Self-adaptive control method, system and equipment of underwater resident robot and medium

Abstract

The invention relates to a self-adaptive control method for an underwater resident robot, and the method is applied to the underwater resident robot, and comprises the steps: carrying out the task planning based on a task demand inputted by a user, and generating a navigation path point sequence; determining a target task type according to the navigation path point sequence, and outputting an expected speed and an expected attitude angle based on the target task type; on the basis of the expected speed and the expected attitude angle, generating a control instruction through an adaptive robust PID (Proportion Integration Differentiation) controller; and executing the control instruction and feeding back sensor data to a control platform.

Description

Technical Field

[0001] This application relates to the field of underwater robot control, and in particular to an adaptive control method, system, device and storage medium for an underwater resident robot. Background Technology

[0002] Control technology for underwater robots (AUVs, ROVs) is a core component of marine engineering, involving PID control, model-based control (such as Model Predictive Control, MPC), underwater navigation (such as Inertial Navigation (INS), Doppler Log (DVL), and Underwater Acoustic Positioning (USBL), and mission planning. Traditional PID control adjusts speed and attitude through error feedback, but it is poorly adaptable to ocean current disturbances and model uncertainties. Model-based control relies on accurate dynamic models, but model errors are large in the complex environment of the deep sea. Navigation technology utilizes Kalman filtering to fuse sensor data, but limitations in sonar resolution and visibility often lead to the loss of pipeline targets. Mission planning often adopts a three-layer architecture (task layer, behavior layer, execution layer), but it lacks dynamic mode switching and is difficult to cope with multi-task requirements.

[0003] Therefore, it is necessary to provide an adaptive control method, system, device, and storage medium for an underwater resident robot, so that the underwater robot can cope with multi-task requirements. Summary of the Invention

[0004] This application provides an adaptive control method, system, and storage medium for a residing underwater robot. By employing adaptive robust PID control, a biomimetic pipeline tracking strategy, and a multi-layer control architecture, it addresses the aforementioned issues, thereby improving the control accuracy and robustness of the residing underwater robot (RAUV) in complex underwater environments, adapting to ocean current disturbances and model uncertainties, increasing the success rate of pipeline inspection tasks, resolving target loss issues through search, tracking, and retrieval strategies, achieving efficient switching between multiple task modes, enhancing the system's flexibility in tasks such as cruising, detailed observation, and obstacle avoidance, and strengthening the system's fault detection and emergency response capabilities to ensure high reliability for long-term residing missions.

[0005] In a first aspect, this application provides an adaptive control method for an underwater resident robot, the method being applied to an underwater resident robot, comprising:

[0006] Based on the user's input task requirements, task planning is performed to generate a sequence of navigation path points;

[0007] Based on the navigation path point sequence, determine the target task type, and output the desired velocity and desired attitude angle based on the target task type;

[0008] Based on the desired speed and desired attitude angle, control commands are generated using an adaptive robust PID controller.

[0009] The control commands are executed and sensor data is fed back to the control platform.

[0010] In some embodiments, the method further includes:

[0011] The observation modes of the underwater resident robot are obtained; wherein the observation modes include a cruise mode and a close-range observation mode.

[0012] When the underwater resident robot is in cruise mode, it switches to close-range observation mode in response to detecting that the state of a feature target meets the preset conditions.

[0013] When the underwater resident robot is in close-range observation mode, it switches to cruise mode in response to the failure to detect the feature target within a preset distance range.

[0014] In some embodiments, the step of planning tasks based on user-inputted task requirements and generating a sequence of navigation path points includes:

[0015] Based on the user's input task requirements, the input data for the task planning algorithm is determined; wherein, the input data includes task type, geographical constraints, environmental constraints, time constraints, and target point set;

[0016] The input data is fed into the task planning algorithm to generate the navigation path point sequence.

[0017] In some embodiments, the task planning algorithm is a spatiotemporal constrained path planning algorithm;

[0018] The spatiotemporal constrained path planning algorithm is shown in the following formula:

[0019]

[0020] Where P is the path curve, P i w represents the coordinates of the path point. t w is the path length weighting factor. r R(s) is the risk cost weighting factor, and R(s) is the risk assessment function at position s; the output is the navigation path point sequence P = [P start ,P1,P2,…,P end ].

[0021] In some embodiments, determining the target task type based on the navigation path point sequence includes:

[0022] Based on the aforementioned task requirements, determine the task instructions;

[0023] Based on the task instruction type, determine the candidate task type;

[0024] Acquire underwater environment information around the underwater resident robot, and determine the task priority corresponding to each candidate task type based on the underwater environment information;

[0025] Based on the priority of the task, the target task type is determined from the candidate task types.

[0026] In some embodiments, the candidate task type includes a pipeline tracking type and an obstacle avoidance mode type; under the pipeline tracking type and the obstacle avoidance mode type, the step of outputting the desired velocity and desired attitude angle based on the target task type includes:

[0027] The desired speed is determined using the following formula:

[0028]

[0029] Among them, v d For the desired speed, v base Based on the base speed, e p The lateral positional deviation from the target point is denoted by , and the vertical distance from the pipeline center to the underwater resident robot is denoted by . Let ψ be the rate of change of position deviation. pipe ψ is the direction angle. d k is the desired direction angle. p For proportional gain, k d Where L is the differential gain and L is the forward look distance;

[0030] The final heading correction is determined based on the following formula, and the desired attitude angle is determined based on the current heading and the final heading correction.

[0031]

[0032] Where Δψ is the final heading correction, Δψ i The suggested avoidance direction for the i-th obstacle is... d is the weighting factor. i The distance to the obstacle.

[0033] In some embodiments, the generation of control commands by an adaptive robust PID controller based on the desired velocity and desired attitude angle includes:

[0034] Based on the desired speed and the desired attitude angle, the attitude deviation and speed preset value are determined by the outer loop controller of the adaptive robust PID controller;

[0035] Based on the preset speed value, the control command is generated by the inner loop controller of the adaptive robust PID controller;

[0036] Based on the status feedback data after the underwater resident robot executes the control command, the adaptive parameter items of the adaptive robust PID controller are updated.

[0037] Secondly, this application provides an adaptive control system for an underwater resident robot, the system being used to control the underwater resident robot, the system comprising:

[0038] The planning module is used to plan tasks based on user-inputted task requirements and generate a sequence of navigation path points.

[0039] The behavior module is used to determine the target task type based on the navigation path point sequence, and output the desired velocity and desired attitude angle based on the target task type;

[0040] The action module is used to generate control commands based on the desired speed and desired attitude angle using an adaptive robust PID controller;

[0041] The device module is used to execute the control commands and feed back sensor data to the control platform.

[0042] Thirdly, an electronic device is provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0043] Memory, used to store computer programs;

[0044] When a processor executes a program stored in a memory, it implements the steps of the adaptive control method for an underwater resident robot as described in any embodiment of the first aspect.

[0045] Fourthly, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the adaptive control method for an underwater resident robot as described in any embodiment of the first aspect.

[0046] Compared with existing technologies, the technical solutions provided in this application have the following advantages: 1. Improved control accuracy and robustness: Traditional PID control is difficult to cope with ocean current disturbances due to fixed parameters, resulting in large errors. This application adopts adaptive robust PID control, adjusting parameters in real time (e.g., through sliding mode variables and adaptive laws) to improve trajectory and attitude control accuracy in deep-sea environments. 2. Strong pipeline tracking continuity: Existing technologies lack target loss recovery mechanisms, easily interrupting tasks. This application uses biomimetic behavior strategies (search, track, recover) to relocate pipelines using comb-shaped or zigzag paths, improving task success rates. 3. Flexible multi-task switching: Traditional architectures are slow and difficult to adapt to multi-task environments. This application dynamically switches between cruise / fine modes based on environmental perception, optimizing response speed and efficiency. 4. High reliability: Fault detection at the action layer and safety modes at the planning layer (e.g., surfacing) ensure system stability under abnormal conditions, meeting long-term dwell requirements. Attached Figure Description

[0047] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 A flowchart illustrating an adaptive control method for an underwater resident robot provided in an embodiment of this application;

[0050] Figure 2 This is a schematic flowchart of the method for generating control instructions provided in an embodiment of this application;

[0051] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0053] Traditional underwater robot control methods have significant shortcomings when facing the following challenges:

[0054] Insufficient adaptability to complex underwater environments: The deep-sea environment is characterized by high pressure, low temperature, strong ocean currents and complex terrain. Traditional control methods (such as fixed parameter PID control) are difficult to effectively cope with dynamic disturbances (such as changes in ocean currents) and model uncertainties (such as changes in fluid resistance), resulting in decreased control accuracy or mission failure.

[0055] Stability and target loss issues in pipeline tracking: In subsea pipeline inspection missions, existing methods rely on sonar imagery for target identification. However, due to limitations in sonar resolution and underwater visibility, pipeline targets are easily lost. Existing tracking strategies lack effective target recovery mechanisms, making it difficult to guarantee mission continuity and reliability.

[0056] Insufficient flexibility in multi-task collaboration and mode switching: Underwater robots need to switch flexibly between different task modes such as cruise observation, fine observation, obstacle avoidance, and docking. However, traditional control system architectures are usually designed for a single mode, which makes it difficult to dynamically adapt to the needs of multiple tasks, resulting in low efficiency or inability to cope with emergencies (such as the approach of obstacles).

[0057] High reliability requirements for long-term deployment: RAUVs ​​need to be deployed in the deep-sea environment for extended periods (e.g., 28 days), requiring the control system to have high robustness and fault self-diagnosis capabilities. However, existing technologies have limited capabilities in online fault detection and emergency handling, making it difficult to meet the reliability requirements of long-term missions.

[0058] The purpose of this invention is to improve control accuracy by combining adaptive robust PID control to compensate for ocean current disturbances in real time; to solve the target loss problem by adopting a biomimetic behavior strategy and designing search, tracking, and recapture stages; and to improve the flexibility and response speed of multi-task scenarios by driving dynamic mode switching through environmental perception, thereby ensuring the efficient and stable operation of RAUV in complex deep-sea environments.

[0059] Figure 1 This is a flowchart illustrating an adaptive control method for an underwater resident robot provided in an embodiment of this application. In some embodiments, the process may include the following operations:

[0060] Step 101: Based on the user's input task requirements, perform task planning and generate a sequence of navigation path points.

[0061] The task requirements input by the user are the operational objectives and environmental constraints specified by the user through the control interface. For example, the objective is to complete the inspection of subsea pipelines, while avoiding coral reefs and maintaining a depth of 200±5 meters.

[0062] Mission planning involves breaking down mission requirements into executable spatiotemporal path logic. For example, path logic: first, descend vertically to 200 meters → cruise along the pipeline → detour around obstacles.

[0063] A navigation waypoint sequence is a set of critical path nodes composed of three-dimensional coordinates and timestamps.

[0064] In some embodiments, the target location and environmental constraints in the task requirements can be parsed, an obstacle avoidance path can be generated by combining the seabed topographic map, and a path point sequence can be generated by interpolation according to the time / energy consumption optimization principle.

[0065] In some embodiments, the underwater resident robot may include multiple control modes. Switching between different modes can be performed via the following operations:

[0066] S10, acquire the observation mode of the underwater resident robot; wherein, the observation mode includes cruise mode and close-range observation mode.

[0067] Observation modes are a set of different operational strategies adopted by RAUVs ​​based on mission requirements and environmental conditions, determining motion parameters and sensor configurations. For example, cruise mode: similar to an aircraft cruising at high altitude, rapidly covering a large area. Cruise mode is a high-speed, long-range operating state, sacrificing local accuracy for operational efficiency.

[0068] Close-range observation mode is a low-speed, high-precision operating mode that prioritizes the quality of target identification.

[0069] In some embodiments, the current mode identifier can be read from the system status register; the mode parameter set (speed threshold, height threshold, etc.) can be parsed; and the controller gain parameters can be updated synchronously.

[0070] S11, when the underwater resident robot is in cruise mode, it switches to close-range observation mode in response to detecting that the state of the characteristic target meets the preset conditions.

[0071] Characteristic targets are key observation objects designated by the mission, possessing identifiable physical features. For example, submarine pipelines: tubular structures with diameters of 0.5-1.2 meters; shipwreck debris: highly reflective metallic objects.

[0072] The characteristic target state is the complete set of attributes of the target object captured by the sensor. For example, presence state: target visible / partially occluded / completely lost; spatial state: distance 3.2 meters, azimuth angle 45°.

[0073] The preset conditions are complex logic rules that trigger mode switching. For example, IF (target distance ≤ 5 meters) AND (target recognition confidence > 80%) THEN switch.

[0074] S12, when the underwater resident robot is in close-range observation mode, it switches to cruise mode in response to the failure to detect the feature target within a preset distance range.

[0075] The preset distance range is the effective operating radius for maintaining close-range observation. For example, pipeline tracking: a spherical space with a radius of 3 meters, centered on the robot's center.

[0076] The absence of a detected feature indicates that the target has continuously disappeared for an extended period. For example, if the disappearance of pipeline features in a sonar image lasts for more than 5 seconds, or if the target moves out of the field of view within the optical field of view.

[0077] During underwater pipeline inspections, the RAUV's control logic is as follows: When the RAUV is in cruise mode, if the detection system identifies a characteristic target or senses an approaching obstacle, the system automatically switches to fine mode and updates the control parameter set to fine parameters to meet the needs of fine inspection. Conversely, when the RAUV is running in fine mode, once it detects no target features in the current area and has successfully maintained a safe distance (i.e., there is no obstacle interference), the system switches back to cruise mode and updates the control parameter set to cruise parameters to ensure efficient and stable operation of the RAUV under normal inspection conditions. This control logic helps the RAUV flexibly respond to various situations in complex underwater environments, achieving accurate and efficient pipeline inspection tasks.

[0078] In some embodiments, the step of planning tasks based on user-input task requirements and generating a sequence of navigation waypoints may include the following operations:

[0079] S20, based on the task requirements input by the user, determine the input data for the task planning algorithm; wherein, the input data includes task type, geographical constraints, environmental constraints, time constraints, and target point set.

[0080] Task planning algorithms are decision logic modules that transform task requirements into executable paths. For example, spatiotemporal constraint path planning algorithms.

[0081] Task type is the core objective category of an operation. For example, pipeline inspection, target search, and geological sampling.

[0082] Geographic constraints limit the spatial scope of task execution. For example, the coordinates of the work area are: longitude 120.1°120.5°E, latitude 24.8°25.2°N, and depth range: 180-220 meters.

[0083] Environmental constraints are rules for avoiding specific hazardous environments. For example, entry into steep slope areas (slope > 30°) is prohibited, and the location of known shipwrecks should be avoided (within a 50-meter radius restricted area).

[0084] Time constraints are the time-sensitive requirements for task execution. For example, if the total task duration is ≤3 hours, the critical node time is: must arrive at point B before 14:00.

[0085] The target point set is a sequence of critical locations that the task must reach. For example, starting point: P0 (120.1°E, 24.8°N, 200m); checkpoint: P1 (120.3°E, 25.0°N, 195m); endpoint: P2 (120.5°E, 25.2°N, 210m).

[0086] S21, input the input data into the task planning algorithm to generate the navigation path point sequence.

[0087] In some embodiments, the task planning algorithm is a spatiotemporal constrained path planning algorithm; the spatiotemporal constrained path planning algorithm is shown in the following formula (1):

[0088]

[0089] Where P is the path curve, P i w represents the coordinates of the path point. t w is the path length weighting factor. r R(s) is the risk cost weighting factor, and R(s) is the risk assessment function at position s; the output is the navigation path point sequence P = [P start ,P1,P2,…,P end ].

[0090] Step 102: Determine the target task type based on the navigation path point sequence, and output the expected velocity and expected attitude angle based on the target task type.

[0091] The target task type is the category of operations that needs to be prioritized at present. For example, cruise observation (large-scale movement), pipeline tracking (precise movement along pipelines), and emergency obstacle avoidance (avoiding sudden obstacles).

[0092] The desired velocity is the ideal velocity vector required to achieve the current task. For example, cruise mode: forward speed 1.5 m / s, fine mode: lateral translation speed 0.3 m / s.

[0093] The desired attitude angle is the three-axis spatial orientation that the robot needs to maintain. For example, tracking a ramp pipeline: pitch angle 10°, resisting ocean currents: yaw angle 15°.

[0094] In some embodiments, the environment around the next waypoint can be detected (sonar / optical sensors), and decisions can be made based on the waypoint spacing and task type: long-distance movement → cruise observation, approaching pipeline → fine tracking; the velocity / attitude angle corresponding to the task can be output.

[0095] In some embodiments, the candidate task type includes a pipeline tracking type and an obstacle avoidance mode type; under the pipeline tracking type and the obstacle avoidance mode type, the step of outputting the desired velocity and desired attitude angle based on the target task type includes:

[0096] The desired speed is determined by the following formula (2):

[0097]

[0098] Among them, v d For the desired speed, v base Based on the base speed, e p The lateral positional deviation from the target point is denoted by , and the vertical distance from the pipeline center to the underwater resident robot is denoted by . Let ψ be the rate of change of position deviation. pipe ψ is the direction angle. d k is the desired direction angle. p For proportional gain, k d Where L is the differential gain and L is the forward look distance;

[0099] The final heading correction is determined based on the following formula (3), and the desired attitude angle is determined based on the current heading and the final heading correction.

[0100]

[0101] Where Δψ is the final heading correction, Δψ i The suggested avoidance direction for the i-th obstacle is... d is the weighting factor. i The distance to the obstacle.

[0102] In some embodiments, determining the target task type based on the navigation waypoint sequence may include the following operations:

[0103] S30, Based on the task requirements, determine the task instructions.

[0104] Task requirements are a set of operational objectives and execution conditions set by the user. For example, completing the detection of submarine fiber optic cable breaks, avoiding areas with strong ocean currents, and searching for the bow and stern positions of shipwreck X within 28 minutes.

[0105] Task instructions are executable control commands generated after task requirements are parsed. For example, fiber optic cable inspection task instruction: {Type: fine observation, target coordinates: [120.8°E, 24.5°N], constraint: depth 200±5 meters}, shipwreck search instruction: {Type: target search, key points: [bow, stern], time limit: 1680 seconds}.

[0106] In some embodiments, action keywords (such as detection and search) in the task requirement text can be parsed, spatial parameters (coordinates and depth) and time parameters (time limit) can be extracted, and mapped to a standard instruction format.

[0107] S31, Based on the task instruction type, determine the candidate task type.

[0108] Candidate task types are the set of alternative operational behaviors required to implement an instruction. For example, candidate tasks for pipeline leak detection include: [pipeline tracking, fixed-point observation, obstacle avoidance], and candidate tasks for shipwreck search include: [cruise scanning, fine-grained identification, real-time collision avoidance].

[0109] S32, acquire underwater environment information around the underwater resident robot, and determine the task priority corresponding to each candidate task type based on the underwater environment information.

[0110] Underwater environmental information is a set of real-time physical environment parameters perceived by the robot. For example, obstacle data: {nearest obstacle distance: 3.2 meters, direction: 45°}, ocean current data: {velocity: 0.8 meters / second, current direction: 120°}.

[0111] Task priority is a quantified value representing the urgency of a candidate task in the current environment. For example, obstacle avoidance task priority: 10 (highest), pipeline tracking priority: 6. The priority of each task can be preset.

[0112] S33, Based on the priority of the task, determine the target task type from the candidate task types.

[0113] The priority of a task is a comparison of the priority values ​​of each candidate task.

[0114] Step 103: Based on the desired speed and desired attitude angle, control commands are generated using an adaptive robust PID controller.

[0115] An adaptive robust PID controller is a motion controller with disturbance compensation and parameter self-tuning capabilities. For example, it can automatically increase control gain to maintain the path in ocean currents.

[0116] Control commands are a set of physical force commands that drive the thrusters. For example, {left front thruster: 120N, right front thruster: 80N, vertical thruster: 50N}.

[0117] Step 104: Execute control commands and feed back sensor data to the control platform.

[0118] Executing control commands is the process by which the thruster system converts electrical signals into physical thrust.

[0119] Sensor data are physical quantities that reflect the robot's motion state and environmental parameters. For example, pose data: pitch angle +2.3°, depth 205.6 meters; environmental data: ocean current velocity 0.8 m / s (direction NE).

[0120] The control platform is the central computing hub for processing decision-making commands. For example, a shipboard control console displays real-time three-dimensional motion trajectories.

[0121] Figure 2 This is a schematic flowchart of a method for generating control instructions provided in an embodiment of this application. In some embodiments, the process may include the following operations:

[0122] Step 201: Based on the desired speed and the desired attitude angle, determine the attitude deviation and speed preset value through the outer loop controller of the adaptive robust PID controller.

[0123] The outer loop controller is the control module responsible for macroscopic pose control. For example, its function is to convert position deviations into speed adjustment commands.

[0124] Attitude deviation is the difference between the actual attitude angle and the desired attitude angle. For example, the desired yaw angle is 45°, the actual yaw angle is 40°, and the deviation is +5°.

[0125] The speed preset value is the baseline speed that the inner loop controller needs to achieve. For example, the outer loop output is: forward preset speed of 1.0 m / s + turning preset angular velocity of 10° / s.

[0126] Step 202: Based on the preset speed value, the control command is generated by the inner loop controller of the adaptive robust PID controller.

[0127] The inner-loop controller is the execution-level control module responsible for precise speed tracking. For example, it translates speed commands into thruster thrust.

[0128] Control commands are a set of physical force commands that drive each thruster.

[0129] Step 203: Based on the status feedback data after the underwater resident robot executes the control command, update the adaptive parameter items of the adaptive robust PID controller.

[0130] Status feedback data consists of motion status parameters collected in real time after the execution of control commands. For example, new pose data: pitch angle 9°, yaw angle 28°; new velocity data: forward 0.95 m / s, vertical 0.25 m / s.

[0131] The adaptive parameter is a compensation coefficient used internally by the controller for dynamic adjustment. For example, ocean current compensation gain: current value 1.2, attitude response coefficient: current value 0.8.

[0132] The core idea of ​​the technical solution provided in this application is to estimate and compensate for unknown disturbances (ocean currents, model uncertainties) in real time, and proposes parameter adaptive rules, such as defining a vector of unknown system parameters.

[0133] The adaptive update rule can be shown in the following formula (4):

[0134]

[0135] in, Here, Γ represents the parameter estimates, γ2 is the adaptive gain matrix (γ2>0), s is the sliding mode variable (defined later), and φ is the regression matrix (system state function). Formula (4) is the adaptive parameter term of the adaptive robust PID controller.

[0136] The sliding mode variable is designed as follows: e = η d -η, Λ=diag (λ1, λ2, λ3, λ4, λ5, λ6), convergence rate matrix λ i >0.

[0137] The overall controller is designed as shown in the following formula (5):

[0138]

[0139] M is the inertia matrix (including added mass), C(v) is the Coriolis matrix, D(v) is the damping matrix, and v r For reference speed, K p K d This is the PID gain matrix.

[0140] Dual-loop adaptive structure: outer loop: pose tracking, inner loop: speed control.

[0141] The technical solution of this application can be implemented based on the four-layer control architecture of MOOS: designing a hierarchical control system including a device layer, an action layer, a behavior layer and a planning layer (which can correspond to the device module, action module, behavior module and planning module respectively), and realizing data communication and task coordination through standard interfaces (such as the MOOS database MOOSDB), supporting autonomous / semi-autonomous control of RAUV (Rapidly Ambient Underwater Vehicle) in complex deep-sea environments.

[0142] The equipment layer mainly manages various devices such as Doppler logs, GPS, inertial measurement units, underwater acoustic communication devices, two-dimensional imaging sonar, and thrusters, and performs basic filtering processing on the raw output data.

[0143] The motion layer implements the carrier's integrated navigation and positioning, as well as basic motion control such as speed and attitude. It distributes the thrust and torque output from the motion control to each thruster according to a vector arrangement, controls the motor speed, records all data, monitors the fault status of all equipment, and performs emergency handling. Furthermore, upon receiving information on the desired speed, attitude angle, and desired position, it combines adaptive robust PID control to complete motion control. In each behavior mode, it dynamically adjusts control parameters based on real-time feedback of control errors; and based on the current behavior mode and control parameters, it calculates the control input and executes control commands through the actuator module.

[0144] The behavior layer implements behaviors such as fixed-point navigation, trajectory tracking, pipeline tracking, real-time collision avoidance, and docking guidance, outputting desired speed and desired attitude angle as control inputs to the action layer. It also determines which task needs to be performed currently (fixed-point navigation, trajectory tracking, pipeline tracking, real-time collision avoidance, or docking guidance), and then outputs the desired speed and desired attitude angle as control inputs to the action layer. Behavioral models for cruise observation mode and close-range observation mode are established at the behavior layer; underwater environmental information is acquired in real time through the environmental perception module; and the RAUV's behavior mode is dynamically switched according to the environmental perception information and operational task requirements.

[0145] The planning layer mainly plans mission tasks based on user-defined missions and monitors the execution of mission tasks throughout the system. It also maintains communication with the surface control platform through a human-machine interface.

[0146] Planning layer: Generates task plans based on requirements. Then, the tasks are distributed to the behavior layer.

[0147] Biomimetic behavior pipeline tracking strategy: A pipeline tracking method based on biomimetic behavior is proposed, which divides the tracking process into three stages: search, track and retrieve. The sonar target loss problem is solved by using comb or zigzag search paths to ensure mission continuity and success rate.

[0148] The dual-loop structure of adaptive robust PID control: adopts a dual-loop structure of outer loop pose tracking and inner loop velocity control, and improves control accuracy and robustness by real-time compensation of ocean current disturbances and model uncertainties through sliding mode variables and parameter adaptive laws (such as ().

[0149] Dynamic behavior mode switching mechanism: At the behavior layer, underwater environmental information is acquired in real time through the environmental perception module, and the cruise observation mode (speed 1.5m / s) and close-range observation mode (speed 0.8m / s) are dynamically switched. The task execution is adjusted according to the behavior priority (such as obstacle avoidance is the highest priority), which improves the flexibility of multi-task scenarios.

[0150] In some embodiments, online fault detection and safety modes may also be included: an online fault detection module is integrated into the action layer to monitor the status of sensors and thrusters, and a safety mode (such as surfacing or hovering) is triggered in conjunction with the state tracking matrix of the planning layer to ensure the reliability of long-term stationary missions.

[0151] Based on the same inventive concept, this application also provides an adaptive control system for an underwater resident robot, the system being used to control the underwater resident robot, the system comprising:

[0152] The planning module is used to plan tasks based on user-inputted task requirements and generate a sequence of navigation path points.

[0153] The behavior module is used to determine the target task type based on the navigation path point sequence, and output the desired velocity and desired attitude angle based on the target task type;

[0154] The action module is used to generate control commands based on the desired speed and desired attitude angle using an adaptive robust PID controller;

[0155] The device module is used to execute the control commands and feed back sensor data to the control platform.

[0156] like Figure 3 As shown in the figure, this application provides an electronic device, including a processor 111, a communication interface 112, a memory 113, and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114.

[0157] Memory 113 is used to store computer programs;

[0158] In one embodiment of this application, processor 111, when executing a program stored in memory 113, implements the adaptive control method for an underwater resident robot provided in any of the foregoing method embodiments. The method is applied to the underwater resident robot and includes:

[0159] Based on the user's input task requirements, task planning is performed to generate a sequence of navigation path points;

[0160] Based on the navigation path point sequence, determine the target task type, and output the desired velocity and desired attitude angle based on the target task type;

[0161] Based on the desired speed and desired attitude angle, control commands are generated using an adaptive robust PID controller.

[0162] The control commands are executed and sensor data is fed back to the control platform.

[0163] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the adaptive control method for an underwater resident robot as provided in any of the foregoing method embodiments.

[0164] It should be noted that, in this document, relational terms such as "first" and "second" are used merely 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 statement "comprising a…" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0165] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. An adaptive control method for an underwater resident robot, characterized in that, The method is applied to underwater resident robots, including: Based on the user's input task requirements, task planning is performed, and a sequence of navigation path points is generated. Based on the navigation path point sequence, determine the target task type, and output the desired velocity and desired attitude angle based on the target task type; Based on the desired speed and desired attitude angle, control commands are generated using an adaptive robust PID controller. The control commands are executed and sensor data is fed back to the control platform.

2. The method according to claim 1, characterized in that, The method further includes: The observation modes of the underwater resident robot are obtained; wherein the observation modes include a cruise mode and a close-range observation mode. When the underwater resident robot is in cruise mode, it switches to close-range observation mode in response to detecting that the state of a feature target meets the preset conditions. When the underwater resident robot is in close-range observation mode, it switches to cruise mode in response to the failure to detect the feature target within a preset distance range.

3. The method according to claim 1, characterized in that, The process of planning tasks based on user-inputted task requirements and generating a sequence of navigation path points includes: Based on the user's input task requirements, the input data for the task planning algorithm is determined; wherein, the input data includes task type, geographical constraints, environmental constraints, time constraints, and target point set; The input data is fed into the task planning algorithm to generate the navigation path point sequence.

4. The method according to claim 3, characterized in that, The task planning algorithm is a spatiotemporally constrained path planning algorithm. The spatiotemporal constrained path planning algorithm is shown in the following formula: Where P is the path curve, P i w represents the coordinates of the path point. t w is the path length weighting factor. r R(s) is the risk cost weighting factor, and R(s) is the risk assessment function at position s; the output is the navigation path point sequence P = [P start ,P1,P2,…,P end ].

5. The method according to claim 1, characterized in that, The step of determining the target task type based on the navigation path point sequence includes: Based on the aforementioned task requirements, determine the task instructions; Based on the task instruction type, determine the candidate task type; Acquire underwater environment information around the underwater resident robot, and determine the task priority corresponding to each candidate task type based on the underwater environment information; Based on the priority of the task, the target task type is determined from the candidate task types.

6. The method according to claim 5, characterized in that, The candidate task types include pipeline tracking type and obstacle avoidance mode type; Under the pipeline tracking type and the obstacle avoidance mode type, the step of outputting the desired velocity and desired attitude angle based on the target task type includes: The desired speed is determined using the following formula: Among them, v d For the desired speed, v base Based on the base speed, e p The lateral positional deviation from the target point is denoted by , and the vertical distance from the pipeline center to the underwater resident robot is denoted by . Let ψ be the rate of change of position deviation. pipe ψ is the direction angle. d k is the desired direction angle. p For proportional gain, k d Where L is the differential gain and L is the forward look distance; The final heading correction is determined based on the following formula, and the desired attitude angle is determined based on the current heading and the final heading correction. Where Δψ is the final heading correction, Δψ i The suggested avoidance direction for the i-th obstacle is... d is the weighting factor. i The distance to the obstacle.

7. The method according to claim 1, characterized in that, The process of generating control commands based on the desired velocity and desired attitude angle using an adaptive robust PID controller includes: Based on the desired speed and the desired attitude angle, the attitude deviation and speed preset value are determined by the outer loop controller of the adaptive robust PID controller; Based on the preset speed value, the control command is generated by the inner loop controller of the adaptive robust PID controller; Based on the status feedback data after the underwater resident robot executes the control command, the adaptive parameter items of the adaptive robust PID controller are updated.

8. An adaptive control system for an underwater resident robot, characterized in that, The system is used to control an underwater resident robot, and the system includes: The planning module is used to plan tasks based on user-inputted task requirements and generate a sequence of navigation path points. The behavior module is used to determine the target task type based on the navigation path point sequence, and output the desired velocity and desired attitude angle based on the target task type; The action module is used to generate control commands based on the desired speed and desired attitude angle using an adaptive robust PID controller; The device module is used to execute the control commands and feed back sensor data to the control platform.

9. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When a processor executes a program stored in a memory, it implements the steps of the adaptive control method for an underwater resident robot according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the adaptive control method for an underwater resident robot as described in any one of claims 1-7.

Images