A precision inspection system and method for seabed facilities using air-submarine collaboration

Abstract

This invention discloses a refined inspection system and method for underwater facilities using a combined air-submarine and underwater approach. The system includes a drone, an underwater robot, and a main control module integrated into the drone. The drone is equipped with a positioning module to receive satellite positioning signals and obtain its absolute coordinates in a global geographic coordinate system. An altitude detection unit is installed below the drone's rotor. A controlled winch is mounted on the drone and connected to the underwater robot via a signal cable, controlling the release and retrieval of the underwater robot. A cable tension detection module is installed at the winch's cable outlet to monitor the tension, elevation angle, and azimuth angle of the signal cable in real time. The underwater robot is equipped with an inertial measurement unit, a depth sensor, and sonar. This invention utilizes a drone as an aerial platform, eliminating the need for costly large inspection mother ships and effectively reducing blind spots that traditional large inspection mother ships cannot cover.

Description

Technical Field

[0001] This invention relates to the field of underwater topographic observation and facility inspection, and in particular to a refined inspection system and method for seabed facilities using a combination of air and underwater capabilities. Background Technology

[0002] Traditional underwater topographic observation or facility inspection typically relies on manned vessels equipped with multibeam sonar or manned remotely operated vehicles (ROVs). The process involves the manned vessel sailing to the target area and anchoring, then using its onboard crane or winch to deploy detection equipment (such as an ROV or side-scan sonar) into the water. The equipment operates underwater along a predetermined route, transmitting the collected data in real-time to the ship's control terminal via an umbilical cable. After the operation is completed, the manned vessel retrieves the equipment and returns to port.

[0003] However, traditional underwater topographic observation techniques have significant drawbacks in practical applications:

[0004] First, large inspection vessels are not only extremely costly to mobilize, but also limited by draft, making it difficult to enter narrow riverbeds, near-shore shoals, and other blind spots for inspection. In some solutions, to overcome the operational costs of large inspection vessels, drones equipped with detection equipment are used for underwater inspections. However, traditional drone inspection solutions lack the ability to dynamically perceive the air-sea interface at multiple points. Single-point altitude radar will experience severe ranging "jumps" under fluctuating sea surfaces, which can easily lead to violent oscillations in the system's attitude or even instantaneous breakage of the signal cable, causing it to crash into the sea.

[0005] Secondly, in terms of data quality, because underwater robots (ROVs) lack absolute geographic coordinate references, the cumulative drift caused by the drag of water currents cannot achieve high-precision spatiotemporal alignment with aerial RTK coordinates, resulting in severe ghosting or deformation in the generated 3D models of facilities; in addition, existing inspection systems mostly use fixed speeds and cannot adaptively adjust the sampling density according to the complexity of the seabed topography, resulting in sparse data in key areas.

[0006] Therefore, developing an inspection system that can sense sea surface dynamics in real time, achieve high-precision cross-medium positioning, and has adaptive inspection capabilities has become an urgent need to improve underwater topographic observation and underwater facility maintenance. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of the prior art by providing a refined inspection system and method for underwater facilities that combines air and submarine operations.

[0008] The objective of this invention is achieved through the following technical solution: a method for refined inspection of seabed facilities using a combination of air and submersible methods, comprising the following specific steps: Step 1: Collect raw sea surface echo data and obtain corresponding ranging values ​​through multiple altitude detection units on the UAV. Convert the ranging values ​​of each altitude detection unit into three-dimensional spatial point coordinates in the UAV coordinate system to form instantaneous point cloud data reflecting the instantaneous shape of the sea surface. Perform plane fitting on the instantaneous point cloud data to extract the mean sea level reference surface. Calculate the instantaneous stable altitude of the UAV relative to the sea level based on the mean sea level reference surface. Step 2: Based on the difference between the instantaneous stable altitude and the preset target inspection altitude, adjust the altitude of the UAV to stabilize it at the target inspection altitude; Step 3: The controlled winch on the UAV releases the signal cable to allow the underwater robot to enter the water to the target inspection depth; the main control module acquires the coordinates of the UAV in the world geographic coordinate system, the attitude of the UAV, the release length of the signal cable, the end depth of the signal cable, the tension vector of the signal cable, and the attitude of the underwater robot in real time. Based on the above data, the underwater robot attitude data is converted from the coordinates of the original terrain data points collected by the sonar to the world geographic coordinate system. Step 4: The system calculates the real-time point cloud coverage density of the target area by the sonar and compares it with the preset density threshold. If the real-time point cloud coverage density is lower than the density threshold, the current area is determined to be a complex seabed topography area or a critical facility area. The underwater robot reduces its inspection speed and triggers a reciprocating scanning mode to increase the point cloud coverage density of the target area and bring it up to the density threshold.

[0009] Preferably, in step one, after extracting the mean sea level reference surface, the normal distance from each height detection unit to the mean sea level reference surface is calculated, and the arithmetic mean of the normal distances corresponding to each height detection unit is used as the instantaneous stable height.

[0010] Preferably, the drone is a multi-rotor drone, with the altitude detection unit installed below each rotor of the drone; The ranging values ​​of each altitude detection unit are converted into three-dimensional spatial point coordinates in the UAV coordinate system based on the spatial vector formula. The spatial vector formula is as follows: ; In the formula, The coordinates of a point in three-dimensional space in the UAV coordinate system. The translation transformation matrix between each altitude detection unit and the UAV is given. The vector representing the direction of the detection axis of the altitude detection unit in the UAV coordinate system. This is the ranging value measured by the altitude detection unit.

[0011] Preferably, in step one, the instantaneous point cloud data is fitted to a plane using a random sampling consensus algorithm. Outliers caused by broken waves, splashes, or floating debris are removed through multiple iterations to obtain the sea level reference surface. If the maximum number of interior points after a single iteration is less than a set threshold or the fitting residual exceeds a preset safety threshold, the fitting is deemed to have failed, and the instantaneous stable altitude calculated in the previous sampling period is taken as the current instantaneous stable altitude of the UAV.

[0012] As a preferred embodiment, the specific method for converting the sonar detection points to the world geographic coordinate system in step three is as follows: The sonar installation bias matrix is ​​constructed based on the relative position of the sonar and the underwater robot; Construct an underwater attitude correction matrix based on the current attitude of the underwater robot; The tension vector includes the tension value, elevation angle, and azimuth angle at the upper end of the signal cable; the signal cable constraint and cross-medium translation matrix is ​​constructed based on the release length of the signal cable, the end depth of the signal cable, and the tension vector of the signal cable. An absolute coordinate mapping matrix is ​​established based on the UAV's own flight attitude data and the UAV's absolute coordinates in the world geographic coordinate system. Based on the sonar installation offset matrix, underwater attitude correction matrix, signal cable constraint and cross-medium translation matrix, and absolute coordinate mapping matrix, the coordinates of the original terrain data points acquired by the sonar are converted into absolute coordinates in the world geographic coordinate system.

[0013] Preferably, when the water flow velocity is lower than a preset velocity threshold, the signal cable is approximated as a straight line. Based on geometric trigonometric relationships, the yaw angle and actual horizontal projection length of the signal cable in the water are calculated, thereby obtaining the distance between the end of the signal cable and the UAV. Horizontal displacement drift Directional horizontal drift, based on Horizontal displacement drift The horizontal drift direction and the depth at the end of the signal cable are used to construct the signal cable constraint and cross-medium translation matrix; When the water flow velocity exceeds a preset velocity threshold, the tension vector of the signal cable is substituted into the segmented catenary mechanical model as a boundary condition. The tension magnitude is used to assess the tension of the signal cable, and the elevation and azimuth angles in the tension vector are used to determine the tangent direction at the upper end of the signal cable. Combined with the depth at the end of the signal cable, the integral path of the signal cable in the water is calculated through numerical iteration, thereby calculating the drag caused by the water flow at the end of the signal cable. Horizontal position drift and Horizontal offset; based on Horizontal displacement drift The direction of horizontal drift and the depth of the signal cable end are used to construct the signal cable constraint and cross-medium translation matrix.

[0014] Preferably, in step four, the point cloud coverage density is the ratio of the number of original terrain data points in the target area of ​​the sonar scan to the area of ​​the target area of ​​the sonar scan; wherein, the area of ​​the target area of ​​the sonar scan is the product of the sonar scan width, the current cruising speed of the underwater robot, and the duration of the statistical window.

[0015] Preferably, after step four is completed, step five is performed: the drone recovers the underwater robot and returns, and a high-precision three-dimensional digital twin model of the seabed facility is generated based on the detection data obtained in step four.

[0016] A sophisticated air-submarine integrated inspection system for seabed facilities includes a drone, an underwater robot, and a main control module integrated on the drone. The drone is equipped with a positioning module to receive satellite positioning signals to obtain the drone's absolute coordinates in the world geographic coordinate system; An altitude detection unit is installed under the rotor of the drone to monitor the instantaneous altitude of the drone relative to the sea surface in real time; a controlled winch is installed on the drone, which is connected to the underwater robot through a signal cable, and the release and retrieval of the underwater robot are controlled by the controlled winch; a cable tension detection module is installed at the cable outlet of the controlled winch to monitor the tension, elevation angle and azimuth angle of the upper end of the signal cable in real time. The underwater robot is equipped with an inertial measurement unit, a depth sensor, and a sonar. The sonar is used to scan the seabed topography. The main control module is electrically connected to the positioning module, the height detection unit, the controlled winch, the cable tension detection module, and the depth sensor.

[0017] Preferably, the UAV is a hexacopter UAV, and the altitude detection unit is a millimeter-wave radar. The UAV is equipped with a total of six altitude detection units, which are arranged in a ring array to form an altitude detection array.

[0018] The beneficial effects of this invention are: 1. This invention, through an altitude detection unit mounted on a UAV, can acquire multi-dimensional sea surface echo data in real time and use an algorithm to fit a dynamic sea level benchmark, thereby eliminating the ranging deviation caused by the jump between the crests and troughs of sea waves in traditional single-point altitude radar. This precise average altitude calculation mechanism ensures that the main control module can output a stable and accurate PWM adjustment signal to the UAV's motor driver, avoiding the risk of severe fuselage vibration or instantaneous cable breakage caused by altitude misjudgment, and ensuring the stability of the relative altitude between the UAV and the sea level.

[0019] 2. This invention achieves centimeter-level precision modeling of underwater facilities through a multi-level spatial coordinate transformation chain. It maps the high-precision geographic location reference from the UAV to the underwater detection unit, effectively compensating for the pose error of the underwater robot caused by water current drag and sensor cumulative drift. This ensures that each frame of seabed point cloud data collected by sonar has accurate global coordinate attributes, solving the technical pain points of underwater models being prone to ghosting and deformation in traditional solutions, and providing a foundation for the restoration of high-fidelity three-dimensional digital twin models of seabed facilities.

[0020] 3. This invention features an adaptive adjustment mechanism for inspection speed, achieving an optimal balance between operational efficiency and sampling accuracy. The system calculates the point cloud coverage density transmitted by the sonar in real time through the main control module and uses it as the core decision indicator for closed-loop speed control. When the real-time point cloud density is lower than a preset threshold, the system determines that the seabed topography is complex or that there are critical facilities. It will automatically reduce the UAV's movement speed and trigger a reciprocating scanning mode to ensure that the detection data density of the target area meets the requirements for fine modeling. This "on-demand allocation" sampling strategy solves the contradiction between data sparsity and sampling redundancy caused by fixed-speed operation in traditional systems.

[0021] 4. This invention greatly expands the environmental adaptability boundary of the inspection system and significantly reduces the economic cost of marine engineering inspection. This invention uses drones as an aerial platform, which eliminates the need for costly large inspection mother ships. The drone platform can flexibly penetrate into narrow riverbeds, near-shore shoals, and river channels with dense obstacles that traditional ships cannot reach, effectively reducing the blind spots that traditional large inspection mother ships cannot reach. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the inspection system of the present invention.

[0023] Figure 2 This is a flowchart of the method of the present invention.

[0024] In the diagram: 1. Unmanned aerial vehicle (UAV); 2. Positioning module; 3. Underwater robot; 4. Controlled winch; 5. Load chamber; 6. Signal cable; 7. Altitude detection unit; 8. Cable tension detection module. Detailed Implementation

[0025] 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. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0026] Those skilled in the art should understand that, in the disclosure of this invention, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the above terms should not be construed as limiting this invention.

[0027] It is understood that the term "a" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0028] like Figure 2 As shown, an air-submarine collaborative precision inspection system for seabed facilities includes a drone 1, an underwater robot 3, and a main control module integrated on the drone; the drone 1 is equipped with a positioning module 2, which is used to receive satellite positioning signals to obtain the absolute coordinates of the drone in the world geographic coordinate system.

[0029] An altitude detection unit 7 is installed below the rotor of the UAV 1 to monitor the instantaneous altitude of the UAV 1 relative to the sea surface in real time. A controlled winch 4 is installed on the UAV 1, which is connected to the underwater robot 3 via a signal cable 6. The controlled winch 4 controls the release and retrieval of the underwater robot 3. A cable tension detection module 8 is installed at the cable outlet of the controlled winch 4 to monitor the tension, elevation angle, and azimuth angle of the upper end of the signal cable 6 in real time. The underwater robot 3 is equipped with an inertial measurement unit, a depth sensor, and a sonar. The sonar is used to perform seabed topographic scanning. The main control module is electrically connected to the positioning module 2, the altitude detection unit 7, the controlled winch 4, the cable tension detection module 8, and the depth sensor.

[0030] The fuselage of drone 1 is made of high-strength carbon fiber.

[0031] The positioning module 2 uses an RTK-GNSS module, which is fixedly installed on the top center axis of the UAV 1. The positioning module 2 is supported by a bracket installed on the top of the UAV 1 so that it is higher than the rotor plane. The RTK-GNSS module receives satellite positioning signals to obtain the absolute geographic coordinates (X,Y,Z) of the UAV at the centimeter level.

[0032] The controlled winch 4 is installed at the lower end of the fuselage of the UAV 1 and is driven by a high-torque brushless motor. One end of the signal cable 6 is wound around the controlled winch 4, and the end of the signal cable 6 passes through the cable tension detection module 8 located at the cable outlet of the controlled winch 4 and is connected to the underwater robot 3.

[0033] The cable tension detection module 8 includes a universal guide bracket, a guide pulley, a pin-type tension sensor, and a dual-axis rotary encoder. The signal cable 6 passes through the guide pulley, and the pin-type tension sensor is mounted on the pulley axle to detect the cable tension in real time. The universal guide bracket has two rotational degrees of freedom: horizontal and pitch. Rotary encoders are connected to its two rotational axes. When the cable is towed by the underwater robot and its attitude changes, the universal guide bracket rotates, and the two rotary encoders synchronously collect the azimuth and elevation angles of the cable. The tension, elevation angle, and azimuth angle together constitute the tension vector of the signal cable 6. The signal cable 6 internally contains a power cable and optical fiber, responsible for supplying power to the UAV 1 and transmitting underwater detection data in real time.

[0034] The underwater robot 3 (ROV) is suspended below the unmanned aerial vehicle 1 via a signal cable 6. The underwater robot 3 (ROV) is equipped with four-axis or six-axis thrusters, enabling it to perform horizontal displacement and attitude adjustment underwater. An internal payload chamber 5 is rigidly fixed to the center of the underwater robot 3. The payload chamber 5 integrates an inertial measurement unit (IMU), a depth sensor, and a sonar. The sonar can be a multi-beam sonar, with its probe extending through the bottom of the chamber to ensure unobstructed acoustic scanning of the seabed topography.

[0035] The main control module is integrated into the host computer within UAV 1 and is connected to the controlled hardware (positioning module 2, altitude detection unit 7, controlled winch 4, cable tension detection module 8, and depth sensor) via a bus. The slave computer within UAV 1 is responsible for outputting PWM (Pulse Width Modulation) signals to the motor driver of UAV 1 according to the instructions from the host computer. By adjusting the duty cycle, the speed of the propellers of UAV 1 is adjusted, thereby stabilizing the altitude of UAV 1.

[0036] The drone can be a multi-rotor drone, with altitude detection units installed below each rotor. In this embodiment, the drone is a hexacopter drone, equipped with six altitude detection units arranged in a circular array to form an altitude detection array.

[0037] In this embodiment, the altitude detection unit is a millimeter-wave radar. The millimeter-wave radar operates at a wavelength of 1-10mm, giving it a strong anti-interference capability against complex marine environments. It has good resistance to weather interference, surface interference, and electromagnetic interference. It can penetrate fog, rain, snow, and salt spray, and is not affected by lighting conditions. It can still work normally under extreme weather conditions such as night, heavy rain, and dense fog.

[0038] like Figure 2 As shown, a method for refined inspection of seabed facilities using a combination of air and submarine operations includes the following specific steps: Step 1: Collect raw sea surface echo data and obtain corresponding ranging values ​​through multiple altitude detection units 7 on UAV 1. Convert the ranging values ​​of each altitude detection unit 7 into three-dimensional spatial point coordinates in the coordinate system of UAV 1 to form instantaneous point cloud data reflecting the instantaneous shape of the sea surface. Perform plane fitting on the instantaneous point cloud data to extract the mean sea level reference surface. Calculate the instantaneous stable altitude of UAV 1 relative to the sea level based on the mean sea level reference surface.

[0039] In this step, the altitude detection units synchronously acquire raw echo data from the sea surface below at a frequency of no less than 50Hz. Within each sampling cycle, the main control module invokes the translation transformation matrix between each altitude detection unit and the UAV. And the direction vector of the detection axis of the altitude detection unit in the UAV coordinate system. .

[0040] The UAV coordinate system is a three-dimensional coordinate system established with the geometric center of the positioning module on the UAV as its origin. This is because the installation coordinates of each altitude detection unit within the UAV coordinate system (…) , The parameters can be pre-calibrated and are known. Therefore, the translation transformation matrix between the altitude detection unit (geometric center) and the UAV (origin of the UAV coordinate system) is... This information can be obtained in advance. Since the altitude detection units are fixedly mounted on the UAV, the direction vector of each altitude detection unit relative to the UAV's detection axis is... This is also known.

[0041] The ranging values ​​of each altitude detection unit are converted into three-dimensional spatial point coordinates in the UAV coordinate system based on the spatial vector formula. The spatial vector formula is as follows: ; In the formula, The coordinates of a point in three-dimensional space in the UAV coordinate system. The translation transformation matrix between each altitude detection unit and the UAV is given. The direction vector of the detection axis of the height detection unit. This is the ranging value measured by the altitude detection unit.

[0042] In this embodiment, the UAV is equipped with six altitude detection units, so it can simultaneously acquire the coordinates of six three-dimensional points in the UAV coordinate system within one detection cycle. - ).

[0043] All 3D space point coordinates will be acquired synchronously. The data is aggregated to form instantaneous point cloud data P that reflects the instantaneous shape of the sea surface.

[0044] A random sampling consensus algorithm is used to fit the instantaneous point cloud data to a plane. Outliers caused by wave fragments, splashes, or floating debris are removed through multiple iterations to obtain the sea level reference surface S. To ensure robustness under undulating sea surface and wave fragment interference, the system sets a fitting tolerance threshold of 5cm-15cm and a maximum number of iterations of 50-100. When the number of interior points conforming to the fitted plane reaches four or more, a valid mean sea level reference surface S is determined, and the iteration is terminated early.

[0045] In extreme cases where large waves may cause plane fitting to fail, if the maximum number of interior points after a single iteration is less than a set threshold (the set threshold is 3 in this embodiment) or the fitting residual exceeds a preset safety threshold, the fitting is deemed to have failed. The instantaneous stable altitude calculated in the previous sampling period is then used as the instantaneous stable altitude of the current UAV to ensure that the system will not cause attitude oscillations due to ranging jumps caused by sudden changes in sea waves.

[0046] After extracting the mean sea level datum, the normal distance from each height detection unit to the mean sea level datum is calculated. The arithmetic mean of the normal distances corresponding to each height detection unit is taken as the instantaneous stable height. The calculation formula is as follows: ; In the above formula, Let N be the normal distance from the k-th altitude sensing unit to the mean sea level datum, and N be the total number of altitude sensing units. This refers to the instantaneous stable altitude.

[0047] Step 2: Based on the difference between the instantaneous stable altitude and the preset target inspection altitude, adjust the altitude of UAV 1 to stabilize it at the target inspection altitude.

[0048] In this step, the main control module calculates the instantaneous stable height in real time. relative to the preset target inspection height The difference , ; like > That is, the difference between the two. If the value is less than zero, the drone's position is too high relative to the target position. The main control module automatically reduces the duty cycle of the PWM signal sent to the drone's rotor motors, lowering the motor terminal voltage and thus reducing rotor lift, allowing the drone to land smoothly at the target inspection altitude. Conversely, if... < That is, the difference between the two. If the value is greater than zero, it indicates that the drone's position is too low relative to the target position. The main control module automatically increases the duty cycle of the PWM signal sent to the drone's rotor motor, increases the voltage at the motor end, thereby increasing the rotor lift and enabling the drone to rise to the target inspection altitude.

[0049] The motor in the controlled winch also performs minute, high-frequency retraction and extension movements based on the wave fluctuations sensed by the radar array, offsetting the traction interference of wave crests and troughs on the underwater robot and ensuring the underwater robot's stability at a constant depth in the water. This closed-loop logic effectively counteracts drastic changes in altitude data caused by wave fluctuations, ensuring the physical safety of the flight platform and signal cable.

[0050] Step 3: The controlled winch 4 on UAV 1 releases the signal cable 6 to allow the underwater robot 3 to enter the water to the target inspection depth; the main control module acquires in real time the coordinates of UAV 1 in the world geographic coordinate system, the attitude of UAV 1, the release length of the signal cable 6, the end depth of the signal cable 6, the tension vector of the signal cable 6, and the attitude of the underwater robot 3. Based on the above data, the attitude data of the underwater robot 3 is used to convert the coordinates of the original terrain data points collected by the sonar to the world geographic coordinate system.

[0051] In this step, after the underwater robot 3 enters the water, it uses the depth information collected by the depth sensor to control the thrusters on the underwater robot 3 to make corresponding adjustments, thereby realizing the fixed-depth inspection of the underwater robot 3.

[0052] The attitude of UAV 1 can be acquired by the inertial measurement unit (IMU) built into UAV 1, and the attitude data of UAV 1 includes pitch angle and roll angle. The coordinates of UAV 1 in the world geographic coordinate system can be acquired by the positioning module 2. The release length of signal cable 6 can be acquired in real time by the motor encoder on the controlled winch.

[0053] The depth of the signal cable's end (i.e., the depth of the underwater robot) can be obtained using a depth sensor mounted on the underwater robot. The tension vector of the signal cable can be obtained through a cable tension detection module, which can simultaneously acquire the tension value F at the upper end of the signal cable and the elevation angle. and azimuth These three elements constitute the tension vector of the signal cable, which is used to construct the signal cable constraint and cross-medium translation matrix. .

[0054] The attitude of an underwater robot can be measured by its onboard inertial measurement unit (IMU). During inspection operations, the underwater robot experiences pitch and roll due to water currents. Therefore, the attitude data collected by the IMU includes the robot's pitch and roll angles. This attitude data is used to construct an underwater attitude correction matrix. .

[0055] The coordinates of the raw terrain data points acquired by sonar are based on the sonar local coordinate system. The obtained sonar local coordinate system. The coordinate system is a coordinate system with its origin located at the center of the sonar probe's transmission.

[0056] underwater robot coordinate system A coordinate system with its origin at the center of the underwater robot's IMU (the geometric measurement center of the inertial measurement unit) is used to describe the underwater robot's motion posture.

[0057] World Geographic Coordinate System Provides absolute latitude, longitude, and altitude references.

[0058] UAV coordinate system It is a three-dimensional coordinate system established with the geometric center of the positioning module on the UAV as the origin.

[0059] The specific method for converting sonar detection points to a world geographic coordinate system is as follows: A sonar installation bias matrix is ​​constructed based on the relative positions of the sonar and the underwater robot. The sonar is equipped with an offset matrix. Used to eliminate physical deviations between the sonar installation location and the underwater robot's center of mass (the geometric measurement center of the inertial measurement unit). Sonar installation offset matrix. Local coordinate system of sonar can be realized To the underwater robot coordinate system Transformation between them.

[0060] Constructing an underwater attitude correction matrix based on the current attitude of the underwater robot Because underwater robots experience pitch and roll during inspections due to water flow, the inertial measurement unit (IMU) onboard the underwater robot collects real-time attitude data (including pitch and roll angles) and constructs an underwater attitude correction matrix. This corrects the underwater robot's tilt in the water. The underwater attitude correction matrix is ​​used to achieve this. The point cloud data is processed in real time to correct the raw terrain data points collected by sonar to data under a horizontal reference system.

[0061] The tension vector of the signal cable is detected by the cable tension detection module, including the tension value, elevation angle and azimuth angle at the upper end of the signal cable; the signal cable constraint and cross-medium translation matrix is ​​constructed based on the release length of the signal cable, the end depth of the signal cable and the tension vector of the signal cable.

[0062] The purpose of this signal cable constraint and cross-medium translation matrix is ​​to determine the three-dimensional offset vector of the underwater robot relative to the unmanned aerial vehicle (UAV), which is used to realize the underwater robot's coordinate system. To UAV coordinate system The transformation.

[0063] When the water flow velocity is lower than a preset velocity threshold, the signal cable is approximated as a straight line. Based on geometric trigonometric relationships, the yaw angle and actual horizontal projection length of the signal cable in the water are calculated, thus obtaining the distance between the end of the signal cable and the drone. Horizontal displacement drift , Horizontal drift ,based on Horizontal displacement drift Directional horizontal drift and depth construction of signal cable ends; signal cable constraints and cross-medium translation matrix. ; .

[0064] When the water flow velocity exceeds a preset velocity threshold, the tension vector of the signal cable is substituted into the segmented catenary mechanical model as a boundary condition. The tension magnitude is used to assess the tension of the signal cable, and the elevation and azimuth angles in the tension vector are used to determine the tangent direction at the upper end of the signal cable. Combined with the depth at the end of the signal cable, the integral path of the signal cable in the water is calculated through numerical iteration, thereby calculating the drag caused by the water flow at the end of the signal cable. Horizontal drift and Horizontal offset of direction ;based on Horizontal displacement drift , Horizontal drift and the depth of the end of the signal cable Constructing signal cable constraints and cross-medium translation matrices ; .

[0065] Using signal cable constraints and cross-medium translation matrix It accurately reflects the spatial position of the underwater robot relative to the drone, thereby realizing the underwater robot's coordinate system. To UAV coordinate system The transformation.

[0066] An absolute coordinate mapping matrix is ​​established based on the UAV's own flight attitude data and the UAV's absolute coordinates in the world geographic coordinate system. The UAV coordinate system was implemented using an absolute coordinate mapping matrix. To the world geographic coordinate system The transformations between these parameters give point cloud data absolute geographical attributes such as longitude, latitude, and altitude.

[0067] Combined with the centimeter-level absolute coordinates provided by the drone's top positioning module ( The data, along with the UAV's own flight attitude data, are transformed at each level using an absolute coordinate mapping matrix. This results in a final global translation and rotation of the transformed data, so that each point cloud data point output contains high-precision three-dimensional geographic information, which can be directly used to construct a three-dimensional digital twin model.

[0068] Based on the sonar installation offset matrix, underwater attitude correction matrix, signal cable constraint and cross-medium translation matrix, and absolute coordinate mapping matrix, the coordinates of the raw terrain data points acquired by the sonar are converted into absolute coordinates in the world geographic coordinate system. The specific transformation formula is as follows: ; in, Install the bias matrix for the sonar; This is the lower attitude correction matrix; For signal cable constraints and cross-medium translation matrices; This is an absolute coordinate mapping matrix. The coordinates of the raw terrain data points acquired by sonar. These are absolute coordinates in the world geographic coordinate system.

[0069] By using the above four-level nested coordinate transformation algorithm, each data point detected by sonar is transformed into the world geographic coordinate system, thus solving the positioning drift problem in underwater operations.

[0070] High-precision satellite timing is obtained through the positioning module (RTK-GNSS module) on the top of the UAV, enabling the UAV's onboard host computer to act as an NTP server and transmit the synchronization signal to the underwater robot via a network link in the signal cable. The inertial measurement unit (IMU) and sonar on the underwater robot act as clients and initiate requests. The UAV's onboard host computer compensates for and calibrates based on network round-trip latency, thereby achieving millisecond-level precise alignment of aerial attitude and underwater detection data on the same global time axis in an underwater environment where satellite signals are blocked. This provides a reliable time consistency basis for high-precision 3D modeling.

[0071] Step 4: The system calculates the real-time point cloud coverage density of the target area by the sonar scan and compares it with a preset density threshold. If the real-time point cloud coverage density is lower than the density threshold, the current area is determined to be a complex seabed topography area or a critical facility area. The underwater robot reduces its inspection speed and triggers a reciprocating scanning mode to increase the point cloud coverage density of the target area and bring it up to the density threshold. If the point cloud coverage density is equal to or greater than the density threshold, it indicates that the density threshold is qualified, and the current efficient inspection speed is maintained.

[0072] Among them, point cloud coverage density The number of raw terrain data points in the target area for sonar scanning. Area of ​​the target region measured by sonar The ratio, that is: ; The area of ​​the target region scanned by the sonar is the scanning width of the sonar. The current cruising speed of the underwater robot and the duration of the statistical window The product of the three is: ; in, The value ranges from 1 to 5 seconds.

[0073] After step four is completed, step five is executed: the drone recovers the underwater robot and returns, and a high-precision three-dimensional digital twin model of the seabed facility is generated based on the detection data obtained in step four.

[0074] This invention has the following advantages: 1. This invention, through an altitude detection unit mounted on a UAV, can acquire multi-dimensional sea surface echo data in real time and use an algorithm to fit a dynamic sea level benchmark, thereby eliminating the ranging deviation caused by the jump between the crests and troughs of sea waves in traditional single-point altitude radar. This precise average altitude calculation mechanism ensures that the main control module can output a stable and accurate PWM adjustment signal to the UAV's motor driver, avoiding the risk of severe fuselage vibration or instantaneous cable breakage caused by altitude misjudgment, and ensuring the stability of the relative altitude between the UAV and the sea level.

[0075] 2. This invention achieves centimeter-level precision modeling of underwater facilities through a multi-level spatial coordinate transformation chain. It maps the high-precision geographic location reference from the UAV to the underwater detection unit, effectively compensating for the pose error of the underwater robot caused by water current drag and sensor cumulative drift. This ensures that each frame of seabed point cloud data collected by sonar has accurate global coordinate attributes, solving the technical pain points of underwater models being prone to ghosting and deformation in traditional solutions, and providing a foundation for the restoration of high-fidelity three-dimensional digital twin models of seabed facilities.

[0076] 3. This invention features an adaptive adjustment mechanism for inspection speed, achieving an optimal balance between operational efficiency and sampling accuracy. The system calculates the point cloud coverage density transmitted by the sonar in real time through the main control module and uses it as the core decision indicator for closed-loop speed control. When the real-time point cloud density is lower than a preset threshold, the system determines that the seabed topography is complex or that there are critical facilities. It will automatically reduce the UAV's movement speed and trigger a reciprocating scanning mode to ensure that the detection data density of the target area meets the requirements for fine modeling. This "on-demand allocation" sampling strategy solves the contradiction between data sparsity and sampling redundancy caused by fixed-speed operation in traditional systems.

[0077] 4. This invention greatly expands the environmental adaptability boundary of the inspection system and significantly reduces the economic cost of marine engineering inspection. This invention uses drones as an aerial platform, which eliminates the need for costly large inspection mother ships. The drone platform can flexibly penetrate into narrow riverbeds, near-shore shoals, and river channels with dense obstacles that traditional ships cannot reach, effectively reducing the blind spots that traditional large inspection mother ships cannot reach.

[0078] This invention is not limited to the preferred embodiments described above. Anyone can derive other products in various forms under the guidance of this invention. However, regardless of any changes in shape or structure, any technical solution that is the same as or similar to this application falls within the protection scope of this invention.

Claims

1. A method for refined inspection of seabed facilities using a combination of air and submersible methods, characterized in that, The specific steps include the following: Step 1: Collect raw sea surface echo data and obtain corresponding ranging values ​​through multiple altitude detection units on the UAV. Convert the ranging values ​​of each altitude detection unit into three-dimensional spatial point coordinates in the UAV coordinate system to form instantaneous point cloud data reflecting the instantaneous shape of the sea surface. Perform plane fitting on the instantaneous point cloud data to extract the mean sea level reference surface. Calculate the instantaneous stable altitude of the UAV relative to the sea level based on the mean sea level reference surface. Step 2: Based on the difference between the instantaneous stable altitude and the preset target inspection altitude, adjust the altitude of the UAV to stabilize it at the target inspection altitude; Step 3: The controlled winch on the UAV releases the signal cable to allow the underwater robot to enter the water to the target inspection depth; the main control module acquires the coordinates of the UAV in the world geographic coordinate system, the attitude of the UAV, the release length of the signal cable, the end depth of the signal cable, the tension vector of the signal cable, and the attitude of the underwater robot in real time. Based on the above data, the underwater robot attitude data is converted from the coordinates of the original terrain data points collected by the sonar to the world geographic coordinate system. Step 4: The system calculates the real-time point cloud coverage density of the target area by the sonar and compares it with the preset density threshold. If the real-time point cloud coverage density is lower than the density threshold, the current area is determined to be a complex seabed topography area or a critical facility area. The underwater robot reduces its inspection speed and triggers a reciprocating scanning mode to increase the point cloud coverage density of the target area and bring it up to the density threshold.

2. The method for refined inspection of seabed facilities using air-submarine collaboration as described in claim 1, characterized in that, In step one, after extracting the mean sea level reference, the normal distance from each height detection unit to the mean sea level reference is calculated, and the arithmetic mean of the normal distances corresponding to each height detection unit is used as the instantaneous stable height.

3. The method for refined inspection of seabed facilities using air-submarine collaboration as described in claim 1, characterized in that, The drone is a multi-rotor drone, and the altitude detection unit is installed under each rotor of the drone; The ranging values ​​of each altitude detection unit are converted into three-dimensional spatial point coordinates in the UAV coordinate system based on the spatial vector formula. The spatial vector formula is as follows: ； In the formula, The coordinates of a point in three-dimensional space in the UAV coordinate system. The translation transformation matrix between each altitude detection unit and the UAV is given. The vector representing the direction of the detection axis of the altitude detection unit in the UAV coordinate system. This is the ranging value measured by the altitude detection unit.

4. The method for refined inspection of seabed facilities using air-submarine collaboration as described in claim 1, characterized in that, In step one, the instantaneous point cloud data is fitted to a plane using a random sampling consensus algorithm. Outliers caused by broken waves, splashes, or floating debris are removed through multiple iterations to obtain the sea level reference surface. If the maximum number of interior points after a single iteration is less than a set threshold or the fitting residual exceeds a preset safety threshold, the fitting is deemed to have failed, and the instantaneous stable altitude calculated in the previous sampling period is taken as the instantaneous stable altitude of the current UAV.

5. The method for refined inspection of seabed facilities using air-submarine collaboration as described in claim 1, characterized in that, In step three, the specific method for converting sonar detection points to the world geographic coordinate system is as follows: A sonar installation bias matrix is ​​constructed based on the relative position of the sonar and the underwater robot; Construct an underwater attitude correction matrix based on the current attitude of the underwater robot; The tension vector includes the tension value, elevation angle, and azimuth angle at the upper end of the signal cable; the signal cable constraint and cross-medium translation matrix is ​​constructed based on the release length of the signal cable, the end depth of the signal cable, and the tension vector of the signal cable. An absolute coordinate mapping matrix is ​​established based on the UAV's own flight attitude data and the UAV's absolute coordinates in the world geographic coordinate system. Based on the sonar installation offset matrix, underwater attitude correction matrix, signal cable constraint and cross-medium translation matrix, and absolute coordinate mapping matrix, the coordinates of the original terrain data points acquired by the sonar are converted into absolute coordinates in the world geographic coordinate system.

6. The method for refined inspection of seabed facilities using air-submarine collaboration as described in claim 5, characterized in that, When the water flow velocity is below a preset velocity threshold, the signal cable is approximated as a straight line. Based on geometric trigonometric relationships, the yaw angle and actual horizontal projection length of the signal cable in the water are calculated, thus obtaining the position of the end of the signal cable relative to the drone. Horizontal displacement drift Directional horizontal drift, based on Horizontal displacement drift The horizontal drift direction and the depth at the end of the signal cable are used to construct the signal cable constraint and cross-medium translation matrix; When the water flow velocity exceeds a preset velocity threshold, the tension vector of the signal cable is substituted into the segmented catenary mechanical model as a boundary condition. The tension magnitude is used to assess the tension of the signal cable, and the elevation and azimuth angles in the tension vector are used to determine the tangent direction at the upper end of the signal cable. Combined with the depth at the end of the signal cable, the integral path of the signal cable in the water is calculated through numerical iteration, thereby calculating the drag caused by the water flow at the end of the signal cable. Horizontal position drift and Horizontal offset; based on Horizontal displacement drift The direction of horizontal drift and the depth of the signal cable end are used to construct the signal cable constraint and cross-medium translation matrix.

7. The method for refined inspection of seabed facilities using air-submarine collaboration as described in claim 1, characterized in that, In step four, the point cloud coverage density is the ratio of the number of original terrain data points in the target area of ​​the sonar scan to the area of ​​the target area of ​​the sonar scan; where the area of ​​the target area of ​​the sonar scan is the product of the sonar scan width, the current cruising speed of the underwater robot, and the duration of the statistical window.

8. The method for refined inspection of seabed facilities using air-submarine collaboration as described in claim 1, characterized in that, After step four is completed, step five is executed: the drone recovers the underwater robot and returns, and a high-precision three-dimensional digital twin model of the seabed facility is generated based on the detection data obtained in step four.

9. A precision inspection system for seabed facilities using a combination of air and submersible methods, characterized in that: This includes drones, underwater robots, and the main control module integrated into the drone; The drone is equipped with a positioning module on its top, which is used to receive satellite positioning signals to obtain the drone's absolute coordinates in the world geographic coordinate system; An altitude detection unit is installed under the rotor of the drone to monitor the drone's instantaneous altitude relative to the sea surface in real time; The drone is equipped with a controlled winch, which is connected to the underwater robot via a signal cable. The release and retrieval of the underwater robot are controlled by the controlled winch. A cable tension detection module is installed at the cable outlet of the controlled winch to monitor the tension, elevation angle and azimuth angle of the upper end of the signal cable in real time. The underwater robot is equipped with an inertial measurement unit, a depth sensor, and a sonar. The sonar is used to scan the seabed topography. The main control module is electrically connected to the positioning module, the height detection unit, the controlled winch, the cable tension detection module, and the depth sensor.

10. A precision inspection system for seabed facilities using air-submarine collaboration as described in claim 9, characterized in that, The drone is a hexacopter drone, and the altitude detection unit is a millimeter-wave radar. The drone is equipped with a total of six altitude detection units, which are arranged in a ring array to form an altitude detection array.

Images