A ship cleaning robot using a cavitation jet device and a cleaning path planning method

Abstract

The application discloses a ship cleaning robot adopting a cavitation jet device and a cleaning path planning method, and belongs to the technical field of ship cleaning robots.The ship cleaning robot comprises a main frame, a buoyancy material, propellers, a track mechanism and a cavitation jet device, the buoyancy material is arranged on the top of the main frame, the number of the propellers is four, four of the propellers are vertically arranged on the main frame, and the other four propellers are horizontally arranged on the main frame at an angle of 45 degrees; the track mechanism is symmetrically fixed on the two sides of the main frame; the cavitation jet device comprises a cavitation disc and a cavitation host, the cavitation disc is fixed on the front part of the main frame, the cavitation host is arranged on the ground of a shore base, the cavitation host is connected with the cavitation disc through a high-pressure pipe, and an inertial navigation cabin and an ultra-short baseline are further arranged on the main frame; the inertial navigation cabin is fixed on the left side of the middle part of the main frame, and the ultra-short baseline is arranged on the right side of the middle part of the main frame.The ship cleaning robot has the self-navigation and obstacle avoidance technologies, and can avoid colliding with obstacles, so that the cleaning efficiency is greatly improved.

Description

Technical Field

[0001] This invention relates to a ship cleaning robot and a cleaning path planning method, belonging to the technical field of ship cleaning robots. Background Technology

[0002] Ships navigate in the marine environment for extended periods, making their underwater hulls highly susceptible to the accumulation of marine organisms (such as barnacles and algae) and corrosion from seawater. These deposits not only increase drag, leading to significantly higher fuel consumption and increased greenhouse gas emissions, but also accelerate the corrosion of the hull's steel plates, jeopardizing the ship's structural safety and navigation performance. Therefore, regular hull cleaning and maintenance are essential components of ship operation.

[0003] Traditional ship cleaning relies on divers or chemical agents, which is inefficient, causes environmental pollution, and poses safety hazards.

[0004] Existing robots use hydraulic pressure to attach to the hull, which is expensive, complex in structure, has low energy loss and efficiency, is noisy, sensitive to oil contamination, requires high maintenance, and is greatly affected by temperature.

[0005] Existing ship cleaning robots have limited functions and rely on manual operation for cleaning ship hulls. They are not equipped with sonar, inertial navigation, ultra-short baseline, altimeter, depth gauge and other equipment, and cannot intelligently detect the underwater environment of the ship hull. This results in inaccurate positioning and unsuitable cleaning trajectories, leading to poor work adaptability and low cleaning efficiency. Summary of the Invention

[0006] This invention proposes a ship cleaning robot using cavitation jet equipment and a cleaning path planning method, which solves the problems of inaccurate positioning and unsuitable cleaning trajectory of existing ship cleaning robots, resulting in poor work adaptability and low cleaning efficiency.

[0007] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0008] A ship cleaning robot employing a cavitation jet device includes a main frame, buoyancy material, propellers, a track mechanism, and a cavitation jet device. The buoyancy material is installed on the top of the main frame. There are eight propellers, four of which are vertically arranged on the main frame, and the other four are horizontally arranged at 45° on the main frame. The track mechanism is symmetrically fixed on both sides of the main frame. The cavitation jet device includes a cavitation disc and a cavitation main unit. The cavitation disc is fixed at the front of the main frame, and the cavitation main unit is placed on a flat shore base. The cavitation main unit and the cavitation disc are connected by a high-pressure pipe. An inertial navigation cabin and an ultra-short baseline are also installed on the main frame. The inertial navigation cabin is fixed at the port side of the midship section of the main frame, and the ultra-short baseline is arranged at the starboard side of the midship section of the main frame.

[0009] Preferably, a power compartment and a control compartment are installed on the main frame. The power compartment is fixed at the upper rear of the main frame to provide power output for the propeller and track mechanism. The control compartment is located directly below the power compartment and is equipped with a depth gauge.

[0010] Preferably, an altimeter is installed below the midship section of the main frame; a laser scale and a forward-looking sonar are arranged at the front of the main frame.

[0011] Preferably, it also includes a cable winding and unwinding device, which transmits electrical energy stably to the power compartment through the power conductor in the umbilical cable, ensuring that the cleaning equipment can work continuously in high-pressure, high-flow mode.

[0012] A cleaning path planning method for a ship cleaning robot using cavitation jet equipment is based on the extended Kalman filter framework to achieve data fusion between the inertial navigation system and the ultra-short baseline positioning system, including the following steps:

[0013] Step S1: Construct a six-degree-of-freedom state vector in the geodetic coordinate system. The state vector is represented as follows:

[0014]

[0015] Where x, y, and z are the robot's absolute positions in the geodetic coordinate system; , , The linear velocity in the geodetic coordinate system; , , For the robot's roll angle, pitch angle, and yaw angle;

[0016] Step S2: Utilize the load system ratio output by the inertial measurement unit. and angular velocity The prior state at time k is obtained through dead reckoning. Specifically, it includes:

[0017] a. Pose update: using unit quaternions The attitude is described, and attitude updates are achieved through quaternion differential equations:

[0018]

[0019] in, For attitude change rate, For IMU gyroscope drift, To correct the angular velocity of the afterload system (subtracting IMU gyroscope drift). The expression for the adjoint matrix of the corrected angular velocity quaternion is as follows:

[0020]

[0021] in, , , Let be the components of the angular velocity along the x, y, and z axes of the loading system;

[0022] b. Velocity Update: Converting the specific force of the load system into the Earth's acceleration, and subtracting the influence of gravity, the velocity update equation is:

[0023]

[0024] in, The direction cosine matrix is ​​derived from quaternions. This is the gravity vector of the Earth system. Zero bias for IMU accelerometer;

[0025] c. Position update: The position is obtained through velocity integration. The position update equation is: , ;

[0026] d. Extended Kalman Filter Prediction Step Discretization: Discretize the continuous quantities of the inertial measurement unit to obtain the prior state and prior covariance at time k.

[0027]

[0028]

[0029] in, Let τ represent the posterior state at time k-1. arrive Any consecutive moments between these points, i.e., time points within the discrete sampling interval. This is a continuous-time state estimate at time τ, including position, velocity, and attitude at time τ. This represents the IMU input vector at time τ. , These represent the angular velocity and specific force output by the IMU at time τ, respectively. The Jacobian matrix is ​​the discretized process model. Let be the discrete posterior covariance matrix at time k-1. This is the discretized IMU process noise covariance matrix;

[0030] Step S3: Correct the inertial navigation accumulated error using the absolute position information provided by USBL, specifically including:

[0031] a. Establish the observation equation: ,in This is the absolute position of the USBL output. For the observation function, USBL observation noise;

[0032] b. Calculate the Kalman gain: ,in To observe the Jacobian matrix, USBL observation noise covariance;

[0033] c. State correction and covariance update: = ,in Let k be the posterior state at time k. Let k be the posterior covariance. It is the identity matrix;

[0034] Step S4: Construct a spiral cleaning trajectory based on the fused positioning results. The trajectory model is as follows:

[0035]

[0036] in, Where is the helix radius. For pitch, Angular velocity of rotation , , The coordinates of the starting point of the spiral;

[0037] Step S5: Output the fusion positioning result and control the ship cleaning robot to complete the underwater ship cleaning operation along the spiral cleaning trajectory.

[0038] The beneficial effects of this invention are as follows:

[0039] 1. Existing technologies in the field of ship cleaning cannot perform precise path planning and lack sensor technologies such as collision avoidance sonar, inertial navigation, and ultra-short baseline, resulting in robots operating in a semi-intelligent state and relying on human intervention. This invention features autonomous navigation and obstacle avoidance technology, avoiding collisions with obstacles and significantly improving cleaning efficiency; it also reduces the need for human intervention by intelligently planning the optimal cleaning route through sensor scanning of the environment, avoiding repeated cleaning and missed areas, achieving a coverage rate of over 90%.

[0040] 2. Existing ship cleaning methods cannot guarantee that the hull will not be damaged, or that the cleaning will not be thorough. This invention utilizes cavitation jet technology, where the impact force of cavitation bubbles generated by a high-pressure water jet powerfully removes stubborn dirt such as shellfish and algae adhering to the hull. This technology causes minimal damage to the hull paint, avoiding scratches and corrosion that may be caused by traditional mechanical cleaning. Attached Figure Description

[0041] Figure 1 This is a top view of the present invention;

[0042] Figure 2 This is the front view of the present invention;

[0043] Figure 3 This is the left view of the present invention;

[0044] Figure 4 A schematic diagram showing the combination of a ship cleaning robot with a cavitation main engine and a line take-up / delay device;

[0045] In the diagram: 1-Main frame, 2-Buoyancy material, 3-Thruster, 4-Track mechanism, 5-Cavitation jet device, 6-Power compartment, 7-Control compartment, 8-Inertial navigation compartment, 9-Ultra-short baseline, 10-Altimeter, 11-Depth gauge, 12-Camera, 13-Lighting light, 14-Laser scale, 15-Forward-looking sonar, 16-Wire retraction and deployment device. Detailed Implementation

[0046] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading the present invention, any modifications of the present invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0047] Specific implementation method one: Combining Figures 1-4 This embodiment describes a ship cleaning robot using a cavitation jet device, comprising a main frame 1, a buoyancy material 2, a propeller 3, a track mechanism 4, and a cavitation jet device 5. The buoyancy material 2 is installed on the top of the main frame 1 and is connected to the main frame 1. It generates positive buoyancy through its low density characteristics, which balances the robot's own weight, the load of the cleaning equipment, and the density of seawater, allowing the robot to maintain a neutral buoyancy state at a specific depth.

[0048] The number of thrusters 3 is 8, of which 4 thrusters 3 are arranged vertically on the main frame 1, and the other 4 thrusters 3 are arranged horizontally at 45° on the main frame 1. The vertical thrusters provide thrust for buoyancy and descent, and the horizontal thrusters provide thrust for forward movement and turning. The 45° arrangement can adjust the attitude more smoothly and avoid deviation from the cleaning path due to water flow or hull curvature, ensuring that the cleaning tool always maintains the best contact angle with the hull surface.

[0049] The track mechanism 4 is symmetrically fixed on both sides of the main frame 1, providing driving force for the robot to walk on the hull. The track mechanism 4 is equipped with a drive motor, and its track is made of rubber. The track mechanism 4 has a certain buffering effect. The track design enables the ROV to easily move forward, backward, turn, and even rotate in place, flexibly cope with the curved surface and weld seams of the hull, and thoroughly clean dead corners.

[0050] The cavitation jet device 5 includes a cavitation disc and a cavitation main unit. The cavitation disc is fixed at the front of the main frame 1, and the cavitation main unit is placed on the flat ground of the shore. The cavitation main unit and the cavitation disc are connected by a high-pressure pipe. The cavitation jet technology forms cavitation bubbles locally through high-pressure water flow. When these bubbles collapse, they release micro-jet and shock wave, which can instantly peel off stubborn stains without damaging the ship's hull substrate.

[0051] An inertial navigation cabin 8 and an ultra-short baseline 9 are also installed on the main frame 1. The inertial navigation cabin 8 is fixed to the port side of the midship section of the main frame 1, and the ultra-short baseline 9 is located to the starboard side of the midship section of the main frame 1. The inertial navigation cabin 8, fixed to the port side of the midship section of the main frame, corrects inertial drift through camera feature point matching, provides absolute position correction using the ultra-short baseline 9 system, and improves planar motion accuracy by fusing track wheel speed information. This fusion strategy reduces long-term positioning errors by more than 60%, ensuring consistency between cleaning path planning and execution. In strong current environments, the rapid response characteristics of the inertial navigation cabin 8 can compensate for attitude deviations caused by water flow disturbances, maintaining a constant distance between the cavitation jet nozzle and the hull.

[0052] The ultra-short baseline 9 is positioned on the starboard side of the midships section of the main frame 1. It is required to be positioned upwards. Through phase difference measurement technology, the positioning error is less than 0.5% of the slant distance within a range of 100 meters, ensuring that the robot can accurately conform to the curved surface of the hull. Under conditions of strong current or ship swaying, the rapid response characteristics of the ultra-short baseline can compensate for the attitude deviation caused by water flow disturbance and maintain a constant distance between the cavitation jet nozzle and the hull.

[0053] The main frame 1 is equipped with a power compartment 6 and a control compartment 7. The power compartment 6 is fixed at the rear upper part of the main frame 1 and provides power output for the thrusters 3 and track mechanism 4, as well as the energy consumption of the lighting and camera equipment. The power is dynamically distributed through the intelligent power distribution module to ensure that the equipment can still maintain stable operation in a strong current environment. When working on the surface of a complex ship, the power compartment needs to respond to control commands in real time and coordinate multiple thrusters to generate vector thrust, so that the robot has the ability to hover, pitch, roll and other movements. Its control accuracy directly affects the planning and execution effect of the cleaning path.

[0054] The control cabin 7 is located directly below the power compartment 6. The integrated display system in the control cabin 7 can simultaneously process multimodal information from the robot, including sonar scans, high-definition camera data, and IMU inertial data. In ship cleaning scenarios, operators can simultaneously monitor the heat map of hull attachment distribution, cavitation jet pressure parameters, track adhesion status, and environmental flow velocity warnings through the control cabin interface.

[0055] In addition, the robot is equipped with an altimeter 10, a depth gauge 11, a camera 12, a lighting lamp 13, a laser ruler 14, a forward-looking sonar 15, and a cable reeling and laying device 16.

[0056] The altimeter 10 is positioned below the midship of the main frame 1. It must be positioned downwards. The altimeter is mainly responsible for measuring the vertical distance between the robot and the hull surface to ensure that the cleaning tool maintains a constant gap with the hull and avoids collisions or incomplete cleaning. It is a key sensor to ensure the safety and effectiveness of the cleaning operation.

[0057] The depth gauge 11 is installed inside the control compartment 7, with its water inlet transducer located on the outside. When the robot is cleaning the bottom of the hull, it ensures that it always stays within the predetermined cleaning depth range, which is crucial for planning the cleaning path and avoiding collisions with the hull or other obstacles.

[0058] Camera 12 is deployed at the front, middle and rear of the robot. Camera 12 can identify the distribution of biological attachments, the degree of coating damage and structural anomalies on the surface of the hull, and provide visual basis for cleaning path planning. Camera 12 is integrated with sensors such as inertial navigation cabin 8 and laser ruler 14 to realize the robot's autonomous navigation and dynamic obstacle avoidance.

[0059] One light fixture 13 is placed at the front, middle and rear of the robot, laying the foundation for robot operation, image acquisition and quality monitoring, and significantly improving the accuracy, safety and environmental adaptability of cleaning operations;

[0060] The laser ruler 14 is positioned directly in front of the robot's main frame 1. When the robot is cleaning the curved surface of the hull, the laser ruler 14 can provide real-time feedback on the distance between the nozzle and the hull, ensuring that the cavitation jet is always within the optimal range of action and avoiding uneven cleaning or coating damage caused by distance deviation.

[0061] The forward-looking sonar 15 is positioned directly in front of the robot's main frame 1. By emitting sound waves and receiving reflected signals, the forward-looking sonar 15 constructs a three-dimensional spatial situation map in front of the robot in real time, providing the robot with high-precision obstacle detection and autonomous navigation capabilities, which can significantly improve the safety of operation and cleaning efficiency in complex ship environments.

[0062] The cable take-up and take-down device 16 transmits electrical energy stably to the robot power compartment 6 through the power conductor in the umbilical cable, ensuring that the cleaning equipment can work continuously in high-pressure, high-flow mode.

[0063] Specifically, the combination of the inertial navigation cabin 8 (equipped with an inertial measurement unit, IMU) and the ultra-short baseline 9 enables the robot to accurately reach the designated position on the hull and maintain a stable attitude, ensuring that the cleaning tools maintain constant contact pressure with the hull surface and improving cleaning uniformity. In dynamic environments such as waves and currents, the autonomous navigation capability of the inertial navigation cabin 8 combined with the absolute positioning of the ultra-short baseline 9 enables the robot to resist interference from ship movement. Stable positioning and navigation enable the robot to autonomously plan cleaning paths, covering hard-to-reach areas such as the bottom and sides of the hull, improving the comprehensiveness and efficiency of cleaning.

[0064] Specifically, the power system combination of the thruster 3 and the track mechanism 4 provides precise motion control through the thruster and a stable working platform through the track. The thruster adjusts the robot's posture to maintain the optimal distance and angle from the hull; the track ensures that the robot does not slip or shift during the cleaning process. This combination allows the robot to operate stably in complex underwater environments, achieving efficient and comprehensive cleaning results.

[0065] As can be seen from the above technical solutions, the automatic ship cleaning tracked robot with cavitation jet equipment provided by the above embodiments of the present invention utilizes multi-sensor fusion and two power combinations to provide the robot with an efficient cleaning method; through the mutual fusion of sensors such as inertial navigation 8, ultra-short baseline 9, altimeter 10, and depth gauge 11, the robot provides a highly efficient operation mode for ship cleaning. In addition, the power system combination of propeller 3 and track mechanism 4 provides strong support for multi-sensor operation, which can maximize the cleaning efficiency of cavitation jet device.

[0066] Specific Implementation Method Two: This implementation method provides a cleaning path planning method for a ship cleaning robot using cavitation jet equipment. It is based on the Extended Kalman Filter (EFK) algorithm framework to achieve data fusion between the inertial navigation system and the ultra-short baseline positioning system, and includes the following steps:

[0067] Step S1: Construct a six-degree-of-freedom state vector in the geodetic coordinate system. The state vector is represented as follows:

[0068]

[0069] Where x, y, and z are the robot's absolute positions in the geodetic coordinate system; , , The linear velocity in the geodetic coordinate system; , , The robot's roll, pitch, and yaw angles are defined; the geodetic coordinate system is an East-North-Sky coordinate system (x-axis points east, y-axis points north, z-axis points to the zenith), and the robot's coordinate system adopts the carrier coordinate system (b-system), using a direction cosine matrix. The transformation between the carrier coordinate system and the geodetic coordinate system is achieved in the following form:

[0070] In the formula, A unit quaternion to describe the robot's attitude (roll, pitch, yaw). Let be the real part of the quaternion. , , The imaginary part of the quaternion corresponds to the rotational components of the carrier coordinate system about the x-axis, y-axis, and z-axis of the geodetic coordinate system, respectively.

[0071] Step S2: Utilize the load system ratio output by the inertial measurement unit. and angular velocity The prior state at time k is obtained through dead reckoning. Specifically, it includes:

[0072] a. Pose update: using unit quaternions The attitude is described, and attitude updates are achieved through quaternion differential equations:

[0073]

[0074] in, For attitude change rate, For IMU gyroscope drift, To correct the angular velocity of the afterload system (subtracting IMU gyroscope drift). The expression for the adjoint matrix of the corrected angular velocity quaternion is as follows:

[0075]

[0076] in, , , Let be the components of the angular velocity along the x, y, and z axes of the loading system;

[0077] The roll angle, pitch angle, and yaw angle obtained through quaternion conversion are obtained using the following formulas:

[0078]

[0079] b. Velocity Update: Converting the specific force of the load system into the Earth's acceleration, and subtracting the influence of gravity, the velocity update equation is:

[0080]

[0081] in, The direction cosine matrix is ​​derived from quaternions. This is the gravity vector of the Earth system. Zero bias for IMU accelerometer;

[0082] c. Position update: The position is obtained through velocity integration. The position update equation is: , ;

[0083] d. Extended Kalman Filter Prediction Step Discretization: Discretize the continuous quantities of the inertial measurement unit to obtain the prior state and prior covariance at time k.

[0084]

[0085]

[0086] in, Let τ represent the posterior state at time k-1. arrive Any consecutive moments between these points, i.e., time points within the discrete sampling interval. This is a continuous-time state estimate at time τ, including position, velocity, and attitude at time τ. This represents the IMU input vector at time τ. , These represent the angular velocity and specific force output by the IMU at time τ, respectively. The Jacobian matrix of the discretized process model. Let be the discrete posterior covariance matrix at time k-1. This is the discretized IMU process noise covariance matrix;

[0087] Step S3: Correct the inertial navigation accumulated error using the absolute position information provided by USBL, specifically including:

[0088] a. Establish the observation equation: ,in This is the absolute position of the USBL output. For the observation function, USBL observation noise;

[0089] The specific form is as follows:

[0090] b. Calculate the Kalman gain: ,in To observe the Jacobian matrix, USBL observation noise covariance;

[0091] c. State correction and covariance update: = ,in Let k be the posterior state at time k. Let k be the posterior covariance. It is the identity matrix;

[0092] d. Robust handling: To address settlement delays in USBL data, timestamp alignment compensation is employed. The compensation formula is as follows: Where d is the delay step, and when the USBL signal is blocked by the hull, the fault is determined by observing the residual threshold. The fault determination formula is: , As the residual threshold, under fault conditions, EKF degenerates into pure inertial navigation mode until the USBL signal is recovered and the observation update is restarted. In this step, a robust fusion mechanism is constructed to address adverse conditions such as underwater USBL signal delay and hull obstruction, ensuring the navigation system remains effective and avoiding trajectory deviation caused by delay.

[0093] Step S4: Construct a spiral cleaning trajectory based on the fused positioning results. The trajectory model is as follows:

[0094]

[0095] in, Where is the helix radius. For pitch, Angular velocity of rotation , , The coordinates of the spiral starting point are used; the spiral cleaning trajectory constructed by this method can achieve adaptive adjustment of the spiral radius with the ship's taper, and the spiral pitch and angular velocity can be flexibly matched with cleaning requirements. It is suitable for ships of different tonnages and line types (such as bulk carriers, tankers, and container ships), without the need to redesign the cleaning trajectory template, and has extremely strong versatility.

[0096] Step S5: Output the fusion positioning result and control the ship cleaning robot to complete the underwater ship cleaning operation along the spiral cleaning trajectory.

[0097] In addition to the steps mentioned above, the navigation performance is also evaluated. Specifically, the fused navigation effect is quantified using root mean square error, maximum absolute error, and error accumulation rate. The specific formula is as follows:

[0098] (1) Root Mean Square Error (RMSE):

[0099] In the formula: N is the total number of data points for performance evaluation, that is, the number of continuously collected positioning result samples (e.g., in a 10-minute cleaning operation, calculated based on an IMU acquisition rate of 100Hz, N=60000). , , This represents the position component of the fused navigation output at time k, which is the corrected posterior state in step S3. , , The true position component of the robot at time k is represented by two methods: (1) the theoretical position of the spiral cleaning trajectory planning in step S4 under simulation scenario, and (2) the position data output by a high-precision reference positioning system (such as DGPS + acoustic beacon) under actual measurement scenario, which serves as the benchmark for positioning accuracy.

[0100] By taking the square root of the sum of the squares of the position errors of all data points (the difference between the fused positioning result and the actual position), the impact of larger errors is highlighted, and the overall deviation level of all samples can be reflected simultaneously. The smaller the RMSE value, the higher the overall positioning accuracy of the fused navigation. By comparing the RMSE of fused navigation (EKF-INS / USBL) with that of pure inertial navigation (IMU dead reckoning only), if the RMSE after fusion is reduced by ≥80% (e.g., pure inertial navigation RMSE=1.0m, fused RMSE≤0.2m), it is proven that data fusion effectively improves the overall positioning accuracy and meets the accuracy requirements of "full coverage and no omissions" for ship cleaning.

[0101] (2) Maximum absolute error (MAE):

[0102]

[0103] By utilizing the maximum absolute error, the maximum position error among all data points is taken, focusing on the positioning deviation under the "worst case" to avoid collisions with the hull or omissions in cleaning operations due to individual extreme errors.

[0104] (3) Error accumulation rate

[0105] Through calculation to The rate of change of the root mean square error over a time period can reflect the rate of accumulation of positioning error over time. The smaller the absolute value of the error accumulation rate, the better the long-term stability of the positioning accuracy. This proves that the fusion algorithm can effectively suppress the accumulated error of inertial navigation and meet the accuracy requirements of long-term continuous cleaning operations.

[0106] This implementation method achieves closed-loop positioning with high-frequency calculation and low-frequency correction by fusing data from the inertial navigation system and the ultra-short baseline positioning system under the EFK framework. This significantly reduces the error accumulation problem of pure inertial navigation and adaptively forms a spiral cleaning trajectory model that can accurately fit the curved surface of the ship, solving the industry pain points of incomplete coverage and repeated cleaning caused by traditional fixed trajectories.

[0107] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A cleaning path planning method for a ship cleaning robot employing cavitation jet equipment, characterized in that, Includes the following steps: Step S1: Construct a six-degree-of-freedom state vector in the geodetic coordinate system. The state vector is represented as follows: Where x, y, and z are the robot's absolute positions in the geodetic coordinate system; , , The linear velocity in the geodetic coordinate system; , , For the robot's roll angle, pitch angle, and yaw angle; Step S2: Utilize the load system ratio output by the inertial measurement unit. and angular velocity The prior state at time k is obtained through dead reckoning. Specifically, it includes: a. Pose update: using unit quaternions The attitude is described, and attitude updates are achieved through quaternion differential equations: in, For attitude change rate, For inertial measurement unit gyroscope drift, To correct the angular velocity of the subsequent load system, The expression for the adjoint matrix of the corrected angular velocity quaternion is as follows: in, , , Let be the components of the angular velocity along the x, y, and z axes of the loading system; b. Velocity Update: Converting the specific force of the load system into the Earth's acceleration, and subtracting the influence of gravity, the velocity update equation is: in, The direction cosine matrix is ​​derived from quaternions. This is the gravity vector of the Earth system. Zero bias for the accelerometer of the inertial measurement unit; c. Position Update: The position is obtained through velocity integration. The position update equation is: , ; d. Extended Kalman Filter Prediction Step Discretization: Discretize the continuous quantities of the inertial measurement unit to obtain the prior state and prior covariance at time k: in, Let τ represent the posterior state at time k-1. arrive Any consecutive moments between these points, i.e., time points within the discrete sampling interval. This is a continuous-time state estimate at time τ, including position, velocity, and attitude at time τ. This represents the inertial measurement unit input vector at time τ. , These represent the angular velocity and specific force output by the IMU at time τ, respectively. The Jacobian matrix of the discretized process model. Let be the discrete posterior covariance matrix at time k-1. This is the discretized process noise covariance matrix of the inertial measurement unit; Step S3: Correct the inertial navigation accumulated error using the absolute position information provided by USBL, specifically including: a. Establish the observation equation: ,in This is the absolute position of the USBL output. For the observation function, USBL observation noise; b. Calculate the Kalman gain: ,in To observe the Jacobian matrix, USBL observation noise covariance; c. State correction and covariance update: = ,in Let k be the posterior state at time k. Let k be the posterior covariance. It is the identity matrix; Step S4: Construct a spiral cleaning trajectory based on the fused positioning results. The trajectory model is as follows: in, Where is the helix radius. For pitch, Angular velocity of rotation , , The coordinates of the starting point of the spiral; Step S5: Output the fusion positioning result and control the ship cleaning robot to complete the underwater ship cleaning operation along the spiral cleaning trajectory.

Images