering coordinate system includes:

[0028] Construct coordinate transformation relationships using the Bursa seven-parameter model;

[0029] Based on the selected common control points, the seven transformation parameters of the Bursa seven-parameter model are solved;

[0030] Based on the seven transformation parameters obtained from the solution, the position information of each GNSS antenna is converted into coordinates in the local engineering coordinate system in real time.

[0031] As can be seen from the above technical solutions, the technical solution of the present invention provides a high-precision positioning and orientation method for waterborne applications based on network RTK, which has the following advantages compared with the prior art:

[0032] 1. It achieves high-precision and high-efficiency single-machine mobile positioning, breaking through the limitations of traditional operation modes.

[0033] This invention employs network RTK technology as the positioning basis, directly utilizing existing Continuously Operating Reference Stations (CORS) services. Compared to traditional RTK technology, which requires users to set up and maintain reference stations near the work site, this method eliminates the need for self-built reference stations, saving construction costs and time, and enabling true stand-alone operation. Simultaneously, the network RTK service boasts a large coverage radius (up to 20km), effectively overcoming the limitations of traditional RTK baseline length. It is particularly suitable for dredgers, survey vessels, and other mobile operations in vast waterways, significantly improving operational convenience and efficiency.

[0034] 2. Through full-chain data quality control, the unique interference of the dynamic water environment is effectively suppressed, ensuring the reliability and stability of the positioning solution.

[0035] To address the data quality issues caused by strong signal reflection from the water surface, ship hull obstruction, and dynamic rolling, this invention implements a full-chain optimization process from data source to solution output. In the data preprocessing stage, the receiver clock skipping is detected and repaired in real time using the inter-epoch difference method, and the MW+GF combination is used to effectively detect carrier phase cycle slips, marking them only to avoid erroneous repair. This eliminates the impact of systematic errors and most cycle slips at the source, providing a clean data foundation for high-precision solution. In the positioning solution stage, robust estimation is introduced, and the observation weight matrix is ​​dynamically adjusted iteratively to effectively resist the influence of residual small cycle slips and random gross errors on parameter estimation. Simultaneously, partial ambiguity fixing technology is applied, prioritizing the fixing of high-quality ambiguity subsets, improving the success rate and speed of ambiguity fixing in complex environments, thereby ensuring the rapid and stable acquisition of fixed solutions with centimeter-level accuracy. In the error filtering stage, Kalman filtering is used to smooth the position sequence, further suppressing residual errors such as multipath effects, making the positioning output results more stable and reliable.

[0036] 3. Through high-precision coordinate transformation, seamless integration of positioning results with engineering applications is ensured.

[0037] This invention employs a rigorous Bursa seven-parameter model for coordinate system transformation. This model can fully describe the translation, rotation, and scaling relationships between two three-dimensional Cartesian coordinate systems. By selecting uniformly distributed common control points within the survey area to solve for the high-precision seven parameters, the WGS-84 geocentric coordinates obtained from network RTK calculations can be losslessly converted to the local plane coordinate system coordinates and orthorium used in engineering design and construction. This step ensures that centimeter-level satellite positioning accuracy can be fully transferred to the final engineering results, allowing the positioning data to be directly used for construction layout, navigation control, and result comparison, meeting the stringent requirements of engineering practice.

[0038] 4. It achieves integrated positioning and orientation functions, and fully describes the motion posture of the carrier.

[0039] This invention deploys at least two GNSS antennas (a main antenna and an auxiliary antenna) on a surface engineering machine. While achieving high-precision positioning, it utilizes the calculated high-precision planar coordinates of the two antennas to calculate the projection direction of the baseline vector onto the horizontal plane in real time, thereby simultaneously obtaining the real-time heading angle of the vehicle. This provides a complete pose (position + orientation) perception solution, offering crucial state input for advanced applications such as precise control of dredger cutter heads and autonomous navigation of unmanned vessels.

[0040] 5. After practical verification, the system has demonstrated superior overall performance and meets high-standard engineering requirements.

[0041] As the experiment shows, the system optimized by the above technology achieves an average accuracy of ±0.55cm (X direction) and ±0.49cm (Y direction) in planar coordinates, and ±1.21cm in elevation. 100% of the points have a planar coordinate accuracy within 2cm. This demonstrates that the system possesses extremely high stability and reliability, and its accuracy and stability fully meet the stringent requirements for centimeter-level positioning and orientation in dredging operations and waterway surveying.

[0042] It should be understood that all combinations of the foregoing concepts and the additional concepts described in more detail below can be considered part of the inventive subject matter of this disclosure, provided that such concepts do not contradict each other.

[0043] The foregoing and other aspects, embodiments, and features of the teachings of the present invention will be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the invention, such as features and / or beneficial effects of exemplary embodiments, will become apparent from the following description or may be learned through practice of specific embodiments according to the teachings of the present invention. Attached Figure Description

[0044] The accompanying drawings are not drawn to scale. In the drawings, each identical or nearly identical component shown in the various figures may be denoted by the same reference numeral. For clarity, not every component is labeled in each figure. Embodiments of various aspects of the invention will now be described by way of example and with reference to the accompanying drawings, wherein:

[0045] Figure 1 The vector attitudes of the two GNSS antennas in this embodiment of the invention;

[0046] Figure 2 This is a schematic diagram of the Bursa seven-parameter model conversion according to an embodiment of the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art.

[0048] The terms "first," "second," and similar words used in the specification and claims of this patent application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular forms of "an," "a," or "the," etc., do not indicate a quantity limitation, but rather indicate the presence of at least one. Terms such as "comprising" or "including" indicate that the element or object preceding "comprising" encompasses the features, wholes, steps, operations, elements, and / or components listed following "comprising" or "including," and do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or sets thereof. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0049] This invention provides a high-precision positioning and orientation method for waterborne applications based on network RTK. It primarily utilizes network RTK technology as the foundation for high-precision real-time positioning, avoiding the cumbersome process of setting up custom reference stations. At least two dual-antenna arrays are deployed, achieving both high-precision positioning and direction finding by calculating the baseline vector of a set of antennas. Furthermore, to address errors during the positioning process, data preprocessing, robust estimation algorithms, and filtering are employed to handle errors caused by satellite clock slips, observation cycle slips, gross errors, and multipath effects, thereby optimizing the accuracy, reliability, and stability of the positioning solution. Finally, to address errors during coordinate transformation, a Bursa seven-parameter model is used to convert the geodetic coordinates obtained from GNSS measurements into local plane coordinates required for engineering applications with high precision.

[0050] In embodiments of the present invention, specific method steps include:

[0051] Step 1: Acquire raw satellite observation data from at least two GNSS antennas mounted on the water surface engineering machinery, and simultaneously receive real-time differential correction data from the Continuously Operating Reference (CORS) website via a communication link.

[0052] Specifically, GNSS receivers are mounted on surface engineering machinery (such as dredgers) and access a continuously operating reference station (CORS) network in the area via wireless communication links (such as mobile networks). This CORS network consists of multiple permanent GNSS reference stations with precisely known coordinates. During positioning, the GNSS receivers on the surface engineering machinery acquire raw satellite observation data in real time, including carrier phase observations and pseudorange observations. Simultaneously, the GNSS receivers receive real-time differential correction data, typically in RTCM format, from the CORS network's data service center via the communication link. This differential correction data is calculated and generated by the data service center based on data from all reference stations in the network and is transmitted in real time to lower-level GNSS receivers.

[0053] Therefore, this method utilizes existing, wide-area CORS network services to directly acquire high-precision network RTK differential data, thereby avoiding the need to set up and maintain one or more fixed reference stations near the work area. Compared with conventional RTK technology that requires users to set up their own reference stations, this method saves the cost and time of on-site station construction, overcomes the limitations of baseline length, and achieves high-precision single-machine mobile positioning in vast waters.

[0054] In some preferred embodiments of the present invention, two GNSS antennas are provided, including a main antenna and an auxiliary antenna, for receiving satellite signals and generating raw satellite observation data to be differentially analyzed, and for receiving real-time differential correction data broadcast by the data service center of the CORS network. After real-time fusion processing of these raw satellite observation data to be differentially analyzed with the received differential correction data, positioning information with centimeter-level accuracy is finally obtained.

[0055] Step 2: Perform data preprocessing on the raw satellite observation data to eliminate or reduce the data quality problems directly caused by the water surface environment (such as ship hull obstruction, dynamic swaying, etc.), so as to provide clean raw observation data for subsequent high-precision calculation.

[0056] Specifically, the raw satellite observation data is preprocessed, including: detecting and repairing clock slips in carrier phase observations and pseudorange observations, and detecting and marking cycle slips in carrier phase observations.

[0057] In this field, clock jump, or receiver clock jump, specifically refers to millisecond-level clock jumps. The main cause of clock jumps is the instability of the GNSS receiver's internal clock. The clock may experience sudden, integer-millisecond jumps due to temperature changes, voltage fluctuations, or component aging. Essentially, this means the receiver's time reference has "jumped." The receiver's local clock is crucial for measuring satellite signal propagation time. A single clock jump means a millisecond-level abrupt change in the timestamp recording the signal's arrival (1 millisecond is equivalent to light traveling approximately 300 kilometers in a vacuum). This directly leads to a significant error in the pseudorange observation calculated based on this incorrect timestamp. If not detected and corrected, this will instantly introduce positioning errors of hundreds of meters or even kilometers, causing the positioning calculation to completely fail.

[0058] Therefore, this embodiment of the invention employs the "epochal difference method" to detect and repair clock jumps in carrier phase observations and pseudorange observations in real time, which is a prerequisite for ensuring the continuous and reliable operation of the positioning system.

[0059] Specifically, this process is performed on the pseudorange observations P and carrier phase observations L generated by the GNSS receiver, aiming to detect and correct millisecond-level errors introduced by jumps in the receiver clock bias.

[0060] First, clock jump detection is performed. For satellite j, a check metric S is constructed at epoch i, and its calculation formula is:

[0061] , formula 1;

[0062] In Formula 1, j represents the satellite number; i represents the epoch; P is the pseudorange observation; L is the carrier phase observation; C is the speed of light; and k1 is the detection threshold. In this embodiment of the invention, considering that clock jumps occur in whole milliseconds, a detection threshold is set. Its value is approximately 300 km. When the verification values ​​of all available satellites at a certain epoch i are greater than the detection threshold k1, it is determined that there is a suspected clock jump at epoch i or that all satellites are experiencing a large cycle jump simultaneously.

[0063] Next, the clock jump value is calculated and determined for epochs suspected of having clock jumps. The specific formula used to calculate the candidate clock jump value m for this epoch is as follows:

[0064] , formula 2;

[0065] In Formula 2, α is a coefficient factor; specifically, in this embodiment of the invention, α = 10. 3 n is the number of available satellites in this epoch; S is the check metric; j represents the satellite number; C is the speed of light.

[0066] After calculating the candidate clock jump value m using Formula 2, the degree of closeness between the m and the nearest integer is used to determine whether a clock jump has occurred and to determine the final clock jump value J. s The unit is milliseconds. The specific formula for judgment is as follows:

[0067] , formula 3;

[0068] In Formula 3, m is the candidate value for clock jump; Round(m) is a positive function; k2 is the judgment threshold, which is 10 in this embodiment of the invention. -7 ~10 -5 .

[0069] According to Formula 3, if the absolute value of the difference between the candidate clock jump value m and the nearest integer is not greater than the judgment threshold k2, then a clock jump with a value of Round(m) milliseconds has occurred; otherwise, it is determined that no clock jump has occurred.

[0070] Finally, clock jump repair is performed. For the epoch at which the clock jump occurred and its subsequent pseudorange and carrier phase observations, the determined clock jump value J is subtracted from the affected pseudorange and carrier phase observations. s The corresponding distance deviation, i.e., J s •C, thereby eliminating the impact of clock jumps on pseudorange and carrier phase observations, providing reliable raw satellite observation data for subsequent high-precision positioning calculations.

[0071] Specifically, cycle slip, also known as integer cycle jump, occurs in carrier phase observations. GNSS receivers receive not only pseudorange measurements but also the phase of the satellite's carrier signal, i.e., carrier phase observations. The receiver continuously counts the integer cycles of the carrier, but this count is fragile. When the satellite signal is temporarily blocked (e.g., passing over bridges or trees), subjected to strong electromagnetic interference, or severely affected by multipath effects, the receiver's tracking of the carrier may be briefly interrupted (lost lock) and then re-locked. After re-locking, the receiver cannot determine how many complete waveform cycles were missed during the interruption, resulting in an integer multiple jump in the integer cycle count. Cycle slips disrupt the continuity of carrier phase observations, causing changes in the integer cycle ambiguity (i.e., the initial integer cycle unknown) before and after the interruption. If cycle slips are not detected, the positioning algorithm will continue to use incorrect integer cycle counts, leading to systematic deviations in the calculated position ranging from decimeters to meters.

[0072] Unlike clock slip handling, this embodiment of the invention does not repair cycle slips that occur in carrier phase observations. This is because incorrect cycle slip repair would have an adverse effect on parameter estimation. Instead, it accurately detects and marks the carrier phase observations that have cycle slips. In subsequent positioning calculations, these marked carrier phase observations are downweighted or deleted to ensure the reliability of the calculation results.

[0073] Furthermore, in this embodiment of the invention, the combined MW observations and GF observations are used to achieve cycle slip detection. The specific method is as follows:

[0074] Test value ΔT for the MW combined observations WL The test value ΔT of the combined observations of GF GF Calculated using the following formulas respectively:

[0075] , formula 4;

[0076] , formula 5;

[0077] In Formula 4, ΔT WL λ represents the test value of the MW combination observations; WL Indicates the wavelength of the wide-lane combined signal; Represents the wide-lane carrier phase observation; ΔP NL ξ represents the pseudorange value of the narrow aisle; ξ is the noise of the MW combined observations, and the standard deviation of ξ is usually between 0.1 and 0.5 cycles.

[0078] In Formula 5, ΔT GF ΔN1 and ΔN2 represent the test value of the GF combined observation; ΔN1 and ΔN2 are the changes in ambiguity of the carrier phase observations of carrier L1 and carrier L2, respectively, i.e., cycle slips; λ1 and λ2 are the wavelengths of carrier L1 and carrier L2, respectively; r represents the ratio of the squares of the two-frequency carrier frequencies, specifically... ΔI represents the ionospheric delay variation between epochs; ε is the noise of the GF combined observations, and the standard deviation of ε is usually between 0.005 and 0.015 m.

[0079] Specifically, for each frequency point of each satellite, its ΔT is continuously monitored. WL and ΔT GF Value, when ΔT WL The absolute value and ΔT GF If the absolute value exceeds the threshold set according to the noise level, it is determined that a cycle slip has occurred in the carrier phase observation value of the corresponding frequency of the satellite at that epoch. The carrier phase observation value is marked. In subsequent positioning calculations, the marked carrier phase observation value will be assigned a lower weight or will be directly removed from the observation equation, thereby avoiding cycle slip contamination of the positioning solution and ensuring the accuracy and stability of the final solution result.

[0080] Step 3: The original satellite observation data after the above clock slip and cycle slip processing is fused with the differential correction data. Based on the fused observation data, a double-difference observation equation is constructed, and the positioning solution is obtained by solving the double-difference observation method. This positioning solution is the first position information of the GNSS antenna, and its coordinates are generally represented in the geocentric coordinate system. Finally, the second position information of the water surface engineering machinery is obtained based on the first position information.

[0081] Furthermore, in solving the double-difference observation equation, a robust estimation strategy is employed for iterative solution to eliminate the influence of residual gross errors. A partial ambiguity fixing method is also applied during the solution process to quickly and reliably determine integer ambiguities, ultimately achieving stable first position information with centimeter-level accuracy. In this embodiment of the invention, the construction method of the double-difference observation equation adopts existing technology, which will not be elaborated further here.

[0082] Specifically, after the aforementioned preprocessing of clock slips and cycle slips, systematic clock slips and most large cycle slips in the original satellite observation data have been eliminated or identified. However, it cannot be ruled out that some small cycle slips or random gross errors may still be mixed into the normal observations. If these abnormal errors are directly used in the solution, they will affect the parameter estimation, leading to jumps in the positioning results or a decrease in accuracy. Therefore, this invention further employs a robust estimation algorithm in the positioning solution to eliminate the influence of such residual gross errors.

[0083] Specifically, this invention employs the IGG III robust estimation method as the core of the robust estimation process. This method constructs an equivalent weight function and dynamically adjusts the weights of each observation during the iterative solution process, thereby eliminating the influence of gross errors on the solution results. The definition of this equivalent weight function is as follows:

[0084] , Formula 6;

[0085] In Formula 6, P i The weights of the observations; The maximum residual; The values ​​are standardized residuals; k1 and k0 are both constants. In this embodiment of the invention, k0 = 1.5 and k1 = 3.0 are taken.

[0086] In practical implementation, the robust estimation algorithm is integrated into the iterative process of localization solution. After each iteration, the standardized residuals of each pseudorange observation and each carrier phase observation are calculated, and their weights are recalculated based on the aforementioned equivalent weight function. Subsequently, the new weight matrix is ​​used for the next iteration. This process is repeated, effectively preventing the gross error observations from affecting the final localization solution by successively reducing their weights. Preferably, the number of iterations in this invention is controlled to be 3 to 5. This robust estimation algorithm can significantly improve the reliability and accuracy of network RTK localization solutions in complex water surface environments.

[0087] After the iterative solution based on robust estimation effectively eliminates gross errors, the solution output is a floating-point solution containing parameters such as position, clock error, and ambiguity. To further achieve and stabilize the positioning solution at the centimeter level of accuracy, it is also necessary to correctly fix the integer ambiguity of the carrier phase. Therefore, this invention also applies a partial ambiguity fixing method during the solution process to improve the success rate of ambiguity fixing and the stability of the solution.

[0088] Specifically, in dynamic water environments, due to signal obstruction and complex errors, the ambiguities of all satellites may not be reliably fixed simultaneously. Partial ambiguity fixing is an effective approach. This method does not seek to fix all satellite ambiguities but instead selects an optimal subset based on reliability indicators such as satellite elevation angle, signal-to-noise ratio, and residual magnitude, while the rest remain floating-point solutions. This method avoids overall solution degradation caused by the failure to fix individual ambiguities, thus significantly improving the accuracy, speed, and overall reliability of the ambiguity-fixed solution. The partial ambiguity fixing method can employ well-established algorithms known in the field.

[0089] Step 4: Locate and filter the solution.

[0090] Due to signal reflection from the water surface and complex electromagnetic interference, the first position information sequence obtained after step three may still contain random errors caused by multipath effects. To obtain more stable and reliable positioning results, this embodiment of the invention further employs a Kalman filter algorithm to perform real-time filtering processing on the first position information sequence.

[0091] Kalman filtering is a recursive algorithm, and its system state equation and measurement equation are as follows:

[0092] Formula 7, System state vector equation;

[0093] Formula 8, Measurement Equation;

[0094] In formulas 7 and 8, X k Represents the state vector of the system; X represents the state transition matrix; k-1，k-1 W represents the optimal state estimate vector of the system in the previous epoch (time k-1), i.e., the posterior state estimate after updating with the previous observation data; k-1 Z represents the system noise matrix; k H represents the measurement vector; k The design matrix representing the system observation equations; V k This represents the measurement noise matrix.

[0095] This algorithm uses the initial position and velocity information output from step three as state variables. Through a prediction-update recursive equation, it continuously optimizes the state estimation using new observations. Its core filtering equations include the state prediction equation, the prediction error variance matrix, the state vector estimation equation, the gain matrix, and the estimation error variance matrix. By combining and optimizing these processes, the algorithm effectively smooths the position trajectory, suppresses high-frequency fluctuations, and outputs optimized, more stable final position information.

[0096] The specific equations above are as follows:

[0097] Formula 9, State prediction equation;

[0098] Formula 10, Prediction error variance matrix;

[0099] Formula 11, State vector estimation equation;

[0100] Formula 12, gain matrix;

[0101] Formula 13, estimate the error variance matrix;

[0102] In formulas 9-13, X k，k-1 This represents the state prediction vector for the current epoch (time k); X represents the state transition matrix; k-1，k-1 P represents the optimal state estimate vector for the previous epoch (time k-1); k，k-1 P represents the prediction error variance matrix; k-1 This represents the estimation error covariance matrix of the previous epoch (time k-1); Represents the state transition matrix Φ k，k-1 The transpose of X k K represents the state vector of the system. k Z represents the gain matrix; k H represents the measurement vector; k The design matrix represents the system observation equations; The system observation matrix, i.e., the design matrix H, represents the system observation matrix. K The transpose of Q; k R represents the dynamic noise variance matrix of the system; k H represents the observation noise variance matrix of the system; I represents the identity matrix; H represents the system noise variance matrix. k，k-1 The observation matrix representing the current epoch (time k) is essentially H. k .

[0103] Step 5: Coordinate system transformation.

[0104] Step 3 directly obtains the high-precision three-dimensional coordinates of the phase centers of the main antenna and auxiliary antenna installed on the hull in a geocentric coordinate system (e.g., WGS-84), which is the first position information.

[0105] It is important to clarify that GNSS positioning technology directly determines the position of the antenna phase center by receiving satellite signals. Since the main antenna's installation position on the hull is precisely determined and fixed in advance, its coordinates are used as the core reference for describing the spatial position of the entire surface engineering machinery. Using known and unchanging antenna installation geometry—specifically, the three-dimensional offset between the antenna phase center and hull design reference points, such as the hull center or cutter head center—the measured antenna coordinates can be converted in real-time and uniquely into the coordinates of any key operating point on the engineering machinery, i.e., the second position information of the surface engineering machinery. Therefore, in subsequent processing, performing coordinate system transformation on the first position essentially transforms the main antenna coordinates, which serve as the position reference. The transformed coordinates can then be directly used for engineering control or, combined with geometric relationships, to derive other required engineering coordinates.

[0106] Since the first position obtained in step three is based on the geocentric coordinate system, while in specific engineering projects such as dredging and tunnel surveying, all design and construction survey results are based on a specified local engineering coordinate system, usually a projected rectangular coordinate system, such as the coordinate system using the Gauss-Kruger projection, the coordinates output by GNSS must be converted to the local engineering coordinate system with high precision in order to seamlessly connect with the engineering design drawings and existing survey results and be directly used for construction guidance and position control.

[0107] Preferably, this invention employs the Bursa seven-parameter model for coordinate system transformation, which is a rigorous spatial coordinate transformation model suitable for spatial rectangular coordinate transformation between different datums and different ellipsoids. Compared with simplified models such as three-parameter and four-parameter models, it offers advantages such as... Figure 2 As shown, the seven-parameter model simultaneously considers three translations, three rotations, and a scale factor, enabling a more complete description of the complex relationship between two three-dimensional Cartesian coordinate systems. This achieves higher accuracy in transformations, making it particularly suitable for engineering projects covering large areas. The specific Bursa seven-parameter model and its transformation steps are as follows:

[0108] S1. Prepare common control point data. The purpose is to obtain known data pairs of the seven parameters used to solve the Bursa seven-parameter model.

[0109] Within the survey area, select no fewer than three common control points. These points must simultaneously possess two sets of known coordinates, including source coordinates and target coordinates. The source coordinates are high-precision WGB-84 geodetic coordinates (B1, L1, H1) obtained through GNSS static surveying or long-term network RTK observations. The target coordinates are the plane coordinates (x, y) of the point in the local engineering coordinate system obtained from existing surveying results, as well as the engineering normal height h obtained through leveling, denoted as (x, y, h).

[0110] S2. Construct the target space rectangular coordinates for solving, and convert the local engineering coordinates (x,y,h) into the target space rectangular coordinate format (x2,y2,z2) required by the Bursa seven-parameter model.

[0111] Specifically, for each common control point, the Gaussian coordinate-geometry inverse calculation formula is used to inversely calculate its plane coordinates (x,y) into the geodetic coordinates (B2,L2) of the corresponding local reference ellipse. At the same time, a theoretical geodetic height H' uniquely determined by geometric relationships is obtained. Therefore, the geodetic coordinates can be recorded as (B2,L2,H').

[0112] The specific formula for the Gaussian coordinate to geodetic coordinate inverse calculation is as follows:

[0113] , Formula 14;

[0114] , Formula 15;

[0115] in, , Formula 16;

[0116] In formulas 14, 15, and 16, L0 represents the longitude of the central meridian; B f The base point latitude represents the latitude corresponding to the meridian arc length when x=X; e is the first eccentricity; and a is the semi-major axis of the ellipsoid. It is worth noting that the Gaussian coordinate-geocentric coordinate inverse calculation method can also employ other existing calculation methods, as long as the coordinate inverse calculation can be achieved.

[0117] Since the project uses normal height, the theoretical geodetic height H' in the geodetic coordinates obtained in the previous step is replaced with the engineering normal height h to form a set of practical geodetic coordinates (B2,L2,h).

[0118] Substituting (B2,L2,h) into the spatial geodetic coordinate-spatial rectangular coordinate conversion formula, we can calculate the spatial rectangular coordinates (x2,y2,z2) of the point under the local reference ellipsoid.

[0119] The specific formula for converting between spatial geodetic coordinates and spatial rectangular coordinates is as follows:

[0120] , Formula 17;

[0121] In Formula 17, N is the radius of the zonal circle. In this embodiment of the invention, ;

[0122] a is the major semi-axis of the Earth ellipsoid, b is the minor semi-axis of the Earth ellipsoid, e is the first eccentricity of the Earth ellipsoid; L is the geodetic longitude; B is the geodetic latitude; H is the geodetic height.

[0123] S3. Construct the source space rectangular coordinates and solve for the seven parameters.

[0124] Using Equation 17, the WGS-84 geodetic coordinates (B1, L1, H1) of the common control points are converted to spatial rectangular coordinates (x1, y1, z1) under the WGS-84 ellipsoid. Thus, each common control point has a pair of coordinates: (x1, y1, z1) and (x2, y2, z2).

[0125] By substituting the coordinate pairs of all common control points into the Bursa seven-parameter model and performing adjustment calculations using the least squares method, the unique and optimal seven transformation parameters can be obtained.

[0126] Among them, such as Figure 2 As shown, the Bursa seven-parameter model is as follows:

[0127] , Formula 18;

[0128] In Formula 18, ΔX0, ΔY0, and ΔZ0 are three translation parameters; ε X ε Y ε Z There are three rotation parameters; m is the scale variation parameter.

[0129] S4. Real-time coordinate transformation.

[0130] Obtain the source coordinates and convert them to spatial rectangular coordinates. Specifically, convert the WGB-84 geodetic coordinates (B0, L0, H0) of the main antenna and auxiliary antenna obtained by real-time network RTK calculation into spatial rectangular coordinates (x0, y0, z0) using formula 17.

[0131] Substitute the (x0, y0, z0) and the seven parameters obtained in the previous step into the Bursa seven-parameter model to calculate the spatial rectangular coordinates (x3, y3, z3) of the antenna under the local reference ellipsoid.

[0132] (x3, y3, z3) is inversely calculated into geodetic coordinates (B3, L3, H3) under the local reference ellipse using formulas 14 and 15. Using the Gauss-Kruger projection forward calculation formula, (B3, L3) is converted into the required plane rectangular coordinates (x4, y4). The theoretical geodetic height is converted into the normal engineering height h using the geoid model. Finally, the complete coordinates (x4, y4, h) of the antenna in the local engineering coordinate system are obtained. (x4, y4) can be used to calculate the heading angle and for planar position control.

[0133] Step 6: Based on the first position information corresponding to two of the GNSS antennas, determine the baseline vector formed by the two antennas, and calculate the projection direction of the baseline vector on the horizontal plane to obtain the real-time heading angle of the water surface engineering machinery.

[0134] The preferred embodiment of this invention uses the baseline vectors of two GNSS antennas as an example, and the specific implementation details are as follows:

[0135] Both GNSS antennas, namely the main antenna and the auxiliary antenna, are located at the rear of the hull of the surface engineering machinery. The main antenna is located on the side closer to the wheelhouse, and the auxiliary antenna is located on the other side farther away from the wheelhouse. A vector pointing to the auxiliary antenna is established with the location of the main antenna as the origin. The projection direction of this vector on the horizontal plane is used to determine the real-time heading angle of the surface engineering machinery.

[0136] First, through the aforementioned positioning calculation process and coordinate transformation, the high-precision three-dimensional coordinates of the main antenna and the auxiliary antenna in the specified local engineering coordinate system are obtained in real time. Here, let the coordinates of the main antenna A be (x1, y1, z1) and the coordinates of the auxiliary antenna B be (x2, y2, z2).

[0137] Next, the baseline vector from the main antenna to the secondary antenna is calculated. ,like Figure 1 As shown, projecting this three-dimensional spatial vector onto the xoy plane (i.e., ignoring the elevation component Z) yields the horizontal projection vector. Its components are represented as (Δx, Δy) = (x2-x1, y2-y1).

[0138] Finally, based on the angle between the horizontal projection vector and the reference direction, such as the positive Y-axis of the coordinate system (usually assuming true north), the real-time heading angle θ is calculated. The specific calculation formula is as follows:

[0139] , Formula 19.

[0140] To verify the positioning accuracy and system stability achieved by this method, the following scheme can be designed:

[0141] Forty-four measuring points and one leveling point were selected as verification points within the survey area. Using a system equipped with this method, data was collected at each point in continuous topographic measurement mode with a sampling interval of 1 second. Data was collected continuously for 60 seconds at each point, and the original positioning results were recorded. Subsequently, the data was re-initialized and collected in the same manner at each point for three sets.

[0142] During data processing, the arithmetic mean of the three-dimensional coordinates (north coordinate X, east coordinate Y, elevation H) of all epochs for each measuring point is first calculated and used as the reference value for that point. Subsequently, the difference between the observed value and the reference value for each epoch is calculated in each direction, and the internal consistency accuracy (arithmetic mean error) of the system in the X, Y, and H directions is calculated using the following formula.

[0143] ;Formula 20;

[0144] In Formula 20, n is the total number of measurements at each point; v is the test value X at each point. i The difference between the system and the corresponding test arithmetic mean in the x, y, and H directions, respectively; M is the mean error of the system's internal compliance accuracy arithmetic mean in the x, y, and H directions, respectively.

[0145] Finally, the results of the internal consistency accuracy calculation for all 44 measuring points were statistically analyzed, and the overall accuracy index is shown in Table 1.

[0146] Table 1 Internal Compliance Accuracy Statistics

[0147]

[0148] The statistical results in Table 1 show that after processing using the entire method (including data preprocessing, robust solution, Kalman filtering, and Bursa-Tropsch seven-parameter coordinate transformation), the average accuracy of the system in the plane coordinates (X, Y directions) is ±0.55cm and ±0.49cm, respectively, and the average accuracy in the elevation (H direction) is ±1.21cm. 100% of the points have a plane coordinate accuracy within 2cm, and 96% have an elevation accuracy within 3cm.

[0149] The above-mentioned accuracy verification results prove that the method can effectively overcome the interference of dynamic environment on the water surface. The network RTK positioning system has extremely high stability and reliability. Its horizontal and vertical accuracy fully meets the stringent requirements of centimeter-level high-precision positioning and orientation for engineering applications such as dredging operations and waterway surveying.

[0150] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A network RTK-based high-precision positioning and orientation method on water, characterized in that, Includes the following steps: Acquire raw satellite observation data from at least two GNSS antennas mounted on the surface engineering machinery, as well as real-time differential correction data from a continuously operating reference station network; The raw satellite observation data is preprocessed. The preprocessed raw satellite observation data is fused with the real-time differential correction data, and the first position information of each GNSS antenna is calculated based on the fused data. The second location information of the water surface engineering machinery is obtained based on the first location information; Based on the first position information corresponding to two of the GNSS antennas, the baseline vector formed by the two is determined, and the projection direction of the baseline vector on the horizontal plane is calculated to obtain the real-time heading angle of the water surface engineering machinery.

2. The network RTK-based high-precision positioning and orientation method on water according to claim 1, characterized in that, The raw satellite observation data includes pseudorange observations and carrier phase observations; The data preprocessing includes detecting and repairing clock jumps in the pseudorange and carrier phase observations.

3. The network RTK-based high-precision positioning and orientation method on water according to claim 1, characterized in that, The detection and repair of clock jumps in the pseudorange and carrier phase observations includes: Based on the pseudorange observations and the carrier phase observations, the epoch in which the clock jump occurs is detected using the inter-epoch difference method, and the clock jump value is calculated. The distance deviation corresponding to the clock jump value is subtracted from the epoch in which the clock jump was detected and its subsequent pseudorange and carrier phase observations.

4. The network RTK-based high-precision positioning and orientation method on water according to claim 2, characterized in that, The data preprocessing also includes detecting and identifying cycle slips in the carrier phase observations.

5. The network RTK-based high-precision positioning and orientation method on water according to claim 4, characterized in that, The detection and identification of cycle slips in the carrier phase observations includes: The fitting method uses the MW combined observations and the GF combined observations for cycle slip detection; When the verification value of any combination of observations exceeds the preset threshold, it is determined that the carrier phase observation of the corresponding satellite has a cycle slip, and the carrier phase observation that has a cycle slip is marked.

6. The high-precision positioning and orientation method for waterborne applications based on network RTK according to claim 1, characterized in that, In the process of solving the first location information, a robust estimation algorithm is used for positioning calculation to eliminate residual gross errors.

7. The high-precision positioning and orientation method for waterborne applications based on network RTK according to claim 1, characterized in that, The first location information is filtered to eliminate multipath effect errors.

8. The high-precision positioning and orientation method for waterborne applications based on network RTK according to claim 1, characterized in that, The first location information obtained from the solution is transformed from the geocentric coordinate system to the specified local engineering coordinate system.

9. The high-precision positioning and orientation method for waterborne applications based on network RTK according to claim 8, characterized in that, The step of converting the calculated first location information from the geocentric coordinate system to the specified local engineering coordinate system includes: Construct coordinate transformation relationships using the Bursa seven-parameter model; Based on the selected common control points, the seven transformation parameters of the Bursa seven-parameter model are solved; Based on the seven transformation parameters obtained from the solution, the position information of each GNSS antenna is converted into coordinates in the local engineering coordinate system in real time.

Images