Navigation and positioning method for underwater robot based on single-beacon ranging and dynamic model

Abstract

The application relates to the technical field of underwater robots, in particular to an underwater robot navigation positioning method based on single-beacon ranging and a dynamics model, which comprises the following steps: determining a dynamic matching window according to a physical installation interval and a water speed, and matching an acoustic feature peak moment of an acoustic difference index and a wake feature peak moment of a current fluctuation index in the dynamic matching window to determine a pile foundation crossing moment; determining a corrected position according to an acoustic ranging value corresponding to the pile foundation crossing moment, installation coordinates of a single acoustic beacon and a straight-line constraint of an inspection channel; determining an environmental flow speed observation value according to a deviation between the corrected position and a calculated position corresponding to the pile foundation crossing moment; updating a flow speed parameter of the dynamics model according to the environmental flow speed observation value; and correcting a current calculated position according to the corrected position. The application provides a method for improving the positioning accuracy of underwater robot inspection operation.

Description

Technical Field

[0001] This application relates to the field of underwater robot technology, specifically to an underwater robot navigation and positioning method based on single beacon ranging and dynamic model. Background Technology

[0002] With the rapid development of the marine economy, the construction scale of underwater infrastructure such as offshore photovoltaic power stations and cross-sea bridges is expanding daily. These facilities typically use dense arrays of steel pile foundations as support structures, and after long-term service, they require regular underwater inspections to assess their structural health. Small remotely operated vehicles (ROVs) have become the primary tool for performing such close-range inspection tasks due to their maneuverability, flexibility, and cost-effectiveness.

[0003] In complex underwater environments, accurate positioning and navigation are fundamental for ROVs to complete automated inspection tasks. Current technologies typically rely on Doppler velocity logs (DVL) to provide velocity information, combined with ultra-short baseline (USBL) or long baseline (LBL) acoustic positioning systems for position correction, forming a complete navigation and positioning scheme. However, when operating in areas with dense pile foundations, the metal piles have strong sound wave reflection characteristics. Multiple reflections of a single beacon ranging signal between piles create a multipath effect, leading to non-line-of-sight jumps in ranging values. This can easily result in accumulated position drift and positioning calculation errors during long-term operations, ultimately reducing positioning accuracy. Summary of the Invention

[0004] To address the technical problems of position drift accumulation and positioning calculation errors that easily occur during long-term operation, leading to a decrease in positioning accuracy, this application aims to provide an underwater robot navigation and positioning method based on single beacon ranging and a dynamic model. The specific technical solution adopted is as follows: Firstly, a navigation and positioning method for underwater robots based on single-beacon ranging and a dynamic model is provided, comprising: acquiring acoustic ranging information, thruster current information, and motion estimation information of the underwater robot; determining acoustic difference indices based on acoustic ranging information and motion estimation information, and determining current fluctuation indices based on thruster current information; determining a dynamic matching window based on the physical installation distance between the underwater robot's acoustic transducer and thruster, and the water velocity determined based on the thruster rotation speed, and matching the peak time of acoustic characteristics of the acoustic difference indices with the peak time of wake characteristics of the current fluctuation indices within the dynamic matching window to determine the pile foundation crossing time; determining a correction position based on the acoustic ranging value corresponding to the pile foundation crossing time, the installation coordinates of the single acoustic beacon, and the linear constraint of the inspection channel, and determining the environmental flow velocity observation value based on the deviation between the correction position and the estimated position corresponding to the pile foundation crossing time; updating the flow velocity parameters of the dynamic model based on the environmental flow velocity observation value, and correcting the current estimated position based on the correction position.

[0005] In one possible design, the correction position is determined based on the acoustic ranging value corresponding to the pile foundation crossing time, the installation coordinates of the single acoustic beacon, and the straight-line constraints of the inspection channel. This includes: constructing a circular equation with the installation coordinates of the single acoustic beacon as the center and the acoustic ranging value corresponding to the pile foundation crossing time as the radius. With the underwater robot in heading-keeping control mode, a straight-line equation is constructed based on the geometric parameters of the channel centerline corresponding to the current inspection task plan. The circular and straight-line equations are solved simultaneously to obtain a set of geometric intersection points. Based on the geometric relationship between the set of geometric intersection points and the calculated position corresponding to the pile foundation crossing time, the correction position is determined from the set of geometric intersection points; if a unique correction position cannot be determined, the current correction is abandoned.

[0006] In one possible design, the peak times of acoustic characteristics of the acoustic difference index and the peak times of wake characteristics of the current fluctuation index are matched within a dynamic matching window to determine the pile foundation crossing time. This includes: performing local peak extraction operations on the time series of the acoustic difference index and the time series of the current fluctuation index respectively to obtain a set of acoustic peak times and a set of wake peak times. A matching cost matrix is ​​constructed, where each element represents the matching cost between any time in the acoustic peak time set and any time in the wake peak time set. If the time difference between the two times falls within the dynamic matching window, the matching cost is the absolute value of the difference between the time difference and the reference delay time; otherwise, the matching cost is infinite. The matching cost matrix is ​​solved using a preset bipartite graph global matching algorithm to obtain the matching pair with the minimum global cost. The acoustic peak times in the matching pair are determined as the pile foundation crossing times.

[0007] In one possible design, a dynamic matching window is determined based on the physical installation distance between the underwater robot's acoustic transducer and thruster, and the water velocity estimated based on the thruster's rotational speed. This includes: determining the underwater robot's longitudinal water velocity along its longitudinal axis based on a preset mapping relationship between thruster rotational speed and thrust, and the underwater robot's hydrodynamic drag characteristics; determining a reference delay time based on the ratio of the physical installation distance to the longitudinal water velocity; and determining the lower and upper limits of the dynamic matching window based on the product of the reference delay time and a preset tolerance coefficient.

[0008] In one possible design, an acoustic difference index is determined based on acoustic ranging information and motion estimation information, including: determining the rate of change of observed distance based on the acoustic ranging values ​​at the current and previous moments; determining the theoretical rate of change of distance based on the azimuth vector of the underwater robot's estimated position relative to the installation coordinates of the single acoustic beacon, combined with the underwater robot's longitudinal travel speed; and determining the acoustic difference index based on the difference between the observed distance change rate and the theoretical distance change rate.

[0009] In one possible design, the current fluctuation index is determined based on the thruster current information, including: performing filtering on the thruster current information; the filtering process is used to retain low-frequency fluctuation components caused by external flow field disturbances and suppress DC components and high-frequency components corresponding to the thruster motor self-noise; calculating the variance of the filtered current signal within a preset statistical window and determining the variance as the current fluctuation index; the duration of the preset statistical window corresponds to the time required for the underwater robot to traverse the diameter of the pile foundation.

[0010] In one possible design, the environmental velocity observation value is determined based on the deviation between the corrected position and the calculated position corresponding to the pile foundation crossing time. This includes: calculating the positional deviation vector between the corrected position and the calculated position corresponding to the pile foundation crossing time; calculating the time interval between the pile foundation crossing time and the previous geometric correction time; and determining the environmental velocity observation value based on the ratio of the positional deviation vector to the time interval.

[0011] In one possible design, the flow velocity parameters of the dynamic model are updated based on environmental flow velocity observations, and the current estimated position is corrected based on the correction position. This includes: using a preset weighted average algorithm to fuse the environmental flow velocity observations with the flow velocity parameters in the dynamic model at the previous moment to obtain the updated flow velocity parameters. These flow velocity parameters are used in dead reckoning calculations to perform vector synthesis with the surface velocity to determine the underwater robot's ground velocity. The position correction amount between the correction position and the estimated position corresponding to the pile foundation crossing moment is calculated. The position correction amount is then added to the estimated position at the current moment to obtain the corrected current position.

[0012] In one possible design, the filtering process is implemented using a bandpass filter. The low-frequency cutoff frequency of the bandpass filter is used to filter out the DC component and the ocean current variation component, while the high-frequency cutoff frequency of the bandpass filter is set to half of the lowest speed frequency of the thruster to filter out the self-noise of the thruster motor.

[0013] In one possible design, the underwater robot acquires acoustic ranging information, thruster current information, and motion estimation information, including: establishing a unified time reference and setting the main control sampling period. At each discrete system moment, the ranging value provided by a single acoustic beacon, the current value fed back by the main thruster motor driver, the real-time speed of the thruster, and the position coordinates and longitudinal ground velocity provided by the navigation filter are acquired. A zero-order hold strategy is used to time-align sensor data with sampling rates lower than the system period. The aligned state vector is stored in a circular historical data buffer.

[0014] The underwater robot navigation and positioning method based on single beacon ranging and dynamic model provided in this application has the following beneficial effects: By fusing acoustic ranging information, thruster current information and motion estimation information, and sequentially performing the extraction of acoustic difference index and current fluctuation index, dynamic matching window construction and peak matching, geometric correction position calculation, environmental flow velocity observation inversion and dynamic model parameter update, this method can automatically identify pile foundation crossing events and complete absolute position correction in dense pile foundation environments with strong acoustic multipath interference without relying on Doppler velocimeters. At the same time, it suppresses the cumulative drift of long-term dead reckoning through flow velocity feedback, thereby significantly improving the positioning accuracy and long-term stability of underwater robot inspection operations. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating an embodiment of an underwater robot navigation and positioning method based on a single beacon ranging and dynamic model, provided for the purposes of this application. Figure 2 This is a schematic diagram of the system architecture of an underwater robot navigation and positioning system based on a single beacon ranging and dynamic model, provided as an embodiment of this application. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the underwater robot navigation and positioning method based on a single beacon ranging and dynamic model proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0019] The specific scheme of the underwater robot navigation and positioning method based on single beacon ranging and dynamic model provided in this application is described in detail below with reference to the accompanying drawings.

[0020] Please see Figure 1 It shows a flowchart of the underwater robot navigation and positioning method based on single beacon ranging and dynamic model provided in this application, such as... Figure 1 As shown, the method includes the following steps: S101-S105.

[0021] S101. Acquire acoustic ranging information, thruster current information, and motion calculation information of the underwater robot.

[0022] The aforementioned acoustic ranging information is used to characterize the distance between the underwater robot and the monoacoustic beacon, provided by the monoacoustic beacon receiver, and represents the straight-line distance from the underwater robot's transducer to the beacon. The thruster current information is used to characterize the operating state of the thruster motor, fed back by the main thruster motor driver, and represents the instantaneous operating current of the thruster motor. The motion estimation information is used to characterize the underwater robot's estimated position and velocity, provided by the underwater robot's underlying navigation filter, including the current estimated position coordinates, longitudinal ground-based estimated velocity, and real-time thruster rotation speed.

[0023] In one specific implementation, the process of acquiring the aforementioned multi-source heterogeneous data in this step includes the following sub-steps: S1011. Establish a unified time reference and set the main control sampling period.

[0024] In some embodiments, the electronic device may allocate a circular historical data buffer of a preset length in memory and set the system's main control sampling period to be [missing information]. ,in This represents the sampling period of the system's master controller, for example, 0.1 seconds. At each current discrete system time... ( (A positive integer, starting from 1), corresponding to the timestamp. The electronic device then performs subsequent data acquisition and alignment operations.

[0025] S1012. At each discrete system moment, acquire the ranging value provided by the monoacoustic beacon, the current value fed back by the main thruster motor driver, the real-time speed of the thruster, and the position coordinates and longitudinal ground velocity provided by the navigation filter.

[0026] In some embodiments, the electronic device at the current system time Acquire the following data: Monoacoustic beacon ranging value, provided by the monoacoustic beacon receiver, representing the straight-line distance from the underwater robot's transducer to the beacon; thruster current value, fed back by the main thruster motor driver, representing the instantaneous operating current of the thruster motor; motion estimation status, including the current estimated position coordinates, longitudinal ground-based estimated velocity, and real-time thruster rotation speed.

[0027] S1013. Employ a zero-order hold strategy to time-align sensor data with sampling rates lower than the system cycle.

[0028] In some embodiments, since the various sensors of an underwater robot often output data at different frequencies, for sensors with a sampling rate lower than the system's main control sampling period... Sensors (such as monoacoustic beacon receivers) at the current moment when no new data arrives. Electronic devices directly reuse the data from the previous time step in the buffer. The values ​​are used to ensure that all data is on the timeline. Strict alignment.

[0029] S1014. Store the aligned state vector into the circular historical data buffer.

[0030] In some embodiments, the aligned full state vector is indexed by the current time step. The data is stored in a circular historical data cache for later backtracking. Furthermore, the circular historical data cache uses a circular structure; when the storage space is full, new data will overwrite the oldest data, ensuring that historical state information from the most recent period is always retained to support subsequent spatiotemporal matching and geometric correction operations.

[0031] The sensors in this embodiment may include: a monoacoustic beacon receiver, a thruster motor driver, a bottom-level navigation filter, and a heading measurement sensor. The monoacoustic beacon receiver can be mounted on the head of the underwater robot to receive acoustic signals emitted by monoacoustic beacons pre-deployed at the end or lateral fixed positions of the inspection channel, and output acoustic ranging information. This acoustic ranging information is the measured straight-line distance between the underwater robot's acoustic transducer and the monoacoustic beacon. The thruster motor driver is electrically connected to the main thruster motor and is used to provide real-time feedback on the instantaneous operating current of the thruster motor, outputting thruster current information. The thruster motor driver can also be connected to its associated speed detection unit to obtain the thruster's rotational speed. The bottom-level navigation filter is integrated into or connected to the underwater robot's navigation computing device, used to fuse raw sensor data from the inertial measurement unit, depth gauge, magnetic compass, etc., and output motion estimation information. The motion estimation information may include estimated position coordinates and longitudinal ground-based estimated velocity. The heading measurement sensor may include an electronic compass or a fiber optic gyroscope, used to measure the heading angle of the underwater robot and send the heading data to the heading hold controller, enabling the underwater robot to travel in a straight line along a predetermined inspection channel.

[0032] S1015. Initialization and boundary condition handling.

[0033] The first discrete moment after the electronic device's system starts This is considered the initialization period. During the initialization period, the electronic device only performs sensor data acquisition and buffering, without calculating acoustic difference indicators or current fluctuation indicators, or performing feature matching and geometric correction.

[0034] For calculations that require data from the previous time step, the following boundary condition is defined: rate of change of observation distance. The calculation (S1021) in It is directly set to 0 because It does not exist. Theoretical rate of change of distance. The calculation in Set directly to 0. Acoustic difference index. (S1023) in Set to 0. From Initially, the system enters its normal workflow, and all formulas... All of these have valid definitions, and the above boundary conditions no longer apply.

[0035] Step S102: Determine the acoustic difference index based on the acoustic ranging information and motion calculation information, and determine the current fluctuation index based on the thruster current information.

[0036] In this step, the acoustic difference index is a feature quantity used to quantify the degree of anomaly in acoustic ranging values, characterizing the intensity of acoustic multipath interference. When sound waves are reflected from the surface of the photovoltaic pile foundation instead of propagating directly, a multipath effect occurs, causing non-physical abrupt changes in the ranging value. By comparing the difference between the observed rate of distance change and the rate of distance change calculated by the motion model, the multipath interference component in the acoustic signal can be extracted. The current fluctuation index is a feature quantity used to quantify the intensity of low-frequency fluctuations in the propeller current. Its physical meaning lies in characterizing the intensity of external flow field disturbances (such as pile foundation wake). When water flows through a cylindrical pile foundation, low-frequency flow field disturbances such as the Karman vortex street are formed behind the pile foundation. When this disturbance acts on the propeller, it causes low-frequency fluctuations in the load torque.

[0037] In one specific implementation, the process of determining the two characteristic indicators, the acoustic difference index and the current fluctuation index, includes the following sub-steps: Step S1021: Determine the rate of change of the observed distance based on the acoustic ranging values ​​at the current time and the previous time.

[0038] In some embodiments, the electronic device utilizes the current time in the buffer. Compared to the previous moment The rate of change of the observed distance can be calculated using the backward difference method from the monoacoustic ranging value. .

[0039] For example, the formula for calculating the rate of change of the observation distance is:

[0040] in, For the current moment The rate of change of the observed distance indicates how fast the distance actually measured by the acoustic receiver changes, and the unit is meters per second (m / s). For the current moment The ranging value of a single acoustic beacon, in meters (m); For the previous moment The ranging value of a single acoustic beacon, in meters (m); The main control sampling period is measured in seconds (s).

[0041] S1022. Based on the azimuth vector of the underwater robot's calculated position relative to the installation coordinates of the monoacoustic beacon, and combined with the underwater robot's longitudinal speed, determine the theoretical distance change rate.

[0042] In some embodiments, the electronic device uses the underwater robot's longitudinal velocity vector and the relative geometric relationship between the current estimated position coordinates and the known coordinates of the beacon to calculate the theoretical rate of change of distance that should theoretically occur under this motion state. Specifically, it first determines the acoustic ranging value from the previous moment. Is it less than a preset minimum threshold (e.g., 0.01 meters)? If If the distance is less than a preset minimum threshold, the theoretical distance change rate at that moment is set to zero to avoid the denominator being zero and rendering the calculation meaningless. Otherwise, the robot's relative position vector with respect to the beacon at the current moment is first defined: .in, The dimension of is meter. For the current discrete time Position vector of underwater robot This is the fixed installation position vector for a single acoustic beacon.

[0043] Theoretical distance change rate The following formula can be used for calculation:

[0044] in, For the current moment The longitudinal velocity vector of an underwater robot, expressed in meters per second (m / s). For the previous moment The ranging value of a single acoustic beacon, in meters (m). The symbol "" in the above formula... The dot product operator ("" indicates a vector product operation). The formula above can be expressed as follows: the rate of change of distance between the underwater robot and the beacon is equal to the projection of the underwater robot's velocity vector onto the line connecting the underwater robot and the beacon. When the underwater robot moves towards the beacon, the rate of change of distance is negative; when it moves away from the beacon, the rate of change of distance is positive. This theoretical rate of change of distance reflects the speed of distance change that should occur during the actual physical motion of the underwater robot.

[0045] S1023. Determine the acoustic difference index based on the difference between the observed distance change rate and the theoretical distance change rate.

[0046] In some embodiments, the electronic device calculates the absolute value of the difference between the observed distance change rate and the theoretical distance change rate as an acoustic difference index. For example, the formula for calculating this acoustic difference index is as follows:

[0047] in, For the current moment The acoustic difference index is expressed in meters per second (m / s). The rate of change of the observed distance at the current time k is expressed in meters per second (m / s). For the current moment The theoretical rate of change of distance, expressed in meters per second (m / s). If The significantly increased value indicates that the observed distance change far exceeds the physical limits achievable by underwater robot dynamics, typically caused by signal jumps due to acoustic multipath reflections. The calculated... The data is appended and stored in chronological order to form an acoustic difference time series, i.e., the first feature sequence.

[0048] S1024. Perform filtering processing on the thruster current information. The filtering processing is used to retain the low-frequency fluctuation component caused by external flow field disturbance and suppress the DC component and the high-frequency component corresponding to the thruster motor self-noise.

[0049] In some embodiments, the electronic device employs a digital bandpass filter. The low-frequency cutoff frequency of this bandpass filter is used to filter out DC components and extremely low-frequency macroscopic ocean current variations (e.g., set to 0.5Hz), while the high-frequency cutoff frequency is set to be less than half the minimum operating frequency of the thruster (e.g., when the frequency corresponding to the minimum thruster speed is 10Hz, the high-frequency cutoff frequency can be half the minimum thruster speed, such as 5Hz). This is used to filter out high-frequency self-noise components such as commutation noise and mechanical vibration generated by the high-speed rotation of the thruster motor. After processing by this filter, a filtered current signal is output, which retains the low-frequency fluctuation components caused by external flow field disturbances (such as pile wake).

[0050] S1025. Calculate the variance of the filtered current signal within the preset statistical window, and determine the variance as the current fluctuation index. The duration of the preset statistical window corresponds to the time required for the underwater robot to traverse the diameter of the pile foundation.

[0051] In some embodiments, the electronic device sets a statistical window length. (For example, 20 sampling points) Extract the nearest sample from the filtered current signal Calculate the statistical variance of each data point. .

[0052] For example, the formula for calculating this current fluctuation index is as follows:

[0053] in, For the current moment The current fluctuation index is the statistical variance of the filtered current signal within the statistical window. The statistical window length is the number of data points involved in the variance calculation. Its value usually corresponds to the time required for an underwater robot to traverse the diameter of a pile foundation. For the statistical window corresponding to discrete time points The filtered current value (where (Indicates the current time); This is the arithmetic mean of the filtered current data within this statistical window. The larger the value, the more severe the low-frequency fluctuations in the thruster's load current over a recent period. These fluctuations are highly correlated with the underwater robot thruster entering the wake turbulence region generated by the pile foundation. The calculated values... The data is appended and stored in chronological order to form a current fluctuation time series, which is the second characteristic sequence.

[0054] Through the above steps S1021 to S1025, the electronic device successfully transforms the raw, heterogeneous, and noisy sensor data into two time-aligned feature sequences: a first feature sequence characterizing acoustic anomalies and a second feature sequence characterizing wake disturbances.

[0055] Step S103: Based on the physical installation distance between the acoustic transducer and the thruster of the underwater robot and the water velocity determined based on the thruster rotation speed, determine the dynamic matching window, and match the peak time of the acoustic characteristics of the acoustic difference index and the peak time of the wake characteristics of the current fluctuation index within the dynamic matching window to determine the pile foundation crossing time.

[0056] In this step, the physical installation spacing is an inherent structural parameter of the underwater robot, referring to the fixed physical distance between the center of the acoustic transducer and the center of the main thruster along the longitudinal axis of the robot body. The water velocity refers to the speed of the water flow relative to the underwater robot's body; it is a key variable determining the physical delay between the acoustic signal and the wake signal. Since the underwater robot's acoustic transducer is mounted at the head and the thruster at the tail, when the underwater robot passes through the pile foundation, the acoustic signal will inevitably appear before the wake signal. The time difference between the two depends on the underwater robot's travel speed and the physical installation spacing. The dynamic matching window defines the physical range within which the reasonable acoustic-wake time difference should fall under the current travel conditions.

[0057] In one specific implementation, the process of determining the pile foundation crossing time in this step includes the following sub-steps: S1031. Perform local peak extraction operations on the time series of acoustic difference index and current fluctuation index respectively to obtain the set of acoustic peak moments and the set of wake peak moments.

[0058] In some embodiments, non-maximum suppression (NMS) is performed on the time series of acoustic difference index and current fluctuation index, as specifically implemented as follows: S1, preset acoustic noise floor threshold.

[0059] In open water without foundations, the underwater robot maintains a constant speed and travels in a straight line, continuously collecting data. seconds (e.g.) Acoustic difference index data (30s). Calculate the mean of this data segment. and standard deviation Acoustic noise floor threshold Set as:

[0060] The acoustic noise threshold is set to three times the standard deviation of the mean background noise, which can effectively filter out 99.7% of conventional noise.

[0061] S2, preset current fluctuation noise threshold.

[0062] Using the same method, in open water without foundations, the underwater robot maintains a constant speed and travels in a straight line, continuously collecting data. seconds (e.g.) Calculate the mean value of the current fluctuation index data (30 seconds). and standard deviation Current noise threshold The calculation formula can be:

[0063] S3, Local Neighborhood Window Settings.

[0064] Set the radius of the local neighborhood window. The local neighborhood window radius should be less than the minimum time interval between characteristic peaks generated by two adjacent pile foundations, but greater than the duration of a single characteristic peak. For example, based on the typical underwater robot's travel speed (0.5 m / s) and the pile diameter (approximately 0.3 m), the duration of a single peak is approximately 0.3 / 0.5 = 0.6 seconds. The spacing between two pile foundations is typically 3 meters, and the time interval between adjacent peaks is approximately 3 / 0.5 = 6 seconds. Therefore, the local neighborhood window radius can be set to 1 second to ensure that local peaks corresponding to the same pile foundation do not inhibit each other, while peaks from different pile foundations do not interfere with each other.

[0065] S4, Non-maximum suppression execution process.

[0066] For each candidate point (i.e., a data point whose amplitude is greater than the preset noise floor threshold), let its timestamp be... The minimum timestamp of the currently analyzed sequence is The maximum timestamp is Define the left and right boundaries of the adaptive local neighborhood window:

[0067]

[0068] in, The preset local neighborhood window radius (e.g., 1 second).

[0069] Checking candidate points within closed intervals Is the value at that moment the maximum value? (If multiple equal maximum values ​​exist, all can be retained or the values ​​can be slightly adjusted according to their amplitudes.) If yes, then that moment is retained as the local peak value; otherwise, it is discarded.

[0070] This processing method ensures that the window is automatically truncated to the valid interval at the start and end boundaries of the time series, and the algorithm can still execute normally.

[0071] S1032. Construct a matching cost matrix. Each element in the matching cost matrix is ​​used to characterize the matching cost between any time in the acoustic peak time set and any time in the wake peak time set.

[0072] In some embodiments, in densely piled foundation areas, signals from multiple piles may exist simultaneously within the observation window, resulting in a complex many-to-many relationship in the candidate node set. The processor constructs a dimension... OK Matching cost matrix of columns The row represents Each acoustic peak represents a column of acoustic peaks. The first wake peak. The first peak in the matrix. Line 1 Column elements Indicates the first Each acoustic peak moment With the Peak wake The matching cost between them. Calculate the time difference for each pair of peaks. .

[0073] If time difference The dynamic matching window that falls into the above calculation Internal (i.e., satisfying) The matching cost is then the absolute value of the difference between the time difference and the baseline delay time, i.e. , The smaller the value, the closer the time difference between the two signals is to the theoretical reference value, and the higher the probability of a match.

[0074] If time difference If the match does not fall within the window (e.g., the time difference is too large, too small, or negative), the matching cost is set to a preset fixed constant. This indicates that the match is prohibited. The value of should be greater than the maximum cost that any valid match might incur. For example, considering the physical range of acoustic ranging values ​​and the speed of the underwater robot, the cost of a valid match typically does not exceed a few seconds (e.g., ...). The magnitude is in seconds, and the window expansion is... (usually less than 1 second), therefore It can be set to 1,000,000 or larger, for example The value is 1,000,000. This value is much larger than the potential maximum value of the cost of all legal matches, which ensures that in subsequent bipartite graph global matching algorithms (such as the Hungarian algorithm), combinations that contain prohibited matches will never be selected as the optimal solution, while avoiding the numerical overflow problem that may be caused by directly using mathematical infinity.

[0075] S1033. Solve the matching cost matrix using a preset bipartite graph global matching algorithm to obtain the matching pair with the minimum global cost.

[0076] In some embodiments, the electronic device will construct a matching cost matrix. Given a pre-defined bipartite graph global matching algorithm (such as the Hungarian algorithm), this algorithm finds the matching pairs with the minimum global cost. This algorithm searches for independent combinations of minimum-cost elements in a matrix and outputs a set of optimal matching index pairs. The output of this bipartite graph global matching algorithm guarantees that, among all possible matching combinations, the sum of the total costs of the selected matching pairs is minimized, which is equivalent to finding the signal interpretation that best conforms to the physical delay law on a global scale.

[0077] The specific implementation of the Hungarian algorithm, matrix initialization, and optimal solution determination rules are as follows: S21. Matching Cost Matrix Initialization: The number of rows in the matrix is ​​the number of acoustic peaks. The number of columns is the number of wake peaks. For elements whose time difference falls into the dynamic matching window, the cost is the absolute value of the time difference and the baseline delay time difference. For elements whose time difference does not fall into the window, the cost is assigned a fixed value much greater than the legal cost (e.g., 1000000), indicating that matching is prohibited.

[0078] S22. The Hungarian algorithm execution steps include: subtracting the minimum value of each row and the minimum value of each column of the matrix; covering all zero elements with the minimum number of horizontal and vertical lines. If the number of covering lines is equal to the matrix order, the optimal match is obtained; otherwise, the minimum value of the uncovered area is adjusted in the matrix, and the above steps are repeated until the condition is met.

[0079] S23. Optimal solution determination rule: Select the matching pair with the minimum total cost as the global optimal solution; if the cost of the optimal matching pair is less than the prohibited matching fixed value, it is determined to be a valid match; if the cost of all matching pairs is the prohibited value, it is determined that there is no valid pile foundation crossing event at present.

[0080] S1034. Determine the acoustic peak time in the matching pair as the pile foundation crossing time.

[0081] In some embodiments, the set of matching pairs output by the electronic device traversal algorithm For each pair If its corresponding cost Less than the preset fixed constant If the window constraint is satisfied, the underwater robot is deemed to have successfully detected a defined pile foundation crossing event. At this point, the electronic equipment extracts the timestamp corresponding to the acoustic peak value in the matching pair. And mark it as the moment of confirmation of pile foundation crossing. It is important to note that the selection of the acoustic peak moment... Instead of the peak wake time, the pile foundation crossing time is used. This is crucial because subsequent geometric correction requires single-beacon ranging data, and the physical meaning of the ranging value is determined at the moment the sound wave arrives (i.e., the moment of acoustic peak). Strictly aligning the time base to the moment the acoustic event occurs eliminates position calculation errors caused by wake hysteresis, ensuring the accuracy of geometric correction.

[0082] S1035. Based on the preset mapping relationship between the thruster speed and thrust, and the hydrodynamic drag characteristics of the underwater robot, determine the longitudinal velocity of the underwater robot along the longitudinal axis.

[0083] In some embodiments, the thrust model and hydrodynamic drag coefficient of the propeller are obtained through steady-state calibration tests in a land-based water tank. Specific implementation may include: Thrust coefficient calibration: The underwater robot is fixed to a six-dimensional force sensor test rig, and the propeller is driven at different rotational speeds in still fresh water, recording the rotational speed and thrust data. Further, the least squares method is used to fit and obtain the thrust model: thrust equals the thrust coefficient multiplied by the square of the rotational speed, thus determining the specific value of the thrust coefficient. Drag coefficient calibration: In a controllable flow field water tank, the underwater robot is pulled at different stable flow velocities, and the horizontal resistance experienced by the underwater robot is recorded. A drag model is fitted and obtained: drag equals the drag coefficient multiplied by the square of the velocity relative to the water, thus determining the specific value of the drag coefficient. Online calculation of velocity relative to the water: When the underwater robot is navigating, the thrust of the propeller and the hydrodynamic drag reach a steady-state equilibrium. Based on the principle that thrust equals drag, the longitudinal velocity relative to the water is calculated from the real-time rotational speed, the calibrated thrust coefficient, and the drag coefficient. Abnormal handling: When the rotational speed is zero or the thrust coefficient and drag coefficient have not been calibrated, the longitudinal velocity against the water is assigned to zero, and subsequent pile foundation crossing tests are not performed.

[0084] S1036. Determine the reference delay time based on the ratio of the physical installation spacing to the longitudinal water velocity.

[0085] In some embodiments, the electronic device reads pre-stored underwater robot structural parameters from non-volatile memory, namely, the fixed physical distance between the center of the acoustic transducer and the center of the main thruster along the longitudinal axis of the underwater robot, denoted as... This distance is a constant, determined by the mechanical design of the underwater robot, and is measured in meters. In this embodiment, the acoustic transducer is mounted on the head of the underwater robot, and the main thruster is mounted on the tail. This is the axial distance from the head to the tail, for example, a value of 1.5 meters.

[0086] Furthermore, the electronic equipment acquires the real-time rotational speed of the thruster, denoted as... (Unit: revolutions per minute). Based on the pre-calibrated correlation between the thrust characteristics and hydrodynamic drag characteristics of the thruster, the longitudinal velocity of the underwater robot along the longitudinal axis is estimated, denoted as... (Unit: m / s). This longitudinal velocity relative to the water indicates the underwater robot's speed relative to the surrounding water current. Unlike the ground velocity provided by the now-defunct Doppler velocity meter, the longitudinal velocity relative to the water is unaffected by unknown ocean currents and directly reflects the underwater robot's motion effect of overcoming water resistance through thruster propulsion.

[0087] In this embodiment, the longitudinal water velocity The estimation is based on the following steady-state equilibrium relationship: thruster thrust Hydrodynamic resistance Equal. Thrust With rotational speed It is proportional to the square of, that is ,in Thrust coefficient; hydrodynamic drag With speed It is proportional to the square of, that is ,in This is the drag coefficient. (From...) achievable After determining the proportional coefficient through pre-calibration, it can be adjusted according to the real-time rotational speed. Calculate .

[0088] Subsequently, as the underwater robot navigates along its longitudinal axis through the pile foundation, the head acoustic transducer first passes through the transverse direction of the pile foundation, causing a sudden change in the acoustic ranging value and generating an acoustic characteristic peak. Then, the tail thruster passes the same location, experiencing wake disturbance from the pile foundation, causing low-frequency fluctuations in the thruster current and generating a wake characteristic peak. The time difference between these two characteristic events depends on the physical installation spacing. longitudinal water velocity of the underwater robot The ratio of .

[0089] Electronic devices calculate the reference delay time according to the following formula. :

[0090] in: This refers to the physical installation spacing, in meters. The longitudinal velocity relative to the water is expressed in meters per second. The base delay time is in seconds. This is a preset positive constant used to prevent the denominator from being zero. The value is 0.001, which is introduced. The resulting calculation error is within the allowable range.

[0091] The above formula represents the underwater robot's speed relative to the water. distance traveled The required time is the theoretical time difference between the acoustic signal and the wake signal.

[0092] The greater the underwater robot's speed in the water, the faster it passes through the pile foundation, and the longer the reference delay time between the acoustic signal and the wake signal. The shorter the time, the longer the reference delay time; conversely, the smaller the water velocity, the longer the reference delay time. This reference delay time will serve as the center value of the subsequent dynamic matching window, used to verify the consistency between the actual observed peak times of acoustic features and the peak times of wake features.

[0093] S1037. Determine the lower limit time and upper limit time of the dynamic matching window based on the product of the reference delay time and the preset tolerance coefficient.

[0094] In some embodiments, the electronic device reads a preset dimensionless tolerance coefficient from the memory, denoted as... This coefficient is used to determine the reference delay time. The time window is expanded vertically to form its boundaries. In this embodiment, The range of values ​​is It can take the value 0.3. The larger the window, the wider it is, and the greater the tolerance for deviation. The smaller the value, the narrower the window, and the higher the required matching accuracy. This coefficient can be pre-calibrated based on engineering experience such as the estimation error of actual sailing speed and the intensity of turbulence in the flow field.

[0095] Furthermore, the product of the baseline delay time and the preset tolerance coefficient is calculated to obtain the window expansion amount. The formula for calculating the window's expansion amount is:

[0096] in: The base delay time is in seconds. This is a preset tolerance coefficient, dimensionless; This represents the expansion of the time window, measured in seconds.

[0097] This extension represents the allowable range of time deviation: allowing deviations upwards and downwards from the baseline delay time. .

[0098] Subsequently, the electronic device calculates the lower limit time of the dynamic matching window based on the baseline delay time and the expansion amount. and upper limit time The calculation formulas are as follows:

[0099]

[0100] in: The lower limit time for dynamically matching windows, in seconds; The upper limit of the dynamic matching window, in seconds; The base delay time is in seconds. This is a preset tolerance coefficient, which is dimensionless.

[0101] The two formulas above define the physical range within which the time difference between the peak acoustic characteristic moment and the peak wake characteristic moment should fall under the current navigation conditions: the time difference must be greater than or equal to... and less than or equal to For example, when Second, At that time, Second, Second.

[0102] Finally, the electronic device will calculate the result. and As the boundary, a complete dynamic matching time window is constructed, denoted as... This window will be used in subsequent steps to match and verify the set of acoustic feature peak times with the set of wake feature peak times: only when the time difference between a pair of acoustic peak times and wake peak times falls within this window will it be considered a candidate matching pair that conforms to the laws of physical motion; otherwise, it will be directly excluded.

[0103] This step constructs a dynamic matching window that adapts to the current navigation speed. This window can automatically adjust its width and center position according to the underwater robot's real-time water velocity, thereby effectively distinguishing between real pile passing events and false noise interference in dense pile foundation areas.

[0104] S104. Based on the acoustic ranging value corresponding to the time of pile foundation crossing, the installation coordinates of the single acoustic beacon, and the straight line constraint of the inspection channel, determine the correction position, and based on the deviation between the correction position and the calculated position corresponding to the time of pile foundation crossing, determine the environmental flow velocity observation value.

[0105] In this step, the installation coordinates of the single acoustic beacon are known quantities that have been precisely measured beforehand. The inspection channel straight line constraint is a geometric prior information, referring to the equation of the channel centerline located between the two rows of piles on which the underwater robot is currently performing its inspection task. In heading-keeping control mode, the lateral position of the underwater robot is constrained to be near this straight line. The core of this step lies in: utilizing the deterministic moment locked at the pile crossing time, combined with the prior geometric constraints (single beacon ranging circle, photovoltaic array channel straight line), to eliminate the ambiguity of single beacon positioning and to reverse-calculate the environmental ocean current parameters.

[0106] In one specific implementation, the process of determining the correction location and environmental flow velocity observations in this step includes the following sub-steps: S1041. Construct a circle equation with the installation coordinates of the single acoustic beacon as the center and the acoustic ranging value corresponding to the moment the pile foundation crosses the circle as the radius.

[0107] In some embodiments, the electronic device is used to measure the time of pile foundation crossing. For indexing, retrieve the state data for that moment from the historical data cache, including historical ranging values ​​and estimated position vectors. The fixed installation coordinates of the underwater monoacoustic beacon are used as ( , For example, an electronic device constructs a distance ring equation. For instance, this ring equation can be:

[0108] in, The coordinates of the underwater robot's actual position are the variables. For the moment of pile foundation crossing The corresponding acoustic ranging value is in meters (m). This circular equation stipulates that the underwater robot's true position must lie on a circular trajectory with the beacon as the center and the historical ranging value as the radius.

[0109] S1042. Under the condition that the underwater robot is in the heading-keeping control mode, construct the straight line equation according to the geometric parameters of the channel centerline corresponding to the current inspection task.

[0110] In some embodiments, this application is implemented on the premise that the underwater robot has activated its underlying heading-keeping control mode. In this mode, the underwater robot corrects its heading in real time using a compass or fiber optic gyroscope to ensure it always points towards the inspection channel of the photovoltaic pile array. Although the underwater robot's longitudinal position may drift due to the influence of unknown ocean currents, its lateral position is always constrained near the channel centerline under the control system, and the lateral deviation is usually small and controllable. Therefore, the electronic equipment reads the geometric parameters of the channel centerline from the electronic chart according to the current task plan (e.g., currently inspecting the kth row of corridors). , , Construct the equation of the line. For example, the equation of the line could be:

[0111] in, , , The geometric parameters for the channel centerline are provided by the electronic chart; Let be the coordinates of the underwater robot's actual position. This equation stipulates that the underwater robot's actual position must lie on a straight line within the current inspection channel.

[0112] The aforementioned heading-keeping control mode employs a proportional-integral-derivative (PID) closed-loop control. The PID control parameters were obtained through steady-state calibration tests in a land-based water tank. Specifically, the calibration conditions included: placing the underwater robot in a still water tank and setting the heading angle of the target inspection channel. The calibration process involved gradually adjusting the proportional, integral, and derivative coefficients to stabilize the underwater robot's heading deviation within the allowable range, without overshoot or oscillation. The calibration results showed a proportional coefficient of 2.0, an integral coefficient of 0.1, and a derivative coefficient of 0.05. The control logic compared the real-time heading angle obtained from the heading measurement sensor with the target heading angle, inputting the deviation into the PID controller. The controller output a control quantity to the thruster, correcting the underwater robot's heading through differential torque to maintain the heading deviation within ±5°.

[0113] S1043. Solve the equations of the circle and the line simultaneously to obtain the set of geometric intersection points.

[0114] In some embodiments, the installation coordinates of a monoacoustic beacon are used. Centered on the pile foundation crossing time Corresponding acoustic ranging value Construct the equation of the circle with radius:

[0115] With the underwater robot in heading-keeping control mode, the geometric parameters of the channel centerline are planned according to the current inspection task. Construct the equation of the straight line:

[0116] Solving the equations of the circle and the line simultaneously yields the set of geometric intersection points. Based on the number of algebraic solutions, It may contain 0, 1, or 2 geometric intersections. When the circle and the line are separate, It is an empty set (0 intersection points); when the circle is tangent to the line, It contains one intersection point; when a circle and a line intersect at two points, It contains 2 intersection points.

[0117] S1044. Determine the corrected position vector based on the geometric relationship between the set of geometric intersections and the calculated position vector.

[0118] In some embodiments, the pile foundation crossing time is set The corresponding calculated position vector is The set of geometric intersections is .according to The number of elements in the vector is determined according to the following rules for the correction position vector. : 1) When Time (tangent) gather There is only one intersection point, denoted as Since there is no ambiguity in the location, the intersection point is directly determined as the correction location:

[0119] 2) When Time (intersection) gather It contains two intersection points, denoted as and Calculate the intersection points and the calculated position vectors respectively. Euclidean distance:

[0120] Because the error in dynamic calculations accumulates over time, the accumulated error is usually much smaller than the macroscopic distance between the two intersection points within the interval between two corrections. Therefore, the closer intersection point is preferred as the correction location.

[0121] However, when the distance difference between the two intersection points and the calculated position is less than the preset ambiguity threshold... When the error is 0.5 meters (e.g., the calculated position is located near the perpendicular bisector of the two intersections, making it impossible to reliably distinguish the actual intersections), it is considered that a unique correction position cannot be determined, and the correction process is abandoned.

[0122] 3) When Time (separation) The absence of an intersection between a circle and a straight line is typically caused by acoustic ranging noise, heading deviation, or beacon coordinate errors. In such cases, no geometric intersection can be obtained, and the correction position cannot be determined.

[0123] Handling of abandoning correction: Discard this correction in any of the following cases: set of geometric intersections An empty set (disjoint); a set of geometric intersections It contains two intersection points, and the difference in distance from the two intersection points to the calculated position is less than the ambiguity threshold. .

[0124] When the correction is abandoned, the correction position vector is not output. Environmental flow velocity observations are not included. The system continues to use the dynamic model parameters from the previous moment, waiting for the next pile foundation crossing event before attempting correction.

[0125] like It contains two intersection points and the distance difference is not less than ,or If a unique intersection point is included, the corrected position vector will be output normally. Then continue to execute S1045 to S1047 (calculate the position deviation and environmental flow velocity observations).

[0126] S1045. Calculate the position deviation vector between the corrected position vector and the calculated position vector corresponding to the moment of pile foundation crossing.

[0127] In some embodiments, the electronic device calculates the vector difference between the corrected position vector and the calculated position vector, denoted as the position deviation vector. For example, the formula for calculating this position deviation vector can be:

[0128] in, The position deviation vector represents the vector pointing from the corrected position to the opposite direction of the calculated position (i.e., the correction direction from the calculated position to the corrected position). The determined correction position vector; For the moment of pile foundation crossing The corresponding calculated position vector. This deviation. The physical meaning is that it represents the time since the last correction moment. up to the current moment During this period, due to environmental ocean current factors not modeled in the dynamic model, the underwater robot's actual trajectory deviated from the calculated trajectory by a cumulative amount of drift.

[0129] S1046. Calculate the time interval between the time of pile foundation crossing and the time of the last geometric correction.

[0130] In some embodiments, the electronic device acquires the current pile foundation crossing time. The time since the last successful geometric correction The time interval between For example, the formula for calculating this time interval is:

[0131] in, The time interval is expressed in seconds (s). This is the current moment when the pile foundation crossing is confirmed; This is the time of the last geometric correction.

[0132] It should be noted that when When the value is zero or close to zero (i.e., two consecutive correction times overlap), the calculation of this ratio loses its physical meaning. In practical applications, a minimum time interval threshold is set, when... If the time interval is less than the minimum time interval threshold, the calculation of the current flow velocity observation is skipped, and the previous flow velocity parameter is used. For example, the minimum time interval threshold can be set to 2.0 seconds to prevent abnormal flow velocity observations caused by two consecutive correction times being too close together. This threshold can be pre-calibrated based on the typical travel speed of the underwater robot (e.g., 0.5 m / s) and the reasonable interval required for positioning correction (generally, it should be greater than the travel time of the underwater robot's own length; for example, at least 3 seconds for an underwater robot 1.5 meters long). In this embodiment, considering both the system sampling period (0.1 seconds) and practical engineering experience, 2.0 seconds is chosen as the minimum time interval threshold.

[0133] S1047. Determine the environmental flow velocity observation value based on the ratio of the position deviation vector to the time interval.

[0134] In some embodiments, assuming that the ambient ocean currents are relatively stable during the time period, the electronic equipment uses positional deviations and time intervals to calculate average ambient current velocity observations. For example, the formula for calculating this environmental flow velocity observation is as follows:

[0135] in, The environmental flow velocity is an observed value, which is a vector and is measured in meters per second (m / s). This is the positional deviation vector, with units in meters (m). The time interval from the previous correction time to the current pile foundation crossing time is expressed in seconds (s). If the flow rate is less than the minimum time threshold (e.g., 2.0 s), the flow rate from the previous time step is used. The dimensional analysis of this formula is: meters (m) divided by seconds (s) yields meters per second (m / s), which is the dimension of velocity, ensuring consistent dimensions. If... A positive value indicates that the ocean current propels the underwater robot forward (with the current); a negative value indicates that the ocean current hinders the underwater robot's movement (against the current).

[0136] S105. Update the velocity parameters of the dynamic model based on the environmental velocity observations, and correct the current estimated position based on the correction position.

[0137] In this step, the flow velocity parameter of the dynamic model is a state variable within the navigation filter, used to characterize the average water flow velocity of the underwater robot's environment. This parameter is vector-synthesized with the water velocity estimated based on thruster rotation speed during dead reckoning to obtain the underwater robot's land velocity. To prevent drastic oscillations in model parameters caused by single observation noise or small lateral deviations, a smoothing filter strategy is introduced in this step to improve the system's engineering robustness.

[0138] In one specific implementation, the process of updating the flow velocity parameters and correcting the estimated location in this step includes the following sub-steps: S1051. A preset weighted average algorithm is used to fuse the observed environmental flow velocity with the flow velocity parameters in the dynamic model at the previous moment to obtain updated flow velocity parameters. These flow velocity parameters are used in dead reckoning calculations to perform vector synthesis with the water velocity to determine the underwater robot's ground velocity.

[0139] In some embodiments, the electronic device reads the flow velocity parameters stored in the dynamic model from the previous moment. Combined with the currently calculated observations New flow velocity parameters are calculated using a preset weighted average algorithm. For example, the calculation formula for this weighted average algorithm is as follows:

[0140] in, The updated flow velocity parameters will be used for velocity synthesis in subsequent dead reckoning. The flow rate parameter before the update represents historical empirical values; These are the currently calculated environmental flow velocity observations; It is a dimensionless smoothing factor, with a value range of 1. . The magnitude of this value determines the system's level of confidence in the new observation: if A larger value (e.g., 0.8) indicates that the system is highly sensitive to the current observation, and the flow velocity parameter updates quickly, but it is easily affected by single-error disturbances and oscillates; if... Smaller values ​​(e.g., 0.1) indicate that the system tends to maintain its historical state, with smoother flow velocity parameter updates and stronger noise resistance, but a slower response to sudden changes in flow velocity. In typical photovoltaic pile foundation inspection scenarios, since ocean current changes are usually relatively gentle, it is preferable to use [a specific value, e.g., 0.1]. Set to a small value (e.g., 0.1 to 0.2) to ensure the smoothness of the navigation system. Updated flow rate parameters. The flow rate parameter is written into the status register of the navigation filter, replacing the original flow rate parameter. In every subsequent system moment... When performing dead reckoning, the navigation filter will use new flow velocity parameters to synthesize velocity, namely the ground velocity. Equal to the velocity of water With updated flow rate parameters The sum of these factors. In this way, the system actively compensates for the influence of ocean currents in the feedforward stage, reducing future prediction errors.

[0141] S1052. Calculate the position correction between the corrected position and the estimated position corresponding to the time of pile foundation crossing.

[0142] In some embodiments, the electronic device will use the position deviation calculated above. As the position correction amount for this calibration That is, the two are numerically equal. Specifically, the formula for calculating the position correction vector is:

[0143] in, This is a position correction vector, with units in meters (m). The calculated position deviation vector is in meters (m). For the moment of pile foundation crossing The corrected position vector; For the moment of pile foundation crossing The corresponding calculated position vector. This position correction vector represents the translation vector required from the calculated position to the corrected position, and is used for subsequent overall translation correction of the calculated positions at the current and future times.

[0144] S1053. Add the position correction amount to the calculated position at the current moment to obtain the corrected current position.

[0145] In some embodiments, due to the time of penetration from the pile foundation up to the current system time A period of time has passed, and the electronic equipment needs to adjust the position from the moment the piles crossed. Recursively apply this to the current moment. The electronic device will adjust the position correction vector. The calculated position vector is directly superimposed on the current time. Up, get the corrected current position For example, the formula for calculating this position correction is:

[0146] in, For the current moment Corrected current position coordinates; For the current moment The calculated position vector; This is the position correction vector, in meters (m). This operation is equivalent to translating and correcting the entire calculated trajectory of the underwater robot on the map, realigning it to its actual physical location.

[0147] It should be noted that the rationale for the above translational correction lies in the fact that at the moment of pile foundation penetration... Based on environmental flow velocity observations The flow velocity parameters of the dynamic model were updated, thus providing feedforward compensation for subsequent dead reckoning. Therefore, from arrive short time interval Within this context, the additional estimation error caused by the unmodeled flow velocity is far smaller than the positional deviation of a single geometric correction. Using overall translation correction can effectively eliminate historical accumulated errors while ensuring the real-time performance of the project, and at the same time avoid the additional noise introduced by complex recursive calculations. The calculation formula is:

[0148] Corrected current position for:

[0149] In this embodiment, the dynamic model used is an underwater robot planar motion dead reckoning model. The core form and implementation of the model are as follows: Model state variables: including the robot's two-dimensional position, longitudinal ground velocity, and environmental flow velocity vector. Core velocity synthesis: The ground velocity is obtained by superimposing the water velocity estimated by the thruster and the environmental flow velocity vector stored in the dynamic model. Dead reckoning integration: The Euler integral method is used to obtain the real-time estimated position by integrating the ground velocity over time, with the integration period consistent with the system's main control sampling period. Model initial conditions: The initial state is the robot's initial position and initial ground velocity at startup, and the initial environmental flow velocity parameter is set to zero by default. Model update mechanism: The smoothed environmental flow velocity is injected into the model in real time to replace the original flow velocity parameter, realizing closed-loop compensation for dead reckoning errors.

[0150] Through steps S101 to S105, this application completes a closed loop from data acquisition, feature extraction, event matching, geometric correction to model updating. The system can achieve robust positioning with high accuracy over a long period of time in dense pile foundation areas where sensor information is limited and environmental interference is strong. Specifically, this application innovatively correlates acoustic multipath interference characteristics with thruster wake disturbance characteristics, enabling reliable identification of the deterministic moment when the underwater robot traverses the pile foundation without relying on DVL; then, using the acoustic ranging value at that moment and the known environmental geometric constraints (straight channel), the absolute position of the underwater robot is calculated, and the environmental ocean current parameters are derived; finally, through a closed-loop feedback mechanism, the ocean current parameters are used to update the dynamic model and correct the current estimated position, effectively overcoming the ambiguity problem of single beacon positioning and the error accumulation problem of dead reckoning.

[0151] The underwater robot navigation and positioning method based on single beacon ranging and dynamic model provided in this application has at least the following beneficial effects: By fusing acoustic ranging information, thruster current information and motion estimation information, and sequentially performing the extraction of acoustic difference index and current fluctuation index, dynamic matching window construction and peak matching, geometric correction position calculation, environmental flow velocity observation inversion and dynamic model parameter update, this method can automatically identify pile foundation crossing events and complete absolute position correction in dense pile foundation environments with strong acoustic multipath interference without relying on Doppler velocimeters. At the same time, it suppresses the cumulative drift of long-term dead reckoning through flow velocity feedback, thereby significantly improving the positioning accuracy and long-term stability of underwater robot inspection operations.

[0152] Please see Figure 2This illustration shows a schematic diagram of the system architecture of an underwater robot navigation and positioning system based on a single beacon ranging and dynamic model, according to an embodiment of the present invention. The underwater robot navigation and positioning system based on a single beacon ranging and dynamic model includes: an information acquisition unit 201, a feature extraction unit 202, a spatiotemporal matching unit 203, a geometric correction unit 204, and a feedback update unit 205. The units can communicate bidirectionally via a communication link to ensure real-time interaction of collected data and analysis results. The communication link can employ wired or wireless transmission methods to meet the communication needs of different monitoring scenarios.

[0153] The information acquisition unit 201 is used to acquire acoustic ranging information, thruster current information and motion calculation information of the underwater robot.

[0154] The feature extraction unit 202 is used to determine the acoustic difference index based on acoustic ranging information and motion estimation information, and to determine the current fluctuation index based on the thruster current information.

[0155] The spatiotemporal matching unit 203 is used to determine the dynamic matching window based on the physical installation distance between the acoustic transducer and the thruster of the underwater robot and the water velocity determined based on the thruster rotation speed, and to match the peak time of the acoustic characteristics of the acoustic difference index and the peak time of the wake characteristics of the current fluctuation index within the dynamic matching window to determine the time of pile foundation crossing.

[0156] The geometric correction unit 204 is used to determine the correction position based on the acoustic ranging value corresponding to the time of pile foundation crossing, the installation coordinates of the single acoustic beacon and the straight line constraint of the inspection channel, and to determine the environmental flow velocity observation value based on the deviation between the correction position and the calculated position corresponding to the time of pile foundation crossing.

[0157] The feedback update unit 205 is used to update the flow velocity parameters of the dynamic model based on the environmental flow velocity observations, and to correct the current estimated position based on the correction position.

[0158] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0159] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A navigation and positioning method for underwater robots based on single beacon ranging and a dynamic model, characterized in that, The method includes: Acquire acoustic ranging information, thruster current information, and motion calculation information of the underwater robot; The acoustic difference index is determined based on the acoustic ranging information and the motion estimation information, and the current fluctuation index is determined based on the thruster current information; Based on the physical installation distance between the acoustic transducer and the thruster of the underwater robot and the water velocity determined based on the thruster rotation speed, a dynamic matching window is determined, and the peak time of the acoustic characteristics of the acoustic difference index and the peak time of the wake characteristics of the current fluctuation index are matched within the dynamic matching window to determine the pile foundation crossing time. Based on the acoustic ranging value corresponding to the time of pile foundation crossing, the installation coordinates of the single acoustic beacon, and the straight line constraint of the inspection channel, the correction position is determined, and based on the deviation between the correction position and the calculated position corresponding to the time of pile foundation crossing, the environmental flow velocity observation value is determined. The flow velocity parameters of the dynamic model are updated based on the environmental flow velocity observations, and the current estimated position is corrected based on the correction position. The step of determining the acoustic difference index based on the acoustic ranging information and motion estimation information includes: The rate of change of the observed distance is determined based on the acoustic ranging values ​​at the current time and the previous time. Based on the azimuth vector of the underwater robot's calculated position relative to the installation coordinates of the monoacoustic beacon, and combined with the underwater robot's longitudinal speed, the theoretical distance change rate is determined. The acoustic difference index is determined based on the difference between the observed distance change rate and the theoretical distance change rate.

2. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 1, characterized in that, The determination of the correction position based on the acoustic ranging value corresponding to the pile foundation crossing time, the installation coordinates of the single acoustic beacon, and the straight-line constraint of the inspection channel includes: Using the installation coordinates of the single acoustic beacon as the center and the acoustic ranging value corresponding to the moment the pile foundation crosses the ground as the radius, construct a circular equation; Under the condition that the underwater robot is in the heading-keeping control mode, a straight line equation is constructed according to the geometric parameters of the channel centerline corresponding to the current inspection task. Solve the equations of the circle and the line simultaneously to obtain the set of geometric intersection points; Based on the geometric relationship between the set of geometric intersections and the calculated position corresponding to the time of pile foundation crossing, the correction position is determined from the set of geometric intersections; if a unique correction position cannot be determined, the correction is abandoned.

3. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 1, characterized in that, The step of matching the peak time of the acoustic characteristics of the acoustic difference index with the peak time of the wake characteristics of the current fluctuation index within the dynamic matching window to determine the pile foundation crossing time includes: Local peak extraction operations are performed on the time series of the acoustic difference index and the time series of the current fluctuation index respectively to obtain the set of acoustic peak moments and the set of wake peak moments. Construct a matching cost matrix, where each element of the matching cost matrix is ​​used to characterize the matching cost between any time in the acoustic peak time set and any time in the wake peak time set. If the time difference between the two times falls within the dynamic matching window, the matching cost is the absolute value of the difference between the time difference and the reference delay time; otherwise, the matching cost is infinite. The matching cost matrix is ​​solved by a preset bipartite graph global matching algorithm to obtain the matching pair with the minimum global cost; The acoustic peak time in the matching pair is determined as the pile foundation crossing time.

4. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 1, characterized in that, The step of determining the dynamic matching window based on the physical installation distance between the acoustic transducer and the thruster of the underwater robot and the water velocity estimated based on the thruster rotation speed includes: Based on the preset mapping relationship between the thruster speed and thrust, and the hydrodynamic drag characteristics of the underwater robot, the longitudinal velocity of the underwater robot along the longitudinal axis is determined. The reference delay time is determined based on the ratio of the physical installation spacing to the longitudinal water velocity. The lower and upper limits of the dynamic matching window are determined based on the product of the baseline delay time and the preset tolerance coefficient.

5. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 1, characterized in that, The step of determining the current fluctuation index based on the thruster current information includes: The thruster current information is filtered to retain the low-frequency fluctuation component caused by external flow field disturbance and to suppress the DC component and the high-frequency component corresponding to the thruster motor self-noise. The variance of the filtered current signal is calculated within a preset statistical window, and the variance is determined as the current fluctuation index. The duration of the preset statistical window corresponds to the time required for the underwater robot to traverse the diameter of the pile foundation.

6. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 1, characterized in that, The step of determining the environmental flow velocity observation value based on the deviation between the corrected position and the calculated position corresponding to the time of pile foundation crossing includes: Calculate the positional deviation vector between the corrected position and the estimated position corresponding to the moment the pile foundation crosses the ground; Calculate the time interval between the time of the pile foundation crossing and the time of the previous geometric correction; The environmental flow velocity observation value is determined based on the ratio of the position deviation vector to the time interval.

7. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 1, characterized in that, The step of updating the flow velocity parameters of the dynamic model based on the observed environmental flow velocity, and correcting the current estimated position based on the correction position, includes: A preset weighted average algorithm is used to fuse the observed environmental flow velocity with the flow velocity parameters in the dynamic model at the previous moment to obtain updated flow velocity parameters; the flow velocity parameters are used to perform vector synthesis with the water velocity in dead reckoning to determine the ground velocity of the underwater robot; Calculate the positional correction between the corrected position and the estimated position corresponding to the time of pile foundation crossing; The position correction is added to the calculated position at the current moment to obtain the corrected current position.

8. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 5, characterized in that, The filtering process is implemented using a bandpass filter. The low-frequency cutoff frequency of the bandpass filter is used to filter out the DC component and the ocean current variation component. The high-frequency cutoff frequency of the bandpass filter is set to half of the lowest speed frequency of the thruster to filter out the self-noise of the thruster motor.

9. The underwater robot navigation and positioning method based on single beacon ranging and dynamic model according to claim 1, characterized in that, The acquisition of acoustic ranging information, thruster current information, and motion estimation information of the underwater robot includes: Establish a unified time reference and set the main control sampling cycle; At each discrete system moment, the ranging value provided by the monoacoustic beacon, the current value fed back by the main thruster motor driver, the real-time speed of the thruster, and the position coordinates and longitudinal ground velocity provided by the navigation filter are acquired. A zero-order hold strategy is used to time-align sensor data with sampling rates lower than the system cycle. The aligned state vector is stored in a circular historical data cache.

Images