Underwater facility generative cavitation cleaning method and system based on diffusion model

Abstract

The application belongs to the technical field of underwater operation and maintenance process, and discloses an underwater facility generative cavitation cleaning method and system based on a diffusion model. The method comprises the following steps: obtaining sonar images, body attitudes and water depth data of the underwater facility; inputting the data into a conditional diffusion model to obtain a cooperative control action sequence, wherein the sequence comprises propeller thrust instructions and cavitation jet instructions, and the feedforward compensation thrust in the thrust instructions is ahead of the jet opening instruction in time; first, the propeller is controlled to output compensation thrust according to the feedforward compensation thrust instruction, and then the cavitation jet device is controlled to be opened; during the jet process, the real-time output thrust of the propeller is monitored, and if it is less than the recoil force estimated according to the current jet parameter, the jet parameter is reduced. The above-mentioned method generates cooperative sequence actions through a diffusion model, combines feedforward thrust pre-compensation and closed-loop safety clamping, realizes active inhibition of recoil force and self-adaptive cleaning of materials, and effectively improves the stability of underwater operation and the cleaning quality.

Description

Technical Field

[0001] This invention relates to the field of underwater operation and maintenance technology, and in particular to a method and system for generating cavitation cleaning of underwater facilities based on a diffusion model. Background Technology

[0002] Underwater structures (such as gates, trash racks, and turbine runners) are constantly submerged in water, and their surfaces are prone to the accumulation of mussels, algae, and calcareous deposits of varying thicknesses, as well as cavitation pits. Existing cleaning methods mainly rely on manual diving or constant-speed cavitation jet cleaning, which has the following drawbacks: First, constant-speed cleaning cannot simultaneously ensure thorough cleaning and protect the anti-corrosion coating; thick, hard dirt is not completely removed, while thin, soft dirt can easily damage the substrate. Second, the recoil force generated by the high-pressure jet can push the robot away from the wall surface. Traditional PID control is a passive response and cannot compensate in advance at the moment the jet is activated, resulting in excessive displacement of the robot or even a fall.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide a generative cavitation cleaning method and system for underwater facilities based on a diffusion model, aiming to solve the technical problem of how to achieve stable, accurate and adaptive generative cavitation cleaning in complex underwater environments.

[0005] To achieve the above objectives, the present invention provides a diffusion-based method for generating cavitation cleaning of underwater facilities, comprising the following steps: Acquire current environmental status data of underwater facilities, wherein the current environmental status data includes at least sonar images, fuselage attitude and water depth data; The current environmental state data is input into the conditional diffusion model, and the cooperative control action sequence output by the conditional diffusion model is obtained. The cooperative control action sequence includes a thruster thrust command sequence and a cavitation jet command sequence. The thruster thrust command sequence and the cavitation jet command sequence have a preset timing phase difference on the time axis. The feedforward compensation thrust command included in the thruster thrust command sequence is ahead of the jet opening command included in the cavitation jet command sequence in time. Before performing cavitation jet cleaning, the thruster outputs compensating thrust according to the feedforward compensating thrust command; After the thruster outputs compensated thrust, the cavitation jet device is controlled to open according to the jet opening command; During the process of controlling the cavitation jet device to start, the real-time output thrust of the propeller is monitored. If the real-time output thrust is less than the recoil force estimated based on the current jet parameters of the cavitation jet device, the current jet parameters are reduced.

[0006] In one embodiment, the formula corresponding to the reverse generation process of the conditional diffusion model is:

[0007] Among them, A t 1 represents the action sequence obtained after denoising, A t Given the current noisy action sequence, α t To preserve the scaling factor of the original signal, t For cumulative product, θ The noise term is the one predicted by the network, t is the current diffusion time, O represents the observation conditions, and σ is the noise term predicted by the network. t Let z be the standard deviation of the random noise at step t, and z be the standard Gaussian random noise.

[0008] In one embodiment, the method further includes: Based on the preset cavitation jet pressure and nozzle area, the recoil force generated at the moment the jet opens is estimated, and the calculation formula is as follows: in, The nozzle flow coefficient is... For nozzle area, The preset cavitation jet pressure is used; the required compensating thrust of the propeller is determined based on the recoil force, wherein the dynamic equilibrium condition satisfied by the recoil force and the compensating thrust is:

[0009] in, To compensate for the thrust, For adsorption force, For recoil force, For buoyancy, For water resistance, This is for the safety factor.

[0010] In one embodiment, the method further includes: Obtain the current water depth and calculate the cavitation number based on the current water depth. The calculation formula is as follows:

[0011] in, For emptying, For environmental absolute pressure, This is the saturated vapor pressure of water. The density of water, It is the acceleration due to gravity. Because of the water depth, The jet exit velocity; The target jet velocity is calculated based on the cavitation number, and the jet velocity after the cavitation jet device is turned on is adjusted to the target jet velocity. The calculation formula is as follows:

[0012] in, The target jet velocity.

[0013] In one embodiment, the process of generating a cooperative control action sequence using the conditional diffusion model includes: Based on the sonar images, the thickness and hardness distribution of the surface deposits of underwater facilities can be identified; Cavitation jet parameters are generated based on the thickness and hardness distribution of the deposits. For deposits with a thickness exceeding a preset threshold and a hardness exceeding a preset hardness, a high-pressure, low-speed pulse stripping strategy is generated. For deposits with a thickness below the preset threshold and a hardness below the preset hardness, a low-pressure, fast sweeping strategy is generated.

[0014] In one embodiment, the method further includes: Acquire sonar echo signals from the surface of the underwater facility after cleaning, and determine whether there are any residual areas based on the sonar echo signals; If a residual area exists, a backwashing trajectory is generated using the conditional diffusion model based on the location and shape of the residual area, and the cavitation jet device is controlled to perform secondary cleaning on the residual area based on the backwashing trajectory.

[0015] In one embodiment, the method further includes: Obtain data on the distribution of cavitation pits on the surface of underwater facilities; The cavitation pit distribution data is input into the conditional diffusion model to generate a force-controlled grinding trajectory; The grinding device is controlled to grind the cavitation pit area according to the grinding trajectory, and the buoyancy of the water is automatically compensated based on the water depth data during the grinding process.

[0016] Furthermore, to achieve the above objectives, this invention also proposes a diffusion-based underwater facility generating cavitation cleaning system, wherein the diffusion-based underwater facility generating cavitation cleaning system includes: An environmental state perception module is used to acquire current environmental state data of underwater facilities. The current environmental state data includes at least sonar images, fuselage attitude, and water depth data. A generative cooperative control module is used to input the current environmental state data into a conditional diffusion model and obtain a cooperative control action sequence output by the conditional diffusion model. The cooperative control action sequence includes a thruster command sequence and a cavitation jet command sequence. The thruster command sequence and the cavitation jet command sequence have a preset timing phase difference on the time axis. The feedforward compensation thrust command included in the thruster command sequence is ahead of the jet opening command included in the cavitation jet command sequence in time. The feedforward thrust execution module is used to control the thruster to output compensating thrust according to the feedforward compensating thrust command before performing cavitation jet cleaning; The jet timing control module is used to control the cavitation jet device to open according to the jet opening command after the thruster outputs compensation thrust; An adaptive adjustment module is used to monitor the real-time output thrust of the propeller during the process of controlling the cavitation jet device to start. If the real-time output thrust is less than the recoil force estimated based on the current jet parameters of the cavitation jet device, the current jet parameters are reduced.

[0017] Furthermore, to achieve the above objectives, the present invention also proposes an underwater facility generating cavitation cleaning device based on a diffusion model. The underwater facility generating cavitation cleaning device based on a diffusion model includes: a memory, a processor, and an underwater facility generating cavitation cleaning program based on a diffusion model stored in the memory and executable on the processor. The underwater facility generating cavitation cleaning program based on a diffusion model is configured to implement the steps of the underwater facility generating cavitation cleaning method based on a diffusion model as described above.

[0018] Furthermore, to achieve the above objectives, the present invention also proposes a storage medium storing a diffusion-based underwater facility generating cavitation cleaning program, wherein when the diffusion-based underwater facility generating cavitation cleaning program is executed by a processor, it implements the steps of the diffusion-based underwater facility generating cavitation cleaning method described above.

[0019] This invention acquires sonar images, fuselage attitude, and water depth data of underwater facilities. The data is input into a conditional diffusion model to obtain a cooperative control action sequence. This sequence includes thruster thrust commands and cavitation jet commands, with the feedforward compensation thrust in the thrust command preceding the jet activation command in time. First, the thruster outputs compensation thrust according to the feedforward compensation thrust command, and then the cavitation jet device is activated. During the jetting process, the real-time output thrust of the thruster is monitored; if it is less than the recoil force estimated based on the current jet parameters, the jet parameters are reduced. This method generates cooperative timing actions through a diffusion model, combined with feedforward thrust pre-compensation and closed-loop safety clamping, achieving active recoil force suppression and material adaptive cleaning, effectively improving the stability and cleaning quality of underwater operations. Attached Figure Description

[0020] Figure 1 This is a schematic flowchart of the first embodiment of the underwater facility generative cavitation cleaning method based on a diffusion model according to the present invention. Figure 2 This is a structural block diagram of the first embodiment of the underwater facility generating cavitation cleaning device based on the diffusion model of the present invention.

[0021] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0023] This invention provides a method for generating cavitation cleaning of underwater facilities based on a diffusion model, referring to... Figure 1 , Figure 1 This is a schematic flowchart of the first embodiment of the underwater facility generation cavitation cleaning method based on a diffusion model according to the present invention.

[0024] In this embodiment, the underwater facility generative cavitation cleaning method based on a diffusion model includes the following steps: Step S10: Obtain the current environmental status data of the underwater facility.

[0025] In this embodiment, the executing entity is an underwater facility generating cavitation cleaning device based on a diffusion model. This underwater facility generating cavitation cleaning device based on a diffusion model has functions such as data processing, data communication, and program execution. The underwater facility generating cavitation cleaning device based on a diffusion model can be a computer terminal device or other network device, or other devices with similar functions. This embodiment does not limit the scope of the application.

[0026] It should be noted that underwater facilities (such as gates, trash racks, and turbine runners) are submerged in water for extended periods, making their surfaces prone to the accumulation of mussels, algae, and calcareous deposits of varying thicknesses, as well as cavitation pits. Existing cleaning methods mainly rely on manual diving or constant-speed cavitation jet cleaning, which has the following drawbacks: First, constant-speed cleaning struggles to balance thorough cleaning with the protection of anti-corrosion coatings; thick, hard dirt is not completely removed, while thin, soft dirt can easily damage the substrate. Second, the recoil force generated by the high-pressure jet can push the robot away from the wall surface; traditional PID control is a passive response and cannot compensate in advance at the moment the jet is activated, leading to excessive displacement of the robot or even a fall.

[0027] To address the aforementioned technical issues, this embodiment acquires sonar images, fuselage attitude, and water depth data of the underwater facility. This data is then input into a conditional diffusion model to obtain a cooperative control action sequence. This sequence includes thruster commands and cavitation jet commands, with the feedforward compensation thrust in the thrust command preceding the jet activation command in time. First, the thruster outputs compensation thrust according to the feedforward compensation thrust command, and then the cavitation jet device is activated. During the jetting process, the real-time output thrust of the thruster is monitored; if it is less than the recoil force estimated based on the current jet parameters, the jet parameters are reduced. This method generates cooperative timing actions through a diffusion model, combined with feedforward thrust pre-compensation and closed-loop safety clamping, achieving active recoil force suppression and material adaptive cleaning, effectively improving the stability and cleaning quality of underwater operations. Specifically, it can be implemented as follows.

[0028] In this embodiment, the current environmental status data includes at least sonar images, fuselage attitude, and water depth data. The sonar images, for example, utilize high-frequency imaging sonar (center frequency 1.2MHz, beam angle 0.5°×20°) to scan the surface of the debris barrier at a rate of 5 frames per second. The sonar echo data is processed to form a two-dimensional grayscale image, where the grayscale value represents the echo intensity. Image segmentation algorithms identify that the barrier surface is covered with a mixed debris layer approximately 3cm thick, exhibiting significant echo attenuation (intensity reduced by 60% compared to bare steel) and localized irregular protrusions (mussel clusters). Further echo texture analysis determines the debris to be of high hardness (calcification degree >80%), thus providing a basis for generating a "high-pressure, low-speed pulsed stripping" strategy for the diffusion model. The fuselage attitude is acquired by a nine-axis inertial measurement unit (IMU, including a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer) mounted on the robot, outputting the attitude angle relative to the wall in real time. For example, when adhering to the vertical wall of the trash rack, the following measurements were taken: pitch angle +2° (slight upward tilt of the robot's nose), roll angle -1.5° (slightly low to the left), yaw angle 0.5° (pointing towards the target), and the distance sensor (laser rangefinder) measured the distance between the robot's body plane and the rack surface as 5.2 cm. Attitude data is used to determine whether the robot is in a stable adhering state. If the pitch angle exceeds ±10° or the roll angle exceeds ±8°, the system determines there is a risk of falling and triggers the safety clamping logic. Water depth data can be obtained by measuring the hydrostatic pressure using a pressure sensor and then calculating the water depth.

[0029] Step S20: Input the current environmental state data into the conditional diffusion model and obtain the cooperative control action sequence output by the conditional diffusion model.

[0030] In this embodiment, the coordinated control action sequence includes a thruster command sequence and a cavitation jet command sequence. The thruster command sequence and the cavitation jet command sequence have a preset timing phase difference on the time axis. The feedforward compensation thrust command included in the thruster command sequence is ahead of the jet opening command included in the cavitation jet command sequence in time.

[0031] It should be noted that in this embodiment, the feedforward compensation thrust command precedes the jet activation command in time. At the execution level, the compensation thrust is first output to "press" the robot against the wall. Once the thrust stabilizes, the jet is activated, and the relationship between thrust and recoil is monitored in real time during the jet process. If the thrust is insufficient, the jet parameters (pressure or flow rate) are immediately reduced, thus forming a dual stabilization mechanism of "feedforward pre-compensation + closed-loop safety clamping." This method can control the robot's displacement at the moment of jet activation to within 1mm, far superior to the displacement of over 10mm achieved by traditional PID control. The preset timing phase difference is 50ms, but it can be adaptively adjusted according to actual needs; this embodiment does not impose any limitations on this.

[0032] In one embodiment, the conditional diffusion model adopts a denoised diffusion probability model architecture. The forward process progressively adds Gaussian noise to the action sequence, while the reverse process (inference process) progressively denoises and recovers the optimal action under given observation conditions (sonar image, fuselage attitude, water depth data). The formula for the reverse generation process is:

[0033] Among them, A t 1 represents the action sequence obtained after denoising, A t Given the current noisy action sequence, α t To preserve the scaling factor of the original signal, t For cumulative product, θ The noise term is the one predicted by the network, t is the current diffusion time, O represents the observation conditions, and σ is the noise term predicted by the network. t Let z be the standard deviation of the random noise at step t, and z be the standard Gaussian random noise.

[0034] In one embodiment, the process of generating a coordinated control action sequence using the conditional diffusion model includes: identifying the thickness and hardness distribution of underwater facility surface deposits based on sonar images; and generating time-varying cavitation jet parameters based on the thickness and hardness of the deposits. For deposits with a thickness exceeding a preset threshold (e.g., 2 cm) and a hardness exceeding a preset hardness (e.g., calcified mussels), a high-pressure, low-speed pulse stripping strategy is generated. This strategy, for example, uses a jet pressure of 20 MPa, a moving speed of 0.05 m / s, and a nozzle distance of 5 cm, switching between 20 MPa and 8 MPa at a frequency of 10 Hz to utilize pulse fatigue effect to shatter the calcified shell. For deposits with a thickness below a preset threshold (e.g., <5 mm) and low hardness (e.g., algae), a low-pressure, rapid sweeping strategy is generated. This strategy, for example, uses a jet pressure of 8 MPa, a moving speed of 0.2 m / s, and continuous scanning. In this embodiment, actual testing shows that this strategy can improve cleaning efficiency by 3 times and reduce the damage rate of the anti-corrosion coating by 90%.

[0035] Step S30: Before performing cavitation jet cleaning, control the thruster to output compensating thrust according to the feedforward compensating thrust command.

[0036] In its implementation, the method also includes recoil force feedforward prediction, which involves estimating the recoil force generated at the moment the jet opens based on preset cavitation jet pressure and nozzle area. The calculation formula is as follows: in, The nozzle flow coefficient is... For nozzle area, The preset cavitation jet pressure is used; the required compensating thrust of the propeller is determined based on the recoil force, wherein the dynamic equilibrium condition satisfied by the recoil force and the compensating thrust is:

[0037] in, To compensate for the thrust, For adsorption force, For recoil force, For buoyancy, For water resistance, This is for the safety factor.

[0038] The above method first calculates the compensation thrust, and then controls the thruster to output compensation thrust according to the feedforward compensation thrust command. This actively suppresses recoil force and greatly reduces the robot's displacement.

[0039] Step S40: After the thruster outputs compensated thrust, the cavitation jet device is activated according to the jet activation command.

[0040] In a specific implementation, this embodiment can also determine the target jet velocity after the cavitation jet device is activated, specifically including: obtaining the current water depth and calculating the cavitation number based on the current water depth, the calculation formula being:

[0041] in, For emptying, For environmental absolute pressure, This is the saturated vapor pressure of water. The density of water, It is the acceleration due to gravity. Because of the water depth, The jet exit velocity; The target jet velocity is calculated based on the cavitation number, and the jet velocity after the cavitation jet device is turned on is adjusted to the target jet velocity. The calculation formula is as follows:

[0042] in, The target jet velocity.

[0043] Step S50: During the process of controlling the cavitation jet device to start, monitor the real-time output thrust of the propeller. If the real-time output thrust is less than the recoil force estimated based on the current jet parameters of the cavitation jet device, then reduce the current jet parameters.

[0044] In one embodiment, the method further includes an adaptive repair step for cavitation pits. For cavitation pits on the surface of turbine runner blades, the system acquires cavitation pit distribution data on the surface of underwater facilities (obtainable via 3D sonar or structured light scanning). The cavitation pit distribution data is input into the conditional diffusion model to generate a force-controlled grinding trajectory. The grinding device is controlled to grind the cavitation pit area according to the grinding trajectory, and during the grinding process, the influence of water buoyancy on the grinding head pressure is automatically compensated based on water depth data. Specifically, for every 10m increase in water depth, the water pressure increases by approximately 0.1MPa, and the influence of buoyancy on the normal force of the grinding head is approximately 0.5N. The model pre-increases the grinding head pressure when generating the grinding trajectory to ensure that the actual grinding depth matches the preset depth. This embodiment was applied to the turbine runner repair of a hydropower station, and the cavitation pit grinding depth error was less than 0.1mm, with a surface roughness Ra ≤ 3.2μm.

[0045] In one embodiment, the safety clamping logic is implemented as follows: during the activation of the cavitation jet device, the actual output thrust of the thruster is monitored in real time by a thrust sensor. If the actual output thrust is less than the recoil force estimated by the current jet parameters of the cavitation jet device, it is determined that the thrust is insufficient. At this time, the system immediately reduces the jet parameters, for example, clamping the jet pressure from 20MPa to 10MPa, or reducing the jet flow rate, until the thrust condition is met. Simultaneously, if the robot tilt angle exceeds a preset threshold (e.g., ±15°), the jet output is automatically cut off and the attitude correction power of the thruster is increased to prevent a fall. In this embodiment, the safety clamping response time is less than 5ms, which is much faster than the main controller software link (typically >50ms), effectively eliminating the physical risks caused by abnormal commands that may be generated by generative control.

[0046] In this embodiment, a piezoelectric hydrophone array with a fish-like lateral line distribution and a flow velocity probe are used to collect flow field data. This data is incrementally modulated and encoded into a pulse sequence. A pre-trained rotational speed noise transfer function filters out the robot's own noise pulses. The data is then input into a pulse neural network to identify turbulent disturbances and instability precursor signals. A hardware bypass bus directly drives the actuator to complete anti-current and anti-fall control. This approach reduces end-to-end control latency, provides high accuracy in instability warnings in turbid water, and enables coordinated control of anti-current disturbances and anti-fall control, avoiding the risk of loss of control and the underwater robot due to sudden changes in the flow field.

[0047] In this embodiment, sonar images, fuselage attitude, and water depth data of the underwater facility are acquired. This data is then input into a conditional diffusion model to obtain a coordinated control action sequence. This sequence includes thruster commands and cavitation jet commands, with the feedforward compensation thrust in the thrust command preceding the jet activation command in time. First, the thruster outputs compensation thrust according to the feedforward compensation thrust command, and then the cavitation jet device is activated. During the jetting process, the real-time output thrust of the thruster is monitored; if it is less than the recoil force estimated based on the current jet parameters, the jet parameters are reduced. This method generates coordinated timing actions through a diffusion model, combined with feedforward thrust pre-compensation and closed-loop safety clamping, achieving active recoil force suppression and material adaptive cleaning, effectively improving the stability and cleaning quality of underwater operations.

[0048] Furthermore, this embodiment of the invention also proposes a storage medium storing a diffusion-based underwater facility generating cavitation cleaning program. When the diffusion-based underwater facility generating cavitation cleaning program is executed by a processor, it implements the steps of the diffusion-based underwater facility generating cavitation cleaning method described above.

[0049] Reference Figure 2 , Figure 2 This is a structural block diagram of the first embodiment of the underwater facility generating cavitation cleaning device based on the diffusion model of the present invention.

[0050] like Figure 2 As shown, the underwater facility generating cavitation cleaning device based on a diffusion model proposed in this embodiment of the invention includes: The environmental state perception module 10 is used to acquire the current environmental state data of the underwater facility, which includes at least sonar images, fuselage attitude and water depth data. Generative cooperative control module 20 is used to input the current environmental state data into the conditional diffusion model and obtain the cooperative control action sequence output by the conditional diffusion model. The cooperative control action sequence includes a thruster command sequence and a cavitation jet command sequence. The thruster command sequence and the cavitation jet command sequence have a preset timing phase difference on the time axis. The feedforward compensation thrust command included in the thruster command sequence is ahead of the jet opening command included in the cavitation jet command sequence in time. The feedforward thrust execution module 30 is used to control the thruster to output compensating thrust according to the feedforward compensating thrust command before performing cavitation jet cleaning; The jet timing control module 40 is used to control the cavitation jet device to open according to the jet opening command after the thruster outputs the compensation thrust. The adaptive adjustment module 50 is used to monitor the real-time output thrust of the propeller during the process of controlling the cavitation jet device to start. If the real-time output thrust is less than the recoil force estimated based on the current jet parameters of the cavitation jet device, the current jet parameters are reduced.

[0051] In this embodiment, sonar images, fuselage attitude, and water depth data of the underwater facility are acquired. This data is then input into a conditional diffusion model to obtain a coordinated control action sequence. This sequence includes thruster commands and cavitation jet commands, with the feedforward compensation thrust in the thrust command preceding the jet activation command in time. First, the thruster outputs compensation thrust according to the feedforward compensation thrust command, and then the cavitation jet device is activated. During the jetting process, the real-time output thrust of the thruster is monitored; if it is less than the recoil force estimated based on the current jet parameters, the jet parameters are reduced. This method generates coordinated timing actions through a diffusion model, combined with feedforward thrust pre-compensation and closed-loop safety clamping, achieving active recoil force suppression and material adaptive cleaning, effectively improving the stability and cleaning quality of underwater operations.

[0052] This application embodiment also provides an underwater facility generating cavitation cleaning device based on a diffusion model, including a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other through the communication bus. The memory is used to store the underwater facility generating cavitation cleaning program based on the diffusion model. When the processor executes the program stored in the memory, it implements the above-mentioned underwater facility generating cavitation cleaning method based on the diffusion model.

[0053] The communication bus mentioned in the above-mentioned diffusion-based underwater facility generating cavitation cleaning equipment can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc.

[0054] The communication interface is used for communication between the aforementioned diffusion-based underwater facility generative cavitation cleaning equipment and other equipment.

[0055] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0056] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0057] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0058] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0059] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0060] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0061] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solutions of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.

[0062] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0063] In addition, for technical details not described in detail in this embodiment, please refer to the underwater facility generation cavitation cleaning method based on diffusion model provided in any embodiment of the present invention, which will not be repeated here.

[0064] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0065] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0066] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0067] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

[0068] It is understood that the system provided in the embodiments of the present invention corresponds to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant content can be referred to the corresponding parts of the above method.

Claims

1. A method for generating cavitation cleaning of underwater facilities based on a diffusion model, characterized in that, The diffusion-model-based underwater facility generative cavitation cleaning method includes: Acquire current environmental status data of underwater facilities, wherein the current environmental status data includes at least sonar images, fuselage attitude and water depth data; The current environmental state data is input into the conditional diffusion model, and the cooperative control action sequence output by the conditional diffusion model is obtained. The cooperative control action sequence includes a thruster thrust command sequence and a cavitation jet command sequence. The thruster thrust command sequence and the cavitation jet command sequence have a preset timing phase difference on the time axis. The feedforward compensation thrust command included in the thruster thrust command sequence is ahead of the jet opening command included in the cavitation jet command sequence in time. Before performing cavitation jet cleaning, the thruster outputs compensating thrust according to the feedforward compensating thrust command; After the thruster outputs compensated thrust, the cavitation jet device is controlled to open according to the jet opening command; During the process of controlling the cavitation jet device to start, the real-time output thrust of the propeller is monitored. If the real-time output thrust is less than the recoil force estimated based on the current jet parameters of the cavitation jet device, the current jet parameters are reduced.

2. The underwater facility generative cavitation cleaning method based on a diffusion model as described in claim 1, characterized in that, The formula corresponding to the reverse generation process of the conditional diffusion model is: Among them, A t 1 represents the action sequence obtained after denoising, A t Given the current noisy action sequence, α t To preserve the scaling factor of the original signal, t For cumulative product, θ The noise term is the one predicted by the network, t is the current diffusion time, O represents the observation conditions, and σ is the noise term predicted by the network. t Let z be the standard deviation of the random noise at step t, and z be the standard Gaussian random noise.

3. The underwater facility generative cavitation cleaning method based on a diffusion model as described in claim 1, characterized in that, The method further includes: Based on the preset cavitation jet pressure and nozzle area, the recoil force generated at the moment the jet opens is estimated, and the calculation formula is as follows: in, The nozzle flow coefficient is... For nozzle area, The preset cavitation jet pressure is used; the required compensating thrust of the propeller is determined based on the recoil force, wherein the dynamic equilibrium condition satisfied by the recoil force and the compensating thrust is: in, To compensate for the thrust, For adsorption force, For recoil force, For buoyancy, For water resistance, This is for the safety factor.

4. The underwater facility generative cavitation cleaning method based on a diffusion model as described in claim 1, characterized in that, The method further includes: Obtain the current water depth and calculate the cavitation number based on the current water depth. The calculation formula is as follows: in, For emptying, For environmental absolute pressure, This is the saturated vapor pressure of water. The density of water, It is the acceleration due to gravity. Because of the water depth, The jet exit velocity; The target jet velocity is calculated based on the cavitation number, and the jet velocity after the cavitation jet device is turned on is adjusted to the target jet velocity. The calculation formula is as follows: in, The target jet velocity.

5. The underwater facility generative cavitation cleaning method based on a diffusion model as described in claim 1, characterized in that, The process by which the conditional diffusion model generates a sequence of cooperative control actions includes: Based on the sonar images, the thickness and hardness distribution of the surface deposits of underwater facilities can be identified; Cavitation jet parameters are generated based on the thickness and hardness distribution of the deposits. For deposits with a thickness exceeding a preset threshold and a hardness exceeding a preset hardness, a high-pressure, low-speed pulse stripping strategy is generated. For deposits with a thickness below the preset threshold and a hardness below the preset hardness, a low-pressure, fast sweeping strategy is generated.

6. The underwater facility generative cavitation cleaning method based on a diffusion model as described in claim 1, characterized in that, The method further includes: Acquire sonar echo signals from the surface of the underwater facility after cleaning, and determine whether there are any residual areas based on the sonar echo signals; If a residual area exists, a backwashing trajectory is generated using the conditional diffusion model based on the location and shape of the residual area, and the cavitation jet device is controlled to perform secondary cleaning on the residual area based on the backwashing trajectory.

7. The underwater facility generative cavitation cleaning method based on a diffusion model as described in claim 1, characterized in that, The method further includes: Obtain data on the distribution of cavitation pits on the surface of underwater facilities; The cavitation pit distribution data is input into the conditional diffusion model to generate a force-controlled grinding trajectory; The grinding device is controlled to grind the cavitation pit area according to the grinding trajectory, and the buoyancy of the water is automatically compensated based on the water depth data during the grinding process.

8. A diffusion-based underwater facility generating cavitation cleaning device, characterized in that, The diffusion-model-based underwater facility generating cavitation cleaning device is applied to the diffusion-model-based underwater facility generating cavitation cleaning method as described in any one of claims 1 to 7, the device comprising: An environmental state perception module is used to acquire current environmental state data of underwater facilities. The current environmental state data includes at least sonar images, fuselage attitude, and water depth data. A generative cooperative control module is used to input the current environmental state data into a conditional diffusion model and obtain a cooperative control action sequence output by the conditional diffusion model. The cooperative control action sequence includes a thruster command sequence and a cavitation jet command sequence. The thruster command sequence and the cavitation jet command sequence have a preset timing phase difference on the time axis. The feedforward compensation thrust command included in the thruster command sequence is ahead of the jet opening command included in the cavitation jet command sequence in time. The feedforward thrust execution module is used to control the thruster to output compensating thrust according to the feedforward compensating thrust command before performing cavitation jet cleaning; The jet timing control module is used to control the cavitation jet device to open according to the jet opening command after the thruster outputs compensation thrust; An adaptive adjustment module is used to monitor the real-time output thrust of the propeller during the process of controlling the cavitation jet device to start. If the real-time output thrust is less than the recoil force estimated based on the current jet parameters of the cavitation jet device, the current jet parameters are reduced.

9. A diffusion-based underwater facility generating cavitation cleaning device, characterized in that, The diffusion-model-based underwater facility generating cavitation cleaning device includes: a memory, a processor, and a diffusion-model-based underwater facility generating cavitation cleaning program stored in the memory and executable on the processor, wherein the diffusion-model-based underwater facility generating cavitation cleaning program is configured to implement the steps of the diffusion-model-based underwater facility generating cavitation cleaning method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a diffusion-based underwater facility generating cavitation cleaning program, which, when executed by a processor, implements the steps of the diffusion-based underwater facility generating cavitation cleaning method as described in any one of claims 1 to 7.

Images