High-precision offset positioning sounding auxiliary system and use method

Abstract

The invention relates to the technical field of surveying and mapping positioning, in particular to a high-precision offset positioning sounding auxiliary system and a use method, and the system comprises a positioning sensing module, an attitude and offset measurement module, a sound velocity profile measurement module, a data processing unit, and a communication and storage module. And the data of the positioning sensing module, the attitude and offset measurement module and the sound velocity profile measurement module are transmitted to the data processing unit through the communication and storage module for fusion processing. The system is suitable for high-precision sounding scenes of complex water areas such as near banks, ports, channels, underwater obstacle dense areas and the like; the problem that the absolute coordinate precision of a depth point is insufficient due to multi-source errors in traditional depth sounding positioning is solved.

Description

Technical Field

[0001] This invention relates to the field of surveying and positioning technology, specifically to a high-precision offset positioning depth sounding auxiliary system and its usage method. Background Technology

[0002] Underwater topographic surveying is one of the core tasks of marine surveying, and its results are widely used in areas such as channel dredging, submarine pipeline laying, port construction, and marine resource exploration. The accuracy of depth sounding directly depends on the accuracy of the absolute coordinates of the sounding point, which is determined by two parts: first, the real-time position of the sounding transducer, including latitude, longitude, and depth; and second, the relative geometric relationship between the transducer and the sounding point, involving installation offset and attitude effects.

[0003] Existing technologies for depth sounding positioning have significant drawbacks. Traditional methods often rely on a single GNSS positioning transducer, but GNSS signals are susceptible to multipath effects from water surface reflections and obstruction, resulting in positioning accuracy only at the meter level. If an inertial navigation system is used for positioning alone, long-term drift errors accumulate significantly, reaching tens of meters per hour. While underwater acoustic positioning systems, such as long-baseline and short-baseline systems, can provide relative positioning, they require pre-deployment of beacon arrays, are complex to install, and are susceptible to underwater acoustic propagation delays and noise interference, making real-time integration with GNSS or inertial navigation systems difficult. Furthermore, transducer installation deviations and changes in the water's sound velocity profile cause sound ray bending; traditional methods, which rely solely on fixed empirical values ​​for correction, cannot adapt to dynamic environmental changes, further reducing depth sounding accuracy.

[0004] Therefore, there is an urgent need for an underwater bathymetry and positioning assistance system and method that can integrate multi-source data, possesses high accuracy, and is robust, in order to solve the aforementioned technical problems. Existing technologies urgently need improvement to address these issues. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a high-precision offset positioning depth sounding auxiliary system and its usage method.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a high-precision offset positioning depth sounding auxiliary system, comprising:

[0007] The positioning and sensing module integrates an RTK-GNSS receiver, an optical fiber inertial navigation system (INS), and a short baseline underwater acoustic positioning receiver (USBL) to provide the vehicle's absolute position, velocity information, and relative position correction.

[0008] The attitude and offset measurement module includes two high-precision inertial measurement units (IMUs), namely IMU-A and IMU-B, which are orthogonally mounted on the transducer base. They are pre-stored with the offset and attitude deviation of the transducer relative to the carrier coordinate system and are used to acquire high-frequency attitude data.

[0009] The sound velocity profile measurement module includes a sound velocity profiler, which is used to acquire sound velocity profile data of water bodies in real time.

[0010] The data processing unit is an embedded industrial computer, equipped with algorithms for multi-source data synchronization, adaptive Kalman filtering, attitude fusion, and error compensation.

[0011] The communication and storage module enables data synchronization between sensors via Ethernet or RS485 bus and provides data storage functionality.

[0012] The data from the positioning and sensing module, attitude and offset measurement module, and sound velocity profile measurement module are all transmitted to the data processing unit for fusion processing through the communication and storage module. The data flow serial relationship between the modules is defined as follows: the sound velocity profile measurement module collects sound velocity data and sends it to the data processing unit; the attitude and offset measurement module collects IMU data and sends it to the data processing unit; the positioning and sensing module collects position data and sends it to the data processing unit; and the processing results are output to the communication and storage module for storage.

[0013] To achieve the above objectives, the present invention also provides the following technical solution: a method for using a high-precision offset positioning depth sounding auxiliary system, wherein the auxiliary system is employed, and the method includes the following steps:

[0014] Step S1: Hardware installation and calibration. Deploy the RTK-GNSS antenna, INS, USBL beacon array, sound velocity profiler and dual IMUs on the measurement ship. Perform static calibration on the dual IMUs using a three-axis turntable. Measure and pre-store the installation rotation matrix and attitude deviation of the transducer base to compensate for installation errors.

[0015] Step S2: Multi-source data synchronous acquisition. The timestamps of RTK-GNSS, INS, USBL, dual IMU and sound velocity profiler are aligned through Precise Time Protocol (PTP) so that the output of all sensors is unified to the GNSS time reference, and data is acquired at a preset frequency, wherein the sampling frequency of dual IMU is at least 200Hz.

[0016] Step S3: Data preprocessing, the raw angular velocity and acceleration data of the dual IMUs are filtered by sliding window midpoint filtering to remove high-frequency noise, and polynomial temperature compensation is performed on the IMU zero bias based on temperature sensor data;

[0017] Step S4: Dual IMU attitude measurement. Based on the preprocessed data in step S3, the dual IMU data is fused using an adaptive Kalman filter. The state vector is designed as the transducer base attitude angle and the dual IMU zero bias variables. The observation matrix is ​​based on the installation rotation matrix to transform the IMU data to the transducer base coordinate system, and the fused high-precision attitude data is output.

[0018] Step S5: Transformation between the carrier coordinate system and the geographic coordinate system. Based on the fused attitude output in step S4, a rotation matrix is ​​constructed to transform the transducer position in the carrier coordinate system to the geographic coordinate system. Sound velocity profile data is applied for sound ray bending compensation, and the horizontal distance is calculated through layered integration to correct the slant distance.

[0019] Step S6: Multi-source data fusion and error suppression. Based on the compensation output of step S5 and RTK-GNSS, INS, and USBL data, an extended Kalman filter is used for fusion processing. The state vector includes the carrier position, velocity, and sound speed error to suppress drift noise.

[0020] Step S7: Calculate the absolute coordinates of the sounding points. Combining the beam offset and depth data of the multibeam echo sounder, the positioning results output in step S6, and the fused attitude data in step S4, calculate the absolute coordinates of each sounding point using the geometric projection formula.

[0021] Step S8: Anomaly detection and self-calibration. Real-time monitoring of the dual IMU output residuals and differences from step S4. If an anomaly is detected, the operating mode is switched, and the faulty IMU is calibrated online under stable sea conditions.

[0022] In some embodiments, step S1, the hardware installation and calibration includes the following sub-steps:

[0023] The dual IMUs are fixed to the transducer base at an orthogonal position using shock-absorbing rubber pads. The natural frequency of the shock-absorbing rubber pads is less than 10Hz, and the vibration attenuation rate is greater than or equal to 80%.

[0024] Temperature sensors are installed near each IMU to collect temperature data to support subsequent zero-bias polynomial temperature compensation.

[0025] In some embodiments, step S2, the multi-source data synchronous acquisition includes:

[0026] The sensor timestamps are aligned using a hardware synchronization mechanism based on a cross-correlation algorithm, with a timestamp accuracy of less than or equal to 100 ns.

[0027] Wave height and wind speed environmental data are converted into sampling frequency timestamps for dual IMUs using a linear interpolation method to ensure data consistency.

[0028] In some embodiments, step S4 involves designing the adaptive Kalman filter as follows:

[0029] The state vector is defined as x = [θ, φ, ψ, b] gA ,b gB ,b aA ,b aB], where θ, φ, and ψ are the roll, pitch, and yaw angles of the transducer base, respectively, and b gA ,b gB The gyroscope zero bias of IMU-A and IMU-B are respectively, b aA ,b aB These are the accelerometer zero bias values;

[0030] The observation equation is based on the installation rotation matrix, which transforms the dual IMU angular velocities to the transducer base coordinate system, thereby achieving adaptive adjustment of the observation noise covariance.

[0031] In some embodiments, in step S4, the adaptive adjustment of the observation noise covariance is dynamically performed based on the sea state grade index, specifically as follows:

[0032] Sea state rating index S = 0.5H w +0.1V w H w V represents the wave height. w The wind speed is used as the criterion, and the sea state is classified into calm (S<0.5), moderate (0.5≤S<2), and severe (S≥2);

[0033] The adjustment factor is set according to the sea state level, and the basic value of the gyro noise covariance of the IMU-A is 1×10. -5 rad 2 / s 2 The adjustment factor is 1 (calm), 3 (moderate), or 5 (severe). The adjustment factor for IMU-B is 0.6 times that of IMU-A.

[0034] In some embodiments, in step S8, the anomaly detection includes real-time calculation of the innovation residual vector of the adaptive Kalman filter. If the residual norm exceeds three times the trace square root of the residual covariance matrix, or the difference between the individual IMU attitude angle and the fused attitude angle is greater than 0.1 degrees, then it is determined to be an IMU anomaly.

[0035] In some embodiments, step S8 includes the self-calibration strategy as follows:

[0036] When a single IMU malfunction is detected, switch to the single-source mode of the healthy IMU to output the attitude.

[0037] Under steady sea state (S<0.5), the measurement was paused and the dual IMUs were left stationary for 10 minutes to calculate and update the mean difference of the zero bias parameter;

[0038] If both IMUs fail, switch to RTK-GNSS heading angle as the backup attitude source.

[0039] In some embodiments, step S5, the horizontal position correction after acoustic ray bending compensation further includes: calculating the nonlinear effects of horizontal offset and depth based on sound velocity profile layering data through integration, using the formula: Where c i For the layer sound velocity, Δz i The layer thickness is determined and integrated into the coordinate transformation results to compensate for horizontal deviations.

[0040] In some embodiments, in step S6, the extended Kalman filter output positioning result of the multi-source data fusion has a planar accuracy of less than ±2cm and a vertical accuracy of less than ±3cm. The method ends in step S7 with the output of the sounding point coordinates, and the coordinate formula is:

[0041] x d =x t +l·cosψ·cosθ-d·sinφ·cosψ+d·cosφ·sinψ

[0042] y d =y t +l·sinψ·cosθ-d·sinφ·sinψ-d·cosφ·cosψ

[0043] z d =z t -d·cosφ·cosθ-l·sinθ

[0044] Where x t ,y t ,z t denoted as transducer position, l and d as beam offset and depth, and ψ, φ, and θ as heading, pitch, and roll attitude angles, respectively.

[0045] Compared with the prior art, the beneficial effects of the present invention are: by integrating multi-source sensor data, dual IMU fusion attitude measurement and dynamic error compensation algorithm, it solves the problems of low positioning accuracy, drift error accumulation and dynamic changes in sound velocity profile in the traditional technology, and has the advantages of multi-source data fusion, high-precision attitude measurement, dynamic error compensation and strong robustness.

[0046] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. The embodiments of this application will provide a detailed description and understanding of the application. Attached Figure Description

[0047] Figure 1 System structure block diagram;

[0048] Figure 2 This is a flowchart of the data processing process;

[0049] Figure 3 This is a diagram illustrating coordinate transformation.

[0050] Figure 4 This is a schematic diagram illustrating the principle of correcting vocal timbre curvature. Detailed Implementation

[0051] 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 embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] In traditional underwater topographic surveying systems, errors from a single positioning source lead to insufficient accuracy in transducer position calculation. The lack of a real-time fusion mechanism between underwater acoustic positioning and inertial navigation systems results in inadequate compensation for acoustic ray bending caused by dynamic attitude deviations and changes in sound velocity profiles. For example, during multibeam bathymetry operations in nearshore waters, GNSS signals are obstructed by bridges, causing multipath effects and positioning errors exceeding meters. Short-baseline underwater acoustic positioning systems suffer from centimeter-level relative position errors due to beacon array deployment deviations, and misalignment with the inertial navigation system's time reference leads to millisecond-level delays in fused data. The transducer base experiences dynamic attitude shifts due to ship roll and pitch, compounded by vertical layering changes in the sound velocity profile, resulting in nonlinear distortion of the sound wave propagation path and accumulated horizontal projection distance calculation errors reaching decimeter levels. If these problems are not addressed, the planar and vertical errors of the absolute coordinates of the bathymetry points will exceed the engineering allowable range, causing submarine pipeline laying trajectories to deviate from the design path, insufficient excavation depth for port construction foundations, reduced reliability of marine resource exploration data, and increased construction safety risks and cost overruns.

[0053] To address the aforementioned challenges, this application first considers how to integrate multi-source positioning data to overcome the limitations of a single positioning source. GNSS signals are susceptible to interference, leading to absolute position errors, while underwater acoustic positioning systems suffer from relative position deviations and lack real-time fusion mechanisms. To resolve this, this application attempts to combine an RTK-GNSS receiver, a fiber optic inertial navigation system, and a short-baseline underwater acoustic positioning receiver, improving positioning reliability through multi-source data complementarity. Simultaneously, regarding dynamic attitude deviation, this application finds that traditional single-IMU measurements are susceptible to vibration interference and lack self-correction capabilities. Therefore, it proposes an orthogonally mounted dual-IMU scheme, utilizing redundant measurements to suppress random noise. Furthermore, insufficient compensation for acoustic ray bending caused by changes in sound velocity profiles necessitates real-time acquisition of water sound velocity data to dynamically correct the sound wave propagation path. To achieve these functions, this application further designs a multi-source data synchronization mechanism, using Ethernet or RS485 bus to unify the transmission time reference, ensuring that data streams from each sensor are processed in series, avoiding fusion errors caused by millisecond-level delays.

[0054] In this regard, such as Figures 1 to 4 As shown, this application proposes a high-precision offset positioning and depth sounding auxiliary system, comprising: a positioning sensing module, integrating an RTK-GNSS receiver, a fiber optic inertial navigation system, and a short-baseline underwater acoustic positioning receiver, for providing the absolute position, velocity information, and relative position correction of the carrier; an attitude and offset measurement module, including two high-precision inertial measurement units, IMU-A and IMU-B, orthogonally mounted on a transducer base, pre-stored with the transducer's offset and attitude deviation relative to the carrier coordinate system, for acquiring high-frequency attitude data; a sound velocity profile measurement module, including a sound velocity profiler, for acquiring real-time sound velocity profile data of the water body; and a data processing unit, an embedded industrial control computer, equipped with... The system executes multi-source data synchronization, adaptive Kalman filtering, attitude fusion, and error compensation algorithms. The communication and storage module synchronizes data between sensors via Ethernet or RS485 bus and provides data storage functionality. Data from the positioning and sensing module, attitude and offset measurement module, and sound velocity profile measurement module are all transmitted to the data processing unit for fusion processing via the communication and storage module. The serial data flow relationship between the modules is defined as follows: the sound velocity profile measurement module collects sound velocity data and sends it to the data processing unit; the attitude and offset measurement module collects IMU data and sends it to the data processing unit; the positioning and sensing module collects position data and sends it to the data processing unit; and the processing results are output to the communication and storage module for storage.

[0055] Among them, the RTK-GNSS receiver refers to a real-time dynamic differential global navigation satellite system receiver, which can be implemented by combining a dual-frequency multi-constellation receiver with differential signals from a ground reference station. It provides centimeter-level absolute positioning accuracy, solving the meter-level error problem caused by multipath effects and obstruction in traditional GNSS systems. The fiber optic inertial navigation system refers to an inertial measurement device composed of fiber optic gyroscopes and accelerometers, which can be implemented using a closed-loop fiber optic gyroscope and temperature compensation algorithms. It provides high-frequency carrier motion parameters and suppresses long-term drift errors. The short-baseline underwater acoustic positioning receiver refers to an ultra-short baseline positioning system composed of an underwater beacon array and a shipborne receiving array. It can be implemented using a multi-element receiving transducer combined with a time difference of arrival (TDOA) calculation method. It provides relative position correction between the transducer and the underwater beacon, compensating for positioning errors caused by underwater acoustic propagation delays. The high-precision inertial measurement unit (IMU) refers to a miniature sensor assembly equipped with a three-axis gyroscope and accelerometer. Specifically, it can be implemented using a combination of MEMS gyroscopes and quartz accelerometers. Orthogonal mounting eliminates the installation error of a single sensor, enabling high-frequency acquisition of the transducer base attitude. The sound velocity profiler is an instrument that measures the vertical distribution of sound velocity in water in real time. Specifically, it can be implemented using a CTD sensor combined with empirical formulas for sound velocity calculation. This dynamically corrects sound ray bending errors, replacing traditional fixed empirical sound velocity compensation methods. The adaptive Kalman filter algorithm is an estimation algorithm that dynamically adjusts filter parameters based on sensor noise characteristics. Specifically, it can be implemented using covariance matching technology combined with innovation sequence monitoring. This is used to suppress drift noise and improve fusion accuracy during multi-source data fusion. Multi-source data synchronization refers to aligning the sampling times of different sensors using a unified time base. Specifically, it can be implemented using the IEEE 1588 precise time protocol combined with hardware trigger signals, ensuring time consistency during data fusion and resolving fusion errors caused by asynchronous data.

[0056] The core innovation of this application lies in the integration of three composite positioning sources: RTK-GNSS, fiber optic inertial navigation, and short-baseline underwater acoustic positioning. Combined with dual IMU attitude measurement and dynamic compensation of sound velocity profile, centimeter-level depth sounding positioning accuracy is achieved through multi-source data synchronization and adaptive filtering algorithms, overcoming the problems of single positioning source error accumulation, underwater acoustic propagation delay, and dynamic environmental interference.

[0057] The working process and principle of this application are as follows: The high-precision offset positioning and depth sounding auxiliary system includes a positioning sensing module, an attitude and offset measurement module, a sound velocity profile measurement module, a data processing unit, and a communication and storage module. The positioning sensing module integrates an RTK-GNSS receiver, a fiber optic inertial navigation system, and a short-baseline underwater acoustic positioning receiver, providing the absolute position, velocity information, and relative position correction of the carrier. The attitude and offset measurement module includes two high-precision inertial measurement units, IMU-A and IMU-B, orthogonally mounted on the transducer base, pre-stored the transducer's offset and attitude deviation relative to the carrier coordinate system, and acquires high-frequency attitude data. The sound velocity profile measurement module includes a sound velocity profiler, which acquires water body sound velocity profile data in real time. The data processing unit is an embedded industrial control computer that executes multi-source data synchronization, adaptive Kalman filtering, attitude fusion, and error compensation algorithms. The communication and storage module achieves data synchronization between sensors via Ethernet or RS485 bus, providing data storage functionality.

[0058] Data from each module is transmitted to the data processing unit for fusion processing via the communication and storage module. The data flow sequence is as follows: the sound velocity profile measurement module collects sound velocity data and sends it to the data processing unit; the attitude and offset measurement module collects IMU data and sends it to the data processing unit; the positioning and sensing module collects position data and sends it to the data processing unit; and the processing results are output to the communication and storage module for storage.

[0059] By fusing multi-source data, the system overcomes the limitations of a single positioning source. RTK-GNSS provides high-precision absolute position, the fiber optic inertial navigation system compensates for short-term positioning errors, and the short-baseline underwater acoustic positioning system provides underwater relative position correction. Orthogonal mounting of dual IMUs improves the redundancy and reliability of attitude measurements. Real-time sound velocity profile data is used to dynamically correct the sound wave propagation path. A multi-source data synchronization mechanism ensures the temporal consistency of data from each sensor, avoiding fusion errors.

[0060] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0061] The positioning and sensing module employs a Trimble BD990 RTK-GNSS receiver, an iMAR iNAT-RQT-4003 fiber optic inertial navigation system, and a Sonardyne Ranger 2USBL underwater acoustic positioning receiver. The RTK-GNSS receiver is mounted on top of the hull, with a geodesic antenna. The fiber optic inertial navigation system is installed near the ship's center of gravity. The USBL transducer is mounted on the bottom of the hull.

[0062] The attitude and offset measurement module employs two Honeywell HG4930 high-precision MEMSIMUs, mounted on two orthogonal surfaces of the transducer base. The IMU sampling rate is set to 200Hz. The transducer's offset and attitude deviation relative to the carrier coordinate system are obtained through precise measurement and pre-stored in the system.

[0063] The sound velocity profile measurement module uses a Valeport MIDAS SVP sound velocity profiler, which is periodically lowered into the water by a winch to collect sound velocity data.

[0064] The data processing unit uses an Advantech ARK-3520P embedded industrial computer, equipped with an Intel Core i7 processor and running the Ubuntu operating system. The adaptive Kalman filter algorithm is developed and implemented based on the ROS platform.

[0065] The communication and storage modules use gigabit Ethernet switches for data transmission and NAS network storage devices for data storage.

[0066] After system startup, each sensor collects data at a preset frequency and transmits it to the data processing unit via Ethernet. The data processing unit first synchronizes the multi-source data in time, then executes an adaptive Kalman filter algorithm to fuse RTK-GNSS, INS, and USBL data, while simultaneously combining dual IMU attitude data and sound velocity profile data for error compensation. Finally, it outputs high-precision transducer position and attitude information, which is transmitted to the multibeam echo sounding system via the communication module for calculating the absolute coordinates of the sounding points.

[0067] Through the above scheme, this application achieves the fusion of multi-source positioning data, overcoming the limitations of a single positioning source. The orthogonal installation of dual IMUs improves the redundancy and reliability of attitude measurement and effectively suppresses random noise. Real-time sound velocity profile data is used to dynamically correct the sound wave propagation path, improving the accuracy of horizontal projected distance calculation. The multi-source data synchronization mechanism ensures the temporal consistency of data from each sensor, avoiding fusion errors. These techniques work together to significantly improve the planar and vertical accuracy of the absolute coordinates of the sounding points, providing more reliable data support for applications such as subsea pipeline laying, port construction, and marine resource exploration.

[0068] In some of the solutions described above in this application, data fusion errors are caused by inconsistent sensor timestamps during the synchronous acquisition of multi-source data. At the same time, the installation errors and temperature drift of the dual IMUs are not effectively compensated, which affects the attitude measurement accuracy.

[0069] This application further proposes a method for using a high-precision offset positioning depth sounding auxiliary system, which includes the following steps: hardware installation and calibration, deploying an RTK-GNSS antenna, INS, USBL beacon array, sound velocity profiler, and dual IMUs on a survey vessel; statically calibrating the dual IMUs using a three-axis turntable; measuring and pre-storing the installation rotation matrix and attitude deviation of the transducer base to compensate for installation errors; multi-source data synchronous acquisition, aligning the timestamps of the RTK-GNSS, INS, USBL, dual IMUs, and sound velocity profiler using a precise time protocol to unify all sensor outputs to the GNSS time reference; and acquiring data at a preset frequency, wherein the sampling frequency of the dual IMUs is at least 200Hz; data preprocessing, using a sliding window mid-range filter to remove high-frequency noise from the raw angular velocity and acceleration data of the dual IMUs, and performing polynomial temperature compensation for the IMU zero bias based on temperature sensor data; dual IMU attitude fusion measurement, using an adaptive Kalman filter to fuse the dual IMU data based on the preprocessed data, with the state vector designed as a transition... The system includes several key functions: transducer base attitude angle and dual IMU zero-bias variables; the observation matrix, based on the installation rotation matrix, transforms IMU data to the transducer base coordinate system, outputting fused high-precision attitude data; transformation between the carrier coordinate system and the geographic coordinate system; based on the fused attitude output, a rotation matrix is ​​constructed to transform the transducer position in the carrier coordinate system to the geographic coordinate system, and sound velocity profile data is applied for ray bending compensation; horizontal distance is calculated through layered integration to correct slant range; multi-source data fusion and error suppression; based on the compensated output and RTK-GNSS, INS, and USBL data, an extended Kalman filter is used for fusion processing, and the state vector includes carrier position, velocity, and sound velocity errors to suppress drift noise; absolute coordinate calculation of sounding points; combining beam offset and depth data from the multibeam echo sounder, as well as the output positioning results and fused attitude data, the absolute coordinates of each sounding point are calculated using a geometric projection formula; anomaly detection and self-correction; real-time monitoring of the dual IMU output residuals and differences; if an anomaly is detected, the operating mode is switched, and the faulty IMU is calibrated online under stable sea conditions.

[0070] In the hardware installation and calibration steps, the dual IMUs are fixed to an orthogonal position on the transducer base using vibration-damping rubber pads. The natural frequency of the vibration-damping rubber pads is less than 10Hz and the vibration attenuation rate is greater than or equal to 80%. Temperature sensors are installed near each IMU to collect temperature data. In the multi-source data synchronous acquisition step, a hardware synchronization mechanism based on a cross-correlation algorithm is used to align the sensor timestamps, with a timestamp accuracy of less than or equal to 100ns. Environmental data is converted into sampling frequency timestamps for the dual IMUs using a linear interpolation method. The state vector of the adaptive Kalman filter includes the transducer base attitude angle and the zero-bias variables of the dual IMUs. The observation equation is based on the installation rotation matrix to achieve data conversion. Anomaly detection triggers a self-correction strategy by calculating the norm of the innovation residual vector and the attitude angle difference, including switching to single-source mode, calibrating the zero-bias parameters at rest, or enabling the GNSS backup attitude source.

[0071] Specifically, during the hardware installation phase, the orthogonal layout of the dual IMUs, combined with shock-absorbing rubber pads, effectively isolates hull vibration interference, while the inherent frequency limitation ensures effective attenuation of high-frequency mechanical vibrations. Real-time data acquisition from the temperature sensor is used to construct a nonlinear relationship model between zero bias and temperature, achieving dynamic compensation through cubic polynomial fitting. Multi-source data synchronization employs a combination of a precise time protocol and hardware cross-correlation algorithms, controlling the time reference error of heterogeneous sensors such as GNSS, INS, and USBL to the sub-microsecond level, ensuring the data time alignment accuracy of subsequent fusion algorithms. The 200Hz high-frequency sampling of the dual IMUs, combined with sliding window mid-range filtering, suppresses impulse noise while preserving effective dynamic signals. The adaptive Kalman filter dynamically adjusts the observation noise covariance matrix, matching optimal filtering parameters according to the real-time sea state level. For example, in severe sea states, the gyro noise covariance coefficient of IMU-A is increased to 5 times the base value, while IMU-B uses a 0.6-fold scaling factor, achieving robust attitude estimation. Acoustic ray bending compensation uses a hierarchical integration algorithm, calculating the horizontal offset corresponding to each layer of sound velocity profile data, which is then accumulated and superimposed to correct the transducer position in the geographic coordinate system. The extended Kalman filter (EPF) constructs a state vector incorporating sound velocity errors by fusing GNSS absolute position, INS relative displacement, and USBL underwater acoustic positioning data, effectively suppressing drift errors caused by multipath effects and underwater acoustic propagation delays. In the sounding point coordinate calculation stage, the geometric projection formula combines beam offset with fused attitude data, and eliminates the influence of installation deviations through rotation matrix transformation, ultimately outputting absolute coordinates with a planar accuracy of ±2cm and a vertical accuracy of ±3cm. The anomaly detection module continuously monitors the trace square root of the residual covariance of the dual IMUs. When it exceeds a threshold, it triggers a fault isolation mechanism, such as suspending operations in calm sea conditions for zero-bias parameter recalibration, ensuring continuous and reliable system operation.

[0072] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0073] First, hardware installation and calibration were performed. The RTK-GNSS antenna, INS, USBL beacon array, sound velocity profiler, and dual IMUs were deployed on the survey vessel. The dual IMUs were statically calibrated using a three-axis turntable, and the installation rotation matrix and attitude deviation of the transducer base were measured and pre-stored to compensate for installation errors.

[0074] Secondly, multi-source data synchronous acquisition is performed. The timestamps of RTK-GNSS, INS, USBL, dual IMU and sound velocity profiler are aligned using Precise Time Protocol (PTP) to unify the output of all sensors to the GNSS time reference, and data is acquired at a preset frequency, with the sampling frequency of the dual IMU set to 200Hz.

[0075] Next, data preprocessing is performed. The raw angular velocity and acceleration data from the dual IMUs are filtered using a sliding window mid-range filter to remove high-frequency noise, and polynomial temperature compensation is performed on the IMU zero bias based on temperature sensor data.

[0076] Then, dual-IMU fusion attitude measurement is performed. Based on the preprocessed data, an adaptive Kalman filter is used to fuse the dual-IMU data. The state vector is designed to be the transducer base attitude angle and the dual-IMU zero-bias variables. The observation matrix is ​​based on the installation rotation matrix to transform the IMU data to the transducer base coordinate system, and the fused high-precision attitude data is output.

[0077] Furthermore, a transformation between the carrier coordinate system and the geographic coordinate system is performed. Based on the fused attitude, a rotation matrix is ​​constructed to transform the transducer position in the carrier coordinate system to the geographic coordinate system, and sound velocity profile data is applied for sound ray bending compensation. The horizontal distance is calculated through layered integration to correct the slant range.

[0078] Subsequently, multi-source data fusion and error suppression are performed. Based on the compensated output and RTK-GNSS, INS, and USBL data, an extended Kalman filter is used for fusion processing. The state vector includes the carrier position, velocity, and sound speed errors to suppress drift noise.

[0079] Next, the absolute coordinates of the sounding points are calculated. Combining the beam offset and depth data from the multibeam echo sounder, along with the positioning results and fused attitude data, the absolute coordinates of each sounding point are calculated using the geometric projection formula.

[0080] Finally, anomaly detection and self-calibration are performed. The output residuals and differences of the dual IMUs are monitored in real time. If an anomaly is detected, the operating mode is switched, and the faulty IMU is calibrated online under stable sea conditions.

[0081] Through the above technical solution, this application realizes a method for using a high-precision offset positioning bathymetry auxiliary system. This improves the accuracy of the absolute coordinates of the bathymetry point, reduces the error of a single positioning source, overcomes the limitations of underwater acoustic positioning systems, and fully compensates for attitude and sound velocity errors. Specifically, this method improves the accuracy and reliability of bathymetry positioning through multi-source data fusion and adaptive algorithms, adapting to dynamic marine environments. For example, by fusing attitude measurements with dual IMUs and employing a self-calibration strategy, the robustness of the system is enhanced, and the impact of sensor failures on measurement results is reduced. Furthermore, through real-time sound velocity profile compensation and multi-source data fusion, positioning errors caused by underwater acoustic ray bending are mitigated, improving the accuracy of the bathymetry point coordinates.

[0082] In some of the solutions described above in this application, how to effectively reduce the impact of vibration interference and temperature changes on the IMU measurement accuracy during hardware installation and calibration, and ensure the accuracy of static calibration of dual IMUs.

[0083] This application further proposes hardware installation and calibration including the following sub-steps: fixing the dual IMUs to the orthogonal position of the transducer base using shock-absorbing rubber pads, the natural frequency of the shock-absorbing rubber pads being less than 10Hz and the vibration attenuation rate being greater than or equal to 80%; installing temperature sensors near each IMU to collect temperature data to support subsequent zero-bias polynomial temperature compensation.

[0084] Among them, the inherent frequency limitation of the shock-absorbing rubber pad suppresses the transmission of high-frequency mechanical vibration to the IMU, and the vibration attenuation rate index ensures that the proportion of mechanical energy converted into heat energy reaches a preset threshold; the orthogonal installation method makes the sensitive axes of the two IMUs form a spatial complementary measurement reference; the physical distance between the temperature sensor and the IMU is controlled within 5 cm to ensure the consistency of temperature acquisition data with the temperature field distribution inside the IMU.

[0085] Specifically, the natural frequency parameters of the damping rubber pads are determined through vibration transfer function calculations, and their upper limit ensures effective isolation of high-frequency vibrations generated by ship engines and wave impacts. The vibration attenuation rate is achieved through a combination of the rubber material's damping coefficient and structural thickness; when the attenuation rate meets the standard, vibration energy can be reduced by more than 80% in the transmission path. The temperature sensor uses a platinum resistance element with a linearity error of less than 0.1%, directly mounted on the IMU housing surface, with thermal resistance reduced using thermally conductive silicone grease. During the static calibration phase, the angular motion excitation frequency applied by the three-axis turntable is limited to below the natural frequency of the damping system to avoid measurement distortion caused by resonance. A timestamp synchronization relationship is established between temperature data acquisition and IMU output data, providing time-aligned sample data for subsequent establishment of a zero-bias-temperature relationship model.

[0086] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0087] The hardware installation and calibration process includes the following sub-steps:

[0088] The dual IMUs are fixed to the transducer base at orthogonal positions using shock-absorbing rubber pads. These pads are made of natural rubber with a natural frequency of 8Hz and a vibration attenuation rate of 85%. During installation, two orthogonal mounting surfaces are pre-machined on the transducer base, each with four M6 threaded holes. The bottom of the shock-absorbing rubber pads has through holes corresponding to the threaded holes on the mounting surfaces, and the pads are secured to the mounting surfaces using M6 bolts. The bottom of the IMU housing also has four pre-drilled M6 threaded holes, and the IMUs are secured to the top of the shock-absorbing rubber pads using bolts.

[0089] Furthermore, a temperature sensor is installed near each IMU. The temperature sensor is a PT100 platinum resistance thermometer with a temperature range of -50℃ to 150℃ and an accuracy of ±0.1℃. The temperature sensor is attached to the surface of the IMU housing using thermally conductive silicone to ensure it is at the same temperature as the IMU's internal temperature. The signal line of the temperature sensor is connected to the analog input port of the data acquisition unit, and the sampling frequency is set to 1Hz.

[0090] Therefore, temperature data is used to support subsequent polynomial temperature compensation for IMU bias. The compensation algorithm employs third-order polynomial fitting, with coefficients obtained through offline calibration. During actual measurements, bias correction values ​​are calculated based on real-time temperature data and applied to the preprocessing of the raw IMU data.

[0091] Through the above technical solutions, this application effectively reduces the impact of hull vibration on IMU measurements, improving the stability and accuracy of attitude data. Simultaneously, real-time temperature compensation reduces the drift error caused by temperature variations in the IMU zero bias, further enhancing attitude measurement accuracy during long-term operation. Furthermore, the orthogonally mounted dual IMU configuration provides redundant measurements, enhancing the system's reliability and robustness.

[0092] In some of the solutions described above in this application, there is a problem of insufficient sensor timestamp alignment accuracy during multi-source data synchronous acquisition, leading to time misalignment during data fusion and thus affecting the accuracy of subsequent filtering and attitude fusion. Furthermore, the inconsistency between the timestamps of environmental data and dual IMU data further exacerbates the data fusion error.

[0093] This application further proposes multi-source data synchronous acquisition including: aligning sensor timestamps using a hardware synchronization mechanism based on a cross-correlation algorithm, with a timestamp accuracy of less than or equal to 100 ns; and converting wave height and wind speed environmental data into sampling frequency timestamps of dual IMUs through a linear interpolation method to ensure data consistency.

[0094] The hardware synchronization mechanism calculates the time delay difference of sensor signals using a cross-correlation algorithm to generate a precise clock correction signal, ensuring that the time reference error of each sensor is controlled within 100ns. Environmental data conversion employs a linear interpolation method, interpolating and reconstructing low-frequency wave height and wind speed data according to the 200Hz sampling frequency of the dual IMUs, generating an environmental parameter sequence aligned with the IMU data time axis.

[0095] Specifically, the cross-correlation algorithm determines the time offset and generates a synchronization trigger pulse by calculating the peak value of the similarity function between different sensor signals, achieving microsecond-level timestamp alignment. This mechanism distributes the GNSS time reference to all sensor nodes, eliminating internal clock drift. For environmental data, linear interpolation is performed on the original low-frequency data using the sampling time of the dual IMUs as a reference, generating environmental parameter values ​​that correspond one-to-one with the IMU data points. For example, when the wave height data sampling rate is 10Hz, 19 intermediate values ​​are inserted between every two adjacent data points, increasing the time resolution of the environmental data sequence to 200Hz. Through the above method, the time consistency error of the sensor data is reduced to within 0.1 milliseconds, ensuring strict time synchronization of the input data for subsequent fusion algorithms.

[0096] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0097] Multi-source data synchronization acquisition includes: aligning sensor timestamps using a hardware synchronization mechanism based on a cross-correlation algorithm, with a timestamp accuracy set to 50ns; and converting wave height and wind speed environmental data into sampling frequency timestamps for dual IMUs using linear interpolation to ensure data consistency. Specifically, a high-precision time synchronization board is used, integrating an atomic clock chip to provide a 10MHz reference clock signal. Each sensor is connected to the synchronization board via Ethernet or serial port, receiving PPS pulse signals for time synchronization.

[0098] Wave height data is collected once per second using wave radar. Wind speed data is collected at a frequency of 1 Hz using an ultrasonic anemometer. This environmental data is then linearly interpolated to correspond to the 200 Hz sampling frequency of the dual IMUs. For example, if the wave heights at times t1 and t2 are h1 and h2 respectively, the interpolated wave height h for any time t between t1 and t2 is calculated as follows:

[0099] h = h1 + (h2 - h1) * (t - t1) / (t2 - t1)

[0100] The interpolation method for wind speed data is similar. In this way, environmental data is temporally aligned with high-frequency IMU data, providing input for subsequent adaptive filtering.

[0101] Through the above technical solution, this application achieves high-precision time synchronization of multi-source heterogeneous sensor data, effectively eliminating measurement errors caused by time inconsistencies between sensors. Simultaneously, the interpolation processing of environmental data ensures data continuity and consistency, providing reliable input for subsequent attitude fusion and error compensation. This synchronization mechanism significantly improves the measurement accuracy and reliability of the entire system, and is particularly significant in applications within dynamic marine environments.

[0102] In some of the above-mentioned schemes in this application, during the dual IMU fusion attitude measurement process, the state vector design does not fully cover the transducer base attitude angle and IMU zero bias variables, and the observation matrix does not make full use of the installation rotation matrix for coordinate system transformation. As a result, the attitude fusion accuracy is limited by sensor zero bias and environmental interference, and cannot dynamically adapt to noise changes under different sea conditions.

[0103] This application further proposes the design of an adaptive Kalman filter, where the state vector is defined as x=[θ,φ,ψ,b] gA ,b gB ,b aA ,b aB ], where θ, φ, and ψ are the roll, pitch, and yaw angles of the transducer base, respectively, and b gA ,b gB The gyroscope zero bias of IMU-A and IMU-B are respectively, b aA ,b aB To achieve zero bias in the accelerometer; the observation equation is based on the installation rotation matrix, which transforms the dual IMU angular velocities to the transducer base coordinate system, thereby enabling adaptive adjustment of the observation noise covariance.

[0104] The state vector uses the three-axis attitude angles of the transducer base and the zero bias of the gyroscopes and accelerometers of the dual IMUs as estimation parameters. By simultaneously modeling attitude and sensor errors, it improves the dynamic tracking capability of the filter. The installation rotation matrix is ​​used to transform the raw angular velocity data of the IMUs to the transducer base coordinate system, eliminating measurement bias caused by differences in installation position. The adaptive adjustment of the observation noise covariance is dynamically executed based on the sea state index, where the sea state index S = 0.5H. w +0.1V w H w V represents the wave height. w For wind speed, sea state is classified as calm, moderate, and severe; the baseline value of the gyroscope noise covariance of the IMU-A is 1×10⁻⁶. -5 rad 2 / s 2 The adjustment ratios are 1, 3, or 5, and the adjustment ratio for IMU-B is 0.6 times that of IMU-A.

[0105] Specifically, the state vector, by simultaneously estimating attitude angles and zero-bias variables, can compensate for IMU drift errors in real time, avoiding the accumulation of zero bias due to temperature changes or long-term operation. The application of the installation rotation matrix ensures that the dual IMU data are fused in a unified coordinate system, reducing attitude calculation errors caused by installation orthogonality deviations. The observation noise covariance is dynamically adjusted according to sea conditions; in severe sea conditions, the noise covariance scaling factor is increased to reduce the impact of high-frequency vibrations on the filtering results, while in calm sea conditions, the scaling factor is decreased to improve the attitude update rate. Through the above design, the attitude fusion accuracy is improved by more than 30% under dynamic sea conditions, and the zero-bias estimation error is controlled within 0.01 degrees / hour.

[0106] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0107] The design of an adaptive Kalman filter includes the following steps:

[0108] First, define the state vector x = [θ, φ, ψ, b] gA ,b gB ,b aA ,b aB ]. Where θ, φ, and ψ represent the roll, pitch, and yaw angles of the transducer base, respectively, and b gA ,b gB These represent the gyroscope zero bias of IMU-A and IMU-B, respectively. aA ,b aB These represent the zero bias of the accelerometers for IMU-A and IMU-B, respectively.

[0109] Next, the observation equations are constructed. These equations, based on the installation rotation matrix, transform the angular velocity data of the dual IMUs to the transducer base coordinate system. Specifically, let the installation rotation matrices of IMU-A and IMU-B be R... A and R B Then the observation equation can be expressed as:

[0110] z A =R A (ω A -b gA )

[0111] z B =R B (ω B -b gB )

[0112] Where ω A and ω B The angular velocities measured by IMU-A and IMU-B are respectively.

[0113] Furthermore, adaptive adjustment of the observation noise covariance is achieved. The observation noise covariance matrix R is dynamically adjusted based on the current sea state. For example, R can be set as a diagonal matrix, with its diagonal elements changing according to sea state. In calm sea states, the diagonal elements of R can take smaller values; in severe sea states, larger values ​​are taken to reduce the confidence level in the observation data.

[0114] Finally, the Kalman filter is executed for prediction and update. In the prediction step, the system model is used to predict the state estimate and error covariance for the next time step; in the update step, the prediction results are corrected by incorporating the observed data to obtain the optimal estimate.

[0115] Through the above technical solutions, this application achieves high-precision attitude measurement and zero-bias estimation. The adoption of a dual IMU fusion strategy improves the reliability and accuracy of attitude measurement. The adaptive Kalman filter can dynamically adjust its filtering parameters according to sea state, adapting to measurement requirements under different environmental conditions. Furthermore, incorporating the IMU zero bias into the state vector enables online estimation and compensation of the zero bias, effectively suppressing the accumulation of IMU drift errors. These combined techniques significantly improve the accuracy and stability of underwater depth sounding and positioning.

[0116] In some of the above-mentioned schemes in this application, a method based on adaptive Kalman filter to fuse dual IMU data to output high-precision attitude data is proposed. However, in this process, the fixed observation noise covariance parameter cannot adapt to dynamic sea state changes, which leads to a significant decrease in attitude fusion accuracy due to environmental interference, thereby affecting the accuracy of subsequent acoustic ray bending compensation and multi-source data fusion.

[0117] This application further proposes an adaptive adjustment of the observation noise covariance based on dynamic execution using a sea state grade index, specifically: the sea state grade index S is determined by wave height H. w Wind speed V w The linear combination calculation is expressed as S = 0.5H. w +0.1V w The sea state is divided into three levels: calm, moderate, and severe. Adjustment factors are set according to the sea state level, and the baseline value of the IMU-A's gyroscope noise covariance is 1×10⁻⁶. -5 rad 2 / s 2 The adjustment ratios are 1, 3, and 5, respectively, and the adjustment ratio of IMU-B is 0.6 times that of IMU-A.

[0118] Among them, the sea state level index is calculated by weighting wave height and wind speed, with weighting coefficients of 0.5 and 0.1 reflecting the dominant influence of waves on carrier vibration and the secondary role of wind speed, respectively; the sea state classification thresholds of 0.5 and 2 are determined by statistical analysis of measured data, corresponding to the intensity difference of vibration interference experienced by the dual IMUs under different levels; the difference in the proportional coefficient between IMU-A and IMU-B stems from the different vibration transmission paths caused by their orthogonal installation positions. Because IMU-B is installed closer to the damping rubber pad, its vibration attenuation rate is higher, so the adjustment coefficient is reduced by 40%.

[0119] Specifically, under dynamic sea conditions, sea state indexes are calculated by collecting wave height and wind speed data in real time. For example, when H... w = 1.2 meters, V w When the speed is 8 m / s, S = 0.5 × 1.2 + 0.1 × 8 = 1.4, which is considered a moderate sea state; at this time, the gyro noise covariance of the IMU-A is adjusted to 3 × 10. -5 rad 2 / s 2 The IMU-B was adjusted to 1.8×10 -5 rad 2 / s 2 The sensitivity of the filter to abnormal vibration data is reduced by increasing the noise covariance parameter; under calm sea conditions, a basic covariance value of 1×10 is used. - 5 rad 2 / s 2 To improve the resolution of attitude angle estimation, the dynamic adjustment mechanism enables the filter to preferentially suppress high-frequency noise in severe vibration environments and improve measurement accuracy in stable environments, thereby ensuring the robustness of attitude fusion results across the entire sea state range.

[0120] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0121] The adaptive adjustment of the observed noise covariance is dynamically implemented based on the sea state grade index. The sea state grade index S is calculated using the formula S = 0.5H. w +0.1V w Calculate H w V represents the wave height. w This represents wind speed. Sea conditions are classified into three categories: calm (S<0.5), moderate (0.5≤S<2), and severe (S≥2).

[0122] The adjustment factor is set according to the sea state level. The basic value of the gyro noise covariance of the IMU-A is 1×10⁻⁶. -5 rad 2 / s 2In calm sea conditions, the adjustment factor is 1; in moderate sea conditions, the adjustment factor is 3; and in severe sea conditions, the adjustment factor is 5. The adjustment factor for IMU-B is 0.6 times that of IMU-A.

[0123] For example, when the wave height is 1 meter and the wind speed is 5 meters per second, the calculated value is S = 0.5 × 1 + 0.1 × 5 = 1, which corresponds to a moderate sea state. In this case, the IMU-A's gyro noise covariance is adjusted to 3 × 10⁻⁶. -5 rad 2 / s 2 The gyroscope noise covariance of the IMU-B was adjusted to 1.8 × 10⁻⁶. -5 rad 2 / s 2 .

[0124] Through the above technical solution, this application achieves dynamic adaptive adjustment of the observation noise covariance. Therefore, the Kalman filter can automatically optimize its filtering parameters according to real-time sea state changes, improving the accuracy and robustness of attitude estimation. Furthermore, by differentiating the adjustment ratios of IMU-A and IMU-B, the complementary characteristics of the two IMUs are fully utilized, enhancing the system's anti-interference capability while ensuring measurement accuracy.

[0125] In some of the solutions described above in this application, the anomaly detection process may be misjudged or missed due to interference from the dynamic environment, making it impossible to accurately identify the instantaneous faults or drifts of the IMU, thereby affecting the reliability of the fused attitude data.

[0126] This application further proposes to calculate the innovation residual vector of the adaptive Kalman filter in real time. If the residual norm exceeds three times the trace square root of the residual covariance matrix, or the difference between the attitude angle of a single IMU and the fused attitude angle is greater than 0.1 degrees, it is determined to be an IMU anomaly.

[0127] The innovation residual vector is generated by the difference between the predicted and actual observed values ​​of the adaptive Kalman filter. The residual norm is set based on the trace square root of the covariance matrix to set a dynamic threshold, and the attitude angle difference threshold is determined through experimental data verification. The trace square root of the residual covariance matrix reflects the statistical characteristics of the current filtering state. The triple coefficient covers 99.7% of the normal distribution confidence interval, and the 0.1-degree threshold balances the sensitivity and false alarm rate in dynamic marine environments.

[0128] Specifically, when fusing dual IMU data, the adaptive Kalman filter calculates the innovation residual vector and updates the covariance matrix in real time. The residual norm is calculated using the Euclidean norm and compared with the trace square root of the covariance matrix; if the difference exceeds three times, it is considered an anomaly. Simultaneously, the attitude angles output by the dual IMUs are compared with the fused result in real time, and an anomaly detection is triggered if the difference in any axial direction exceeds 0.1 degrees. For example, when the roll angle of IMU-A deviates from the fused roll angle by 0.12 degrees, the system immediately marks that IMU as an anomaly. Through this dual criterion, both statistical characteristics are used to suppress random noise interference, and attitude angle differences are used to capture sudden hardware failures, ensuring the robustness of anomaly detection.

[0129] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0130] Anomaly detection involves real-time computation of the innovation residual vector of the adaptive Kalman filter. The innovation residual vector is obtained by comparing the differences between observed and predicted values. The residual norm is obtained by calculating the Euclidean norm of the innovation residual vector. The residual covariance matrix is ​​jointly determined by the filter's state estimation error covariance matrix and the observation noise covariance matrix.

[0131] Specifically, first, the innovation residual vector r = z - Hx is calculated, where z is the observed value, H is the observation matrix, and x is the state estimate. Then, the residual norm ||r|| = sqrt(r T *r). Next, calculate the residual covariance matrix S = HPH. T +R, where P is the state estimation error covariance matrix and R is the observation noise covariance matrix. Finally, the decision threshold is calculated as threshold = 3*sqrt(trace(S)).

[0132] If the residual norm exceeds the judgment threshold, or the difference between the attitude angle of a single IMU and the fused attitude angle is greater than 0.1 degrees, it is judged as an IMU anomaly. The attitude angle difference is obtained by calculating the Euclidean distance between the roll, pitch, and yaw angles output by a single IMU and the fused attitude angle.

[0133] For example, assuming the residual norm calculated at a certain moment is 0.15 and the trace of the residual covariance matrix is ​​0.01, the judgment threshold is 3*sqrt(0.01) = 0.3. Since 0.15 < 0.3, the first anomaly judgment condition is not met. Further calculation shows that the difference between the attitude angle of IMU-A and the fused attitude angle is 0.08 degrees, and the difference between IMU-B and IMU-B is 0.12 degrees. Since the difference between IMU-B and IMU-B exceeds 0.1 degrees, IMU-B is judged to be an anomaly.

[0134] Through the above technical solutions, this application can promptly detect IMU anomalies, improving system reliability. Real-time monitoring of the output residuals and differences of the dual IMUs allows for rapid identification of sensor faults or performance degradation. Setting reasonable judgment thresholds can capture obvious anomalies while avoiding false alarms caused by oversensitivity. Combining residual statistical characteristics and attitude angle difference values ​​for dual judgment improves the accuracy and robustness of anomaly detection. Through the anomaly detection mechanism, the system can promptly switch operating modes or perform self-calibration, ensuring the continuity and accuracy of depth sounding and positioning.

[0135] In some of the solutions described above in this application, attitude data may fail due to sensor malfunctions during dual IMU fusion attitude measurement, affecting the reliability of depth sounding point coordinate calculation. Especially in dynamic marine environments, IMU zero-bias drift or sudden failures can cause the accumulation of fusion attitude errors. It is necessary to effectively detect anomalies and achieve rapid self-correction to maintain continuous system operation.

[0136] This application further proposes a self-calibration strategy including: when a single IMU anomaly is detected, switching to the single-source mode output attitude of the healthy IMU; under stable sea conditions, pausing the measurement and placing both IMUs stationary for 10 minutes to calculate and update the mean difference of the zero bias parameter; if both IMUs fail, switching to the RTK-GNSS heading angle as a backup attitude source.

[0137] In this mode, switching to the single-source mode of the healthy IMU disables the data input of the abnormal IMU and retains only the attitude calculation of the healthy IMU, thus avoiding erroneous data from contaminating the fusion results. Stable sea state is defined as a sea state index S less than 0.5. At this time, the measurement is paused and both IMUs are kept stationary. The zero bias parameter is recalibrated using the difference in the mean output of the IMUs in the stationary state to eliminate zero bias drift caused by temperature or mechanical stress. When both IMUs are faulty, the heading angle output by RTK-GNSS is used as the attitude source, and the roll and pitch data from INS are combined to maintain the basic attitude calculation capability.

[0138] Specifically, an anomaly detection mechanism is triggered when the innovation residual norm of the adaptive Kalman filter exceeds a threshold or when the difference between the attitude angle of a single IMU and the fusion result is greater than 0.1 degrees. After switching to single-source mode, the system relies solely on healthy IMUs for attitude calculation, reducing the impact of abnormal data on the fusion process. Under stable sea conditions, the IMUs are stationary for 10 minutes to ensure they are free from dynamic interference, and the mean difference of the zero-bias parameters is collected and updated in the filter to compensate for long-term drift errors. If both IMUs fail simultaneously, the system switches to RTK-GNSS heading angle as a backup, combining roll and pitch data from INS to maintain the basic function of depth sounding point coordinate calculation through geometric projection. Through multi-level fault tolerance mechanisms, the continuity and reliability of attitude data are maintained even in the event of sensor anomalies or sudden environmental changes, avoiding interruptions or significant decreases in accuracy during depth sounding.

[0139] As a preferred embodiment, the solution of this application is implemented as follows: During the navigation of the measurement vessel, if the norm of the innovation residual vector of the adaptive Kalman filter exceeds three times the trace square root of the residual covariance matrix for five consecutive times, or if the difference between the roll angle and the fused attitude angle of a single IMU exceeds 0.1 degrees for 10 seconds, an anomaly detection mechanism is triggered. At this time, the system automatically switches to the single-source mode of the healthy IMU. For example, when IMU-A experiences a zero-bias mutation, only the raw data of IMU-B is used for attitude calculation. If the sea state index S is less than 0.5 and the vessel is in a stable state, the system suspends the measurement task and controls both IMUs to remain stationary, updating the gyro zero-bias parameters of IMU-A by calculating the zero-bias mean over 10 minutes. In extreme cases, when the accelerometer outputs of both IMUs simultaneously show an abnormal value exceeding 5g and do not recover, the system automatically switches to the heading angle output by RTK-GNSS as the attitude reference, and simultaneously calls the transducer installation deviation matrix pre-stored in the storage module for heading compensation.

[0140] Through the above technical solutions, this application effectively solves the problem of attitude data interruption or deviation accumulation caused by sudden IMU failure in traditional bathymetry positioning systems. Through a multi-level fault-tolerant mechanism, high-precision attitude output can still be maintained even when a single IMU malfunctions, avoiding bathymetry point coordinate errors caused by sensor failure; online calibration in a static state eliminates the impact of zero-bias drift on subsequent measurements; and the backup application of GNSS heading angle ensures basic positioning functions even when both IMUs fail simultaneously, significantly improving the system's operational reliability in complex marine environments.

[0141] This application further proposes that step S5, the horizontal position correction after acoustic ray bending compensation, further includes: calculating the nonlinear effects of horizontal offset and depth based on sound velocity profile layering data through integration, using the following formula: Where c i For the layer sound velocity, Δz i The layer thickness is determined and integrated into the coordinate transformation results to compensate for horizontal deviations.

[0142] As a preferred embodiment, the specific implementation of this application is as follows: In a water depth measurement operation in a certain sea area of ​​the South China Sea, a sound velocity profiler acquires real-time sound velocity distribution data at water depths of 0-200 meters, determining that there are a total of 8 sound velocity layers from the surface to the bottom. The data processing unit substitutes the sound velocity data of each layer into the horizontal offset integral formula to calculate the sound velocity propagation path deviation of each layer. Specifically, for the third layer at a water depth of 15-25 meters, the measured sound velocity value is 1520 m / s, and the thickness of this layer is Δz. i The height is 10 meters, and the sound velocity difference between this layer and the upper layer, v(z), is 3 m / s. The horizontal offset Δs of this layer is calculated using the formula. iThe value is 0.28 meters. The horizontal offsets of all sound velocity layers are accumulated and then vector-superimposed with the original slant range data of the multibeam echo sounder. Finally, dynamic compensation for the horizontal position deviation is achieved during the transformation from the transducer base coordinate system to the geographic coordinate system.

[0143] Through the above technical solution, this application effectively solves the problem of cumulative error caused by the reliance on static empirical values ​​in traditional sound ray bending compensation methods. By calculating the horizontal offset of the sound speed propagation path through layered integration, it realizes real-time sound ray bending compensation for dynamically changing water bodies, which significantly improves the positioning accuracy of the horizontal coordinates of the sounding point. Especially in marine environments where the sound speed gradient changes significantly, it can avoid horizontal position offset errors caused by sound ray refraction.

[0144] In some of the solutions described above in this application, the calculation of the absolute coordinates of the sounding point depends on the geometric transformation of the transducer position, attitude data, and beam parameters. However, traditional methods do not fully consider the nonlinear coupling effect of attitude angle and beam offset in three-dimensional space, which leads to the accumulation of horizontal and vertical position errors and affects the sounding accuracy.

[0145] This application further proposes a specific formula for calculating the absolute coordinates of sounding points, the expression of which is:

[0146] x d =x t +l·cosψ·cosθ-d·sinφ·cosψ+d·cosφ·sinψ

[0147] y d =y t +l·sinψ·cosθ-d·sinφ·sinψ-d·cosφ·cosψ

[0148] z d =z t -d·cosφ·cosθ-l·sinθ

[0149] Where x t ,y t ,z t denoted as transducer position, l and d as beam offset and depth, and ψ, φ, and θ as heading, pitch, and roll attitude angles, respectively.

[0150] The formula decomposes the three-dimensional influence of attitude angles on beam offset, using the heading angle ψ as the principal rotation axis, and the roll angle θ and pitch angle φ acting on the horizontal and vertical components, respectively. For example, the cosine term cosψ and the sine term sinψ of the heading angle are used to project the beam offset l onto the east and north coordinate axes, while the cosine term cosθ of the roll angle θ is used to correct the horizontal projection scale, and the sine term sinφ and the cosine term cosφ of the pitch angle φ are used to compensate for beam tilt errors in the depth direction caused by hull undulations. The beam depth d achieves vertical attenuation through the cosφ·cosθ term, avoiding depth overestimation caused by attitude angles.

[0151] Specifically, during the transformation from the transducer base coordinate system to the geographic coordinate system, the fused attitude data is constructed using a rotation matrix. However, directly applying matrix multiplication increases computational complexity. This scheme decomposes the rotation matrix into a linear combination of three independent Euler angles: heading, roll, and pitch, simplifying the calculation steps using trigonometric functions. For example, when the transducer tilts due to the ship's roll, the horizontal beam offset *l* needs to be multiplied by *cosθ* to eliminate the projection shortening effect caused by the roll; when the ship's pitch causes the transducer to pitch, the depth *d* needs to be multiplied by *cosφ* to correct the vertical component of the beam path. Furthermore, the horizontal distance after sound velocity profile compensation is calculated through integration and combined with the beam offset *l* in the geometric projection formula, ultimately outputting a planar accuracy of less than ±2cm and a vertical accuracy of less than ±3cm for the sounding point.

[0152] As a preferred embodiment, the solution of this application is specifically implemented as follows: When the survey vessel performs multibeam echo sounding operations, the transducer base acquires the three-dimensional position coordinates (x, y, y) in the geographic coordinate system in real time. t =121.5432°E, y t =31.2356°N, z t =2.15m). The multibeam echo sounder measured the lateral offset of a certain beam as l = 15.6m and the depth as d = 28.4m. Simultaneously, it fused attitude data to output the heading angle ψ = 45.3°, pitch angle φ = 1.2°, and roll angle θ = 0.8°. Based on the geometric projection formula, the beam offset and attitude parameters were substituted into the coordinate transformation equation to calculate the absolute coordinates x of the sounding point. d =121.5432°E+15.6×cos45.3°×cos0.8°-28.4×sin1.2°×cos45.3°+28.4×cos1.2°×sin453.y d =31.2356°N+15.6×sin45.3°×cos0.8°-28.4×sin1.2°×sin45.3°-28.4×cos1.2°×cos45.3°z d=2.15-28.4×cos1.2°×cos0.8°-15.6×sin0.8°.

[0153] The calculation results automatically compensate for the geometric projection deviation caused by the transducer attitude, achieving accurate calculation of the geographic coordinates of the sounding point.

[0154] Through the above technical solution, this application effectively solves the problem of coordinate calculation deviation caused by insufficient compensation for transducer attitude angles and beam offset in traditional depth sounding and positioning. By establishing a three-dimensional geometric projection model that includes heading, pitch, and roll angles, the beam spatial offset is accurately mapped to the geographic coordinate system, eliminating horizontal position drift and depth measurement errors caused by changes in carrier attitude, and ensuring that the coordinate accuracy of underwater topographic survey results meets the requirements of high-precision marine engineering applications.

[0155] The above embodiments merely illustrate several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

[0156] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A high-precision offset positioning depth sounding auxiliary system, characterized in that: include: The positioning and sensing module integrates an RTK-GNSS receiver, an optical fiber inertial navigation system (INS), and a short baseline underwater acoustic positioning receiver (USBL) to provide the vehicle's absolute position, velocity information, and relative position correction. The attitude and offset measurement module includes two high-precision inertial measurement units (IMUs), namely IMU-A and IMU-B, which are orthogonally mounted on the transducer base. They are pre-stored with the offset and attitude deviation of the transducer relative to the carrier coordinate system and are used to acquire high-frequency attitude data. The sound velocity profile measurement module includes a sound velocity profiler, which is used to acquire sound velocity profile data of water bodies in real time. The data processing unit is an embedded industrial computer, equipped with algorithms for multi-source data synchronization, adaptive Kalman filtering, attitude fusion, and error compensation. The communication and storage module enables data synchronization between sensors via Ethernet or RS485 bus and provides data storage functionality. The data from the positioning and sensing module, attitude and offset measurement module, and sound velocity profile measurement module are all transmitted to the data processing unit for fusion processing through the communication and storage module. The data flow serial relationship between the modules is defined as follows: the sound velocity profile measurement module collects sound velocity data and sends it to the data processing unit; the attitude and offset measurement module collects IMU data and sends it to the data processing unit; the positioning and sensing module collects position data and sends it to the data processing unit; and the processing results are output to the communication and storage module for storage.

2. A method for using a high-precision offset positioning depth sounding auxiliary system, characterized in that: Using the auxiliary system of claim 1, the method includes the following steps: Step S1: Hardware installation and calibration. Deploy the RTK-GNSS antenna, INS, USBL beacon array, sound velocity profiler and dual IMUs on the measurement ship. Perform static calibration on the dual IMUs using a three-axis turntable. Measure and pre-store the installation rotation matrix and attitude deviation of the transducer base to compensate for installation errors. Step S2: Multi-source data synchronous acquisition. The timestamps of RTK-GNSS, INS, USBL, dual IMU and sound velocity profiler are aligned through Precise Time Protocol (PTP) so that the output of all sensors is unified to the GNSS time reference, and data is acquired at a preset frequency, wherein the sampling frequency of dual IMU is at least 200Hz. Step S3: Data preprocessing, the raw angular velocity and acceleration data of the dual IMUs are filtered by sliding window midpoint filtering to remove high-frequency noise, and polynomial temperature compensation is performed on the IMU zero bias based on temperature sensor data; Step S4: Dual IMU attitude measurement. Based on the preprocessed data in step S3, the dual IMU data is fused using an adaptive Kalman filter. The state vector is designed as the transducer base attitude angle and the dual IMU zero bias variables. The observation matrix is ​​based on the installation rotation matrix to transform the IMU data to the transducer base coordinate system, and the fused high-precision attitude data is output. Step S5: Transformation between the carrier coordinate system and the geographic coordinate system. Based on the fused attitude output in step S4, a rotation matrix is ​​constructed to transform the transducer position in the carrier coordinate system to the geographic coordinate system. Sound velocity profile data is applied for sound ray bending compensation, and the horizontal distance is calculated through layered integration to correct the slant distance. Step S6: Multi-source data fusion and error suppression. Based on the compensation output of step S5 and RTK-GNSS, INS, and USBL data, an extended Kalman filter is used for fusion processing. The state vector includes the carrier position, velocity, and sound speed error to suppress drift noise. Step S7: Calculate the absolute coordinates of the sounding points. Combining the beam offset and depth data of the multibeam echo sounder, the positioning results output in step S6, and the fused attitude data in step S4, calculate the absolute coordinates of each sounding point using the geometric projection formula. Step S8: Anomaly detection and self-calibration. Real-time monitoring of the dual IMU output residuals and differences from step S4. If an anomaly is detected, the operating mode is switched, and the faulty IMU is calibrated online under stable sea conditions.

3. The method of using the high-precision offset positioning depth sounding auxiliary system according to claim 2, characterized in that: In step S1, the hardware installation and calibration includes the following sub-steps: The dual IMUs are fixed to the transducer base at an orthogonal position using shock-absorbing rubber pads. The natural frequency of the shock-absorbing rubber pads is less than 10Hz, and the vibration attenuation rate is greater than or equal to 80%. Temperature sensors are installed near each IMU to collect temperature data to support subsequent zero-bias polynomial temperature compensation.

4. The method of using the high-precision offset positioning depth sounding auxiliary system according to claim 2, characterized in that: In step S2, the multi-source data synchronous acquisition includes: The sensor timestamps are aligned using a hardware synchronization mechanism based on a cross-correlation algorithm, with a timestamp accuracy of less than or equal to 100 ns. Wave height and wind speed environmental data are converted into sampling frequency timestamps for dual IMUs using a linear interpolation method to ensure data consistency.

5. The method of using a high-precision offset positioning depth sounding auxiliary system according to claim 2, characterized in that: In step S4, the design of the adaptive Kalman filter includes: The state vector is defined as x = [θ, φ, ψ, b] gA ,b gB ,b aA ,b aB ], where θ, φ, and ψ are the roll, pitch, and yaw angles of the transducer base, respectively, and b gA ,b gB The gyroscope zero bias of IMU-A and IMU-B are respectively, b aA ,b aB These are the accelerometer zero bias values; The observation equation is based on the installation rotation matrix, which transforms the dual IMU angular velocities to the transducer base coordinate system, thereby achieving adaptive adjustment of the observation noise covariance.

6. The method of using a high-precision offset positioning depth sounding auxiliary system according to claim 5, characterized in that: In step S4, the adaptive adjustment of the observation noise covariance is dynamically performed based on the sea state level index, specifically as follows: Sea state rating index S = 0.5H w +0.1V w H w V represents the wave height. w The wind speed is used as the criterion, and the sea state is classified into calm (S<0.5), moderate (0.5≤S<2), and severe (S≥2); The adjustment factor is set according to the sea state level, and the basic value of the gyro noise covariance of the IMU-A is 1×10. -5 rad 2 / s 2 The adjustment factor is 1 (calm), 3 (moderate), or 5 (severe). The adjustment factor for IMU-B is 0.6 times that of IMU-A.

7. The method of using a high-precision offset positioning depth sounding auxiliary system according to claim 2, characterized in that: In step S8, the anomaly detection includes real-time calculation of the innovation residual vector of the adaptive Kalman filter. If the residual norm exceeds three times the trace square root of the residual covariance matrix, or the difference between the attitude angle of a single IMU and the fused attitude angle is greater than 0.1 degrees, then it is determined to be an IMU anomaly.

8. The method of using a high-precision offset positioning depth sounding auxiliary system according to claim 7, characterized in that: In step S8, the self-calibration strategy includes: When a single IMU malfunction is detected, switch to the single-source mode of the healthy IMU to output the attitude. Under steady sea state (S<0.5), the measurement was paused and the dual IMUs were left stationary for 10 minutes to calculate and update the mean difference of the zero bias parameter; If both IMUs fail, switch to RTK-GNSS heading angle as the backup attitude source.

9. The method of using a high-precision offset positioning depth sounding auxiliary system according to claim 2, characterized in that: In step S5, the horizontal position correction after acoustic ray bending compensation further includes: calculating the nonlinear effects of horizontal offset and depth based on layered acoustic velocity profile data through integration, using the following formula: Where c i For the layer sound velocity, Δz i The layer thickness is determined and integrated into the coordinate transformation results to compensate for horizontal deviations.

10. The method of using a high-precision offset positioning depth sounding auxiliary system according to claim 2, characterized in that: In step S6, the extended Kalman filter output positioning result of the multi-source data fusion has a planar accuracy of less than ±2cm and a vertical accuracy of less than ±3cm. The method ends with the output of the sounding point coordinates in step S7, and the coordinate formula is: x d =x t +l·cosψ·cosθ-d·sinφ·cosψ+d·cosφ·sinψ y d =y t +l·sinψ·cosθ-d·sinφ·sinψ-d·cosφ·cosψ z d =z t -d·cosφ·cosθ-l·sinθ Where x t ,y t ,z t denoted as transducer position, l and d as beam offset and depth, and ψ, φ, and θ as heading, pitch, and roll attitude angles, respectively.

Images